From myles at myles.id.au  Fri Aug  1 02:27:23 2008
From: myles at myles.id.au (Myles Byrne)
Date: Fri, 1 Aug 2008 16:27:23 +1000
Subject: [ditz-talk] First impressions
In-Reply-To: <200807310949.39086.ninja@slaphack.com>
References: <200807290315.16589.ninja@slaphack.com>
	<200807292325.27659.ninja@slaphack.com> <1217449815-sup-4025@entry>
	<200807310949.39086.ninja@slaphack.com>
Message-ID: <128508630807312327m6746d9e6ud26f91a57d1ab803@mail.gmail.com>

On Fri, Aug 1, 2008 at 12:49 AM, David Masover <ninja at slaphack.com> wrote:

> Bugfix/Feature/Task should probably be set to some default, as I'm assuming it
> must be provided.

Second this. Task is what I used 90% of the time. Bugfix for defects
and Feature for "ideas".

> Changing who submitted it isn't a common operation, and is deceptive enough
> that it might be worthwhile to only allow this through either hacking Ditz,
> or changing the username in setup.

Or even via the command line:

  $ ditz add different at submitter.com
  Submitter: different at submitter.com
  Title: <- cursor here

by default it could show your address as the submitter. So if you
actually intend to add an issue as a different user you'd see straight
away that you forgot add the user arg and could hit ctrl+d and start
again.

> One more thing: For a one-project setup, having to type <project>-123 at the
> commandline is annoying. That could be implied.

This would be awesome too.

Thanks for ditz!

-- Myles

From wmorgan-ditz at masanjin.net  Sun Aug  3 13:26:12 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 03 Aug 2008 10:26:12 -0700
Subject: [ditz-talk] Some work on ditz
In-Reply-To: <1217545108-sup-6970@entry>
References: <8e50e92e0807291017q54a32a28y81e205aede8e9162@mail.gmail.com>
	<1217354174-sup-9459@entry>
	<8e50e92e0807291321w7dac21aal90cf9ddfdd868c4@mail.gmail.com>
	<1217448631-sup-803@entry>
	<8e50e92e0807301433j18472c58o25d241df4c8be0a@mail.gmail.com>
	<1217454686-sup-8609@entry>
	<20080730224258.GA1931@yasmin.technion.ac.il>
	<1217458164-sup-6885@entry>
	<8e50e92e0807310250s45e7b979v45e5e9785fd24e0f@mail.gmail.com>
	<1217535095-sup-9161@entry>
	<8e50e92e0807311403u7e3cd8d6s19d96f7dbd9facc@mail.gmail.com>
	<1217545108-sup-6970@entry>
Message-ID: <1217784239-sup-9137@entry>

Reformatted excerpts from William Morgan's message of 2008-07-31:
> I don't think so. I'm probably going to take your branch, squash the
> earlier and later commits together, and elide the trollop changes.
> Someone else just submitted some trollop patches, so I'd like to keep
> trollop development out of ditz as much as possible.

This is merged in, I've updated lib/trollop.rb with the latest version,
ditz help --cow works, and I've marked the issue as closed. Thanks!
-- 
William <wmorgan-ditz at masanjin.net>

From matt at tplus1.com  Sun Aug  3 15:31:49 2008
From: matt at tplus1.com (Matthew Wilson)
Date: Sun, 3 Aug 2008 15:31:49 -0400
Subject: [ditz-talk] I don't know how to use git, so here's my patch
Message-ID: <f6f643a20808031231l73681a35n19a0df2ee1b7b2ca@mail.gmail.com>

I'll read the git tutorial later and learn how to do this the cool way.

This patch adds DOCTYPES to the tops of the HTML templates.  It closes
a few <p> tags and puts the link at the top of all the pages inside a
div.

The pages appear the same as before, but they're valid now (at least
according to my validator).

Anyhow, ditz looks really neat after playing around with this morning.

Matt

--
Matthew Wilson
matt at tplus1.com
http://tplus1.com
216-470-6058
-------------- next part --------------
A non-text attachment was scrubbed...
Name: my_patch.diff
Type: text/x-diff
Size: 4021 bytes
Desc: not available
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080803/d3edf0de/attachment-0001.bin>

From wmorgan-ditz at masanjin.net  Sun Aug  3 18:39:31 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 03 Aug 2008 15:39:31 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <200807310949.39086.ninja@slaphack.com>
References: <200807290315.16589.ninja@slaphack.com>
	<200807292325.27659.ninja@slaphack.com> <1217449815-sup-4025@entry>
	<200807310949.39086.ninja@slaphack.com>
Message-ID: <1217801806-sup-5317@entry>

Reformatted excerpts from David Masover's message of 2008-07-31:
> Couple things I'm starting to notice, though -- one thing I miss from
> Trac is the ability to go back and edit other parts of the issue
> before I submit.

There's ditz edit, but that generates a log message. I've just added an
issue requesting a --silent flag to turn off logging, so that would be
useful in this case.

> Could probably get the same effect by providing an editor-friendly
> format (not yaml) in which to enter a complete ticket. 

That seems like a different use case to me. Personally I consider YAML
to be editor-friendly, but in general I'm not adverse to other input
mechanisms besides what's there now. So, make a proposal.

> Bugfix/Feature/Task should probably be set to some default, as I'm
> assuming it must be provided.

If you can come up with a default that satisfies everyone, sure. I tend
to be about 50% between bugs and features. We could do something like
"default to the last type the user picked", and then store that
somewhere... dunno.

> One more thing: For a one-project setup, having to type <project>-123
> at the commandline is annoying. That could be implied.

With tab completion I don't see this being that big of a deal. And in
issue titles, comments and descriptions, the issue names have to be
identifiable (because we replace them with issue ids within the yaml),
and breaking consistency between "what I type on the commandline" and
"what I type in the issue itself to refer to another feature" seems bad.
-- 
William <wmorgan-ditz at masanjin.net>

From ninja at slaphack.com  Sun Aug  3 19:02:56 2008
From: ninja at slaphack.com (David Masover)
Date: Sun, 3 Aug 2008 18:02:56 -0500
Subject: [ditz-talk] First impressions
In-Reply-To: <1217801806-sup-5317@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200807310949.39086.ninja@slaphack.com> <1217801806-sup-5317@entry>
Message-ID: <200808031802.56706.ninja@slaphack.com>

On Sunday 03 August 2008 17:39:31 William Morgan wrote:
> Reformatted excerpts from David Masover's message of 2008-07-31:
> > Couple things I'm starting to notice, though -- one thing I miss from
> > Trac is the ability to go back and edit other parts of the issue
> > before I submit.
> 
> There's ditz edit, but that generates a log message. I've just added an
> issue requesting a --silent flag to turn off logging, so that would be
> useful in this case.
> 
> > Could probably get the same effect by providing an editor-friendly
> > format (not yaml) in which to enter a complete ticket. 
> 
> That seems like a different use case to me. Personally I consider YAML
> to be editor-friendly, but in general I'm not adverse to other input
> mechanisms besides what's there now. So, make a proposal.

Something like:


Title:

Type: (b)ugfix, (f)eature, (t)ask?
b
Description:

Issue creator:
My Name <autofilled at example.com>


Throw it all in an editor, and the result is: Press the down arrow once to 
start typing a title, then twice to edit bugfix/feature/task (and I am 
proposing a default), twice more to fill in a description. Filled out, it'd 
look like this:


Title:
Some Ticket
Type: (b)ugfix, (f)eature, (t)ask?
b
Description:
Filling in some description
Issue creator:
My Name <autofilled at example.com>


The idea is, up and down arrows in the editor are going to be a lot faster 
than save/quit + reopen with "ditz edit". I'm speaking mostly of the process 
of initially posting the issue, though it might also help with editing.

I think that's almost as quick as tabbing between fields in a GUI, and it has 
the added benefit of using your favorite editor to edit the description.

And I agree, YAML is editor-friendly -- but in this simple case, I think the 
special-purpose format is both faster and more readable.

> > Bugfix/Feature/Task should probably be set to some default, as I'm
> > assuming it must be provided.
> 
> If you can come up with a default that satisfies everyone, sure.

Point is not that there needs to be one in Ditz itself, but that it should be 
a config option.

> > One more thing: For a one-project setup, having to type <project>-123
> > at the commandline is annoying. That could be implied.
> 
> With tab completion I don't see this being that big of a deal.

Where can I get Bash tab-completion for Ditz? Or was that a proposal?

Could imply it everywhere in that project, including editing issues -- kind of 
nice how, in Trac, I only need to refer to issue #384 (say) and it becomes a 
link.

It'd still be bad if that one Ditz installation had a separate project added, 
maybe -- either you'd have errors, or you'd have to always assume the old 
default project.

From wmorgan-ditz at masanjin.net  Sun Aug  3 19:07:08 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 03 Aug 2008 16:07:08 -0700
Subject: [ditz-talk] I don't know how to use git, so here's my patch
In-Reply-To: <f6f643a20808031231l73681a35n19a0df2ee1b7b2ca@mail.gmail.com>
References: <f6f643a20808031231l73681a35n19a0df2ee1b7b2ca@mail.gmail.com>
Message-ID: <1217804737-sup-1247@entry>

Reformatted excerpts from Matthew Wilson's message of 2008-08-03:
> I'll read the git tutorial later and learn how to do this the cool way.

For this type of thing, it's "very simple". git commit -a, add a
one-line description followed by a blank line and more commentary, and
then git format-patch HEAD^.

> This patch adds DOCTYPES to the tops of the HTML templates.  It closes
> a few <p> tags and puts the link at the top of all the pages inside a
> div.

Applied. Thanks!

> Anyhow, ditz looks really neat after playing around with this morning.

Glad to hear it.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Sun Aug  3 21:58:27 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 03 Aug 2008 18:58:27 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <200808031802.56706.ninja@slaphack.com>
References: <200807290315.16589.ninja@slaphack.com>
	<200807310949.39086.ninja@slaphack.com> <1217801806-sup-5317@entry>
	<200808031802.56706.ninja@slaphack.com>
Message-ID: <1217814161-sup-4774@entry>

Reformatted excerpts from David Masover's message of 2008-08-03:
> Throw it all in an editor, and the result is: Press the down arrow
> once to start typing a title, then twice to edit bugfix/feature/task
> (and I am proposing a default), twice more to fill in a description.

That sounds perfectly reasonable to me. I would accept a patch.

> Point is not that there needs to be one in Ditz itself, but that it
> should be a config option.

Sure, I'd accept a patch for this.

> Where can I get Bash tab-completion for Ditz? Or was that a proposal?

It's in contrib/completion/ditz.bash. It works really well.

> Could imply it everywhere in that project, including editing issues --
> kind of nice how, in Trac, I only need to refer to issue #384 (say)
> and it becomes a link.

I'm not entirely sure I understand, but if currently if you refer to any
issue by its name (e.g. ditz-1) in the title, description, or comments
of an issue, that will get turned into a HTML link by "ditz html".

> It'd still be bad if that one Ditz installation had a separate project
> added, maybe -- either you'd have errors, or you'd have to always
> assume the old default project.

Multiple projects is definitely not supported by ditz, except to the
extent that you can specify --issue-dir.
-- 
William <wmorgan-ditz at masanjin.net>

From ninja at slaphack.com  Sun Aug  3 23:17:58 2008
From: ninja at slaphack.com (David Masover)
Date: Sun, 3 Aug 2008 22:17:58 -0500
Subject: [ditz-talk] First impressions
In-Reply-To: <1217814161-sup-4774@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200808031802.56706.ninja@slaphack.com> <1217814161-sup-4774@entry>
Message-ID: <200808032217.58787.ninja@slaphack.com>

On Sunday 03 August 2008 20:58:27 William Morgan wrote:
> Reformatted excerpts from David Masover's message of 2008-08-03:

> It's in contrib/completion/ditz.bash. It works really well.

Cool, thanks. I'll give it a try, next time I make any progress on a 
Ditz-enabled project.

> > Could imply it everywhere in that project, including editing issues --
> > kind of nice how, in Trac, I only need to refer to issue #384 (say)
> > and it becomes a link.
> 
> I'm not entirely sure I understand, but if currently if you refer to any
> issue by its name (e.g. ditz-1) in the title, description, or comments
> of an issue, that will get turned into a HTML link by "ditz html".

My point is that if there is only one component, it makes more sense to make 
this something short. I like Trac's notation -- anything that looks like a 
pound-sign followed by an integer (as in #123) is treated as a link to the 
ticket identified by that number (ditz uses "issue" instead of "ticket", but 
the concept is the same).

Another concept from Trac is having issue numbers be globally unique, so that 
even with multiple projects, you never have to type more than the issue 
number. I think this makes a lot of sense, as it's not as though we're going 
to memorize these numbers -- and in the process of looking them up, we can 
readily see what project they're from.

The only problem with this is that it might suggest a way of escaping things, 
so it could get complicated, fast. How would you feel about adapting some 
existing text-to-HTML engine, like Markdown or Textile? Or just backslashing 
them...

I'd almost just let it be -- false positives aren't particularly bad here. If 
I want to say "I'm #1!" and it links to issue #1, so what?

> > It'd still be bad if that one Ditz installation had a separate project
> > added, maybe -- either you'd have errors, or you'd have to always
> > assume the old default project.
> 
> Multiple projects is definitely not supported by ditz, except to the
> extent that you can specify --issue-dir.

Ah... Looks like I meant "components". See above, then.


I probably won't be able to work on this till next weekend -- unfortunately, 
we're still using Trac+SVN at work, and the migration path looks to be 
Trac+Git. But it's an immediate personal itch, so sometime this month...

From ohad at lutzky.net  Mon Aug  4 07:15:09 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Mon, 4 Aug 2008 14:15:09 +0300
Subject: [ditz-talk] [merge] git options for ditz commit
Message-ID: <8e50e92e0808040415u486c1c5ci78686893d82783de@mail.gmail.com>

Hi,

Now that subcommand options work properly, I've implemented the -a, -m
and -v flags for ditz commit.

(How's this [merge] mail tag for merge requests?)

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From wmorgan-ditz at masanjin.net  Mon Aug  4 11:17:22 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Mon, 04 Aug 2008 08:17:22 -0700
Subject: [ditz-talk] [merge] git options for ditz commit
In-Reply-To: <8e50e92e0808040415u486c1c5ci78686893d82783de@mail.gmail.com>
References: <8e50e92e0808040415u486c1c5ci78686893d82783de@mail.gmail.com>
Message-ID: <1217863001-sup-9073@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-04:
> (How's this [merge] mail tag for merge requests?)

Just fine. It would be helpful if you could include the branch name too.

Merged, thanks!
-- 
William <wmorgan-ditz at masanjin.net>

From ohad at lutzky.net  Mon Aug  4 18:04:30 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Tue, 5 Aug 2008 01:04:30 +0300
Subject: [ditz-talk] [issue 535683] the previous reporter was me
Message-ID: <8e50e92e0808041504o6d1dfb65j6ba0ab289f91a461@mail.gmail.com>

I previously reported this issue (using ditz, in my own branch), but
it was never merged so I figured it was a dupe and deleted the branch.

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From wmorgan-ditz at masanjin.net  Mon Aug  4 18:51:27 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Mon, 04 Aug 2008 15:51:27 -0700
Subject: [ditz-talk] [issue 535683] the previous reporter was me
In-Reply-To: <8e50e92e0808041504o6d1dfb65j6ba0ab289f91a461@mail.gmail.com>
References: <8e50e92e0808041504o6d1dfb65j6ba0ab289f91a461@mail.gmail.com>
Message-ID: <1217890205-sup-6532@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-04:
> I previously reported this issue (using ditz, in my own branch), but
> it was never merged so I figured it was a dupe and deleted the branch.

Wow, I woke up this morning with that bug on my mind, but I couldn't
find it anywhere. I must've seen it when I was looking through your
repo.

Was it on your master? Was I supposed to merge that in?
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Wed Aug  6 12:24:46 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 06 Aug 2008 09:24:46 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <200808032217.58787.ninja@slaphack.com>
References: <200807290315.16589.ninja@slaphack.com>
	<200808031802.56706.ninja@slaphack.com> <1217814161-sup-4774@entry>
	<200808032217.58787.ninja@slaphack.com>
Message-ID: <1218039422-sup-6792@entry>

Reformatted excerpts from David Masover's message of 2008-08-03:
> My point is that if there is only one component, it makes more sense
> to make this something short. I like Trac's notation -- anything that
> looks like a pound-sign followed by an integer (as in #123) is treated
> as a link to the ticket identified by that number (ditz uses "issue"
> instead of "ticket", but the concept is the same).

I see. Yeah, this might be nice, because it does make things much
shorter in the single-component case. But I do like the current issue
name format for when you have multiple components (maybe only because
JIRA managed to burn itself into my brain), so if this were to happen, I
would want the transition to be seamless if you add components.

I've added this as a feature request issue. (2cd9e99c)

> I'd almost just let it be -- false positives aren't particularly bad
> here. If I want to say "I'm #1!" and it links to issue #1, so what?

Yeah, I think that's fine too, and prefereable to having to write
everything in yet another wiki syntax. Especially since the primary use
case for ditz is on the console, not on the web.
-- 
William <wmorgan-ditz at masanjin.net>

From ohad at lutzky.net  Wed Aug  6 13:01:56 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Wed, 6 Aug 2008 20:01:56 +0300
Subject: [ditz-talk] First impressions
In-Reply-To: <1218039422-sup-6792@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200808031802.56706.ninja@slaphack.com> <1217814161-sup-4774@entry>
	<200808032217.58787.ninja@slaphack.com> <1218039422-sup-6792@entry>
Message-ID: <8e50e92e0808061001x76e7b064hff761955ad4cc129@mail.gmail.com>

On Wed, Aug 6, 2008 at 7:24 PM, William Morgan
<wmorgan-ditz at masanjin.net> wrote:
> Reformatted excerpts from David Masover's message of 2008-08-03:
>> My point is that if there is only one component, it makes more sense
>> to make this something short. I like Trac's notation -- anything that
>> looks like a pound-sign followed by an integer (as in #123) is treated
>> as a link to the ticket identified by that number (ditz uses "issue"
>> instead of "ticket", but the concept is the same).
>
> I see. Yeah, this might be nice, because it does make things much
> shorter in the single-component case. But I do like the current issue
> name format for when you have multiple components (maybe only because
> JIRA managed to burn itself into my brain), so if this were to happen, I
> would want the transition to be seamless if you add components.
>
> I've added this as a feature request issue. (2cd9e99c)
>
>> I'd almost just let it be -- false positives aren't particularly bad
>> here. If I want to say "I'm #1!" and it links to issue #1, so what?
>
> Yeah, I think that's fine too, and prefereable to having to write
> everything in yet another wiki syntax. Especially since the primary use
> case for ditz is on the console, not on the web.

I don't see how all of this can work, as issues change numbers once in
a while (and only the sha1s remain).

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From wmorgan-ditz at masanjin.net  Wed Aug  6 14:22:44 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 06 Aug 2008 11:22:44 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <8e50e92e0808061001x76e7b064hff761955ad4cc129@mail.gmail.com>
References: <200807290315.16589.ninja@slaphack.com>
	<200808031802.56706.ninja@slaphack.com> <1217814161-sup-4774@entry>
	<200808032217.58787.ninja@slaphack.com> <1218039422-sup-6792@entry>
	<8e50e92e0808061001x76e7b064hff761955ad4cc129@mail.gmail.com>
Message-ID: <1218046853-sup-3624@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-06:
> On Wed, Aug 6, 2008 at 7:24 PM, William Morgan
> <wmorgan-ditz at masanjin.net> wrote:
> >> I'd almost just let it be -- false positives aren't particularly bad
> >> here. If I want to say "I'm #1!" and it links to issue #1, so what?
> >
> > Yeah, I think that's fine too, and prefereable to having to write
> > everything in yet another wiki syntax. Especially since the primary use
> > case for ditz is on the console, not on the web.
> 
> I don't see how all of this can work, as issues change numbers once in
> a while (and only the sha1s remain).

Issue names are replaced with ids before being stored to disk. So you
can safely refer to issue names within issue titles, descriptions and
comments. As long as the issue names take some detectable format, it
doesn't really matter what that format is.
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Sat Aug  9 09:20:40 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Sat,  9 Aug 2008 15:20:40 +0200
Subject: [ditz-talk] [PATCH] fix completion bug on strings like '-n'
Message-ID: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>

This fixes a bug in the bash completion that can be triggered by entering
something like "ditz -n" and pressing tab.

Regards,
Christian

---
 contrib/completion/ditz.bash |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/contrib/completion/ditz.bash b/contrib/completion/ditz.bash
index e2c6dc1..4f43e21 100644
--- a/contrib/completion/ditz.bash
+++ b/contrib/completion/ditz.bash
@@ -8,14 +8,14 @@ _ditz()
 {
     cur=${COMP_WORDS[COMP_CWORD]}
     if [ $COMP_CWORD -eq 1 ]; then
-	COMPREPLY=( $( compgen -W "$(ditz --commands)" $cur ) )
+	COMPREPLY=( $( compgen -W "$(ditz --commands)" -- $cur ) )
     elif [ $COMP_CWORD -eq 2 ]; then
 	cmd=${COMP_WORDS[1]}
-	COMPREPLY=( $( compgen -W "$(ditz "$cmd" '<options>' 2>/dev/null)" $cur ) )
+	COMPREPLY=( $( compgen -W "$(ditz "$cmd" '<options>' 2>/dev/null)" -- $cur ) )
     elif [ $COMP_CWORD -eq 3 ]; then
 	cmd=${COMP_WORDS[1]}
 	parm1=${COMP_WORDS[2]}
-	COMPREPLY=( $( compgen -W "$(ditz "$cmd" "$parm1" '<options>' 2>/dev/null)" $cur ) )
+	COMPREPLY=( $( compgen -W "$(ditz "$cmd" "$parm1" '<options>' 2>/dev/null)" -- $cur ) )
     fi 
 }
 
-- 
1.5.6.3



From wmorgan-ditz at masanjin.net  Mon Aug 11 16:08:22 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Mon, 11 Aug 2008 13:08:22 -0700
Subject: [ditz-talk] [PATCH] fix completion bug on strings like '-n'
In-Reply-To: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
References: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
Message-ID: <1218485251-sup-9896@entry>

Hi Christian,

Reformatted excerpts from Christian Garbs's message of 2008-08-09:
> This fixes a bug in the bash completion that can be triggered by
> entering something like "ditz -n" and pressing tab.

After applying this patch and re-sourcing the completion file, "ditz -n
<tab>" still falls back to listing the files in the current directory.
Am I doing something wrong?
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Mon Aug 11 16:47:48 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Mon, 11 Aug 2008 22:47:48 +0200
Subject: [ditz-talk] [PATCH] fix completion bug on strings like '-n'
In-Reply-To: <1218485251-sup-9896@entry>
References: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
	<1218485251-sup-9896@entry>
Message-ID: <20080811204747.GA9784@cgarbs.de>

On Mon, Aug 11, 2008 at 01:08:22PM -0700, William Morgan wrote:
> Hi Christian,
> 
> Reformatted excerpts from Christian Garbs's message of 2008-08-09:
> > This fixes a bug in the bash completion that can be triggered by
> > entering something like "ditz -n" and pressing tab.
> 
> After applying this patch and re-sourcing the completion file, "ditz -n
> <tab>" still falls back to listing the files in the current directory.
> Am I doing something wrong?

Before my change, "ditz -n<tab" (no space!) brought this error:

$ ditz -nbash: compgen: -n: invalid option
compgen: usage: compgen [-abcdefgjksuv] [-o option] [-A action] [-G
globpat] [-W wordlist] [-P prefix] [-S suffix] [-X filterpat] [-F
function] [-C command] [word]

The patch expands "-n<tab>" correctly to an empty list, as there is no
ditz command or option that starts with "-n".


"ditz -n <tab>" (with space) would expand to the first parameter of the
"-n" option.  "-n" is unkown to ditz, so falling back to the bash
default should be the right thing to do.
If you typed ">" instead of "-n" which is also unkown to ditz,
the default expansion to filenames and directories is quite useful.

This behaviour is unchanged by my patch.

Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Corry's Law:
        Paper is always strongest at the perforations.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080811/98377393/attachment.bin>

From mitch at cgarbs.de  Mon Aug 11 16:50:47 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Mon, 11 Aug 2008 22:50:47 +0200
Subject: [ditz-talk] [PATCH] don't pollute global namespace with local
	variables
In-Reply-To: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
References: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
Message-ID: <1218487847-11259-1-git-send-email-mitch@cgarbs.de>

---
 contrib/completion/ditz.bash |    8 ++++----
 1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/contrib/completion/ditz.bash b/contrib/completion/ditz.bash
index 4f43e21..4a17c72 100644
--- a/contrib/completion/ditz.bash
+++ b/contrib/completion/ditz.bash
@@ -6,15 +6,15 @@
 
 _ditz() 
 {
-    cur=${COMP_WORDS[COMP_CWORD]}
+    local cur=${COMP_WORDS[COMP_CWORD]}
     if [ $COMP_CWORD -eq 1 ]; then
 	COMPREPLY=( $( compgen -W "$(ditz --commands)" -- $cur ) )
     elif [ $COMP_CWORD -eq 2 ]; then
-	cmd=${COMP_WORDS[1]}
+	local cmd=${COMP_WORDS[1]}
 	COMPREPLY=( $( compgen -W "$(ditz "$cmd" '<options>' 2>/dev/null)" -- $cur ) )
     elif [ $COMP_CWORD -eq 3 ]; then
-	cmd=${COMP_WORDS[1]}
-	parm1=${COMP_WORDS[2]}
+	local cmd=${COMP_WORDS[1]}
+	local parm1=${COMP_WORDS[2]}
 	COMPREPLY=( $( compgen -W "$(ditz "$cmd" "$parm1" '<options>' 2>/dev/null)" -- $cur ) )
     fi 
 }
-- 
1.5.6.3



From wmorgan-ditz at masanjin.net  Mon Aug 11 19:23:16 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Mon, 11 Aug 2008 16:23:16 -0700
Subject: [ditz-talk] [PATCH] fix completion bug on strings like '-n'
In-Reply-To: <20080811204747.GA9784@cgarbs.de>
References: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
	<1218485251-sup-9896@entry> <20080811204747.GA9784@cgarbs.de>
Message-ID: <1218488374-sup-7066@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-11:
> The patch expands "-n<tab>" correctly to an empty list, as there is no
> ditz command or option that starts with "-n".

Got it. Applied, thanks!

There actually is a ditz -n, but we're not doing completions on
flags. (Yet!)
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Tue Aug 12 02:49:14 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Tue, 12 Aug 2008 08:49:14 +0200
Subject: [ditz-talk] [PATCH] fix completion bug on strings like '-n'
In-Reply-To: <1218488374-sup-7066@entry>
References: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
	<1218485251-sup-9896@entry> <20080811204747.GA9784@cgarbs.de>
	<1218488374-sup-7066@entry>
Message-ID: <20080812064914.GA25031@cgarbs.de>

On Mon, Aug 11, 2008 at 04:23:16PM -0700, William Morgan wrote:

> There actually is a ditz -n, but we're not doing completions on
> flags. (Yet!)

Oh, didn't know that.  Sorry for the confusion :-)

I'm pretty much sticking with 0.4 stable.  Before the 0.5 release, I
can update the completion if I know what's new.

Regards,
Christian

PS: I want to do some more testing, but yesterday it looked like
    ditz assign and ditz set-component couldn't handle the "long"
    expanded issue names (ditz-1-description-of-issue).  If this were
    indeed an error in 0.4, would you release a 0.4.1 or would this
    have to wait for 0.5?
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Wenn man sagt, dass Waffen Menschen t?ten,
dann machen Stifte Schreibfehler,
Autos lassen Leute betrunken fahren
und Gabeln machen dick!  
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080812/8fe2329b/attachment.bin>

From wmorgan-ditz at masanjin.net  Tue Aug 12 12:43:08 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Tue, 12 Aug 2008 09:43:08 -0700
Subject: [ditz-talk] [PATCH] fix completion bug on strings like '-n'
In-Reply-To: <20080812064914.GA25031@cgarbs.de>
References: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
	<1218485251-sup-9896@entry> <20080811204747.GA9784@cgarbs.de>
	<1218488374-sup-7066@entry> <20080812064914.GA25031@cgarbs.de>
Message-ID: <1218559164-sup-2886@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-11:
> I'm pretty much sticking with 0.4 stable.  Before the 0.5 release, I
> can update the completion if I know what's new.

I don't think there's anything new that requires updating completion,
unless we want to complete --arguments as well. (I think that would be
too much, personally.)

> PS: I want to do some more testing, but yesterday it looked like ditz
> assign and ditz set-component couldn't handle the "long" expanded
> issue names (ditz-1-description-of-issue).

This works for me in both 0.4 and in git master. Weird. Is it only those
two commands for you? Is it one particular issue, or all issues? It
should be "ditz-1_description-of-issue" (underscore is the separator
between name and title); did that somehow get changed in your repo?
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Sun Aug 17 04:35:37 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Sun, 17 Aug 2008 10:35:37 +0200
Subject: [ditz-talk] [PATCH] fix completion bug on strings like '-n'
In-Reply-To: <1218559164-sup-2886@entry>
References: <1218288040-20109-1-git-send-email-mitch@cgarbs.de>
	<1218485251-sup-9896@entry> <20080811204747.GA9784@cgarbs.de>
	<1218488374-sup-7066@entry> <20080812064914.GA25031@cgarbs.de>
	<1218559164-sup-2886@entry>
Message-ID: <20080817083537.GB12351@cgarbs.de>

On Tue, Aug 12, 2008 at 09:43:08AM -0700, William Morgan wrote:

> > PS: I want to do some more testing, but yesterday it looked like
> > ditz assign and ditz set-component couldn't handle the "long"
> > expanded issue names (ditz-1-description-of-issue).
 
> This works for me in both 0.4 and in git master. Weird. Is it only
> those two commands for you? Is it one particular issue, or all
> issues? It should be "ditz-1_description-of-issue" (underscore is
> the separator between name and title); did that somehow get changed
> in your repo?

Invoking ditz on several occasions in the last day I was not able to
reproduce this.  Please go on, there's nothing to see here...

Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Lightspeed Briefs - as seen in your dreams
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080817/da58674c/attachment.bin>

From ohad at lutzky.net  Sun Aug 17 13:23:27 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Sun, 17 Aug 2008 20:23:27 +0300
Subject: [ditz-talk] HTML rendering bug on trivially malformed project.yaml
Message-ID: <8e50e92e0808171023w1cce2a39q436b2a5dd843bd72@mail.gmail.com>

Hello,

For some reason, my project.yaml file became corrupted in a minute
way: In one of the releases' log_events, the terminating '- ""' line
was missing. This can be reproduced like so:

mkdir test
ditz init
ditz add-release

Edit bugs/project.yaml and remove the last line

ditz html

The error I get is:

$ ditz html
(erb):56:in `render_template': undefined method `empty?' for
nil:NilClass (NoMethodError)
	from (erb):68:in `each_with_index'
	from (erb):45:in `each'
	from (erb):45:in `each_with_index'
	from (erb):45:in `render_template'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/html.rb:23:in `render_template'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/views.rb:99:in `render_all'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/views.rb:98:in `open'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/views.rb:98:in `render_all'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/views.rb:95:in `each'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/views.rb:95:in `render_all'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/operator.rb:425:in `html'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/operator.rb:100:in `send'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/lib/operator.rb:100:in `do'
	from /usr/lib/ruby/gems/1.8/gems/ditz-0.4/bin/ditz:159
	from /usr/bin/ditz:19:in `load'
	from /usr/bin/ditz:19

This also occurs for current mainline git ditz.

I believe that ditz html should be resilient to this problem, as the
rest of ditz is. If not, a more intelligible error message should be
presented (it took me way too long to figure this one out).

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From wmorgan-ditz at masanjin.net  Tue Aug 19 02:05:46 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Mon, 18 Aug 2008 23:05:46 -0700
Subject: [ditz-talk] HTML rendering bug on trivially malformed
	project.yaml
In-Reply-To: <8e50e92e0808171023w1cce2a39q436b2a5dd843bd72@mail.gmail.com>
References: <8e50e92e0808171023w1cce2a39q436b2a5dd843bd72@mail.gmail.com>
Message-ID: <1219125909-sup-1411@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-17:
> I believe that ditz html should be resilient to this problem, as the
> rest of ditz is. If not, a more intelligible error message should be
> presented (it took me way too long to figure this one out).

Agreed. Issue c83a8f40 created.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Tue Aug 19 12:17:01 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Tue, 19 Aug 2008 09:17:01 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <1218039422-sup-6792@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200808031802.56706.ninja@slaphack.com> <1217814161-sup-4774@entry>
	<200808032217.58787.ninja@slaphack.com> <1218039422-sup-6792@entry>
Message-ID: <1219162366-sup-4593@entry>

Reformatted excerpts from William Morgan's message of 2008-08-06:
> Reformatted excerpts from David Masover's message of 2008-08-03:
> > My point is that if there is only one component, it makes more sense
> > to make this something short. I like Trac's notation -- anything
> > that looks like a pound-sign followed by an integer (as in #123) is
> > treated as a link to the ticket identified by that number (ditz uses
> > "issue" instead of "ticket", but the concept is the same).
> 
> I see. Yeah, this might be nice, because it does make things much
> shorter in the single-component case. But I do like the current issue
> name format for when you have multiple components (maybe only because
> JIRA managed to burn itself into my brain), so if this were to happen,
> I would want the transition to be seamless if you add components.

I've been playing around with this idea a little and the more I use it,
the more I like it. I'm thinking of just turning it on by default,
independent of the number of components, and adding a configuration
option (or maybe a plugin) that reverts to the old behavior.

The only irritation is that (in bash, at least) # is a comment, so
something like "ditz show #77" is as if you had just typed "ditz show".
You have to backslash the #.

So I've been thinking about "+77" instead, which isn't quite as nice,
but at least doesn't require a backslash...
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Tue Aug 19 14:07:28 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Tue, 19 Aug 2008 11:07:28 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <1217801806-sup-5317@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200807292325.27659.ninja@slaphack.com> <1217449815-sup-4025@entry>
	<200807310949.39086.ninja@slaphack.com> <1217801806-sup-5317@entry>
Message-ID: <1219169234-sup-5240@entry>

Reformatted excerpts from William Morgan's message of 2008-08-03:
> Reformatted excerpts from David Masover's message of 2008-07-31:
> > Bugfix/Feature/Task should probably be set to some default, as I'm
> > assuming it must be provided.
> 
> If you can come up with a default that satisfies everyone, sure. I tend
> to be about 50% between bugs and features. We could do something like
> "default to the last type the user picked", and then store that
> somewhere... dunno.

I'm actually wondering if I should just remove this categorization
entirely. (Or relegate it to a plugin.)

Do people find it helpful to divide things into features, bugs and
tasks? There aren't really any tools right now to list things by type
(though that will be easy now that subcommand arguments is in place).

Personally I can mostly categorize things, but I never do anything with
that information, so I'm wondering if this is a JIRA-ism left in my
brain that only makes sense when you think like a Java programmer.
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Tue Aug 19 14:29:53 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Tue, 19 Aug 2008 20:29:53 +0200
Subject: [ditz-talk] First impressions
In-Reply-To: <1219169234-sup-5240@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200807292325.27659.ninja@slaphack.com> <1217449815-sup-4025@entry>
	<200807310949.39086.ninja@slaphack.com>
	<1217801806-sup-5317@entry> <1219169234-sup-5240@entry>
Message-ID: <20080819182953.GB13227@cgarbs.de>

On Tue, Aug 19, 2008 at 11:07:28AM -0700, William Morgan wrote:

> Do people find it helpful to divide things into features, bugs and
> tasks? There aren't really any tools right now to list things by type
> (though that will be easy now that subcommand arguments is in place).

I'm currently using it, but I propably won't miss it if it's gone.

I think it's good to look at the list of last changes and see that a
bug has been fixed so you might want to get that update.  On the other
hand it's propably just enough to prefix bugs with "bug: " in their
subject line.

Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

<<And now, our special guest star, THE CHORD FROM HELL>>
     (The Long Road - Book Two of Undocumented Features)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080819/5809074d/attachment.bin>

From ninja at slaphack.com  Tue Aug 19 23:29:06 2008
From: ninja at slaphack.com (David Masover)
Date: Tue, 19 Aug 2008 22:29:06 -0500
Subject: [ditz-talk] First impressions
In-Reply-To: <1219162366-sup-4593@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<1218039422-sup-6792@entry> <1219162366-sup-4593@entry>
Message-ID: <200808192229.06622.ninja@slaphack.com>

On Tuesday 19 August 2008 11:17:01 William Morgan wrote:

> The only irritation is that (in bash, at least) # is a comment, so
> something like "ditz show #77" is as if you had just typed "ditz show".
> You have to backslash the #.
> 
> So I've been thinking about "+77" instead, which isn't quite as nice,
> but at least doesn't require a backslash...

Is there a reason you need a symbol there at all? What about:

dits show 77

It should work with a # too, but I think it could be implicit. (What else 
would you be using a number for in that context?)

From ninja at slaphack.com  Tue Aug 19 23:40:29 2008
From: ninja at slaphack.com (David Masover)
Date: Tue, 19 Aug 2008 22:40:29 -0500
Subject: [ditz-talk] First impressions
In-Reply-To: <1219169234-sup-5240@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<1217801806-sup-5317@entry> <1219169234-sup-5240@entry>
Message-ID: <200808192240.29726.ninja@slaphack.com>

On Tuesday 19 August 2008 13:07:28 William Morgan wrote:

> Personally I can mostly categorize things, but I never do anything with
> that information, so I'm wondering if this is a JIRA-ism left in my
> brain that only makes sense when you think like a Java programmer.

I think I am the same way. Not JIRA, but Trac.

About the only reason I can think of is if you have a different development 
cycle for bugfixes and enhancements -- at the very least, it would be healthy 
to have periodic feature freezes, when only bugfixes may be dealt with.

Even then, it's not as useful as you'd think -- for a small project, you can 
look at all ten or twenty tickets and actually read through them, and decide 
on a case by case basis.

For a larger project, you never know if a bugfix is going to depend on an 
enhancement until you actually read the ticket, and possibly until you 
actually tackle that bug. It'd still probably be a useful starting point -- 
at least that way, there's a fair chunk of tickets you can simply ignore.

I think what might be more generally useful is a generic tagging system. After 
all, as I've just described, it's possible for something to be both a bugfix 
and an enhancement.

I do think priorities make sense, because then you can sort by priority, 
highest first -- or choose to only view tasks above a certain priority. 
(Hypothetically -- I don't know if ditz actually provides these features.)

From wmorgan-ditz at masanjin.net  Wed Aug 20 00:09:10 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Tue, 19 Aug 2008 21:09:10 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <200808192229.06622.ninja@slaphack.com>
References: <200807290315.16589.ninja@slaphack.com>
	<1218039422-sup-6792@entry> <1219162366-sup-4593@entry>
	<200808192229.06622.ninja@slaphack.com>
Message-ID: <1219205251-sup-7036@entry>

Reformatted excerpts from David Masover's message of 2008-08-19:
> Is there a reason you need a symbol there at all? What about:
> 
> dits show 77

These issue names have to be detectable in title, description and
comment strings, so that they can be replaced with the 40-byte issue
ids. I suspect just using a number will result in too many false
positives.
-- 
William <wmorgan-ditz at masanjin.net>

From ninja at slaphack.com  Wed Aug 20 01:22:31 2008
From: ninja at slaphack.com (David Masover)
Date: Wed, 20 Aug 2008 00:22:31 -0500
Subject: [ditz-talk] First impressions
In-Reply-To: <1219205251-sup-7036@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200808192229.06622.ninja@slaphack.com> <1219205251-sup-7036@entry>
Message-ID: <200808200022.31134.ninja@slaphack.com>

On Tuesday 19 August 2008 23:09:10 William Morgan wrote:
> Reformatted excerpts from David Masover's message of 2008-08-19:
> > Is there a reason you need a symbol there at all? What about:
> > 
> > dits show 77
> 
> These issue names have to be detectable in title, description and
> comment strings, so that they can be replaced with the 40-byte issue
> ids. I suspect just using a number will result in too many false
> positives.

Yes, when used in title, description, and comment strings. I'm not suggesting 
you only use a number there.

I'm suggesting you allow the pound sign to be omitted in places where the 
number is unambiguous -- specifically, the commandline.

From wmorgan-ditz at masanjin.net  Wed Aug 20 01:37:47 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Tue, 19 Aug 2008 22:37:47 -0700
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
Message-ID: <1219210549-sup-3129@entry>

Ditz version 0.5 has been released!

This release features two new plugins and a host of bugfixes and
enhancements.

http://ditz.rubyforge.org

Ditz is a simple, light-weight distributed issue tracker designed to
work with distributed version control systems like git, darcs,
Mercurial, and Bazaar. It can also be used with centralized systems like
SVN.

Ditz maintains an issue database directory on disk, with files written
in a line-based and human-editable format. This directory can be kept
under version control, alongside project code.

Ditz provides a simple, console-based interface for creating and
updating the issue database files, and some basic static HTML generation
capabilities for producing world-readable status pages (for a demo, see
the ditz ditz page).

Ditz includes a robust plugin system for adding commands, model fields,
and modifying output. See PLUGINS.txt for documentation on the
pre-shipped plugins.

Ditz currently offers no central public method of bug submission. 

## USING DITZ

There are several different ways to use Ditz:

1. Treat issue change the same as code change: include it as part of
   commits, and merge it with changes from other developers, resolving
   conflicts in the usual manner.
2. Keep the issue database in the repository but in a separate branch.
   Issue changes can be managed by your VCS, but is not tied directly to
   code commits.
3. Keep the issue database separate and not under VCS at all.

Changes:

## 0.5 / 2008-08-20
* ditz tarball now installable without requiring rubygems
* commands now take their own arguments. use 'ditz help' to see them.
* 'todo-full' has been replaced with 'todo --all'
* all commands that required comments now take --comment and --no-comment
* new git-sync hooks git-sync for synchronized git usage
* new issue-claiming hook for claiming issues for oneself
* update hook now called on project.yaml as well as changed issue yaml files
* hooks no longer run twice if a project devel root is under a home directory
* when editing a field in an editor, comments are now at the bottom of the buffer
* ditz now acceps issue ids as well as issue names on the commandline
* ditz edit takes a --silent option to avoid generating a log message
* a 'reconfigure' command has been added
* various webpage tweaks
* more unix-centrisms fixed

-- 
William <wmorgan-ditz at masanjin.net>

From dcalford at distributel.ca  Wed Aug 20 11:16:10 2008
From: dcalford at distributel.ca (Dalton Calford)
Date: Wed, 20 Aug 2008 11:16:10 -0400
Subject: [ditz-talk] Errors with new release
Message-ID: <1219245370.6911.12.camel@dcalford-laptop.corp.distributel.ca>

Hi all,

I installed ditz yesterday, then again today (aren't new releases great)

With the new version, (installed via "gem install ditz")

I get the following error while attempting to perform a ditz init on a
new git repository

ditz:37:in 'cloaker_':undefined local variable or method
'stop_on_unknown' for #<Trollop::Parser:0xb78245bc> (NameError)
from trollop.rb:58: in 'call'
from trollop.rb:58: in 'initialize'
from trollop.rb:441: in 'new'
from trollop.rb:441: in 'options'
from ditz:30
from ditz:16 in 'load'
from ditz:16

I am hoping there is a simple fix with this (I am using the .5 version
of ditz)

Thanks




From wmorgan-ditz at masanjin.net  Wed Aug 20 12:46:22 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 20 Aug 2008 09:46:22 -0700
Subject: [ditz-talk] Errors with new release
In-Reply-To: <1219245370.6911.12.camel@dcalford-laptop.corp.distributel.ca>
References: <1219245370.6911.12.camel@dcalford-laptop.corp.distributel.ca>
Message-ID: <1219250605-sup-8755@entry>

Reformatted excerpts from Dalton Calford's message of 2008-08-20:
> I get the following error while attempting to perform a ditz init on a
> new git repository
> 
> ditz:37:in 'cloaker_':undefined local variable or method
> 'stop_on_unknown' for #<Trollop::Parser:0xb78245bc> (NameError)

Whoops!

You need to upgrade to Trollop 1.9. I've just published that gem, so it
might take a few hours to percolate through the RubyForge gem servers.
You can always download it directly here:
http://rubyforge.org/projects/trollop/

I've re-released ditz 0.5 in place, with this dependency now expressed.
Hopefully that will work. If not, I'll release a 0.5.1.
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Wed Aug 20 13:48:13 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Wed, 20 Aug 2008 19:48:13 +0200
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <1219210549-sup-3129@entry>
References: <1219210549-sup-3129@entry>
Message-ID: <20080820174813.GA29201@cgarbs.de>

On Tue, Aug 19, 2008 at 10:37:47PM -0700, William Morgan wrote:

> Ditz version 0.5 has been released!

I'm currently packaging new Debian packages.  I made a note to use
setup.rb for the build process, but it seems to be missing from the
tarball.

Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Die Leute in der wirklichen Welt verstehen das oft nicht.
                Prof J. Buchmann, Theoretische Informatik
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080820/58dcf509/attachment.bin>

From mitch at cgarbs.de  Wed Aug 20 14:21:40 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Wed, 20 Aug 2008 20:21:40 +0200
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <1219210549-sup-3129@entry>
References: <1219210549-sup-3129@entry>
Message-ID: <20080820182140.GA8292@cgarbs.de>

On Tue, Aug 19, 2008 at 10:37:47PM -0700, William Morgan wrote:

> Ditz version 0.5 has been released!

My first call of ditz 0.5 in a ditz-enabled directory results in a
warning that I can't figure out:

mitch at yggdrasil:~/debian/own.other/ditz_0.5/ditz-0.5$ ditz
ditz-0.5 (unreleased):
_ ditz-1: use setup.rb

Unassigned:
_ qtpfsgui-1: mention NMU
_    hugin-1: mention NMU
_  enblend-1: mention NMU
_  libpano-1: mention NMU

You may have to inform your SCM that the following files have been modified:
/usr/share/ditz/file-storage.rb:8: warning: Object#id will be deprecated; use Object#object_id
  ../../../bugs/issue--70605488.yaml

The mentioned file does not exist:

mitch at yggdrasil:~/debian/own.other/ditz_0.5/ditz-0.5$ ls -lh ../../../bugs/issue--70605488.yaml
ls: ../../../bugs/issue--70605488.yaml: No such file or directory


How do I fix this?
Apparently thie file has not been modified because the warning does
not go away on subsequent calls of ditz.

Regards,
Christian

PS: I've also run into a bug with the bash completion:
    Type "ditz --v" and press tab...
    I'll look into that.
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

If you took all of the grains of sand in the world, and lined
them up end to end in a row, you'd be working for the government!
                -- Mr. Interesting
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080820/f1efdddc/attachment-0001.bin>

From ohad at lutzky.net  Wed Aug 20 14:36:01 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Wed, 20 Aug 2008 21:36:01 +0300
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <20080820174813.GA29201@cgarbs.de>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
Message-ID: <8e50e92e0808201136u1f00882fqcda165c3ff3985a4@mail.gmail.com>

On Wed, Aug 20, 2008 at 8:48 PM, Christian Garbs <mitch at cgarbs.de> wrote:
> On Tue, Aug 19, 2008 at 10:37:47PM -0700, William Morgan wrote:
>
>> Ditz version 0.5 has been released!
>
> I'm currently packaging new Debian packages.  I made a note to use
> setup.rb for the build process, but it seems to be missing from the
> tarball.

I'd like to point out that it's very easy to do that. As an example,
you can take a look at ttime (a project of mine) at
http://github.com/lutzky/ttime (sorry if I'm underestimating you, but
you said you want to "use it for the build process", whereas in my
method it pretty much IS the build process)

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From wmorgan-ditz at masanjin.net  Wed Aug 20 16:08:38 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 20 Aug 2008 13:08:38 -0700
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <20080820182140.GA8292@cgarbs.de>
References: <1219210549-sup-3129@entry> <20080820182140.GA8292@cgarbs.de>
Message-ID: <1219262852-sup-7236@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-20:
> You may have to inform your SCM that the following files have been
> modified: /usr/share/ditz/file-storage.rb:8: warning: Object#id will
> be deprecated; use Object#object_id
>   ../../../bugs/issue--70605488.yaml

I've released a new 0.5 that fixes this. Since the old 0.5 might be in
your gem cache, I recommend downloading it directly:

http://rubyforge.org/frs/download.php/41560/ditz-0.5.gem

Sorry for the false starts.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Wed Aug 20 16:12:40 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 20 Aug 2008 13:12:40 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <200808200022.31134.ninja@slaphack.com>
References: <200807290315.16589.ninja@slaphack.com>
	<200808192229.06622.ninja@slaphack.com> <1219205251-sup-7036@entry>
	<200808200022.31134.ninja@slaphack.com>
Message-ID: <1219263109-sup-3587@entry>

Reformatted excerpts from David Masover's message of 2008-08-19:
> I'm suggesting you allow the pound sign to be omitted in places where
> the number is unambiguous -- specifically, the commandline.

That would be an option, but I'd prefer it to be consistent everywhere.
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Wed Aug 20 17:47:14 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Wed, 20 Aug 2008 23:47:14 +0200
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <8e50e92e0808201136u1f00882fqcda165c3ff3985a4@mail.gmail.com>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
	<8e50e92e0808201136u1f00882fqcda165c3ff3985a4@mail.gmail.com>
Message-ID: <20080820214714.GB7057@cgarbs.de>

On Wed, Aug 20, 2008 at 09:36:01PM +0300, Ohad Lutzky wrote:
> On Wed, Aug 20, 2008 at 8:48 PM, Christian Garbs <mitch at cgarbs.de> wrote:

>> [using setup.rb in debian/rules]

> I'd like to point out that it's very easy to do that.

I didn't know there's a cdbs include for that.  Thanks!

Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Raketentriebwerk:
  gibt es in Schubladen zu kaufen
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080820/1781695d/attachment.bin>

From wmorgan-ditz at masanjin.net  Thu Aug 21 13:31:16 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Thu, 21 Aug 2008 10:31:16 -0700
Subject: [ditz-talk] Errors with new release
In-Reply-To: <1219322794.10827.3.camel@dcalford-laptop.corp.distributel.ca>
References: <1219245370.6911.12.camel@dcalford-laptop.corp.distributel.ca>
	<1219250605-sup-8755@entry>
	<1219255231.6911.23.camel@dcalford-laptop.corp.distributel.ca>
	<1219262947-sup-6657@entry>
	<1219322794.10827.3.camel@dcalford-laptop.corp.distributel.ca>
Message-ID: <1219338858-sup-3907@entry>

[cc'ing list]

Reformatted excerpts from Dalton Calford's message of 2008-08-21:
> Is there a method to have ditz generate the html file with a different
> name/location?

'ditz html' takes an optional argument, which is the directory in which
to place files.

> I am integrating ditz with gitweb along with git:pserver

Cool!

> BTW, I am looking at making web forms for submitting issues with ditz.
> 
> We are using perl and calling ditz as an external process.
> 
> Have you done anything like this?

You can look at sheila, Why the lucky stiff's camping+ditz app:
  http://github.com/why/sheila/tree/master
It's not entirely complete and doesn't quite work with latest releases,
but I'm working on fixing it up.

But that's in Ruby, so it can use the Ditz model objects directly. The
commandline stuff is structured to work in an interactive environment,
so it might be kinda irritating to call non-interactively. I've been
thinking a little about how to make a non-interactive commandline mode
but don't have any great ideas so far.
-- 
William <wmorgan-ditz at masanjin.net>

From dcalford at distributel.ca  Thu Aug 21 14:02:26 2008
From: dcalford at distributel.ca (Dalton Calford)
Date: Thu, 21 Aug 2008 14:02:26 -0400
Subject: [ditz-talk] Errors with new release
In-Reply-To: <1219338858-sup-3907@entry>
References: <1219245370.6911.12.camel@dcalford-laptop.corp.distributel.ca>
	<1219250605-sup-8755@entry>
	<1219255231.6911.23.camel@dcalford-laptop.corp.distributel.ca>
	<1219262947-sup-6657@entry>
	<1219322794.10827.3.camel@dcalford-laptop.corp.distributel.ca>
	<1219338858-sup-3907@entry>
Message-ID: <1219341746.10827.34.camel@dcalford-laptop.corp.distributel.ca>

Hey Guys,

I figured I may as well let you know what we are up to, in order to stop
re-inventing the wheel as well as allowing for others to benefit from
our work.

First off, we are replacing an existing cvs server with the git.
This has alot of issues including the fact that most of the IDE's don't
support the git directly.
So, we are using the git-pserver environment to give support to our
legacy developers.
We have edited the gitweb code to allow for windows domain
authentication and we restrict our users rights based upon their
membership in groups on the windows domain controller.

Our host machine is a Ubuntu 8.4 server.

We place all of our projects in the /home/git/ directory.

Due to special needs/configuration requirements to support pserver under
git, we have created our own make project script that creates the bare
repository, sets up the rights, initilizes ditz etc.

Since most of the users are windows based, we created a web front end
for doing all of this and we even support the creation of a new project
from existing cvs projects.

This allows the users to remotely create/manage the repository while
using git or cvs to update the code within the repository.

We are currently integrating the html forms created by ditz into the
system.

I am currently reading the rhtml files in the ditz-0.5/lib directory to
see what it would take to have the ditz add a few buttons to the forms
for me.

I am not familiar with ruby, (any link to beginners guides/reference
materials would be appreciated) but I am thinking of just having the
buttons link to a php file that confirms the users rights and calls ditz
via shell calls.

best regards

Dalton





Currently our project has the following elements



On Thu, 2008-08-21 at 10:31 -0700, William Morgan wrote:
> [cc'ing list]
> 
> Reformatted excerpts from Dalton Calford's message of 2008-08-21:
> > Is there a method to have ditz generate the html file with a different
> > name/location?
> 
> 'ditz html' takes an optional argument, which is the directory in which
> to place files.
> 
> > I am integrating ditz with gitweb along with git:pserver
> 
> Cool!
> 
> > BTW, I am looking at making web forms for submitting issues with ditz.
> > 
> > We are using perl and calling ditz as an external process.
> > 
> > Have you done anything like this?
> 
> You can look at sheila, Why the lucky stiff's camping+ditz app:
>   http://github.com/why/sheila/tree/master
> It's not entirely complete and doesn't quite work with latest releases,
> but I'm working on fixing it up.
> 
> But that's in Ruby, so it can use the Ditz model objects directly. The
> commandline stuff is structured to work in an interactive environment,
> so it might be kinda irritating to call non-interactively. I've been
> thinking a little about how to make a non-interactive commandline mode
> but don't have any great ideas so far.


From mitch at cgarbs.de  Thu Aug 21 14:42:30 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Thu, 21 Aug 2008 20:42:30 +0200
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <20080820174813.GA29201@cgarbs.de>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
Message-ID: <20080821184230.GA24381@cgarbs.de>

On Wed, Aug 20, 2008 at 07:48:13PM +0200, Christian Garbs wrote:
> On Tue, Aug 19, 2008 at 10:37:47PM -0700, William Morgan wrote:
> 
> > Ditz version 0.5 has been released!
> 
> I'm currently packaging new Debian packages.  I made a note to use
> setup.rb for the build process, but it seems to be missing from the
> tarball.

I don't want to be nitpicky, now that setup.rb is included, but I've
come across trollop.rb which is included in the ditz 0.5 gem and tgz.

As trollop is a separate package, shouldn't it be excluded from the
ditz release?

Also, trollop.rb claims VERSION = "1.8.2", I thought that ditz 0.5
needed trollop 1.9.
 
Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Westheimer's Discovery:
        A couple of months in the laboratory can frequently save a
couple of hours in the library.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080821/184e2284/attachment.bin>

From mitch at cgarbs.de  Thu Aug 21 14:48:16 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Thu, 21 Aug 2008 20:48:16 +0200
Subject: [ditz-talk] using setup.rb (was: Re: [ANN] ditz 0.5 distributed
	issue tracker released)
In-Reply-To: <8e50e92e0808201136u1f00882fqcda165c3ff3985a4@mail.gmail.com>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
	<8e50e92e0808201136u1f00882fqcda165c3ff3985a4@mail.gmail.com>
Message-ID: <20080821184816.GB24381@cgarbs.de>

On Wed, Aug 20, 2008 at 09:36:01PM +0300, Ohad Lutzky wrote:

> I'd like to point out that it's very easy to do that. As an example,
> you can take a look at ttime (a project of mine) at
> http://github.com/lutzky/ttime (sorry if I'm underestimating you, but
> you said you want to "use it for the build process", whereas in my
> method it pretty much IS the build process)

I don't know much about setup.rb and ruby in general or the specifics
of ruby integration to Debian.

Is there an easy way to make setup.rb install to /usr/share/ditz
instead of /usr/lib/ruby1.8?

I don't think that files like "style.css" or "green-bar.png" should
end up in a global library path :-)

Currently my debian/rules looks like this, but I want to change it to use
setup.rb:

#!/usr/bin/make -f

include /usr/share/cdbs/1/rules/debhelper.mk

DEB_INSTALL_DOCS_ditz = README.txt ReleaseNotes
DEB_INSTALL_EXAMPLES_ditz = contrib/completion/_ditz.zsh
DEB_INSTALL_MANPAGES_ditz = debian/ditz.1

install/ditz::

#       binary: add ditz lib include directory (taken from apt-listbugs - is there a better way?)
        sed -e 's,^#!/usr/bin/env ruby,#!/usr/bin/ruby -I/usr/share/ditz,' \
                < bin/ditz \
                > $(DEB_DESTDIR)/usr/bin/ditz
        chmod 755 $(DEB_DESTDIR)/usr/bin/ditz

#       libraries
        install -m 0644 lib/*.css           $(DEB_DESTDIR)/usr/share/ditz
        install -m 0644 lib/*.rhtml         $(DEB_DESTDIR)/usr/share/ditz
        install -m 0644 lib/*.rb            $(DEB_DESTDIR)/usr/share/ditz
        install -m 0644 lib/plugins/*.rb    $(DEB_DESTDIR)/usr/share/ditz/plugins

#       additional files
        install -m 0644 contrib/completion/ditz.bash   $(DEB_DESTDIR)/etc/bash_completion.d/ditz



Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Kilroe hic erat!
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080821/876127ff/attachment.bin>

From wmorgan-ditz at masanjin.net  Thu Aug 21 15:47:49 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Thu, 21 Aug 2008 12:47:49 -0700
Subject: [ditz-talk] using setup.rb (was: Re: [ANN] ditz 0.5 distributed
	issue tracker released)
In-Reply-To: <20080821184816.GB24381@cgarbs.de>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
	<8e50e92e0808201136u1f00882fqcda165c3ff3985a4@mail.gmail.com>
	<20080821184816.GB24381@cgarbs.de>
Message-ID: <1219347708-sup-4912@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-21:
> I don't know much about setup.rb and ruby in general or the specifics
> of ruby integration to Debian.

Well setup.rb is a very old thing that people used to use in the dark
ages. I've only included it to make the tarball actually usable.

> Is there an easy way to make setup.rb install to /usr/share/ditz
> instead of /usr/lib/ruby1.8?

Look at INSTALL. There's a way to configure setup.rb by running it
twice.

> I don't think that files like "style.css" or "green-bar.png" should
> end up in a global library path :-)

I'd be happy to move the images and the .rhtml files into a share/
directory since that's probably a better location. They're only in lib
for historical reasons.

For packaging this release, you may want to do that manually, or to
explictly mkdir share; mv lib/*.{png,rhtml} share/.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Thu Aug 21 15:52:26 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Thu, 21 Aug 2008 12:52:26 -0700
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <20080821184230.GA24381@cgarbs.de>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
	<20080821184230.GA24381@cgarbs.de>
Message-ID: <1219348074-sup-5038@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-21:
> I don't want to be nitpicky, now that setup.rb is included, but I've
> come across trollop.rb which is included in the ditz 0.5 gem and tgz.
> 
> As trollop is a separate package, shouldn't it be excluded from the
> ditz release?

I included it because I want the tarball to be useful for people without
rubygems. Ideally it should be included by the tarball and excluded by
the gem, but that's hard.

> Also, trollop.rb claims VERSION = "1.8.2", I thought that ditz 0.5
> needed trollop 1.9.

That's a fuckup on my part. (I've corrected it in git.) I copied that
file over from trollop before bumping the version number.  The code is
1.9; it's just the version number in there that's wrong.

I appreciate your help ironing out all these issues. It sounds like by
0.6 we should have a tarball that you can automatically Debianize. Maybe
next time I'll actually try and see that everything's working before
just releasing it into the wild. :)
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Fri Aug 22 08:11:56 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Fri, 22 Aug 2008 14:11:56 +0200
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <1219348074-sup-5038@entry>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
	<20080821184230.GA24381@cgarbs.de> <1219348074-sup-5038@entry>
Message-ID: <20080822121156.GB14345@cgarbs.de>

On Thu, Aug 21, 2008 at 12:52:26PM -0700, William Morgan wrote:
> Reformatted excerpts from Christian Garbs's message of 2008-08-21:

> > As trollop is a separate package, shouldn't it be excluded from
> > the ditz release?
 
> I included it because I want the tarball to be useful for people
> without rubygems. Ideally it should be included by the tarball and
> excluded by the gem, but that's hard.

Ok, so it should stay.  I can very easily work around it in the Debian
package, so no problem.

 
> It sounds like by 0.6 we should have a tarball that you can
> automatically Debianize.

The first time I read this I thought you wanted to include the debian/
directory directly into the ditz git and went "noooooooo!" :-)

But if it's only "making the packager's life easier", then it's ok -
but ditz is already very easy to package.  I've never done a ruby
package before, but everythings seems to work fine and there was
nearly nothing for me to do to update it from 0.4 to 0.5.

I think I'll release the 0.5 Debian packages later this day, as my
plans of using setup.rb have been thwarted for now.


> Maybe next time I'll actually try and see that everything's working
> before just releasing it into the wild. :)

I was in fact a little surprised of the 0.5 release announcement
because there were no calls for "I want to release, please check for
last bugs or problems anyone" beforehand :-)

Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

"Ober, zahlen!"  --  "Gast, Buchstaben!"
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080822/9d529854/attachment.bin>

From mitch at cgarbs.de  Fri Aug 22 08:01:26 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Fri, 22 Aug 2008 14:01:26 +0200
Subject: [ditz-talk] using setup.rb
In-Reply-To: <1219347708-sup-4912@entry>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
	<8e50e92e0808201136u1f00882fqcda165c3ff3985a4@mail.gmail.com>
	<20080821184816.GB24381@cgarbs.de> <1219347708-sup-4912@entry>
Message-ID: <20080822120126.GA14345@cgarbs.de>

On Thu, Aug 21, 2008 at 12:47:49PM -0700, William Morgan wrote:
> Reformatted excerpts from Christian Garbs's message of 2008-08-21:

> > I don't know much about setup.rb and ruby in general or the specifics
> > of ruby integration to Debian.
> 
> Well setup.rb is a very old thing that people used to use in the dark
> ages. I've only included it to make the tarball actually usable.

Ok.
 
> > Is there an easy way to make setup.rb install to /usr/share/ditz
> > instead of /usr/lib/ruby1.8?
> 
> Look at INSTALL. There's a way to configure setup.rb by running it
> twice.

Is it possible to change the setup.rb so that the ditz files don't get
installed directly into the ruby library directory but into a "ditz"
subfolder?  Something like a package name in Perl (e.g. calling the
html module Ditz::Html instead of just Html)?

For example, I already have a file named html.rb on my system, but
it's in a package specific subdirectory:

/usr/lib/ruby/1.8/rdoc/generators/template/html/html.rb

Using something like a "ditz" namespace should prevent problems with
other packages.


Looking at the automagic Debian cdbs snippet
/usr/share/ruby-pkg-tools/1/class/ruby-setup-rb.mk, setup.rb is
already called with --prefix and --siterubyvers.  I don't see an easy
way to move everything to a ditz subdirectory using this snippet.


> I'd be happy to move the images and the .rhtml files into a share/
> directory since that's probably a better location. They're only in
> lib for historical reasons.
>
> For packaging this release, you may want to do that manually, or to
> explictly mkdir share; mv lib/*.{png,rhtml} share/.

Currently I'm doing exactly that for all files (also *.rb) which
prevents the namespace problems mentioned above.


I thought using the setup.rb was the "cleaner" method of install, but
looking at this, it's no problem for me if everything stays as-is.

Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

"Wir sind RFC822 von Textmail. Attachments sind zwecklos. Bereiten Sie sich
 auf deren L?schung vor."
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080822/200c21ff/attachment.bin>

From mitch at cgarbs.de  Fri Aug 22 08:30:01 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Fri, 22 Aug 2008 14:30:01 +0200
Subject: [ditz-talk] (no subject)
In-Reply-To: <20080820182140.GA8292@cgarbs.de>
References: <20080820182140.GA8292@cgarbs.de>
Message-ID: <1219408202-29411-1-git-send-email-mitch@cgarbs.de>


> PS: I've also run into a bug with the bash completion:
>     Type "ditz --v" and press tab...
>     I'll look into that.

I can't currently reproduce this, but as the parameters --comment and
--no-comment should be included in the bash completion, I'll come across
this again if it really is an error.

The patch below adds a task to update the bash completion.

Regards,
Christian


From mitch at cgarbs.de  Fri Aug 22 08:30:02 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Fri, 22 Aug 2008 14:30:02 +0200
Subject: [ditz-talk] [PATCH] issue update
In-Reply-To: <1219408202-29411-1-git-send-email-mitch@cgarbs.de>
References: <20080820182140.GA8292@cgarbs.de>
	<1219408202-29411-1-git-send-email-mitch@cgarbs.de>
Message-ID: <1219408202-29411-2-git-send-email-mitch@cgarbs.de>

---
 ...e-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml |   24 ++++++++++++++++++++
 1 files changed, 24 insertions(+), 0 deletions(-)
 create mode 100644 bugs/issue-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml

diff --git a/bugs/issue-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml b/bugs/issue-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml
new file mode 100644
index 0000000..6989b8a
--- /dev/null
+++ b/bugs/issue-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml
@@ -0,0 +1,24 @@
+--- !ditz.rubyforge.org,2008-03-06/issue 
+title: "bash completion: support for --comment and --no-comment"
+desc: add bash completion support for the feature new in 0.5 "all commands that required comments now take --comment and --no-comment"
+type: :task
+component: ditz
+release: "0.6"
+reporter: Christian Garbs <mitch at cgarbs.de>
+status: :in_progress
+disposition: 
+creation_time: 2008-08-22 12:19:45.501116 Z
+references: []
+
+id: b3468b77d26514af3c7667f0d644b3f38f9e29f0
+log_events: 
+- - 2008-08-22 12:19:47.334102 Z
+  - Christian Garbs <mitch at cgarbs.de>
+  - created
+  - ""
+- - 2008-08-22 12:25:51.489908 Z
+  - Christian Garbs <mitch at cgarbs.de>
+  - changed status from unstarted to in_progress
+  - ""
+git_branch: 
+claimer: 
-- 
1.5.6.3



From ohad at lutzky.net  Fri Aug 22 08:46:37 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Fri, 22 Aug 2008 15:46:37 +0300
Subject: [ditz-talk] Hoe dependency?
Message-ID: <8e50e92e0808220546l51c480aeve0dd371fcc75d901@mail.gmail.com>

I'm not sure why, but it seems that the ditz gem has a dependency on
hoe. If so, it is not automatically installed (unlike trollop). Before
installing hoe I get the following error:

/usr/local/lib/site_ruby/1.8/rubygems.rb:578:in
`report_activate_error': Could not find RubyGem hoe (>= 1.7.0)
(Gem::LoadError)
	from /usr/local/lib/site_ruby/1.8/rubygems.rb:134:in `activate'
	from /usr/local/lib/site_ruby/1.8/rubygems.rb:158:in `activate'
	from /usr/local/lib/site_ruby/1.8/rubygems.rb:157:in `each'
	from /usr/local/lib/site_ruby/1.8/rubygems.rb:157:in `activate'
	from /usr/local/lib/site_ruby/1.8/rubygems.rb:158:in `activate'
	from /usr/local/lib/site_ruby/1.8/rubygems.rb:157:in `each'
	from /usr/local/lib/site_ruby/1.8/rubygems.rb:157:in `activate'
	from /usr/local/lib/site_ruby/1.8/rubygems.rb:49:in `gem'
	from /usr/bin/ditz:18

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From dcalford at distributel.ca  Fri Aug 22 09:42:47 2008
From: dcalford at distributel.ca (Dalton Calford)
Date: Fri, 22 Aug 2008 09:42:47 -0400
Subject: [ditz-talk] command line ditz
Message-ID: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>

Is it possible to use ditz in a non-interactive mode?
I can not find any documentation for it.

I would like to be able to pass all the ditz command information on the
command line in order to ease the use of ditz in scripts.

Where would I go to find this section in the code?

best regards

Dalton


From nicolas.pouillard at gmail.com  Sat Aug 23 05:49:01 2008
From: nicolas.pouillard at gmail.com (Nicolas Pouillard)
Date: Sat, 23 Aug 2008 11:49:01 +0200
Subject: [ditz-talk] First impressions
In-Reply-To: <200808200022.31134.ninja@slaphack.com>
References: <200807290315.16589.ninja@slaphack.com>
	<200808192229.06622.ninja@slaphack.com> <1219205251-sup-7036@entry>
	<200808200022.31134.ninja@slaphack.com>
Message-ID: <cd67f63a0808230249r5e16158do8ddd7b791ea1d30a@mail.gmail.com>

On Wed, Aug 20, 2008 at 7:22 AM, David Masover <ninja at slaphack.com> wrote:
> On Tuesday 19 August 2008 23:09:10 William Morgan wrote:
>> Reformatted excerpts from David Masover's message of 2008-08-19:
>> > Is there a reason you need a symbol there at all? What about:
>> >
>> > dits show 77
>>
>> These issue names have to be detectable in title, description and
>> comment strings, so that they can be replaced with the 40-byte issue
>> ids. I suspect just using a number will result in too many false
>> positives.
>
> Yes, when used in title, description, and comment strings. I'm not suggesting
> you only use a number there.
>
> I'm suggesting you allow the pound sign to be omitted in places where the
> number is unambiguous -- specifically, the commandline.

+1

-- 
Nicolas Pouillard

From nicolas.pouillard at gmail.com  Sat Aug 23 05:50:26 2008
From: nicolas.pouillard at gmail.com (Nicolas Pouillard)
Date: Sat, 23 Aug 2008 11:50:26 +0200
Subject: [ditz-talk] First impressions
In-Reply-To: <1219169234-sup-5240@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200807292325.27659.ninja@slaphack.com> <1217449815-sup-4025@entry>
	<200807310949.39086.ninja@slaphack.com> <1217801806-sup-5317@entry>
	<1219169234-sup-5240@entry>
Message-ID: <cd67f63a0808230250n3134e896rf439cf8ff4f7c786@mail.gmail.com>

On Tue, Aug 19, 2008 at 8:07 PM, William Morgan
<wmorgan-ditz at masanjin.net> wrote:
> Reformatted excerpts from William Morgan's message of 2008-08-03:
>> Reformatted excerpts from David Masover's message of 2008-07-31:
>> > Bugfix/Feature/Task should probably be set to some default, as I'm
>> > assuming it must be provided.
>>
>> If you can come up with a default that satisfies everyone, sure. I tend
>> to be about 50% between bugs and features. We could do something like
>> "default to the last type the user picked", and then store that
>> somewhere... dunno.
>
> I'm actually wondering if I should just remove this categorization
> entirely. (Or relegate it to a plugin.)
>
> Do people find it helpful to divide things into features, bugs and
> tasks? There aren't really any tools right now to list things by type
> (though that will be easy now that subcommand arguments is in place).
>
> Personally I can mostly categorize things, but I never do anything with
> that information, so I'm wondering if this is a JIRA-ism left in my
> brain that only makes sense when you think like a Java programmer.

I also use them, But labels could perhaps make them useless.

-- 
Nicolas Pouillard

From ohad at lutzky.net  Sun Aug 24 11:05:51 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Sun, 24 Aug 2008 18:05:51 +0300
Subject: [ditz-talk] Less-than-friendly error on missing branch for ditz html
Message-ID: <8e50e92e0808240805q254c84fdl90bad13c602d2510@mail.gmail.com>

Working on a different machine than I usually do, the "rating" branch
is missing here. One of the bugs has that assigned as its git branch,
and on "ditz html" I get the following, uninformative message:

$ ditz html
fatal: ambiguous argument 'master..rating': unknown revision or path
not in the working tree.
Use '--' to separate paths from revisions
Local generated URL: file:///home/ohad/src/ttime/ditz/index.html

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From wmorgan-ditz at masanjin.net  Sun Aug 24 19:58:18 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 24 Aug 2008 16:58:18 -0700
Subject: [ditz-talk] Less-than-friendly error on missing branch for ditz
	html
In-Reply-To: <8e50e92e0808240805q254c84fdl90bad13c602d2510@mail.gmail.com>
References: <8e50e92e0808240805q254c84fdl90bad13c602d2510@mail.gmail.com>
Message-ID: <1219622168-sup-5639@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-24:
> $ ditz html
> fatal: ambiguous argument 'master..rating': unknown revision or path
> not in the working tree.
> Use '--' to separate paths from revisions

This is a message from git itself, during the shell call. I could
probably just redirect all stderr output to /dev/null now that it's
stable.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Sun Aug 24 20:37:31 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 24 Aug 2008 17:37:31 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <cd67f63a0808230249r5e16158do8ddd7b791ea1d30a@mail.gmail.com>
References: <200807290315.16589.ninja@slaphack.com>
	<200808192229.06622.ninja@slaphack.com> <1219205251-sup-7036@entry>
	<200808200022.31134.ninja@slaphack.com>
	<cd67f63a0808230249r5e16158do8ddd7b791ea1d30a@mail.gmail.com>
Message-ID: <1219624293-sup-3478@entry>

Reformatted excerpts from nicolas.pouillard's message of 2008-08-23:
> > Yes, when used in title, description, and comment strings. I'm not
> > suggesting you only use a number there.
> >
> > I'm suggesting you allow the pound sign to be omitted in places
> > where the number is unambiguous -- specifically, the commandline.
> 
> +1

Interesting. That's two in favor and one against. Any more opinions?

Informal poll:

Reply with +1 if you think it's fine for ditz to take unadorned issue
numbers like "23" on the commandline, while requiring that you mark them
up somehow (like "#23") when using them in a comment, or an issue title
or description.

Reply with -1 if you think the short issue name format should be
consistent on the commandline and within comments, issue titles and
issue descriptions. Note that using a # is interpreted a comment in most
shells, so it will have to be something like "@23", "i23", "+23", "^23",
etc. Feel free to include your favorite prefix.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Sun Aug 24 20:43:01 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 24 Aug 2008 17:43:01 -0700
Subject: [ditz-talk] command line ditz
In-Reply-To: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>
References: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>
Message-ID: <1219624940-sup-6459@entry>

Reformatted excerpts from Dalton Calford's message of 2008-08-22:
> Is it possible to use ditz in a non-interactive mode?
> I can not find any documentation for it.

There's no specific non-interactive mode, but you should be able to pipe
input into it, even in interactive mode.

Depending on what you're doing, you can also create the YAML files
directly.

Suggestions for what a good non-interactive interface would look like
are welcome. One big special commandline argument? A structured text
blob piped into to stdin? Etc.

> Where would I go to find this section in the code?

Model#create_interactively in model.rb interprets the "field" calls in
the model objects in model-object.rb, and both of those places call the
input routines in lowline.rb.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Sun Aug 24 20:47:19 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 24 Aug 2008 17:47:19 -0700
Subject: [ditz-talk] Hoe dependency?
In-Reply-To: <8e50e92e0808220546l51c480aeve0dd371fcc75d901@mail.gmail.com>
References: <8e50e92e0808220546l51c480aeve0dd371fcc75d901@mail.gmail.com>
Message-ID: <1219624990-sup-5878@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-22:
> I'm not sure why, but it seems that the ditz gem has a dependency on
> hoe. If so, it is not automatically installed (unlike trollop).

Curious for two reasons:
1. There's code in Rakefile to explicitly remove the dependency on hoe
   (since it's a bogus one). Maybe that trick doesn't work in recent
   Hoes.

2. If it's a dependency, then why isn't it installed? "gem install"
   always checks deps before installing, and refuses to proceed if
   they're not met.

Has anyone else seen anything like this?
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Sun Aug 24 20:53:33 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 24 Aug 2008 17:53:33 -0700
Subject: [ditz-talk] using setup.rb
In-Reply-To: <20080822120126.GA14345@cgarbs.de>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
	<8e50e92e0808201136u1f00882fqcda165c3ff3985a4@mail.gmail.com>
	<20080821184816.GB24381@cgarbs.de> <1219347708-sup-4912@entry>
	<20080822120126.GA14345@cgarbs.de>
Message-ID: <1219625312-sup-7647@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-22:
> Is it possible to change the setup.rb so that the ditz files don't get
> installed directly into the ruby library directory but into a "ditz"
> subfolder?

Actually I think that better than changing setup.rb would be to move the
.rb files from lib/ to lib/ditz/ within git. (Except for lib/ditz.rb.)
That's the standard Ruby layout and something I should've done in the
first place. 

That sounds like what you're doing manually anyways.

I've created issue cfbb for this, assigned to issue 0.6.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Sun Aug 24 20:55:19 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 24 Aug 2008 17:55:19 -0700
Subject: [ditz-talk] [ANN] ditz 0.5 distributed issue tracker released
In-Reply-To: <20080822121156.GB14345@cgarbs.de>
References: <1219210549-sup-3129@entry> <20080820174813.GA29201@cgarbs.de>
	<20080821184230.GA24381@cgarbs.de> <1219348074-sup-5038@entry>
	<20080822121156.GB14345@cgarbs.de>
Message-ID: <1219625638-sup-4735@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-22:
> But if it's only "making the packager's life easier", then it's ok -

Yep, that's the plan. :)

> I was in fact a little surprised of the 0.5 release announcement
> because there were no calls for "I want to release, please check for
> last bugs or problems anyone" beforehand :-)

But I was just making a few trivial fixes! There's no way I could've
screwed something up!
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Sun Aug 24 20:59:44 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sun, 24 Aug 2008 17:59:44 -0700
Subject: [ditz-talk] (no subject)
In-Reply-To: <1219408202-29411-1-git-send-email-mitch@cgarbs.de>
References: <20080820182140.GA8292@cgarbs.de>
	<1219408202-29411-1-git-send-email-mitch@cgarbs.de>
Message-ID: <1219625925-sup-6342@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-22:
> The patch below adds a task to update the bash completion.

Applied.

This might be the first issue not created by me in the ditz issue db.
-- 
William <wmorgan-ditz at masanjin.net>

From ohad at lutzky.net  Mon Aug 25 04:47:59 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Mon, 25 Aug 2008 11:47:59 +0300
Subject: [ditz-talk] [merge] lutzky/master Easier running of non-installed
	version, fix for unavailable rubygems
Message-ID: <8e50e92e0808250147yb0d25a2h63011486bba25310@mail.gmail.com>

Merge request for git://github.com/lutzky/ditz.git

Instead of running ruby -I"wherever_ditz_is/lib"
"wherever_ditz_is/bin/ditz", this patch allows you to run just
"wherever_ditz_is/bin/ditz".
Also, the rescue-in-case-rubygems-aren't-there syntax didn't work for
me (ruby 1.8.6 (2007-09-24 patchlevel 111) [i486-linux], ubuntu)

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From matt at tplus1.com  Mon Aug 25 11:24:42 2008
From: matt at tplus1.com (Matthew Wilson)
Date: Mon, 25 Aug 2008 11:24:42 -0400
Subject: [ditz-talk] ditz man page install not in path?
Message-ID: <f6f643a20808250824tbd98747m5be48496a58e9425@mail.gmail.com>

I'm running ubuntu hardy heron.  I just installed ditz by typing sudo
ruby setup.rb, and I see it copied the ditz man page into
/usr/share/man/ditz.1.

When I run $ man ditz, that page isn't found.

I'm not really up on man page plumbing, but I think I need to have the
ditz.1 file inside /usr/share/man/man1 instead.  A few questions:

1. Do I need to pass some special flag to setup.rb when I run it?
2. Is there something wrong with my man page configuration?  I haven't
(knowingly, anyway) tweaked it whatsoever.

Matt

-- 
Matthew Wilson
matt at tplus1.com
http://tplus1.com
216-470-6058

From mitch at cgarbs.de  Mon Aug 25 13:23:26 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Mon, 25 Aug 2008 19:23:26 +0200
Subject: [ditz-talk] First impressions
In-Reply-To: <1219624293-sup-3478@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200808192229.06622.ninja@slaphack.com> <1219205251-sup-7036@entry>
	<200808200022.31134.ninja@slaphack.com>
	<cd67f63a0808230249r5e16158do8ddd7b791ea1d30a@mail.gmail.com>
	<1219624293-sup-3478@entry>
Message-ID: <20080825172326.GA3662@cgarbs.de>

On Sun, Aug 24, 2008 at 05:37:31PM -0700, William Morgan wrote:

> Informal poll:

I'm not sure.

I'd say "+1" straight away, but then we'd have "#" as a prefix.  This
could clash with e.g. Debian bugs.  I maintain a todo list for my
Debian packages with ditz and I might say "include a fix for #23456"
where #23456 is a Debian bug and no ditz bug (# is the common prefix
for bugs in Debian changelogs).  OTOH this is a rare corner case and I
could work around it.

I wanted to propose be "?":  It's pretty much unused I think and it
does not need to be quoted in the shell, but it's no ASCII character
so it's not acceptable.

"^23" looks like a negation, "i23" looks like an imaginary number.
"d23" (d for ditz) might work, but it looks like a 23-sided dice...

So, in the lack of proper alternatives, I think I'll stick with:

  +1

Regards,
Christian
(so much to your short poll with those two simple valid answers...)
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Die beste Methode, Windows zu beschleunigen, arbeitet mit 9,81 m/s?.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080825/46f62cf4/attachment.bin>

From ohad at lutzky.net  Mon Aug 25 14:44:13 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Mon, 25 Aug 2008 21:44:13 +0300
Subject: [ditz-talk] First impressions
In-Reply-To: <20080825172326.GA3662@cgarbs.de>
References: <200807290315.16589.ninja@slaphack.com>
	<200808192229.06622.ninja@slaphack.com> <1219205251-sup-7036@entry>
	<200808200022.31134.ninja@slaphack.com>
	<cd67f63a0808230249r5e16158do8ddd7b791ea1d30a@mail.gmail.com>
	<1219624293-sup-3478@entry> <20080825172326.GA3662@cgarbs.de>
Message-ID: <8e50e92e0808251144j49f33604la705d526b38cff95@mail.gmail.com>

On Mon, Aug 25, 2008 at 8:23 PM, Christian Garbs <mitch at cgarbs.de> wrote:
> On Sun, Aug 24, 2008 at 05:37:31PM -0700, William Morgan wrote:
>
>> Informal poll:
>
> I'm not sure.
>
> I'd say "+1" straight away, but then we'd have "#" as a prefix.  This
> could clash with e.g. Debian bugs.  I maintain a todo list for my
> Debian packages with ditz and I might say "include a fix for #23456"
> where #23456 is a Debian bug and no ditz bug (# is the common prefix
> for bugs in Debian changelogs).  OTOH this is a rare corner case and I
> could work around it.
>
> I wanted to propose be "?":  It's pretty much unused I think and it
> does not need to be quoted in the shell, but it's no ASCII character
> so it's not acceptable.
>
> "^23" looks like a negation, "i23" looks like an imaginary number.
> "d23" (d for ditz) might work, but it looks like a 23-sided dice...

Given this reasoning, how about #d23 ?

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From ninja at slaphack.com  Tue Aug 26 23:22:41 2008
From: ninja at slaphack.com (David Masover)
Date: Tue, 26 Aug 2008 22:22:41 -0500
Subject: [ditz-talk] First impressions
In-Reply-To: <20080825172326.GA3662@cgarbs.de>
References: <200807290315.16589.ninja@slaphack.com>
	<1219624293-sup-3478@entry> <20080825172326.GA3662@cgarbs.de>
Message-ID: <200808262222.41473.ninja@slaphack.com>

On Monday 25 August 2008 12:23:26 Christian Garbs wrote:
> On Sun, Aug 24, 2008 at 05:37:31PM -0700, William Morgan wrote:
> 
> > Informal poll:
> 
> I'm not sure.
> 
> I'd say "+1" straight away, but then we'd have "#" as a prefix.  This
> could clash with e.g. Debian bugs.

Well, I was going to say that false positives are OK, but this seems like a 
legitimate case.

What if it was configurable? It only needs to be consistent within a given 
project.

From wmorgan-ditz at masanjin.net  Wed Aug 27 11:37:26 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 27 Aug 2008 08:37:26 -0700
Subject: [ditz-talk] First impressions
In-Reply-To: <200808192240.29726.ninja@slaphack.com>
References: <200807290315.16589.ninja@slaphack.com>
	<1217801806-sup-5317@entry> <1219169234-sup-5240@entry>
	<200808192240.29726.ninja@slaphack.com>
Message-ID: <1219851392-sup-6350@entry>

Reformatted excerpts from David Masover's message of 2008-08-19:
> I think what might be more generally useful is a generic tagging
> system. After all, as I've just described, it's possible for something
> to be both a bugfix and an enhancement.

Nicolas was thinking about this for a while (as a replacement for
categories). Now that the plugin system is stable, this would be pretty
easy to do.
-- 
William <wmorgan-ditz at masanjin.net>

From nicolas.pouillard at gmail.com  Wed Aug 27 12:02:11 2008
From: nicolas.pouillard at gmail.com (Nicolas Pouillard)
Date: Wed, 27 Aug 2008 18:02:11 +0200
Subject: [ditz-talk] First impressions
In-Reply-To: <1219624293-sup-3478@entry>
References: <200807290315.16589.ninja@slaphack.com>
	<200808192229.06622.ninja@slaphack.com> <1219205251-sup-7036@entry>
	<200808200022.31134.ninja@slaphack.com>
	<cd67f63a0808230249r5e16158do8ddd7b791ea1d30a@mail.gmail.com>
	<1219624293-sup-3478@entry>
Message-ID: <1219852861-sup-6440@ausone.local>

Excerpts from William Morgan's message of Mon Aug 25 02:37:31 +0200 2008:
> Reformatted excerpts from nicolas.pouillard's message of 2008-08-23:
> > > Yes, when used in title, description, and comment strings. I'm not
> > > suggesting you only use a number there.
> > >
> > > I'm suggesting you allow the pound sign to be omitted in places
> > > where the number is unambiguous -- specifically, the commandline.
> > 
> > +1
> 
> Interesting. That's two in favor and one against. Any more opinions?
> 
> Informal poll:
> 
> Reply with +1 if you think it's fine for ditz to take unadorned issue
> numbers like "23" on the commandline, while requiring that you mark them
> up somehow (like "#23") when using them in a comment, or an issue title
> or description.

+1

> Reply with -1 if you think the short issue name format should be
> consistent on the commandline and within comments, issue titles and
> issue descriptions. Note that using a # is interpreted a comment in most
> shells, so it will have to be something like "@23", "i23", "+23", "^23",
> etc. Feel free to include your favorite prefix.

The i23 is also a nice choice.

-- 
Nicolas Pouillard aka Ertai
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 194 bytes
Desc: not available
URL: <http://rubyforge.org/pipermail/ditz-talk/attachments/20080827/dbc50352/attachment-0001.bin>

From wmorgan-ditz at masanjin.net  Wed Aug 27 14:04:31 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 27 Aug 2008 11:04:31 -0700
Subject: [ditz-talk] [merge] lutzky/master Easier running of
	non-installed version, fix for unavailable rubygems
In-Reply-To: <8e50e92e0808250147yb0d25a2h63011486bba25310@mail.gmail.com>
References: <8e50e92e0808250147yb0d25a2h63011486bba25310@mail.gmail.com>
Message-ID: <1219860044-sup-1020@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-25:
> Instead of running ruby -I"wherever_ditz_is/lib"
> "wherever_ditz_is/bin/ditz", this patch allows you to run just
> "wherever_ditz_is/bin/ditz".

This part, I think, is fine, but this part:

> Also, the rescue-in-case-rubygems-aren't-there syntax didn't work for
> me (ruby 1.8.6 (2007-09-24 patchlevel 111) [i486-linux], ubuntu)

conflicts with a fix that's been in there on the 12th (commit 1cb0995b).
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Wed Aug 27 14:19:20 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 27 Aug 2008 11:19:20 -0700
Subject: [ditz-talk] ditz man page install not in path?
In-Reply-To: <f6f643a20808250824tbd98747m5be48496a58e9425@mail.gmail.com>
References: <f6f643a20808250824tbd98747m5be48496a58e9425@mail.gmail.com>
Message-ID: <1219861098-sup-9727@entry>

Reformatted excerpts from Matthew Wilson's message of 2008-08-25:
> I'm not really up on man page plumbing, but I think I need to have the
> ditz.1 file inside /usr/share/man/man1 instead.  A few questions:

I think you're right. I've fixed this in git master (issue a9ffc7) and
it will go out in 0.6.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Wed Aug 27 14:21:45 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 27 Aug 2008 11:21:45 -0700
Subject: [ditz-talk] command line ditz
In-Reply-To: <1219668763.7205.1.camel@dcalford-laptop.corp.distributel.ca>
References: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>
	<1219624940-sup-6459@entry>
	<1219668763.7205.1.camel@dcalford-laptop.corp.distributel.ca>
Message-ID: <1219861183-sup-3519@entry>

Reformatted excerpts from Dalton Calford's message of 2008-08-25:
> Is there a graphical interface for Ditz?

Not yet. I'm working on getting a webapp (called sheila) working with
recent ditz versions. That's tentatively scheduled for 0.6 but no
guarantees.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Wed Aug 27 14:25:04 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 27 Aug 2008 11:25:04 -0700
Subject: [ditz-talk] Hoe dependency?
In-Reply-To: <8e50e92e0808250123yc4086ebge12fd52317d4d5fe@mail.gmail.com>
References: <8e50e92e0808220546l51c480aeve0dd371fcc75d901@mail.gmail.com>
	<1219624990-sup-5878@entry>
	<8e50e92e0808250123yc4086ebge12fd52317d4d5fe@mail.gmail.com>
Message-ID: <1219861328-sup-1734@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-25:
> Is it listed as an actual dependency in the gemspec?

Apparently so, despite my attempts to remove it.
-- 
William <wmorgan-ditz at masanjin.net>

From ohad at lutzky.net  Wed Aug 27 15:56:33 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Wed, 27 Aug 2008 22:56:33 +0300
Subject: [ditz-talk] Hoe dependency?
In-Reply-To: <1219861328-sup-1734@entry>
References: <8e50e92e0808220546l51c480aeve0dd371fcc75d901@mail.gmail.com>
	<1219624990-sup-5878@entry>
	<8e50e92e0808250123yc4086ebge12fd52317d4d5fe@mail.gmail.com>
	<1219861328-sup-1734@entry>
Message-ID: <8e50e92e0808271256x639a2fferdb2814614c9f2f8d@mail.gmail.com>

On Wed, Aug 27, 2008 at 9:25 PM, William Morgan
<wmorgan-ditz at masanjin.net> wrote:
> Reformatted excerpts from Ohad Lutzky's message of 2008-08-25:
>> Is it listed as an actual dependency in the gemspec?
>
> Apparently so, despite my attempts to remove it.

...so... should it be a dependency? Should it be in the gemspec?



-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From ohad at lutzky.net  Wed Aug 27 16:05:16 2008
From: ohad at lutzky.net (Ohad Lutzky)
Date: Wed, 27 Aug 2008 23:05:16 +0300
Subject: [ditz-talk] [merge] lutzky/master Easier running of
	non-installed version, fix for unavailable rubygems
In-Reply-To: <1219860044-sup-1020@entry>
References: <8e50e92e0808250147yb0d25a2h63011486bba25310@mail.gmail.com>
	<1219860044-sup-1020@entry>
Message-ID: <8e50e92e0808271305ufb7b556of6a5c87ea4765ef0@mail.gmail.com>

On Wed, Aug 27, 2008 at 9:04 PM, William Morgan
<wmorgan-ditz at masanjin.net> wrote:
> Reformatted excerpts from Ohad Lutzky's message of 2008-08-25:
>> Instead of running ruby -I"wherever_ditz_is/lib"
>> "wherever_ditz_is/bin/ditz", this patch allows you to run just
>> "wherever_ditz_is/bin/ditz".
>
> This part, I think, is fine, but this part:
>
>> Also, the rescue-in-case-rubygems-aren't-there syntax didn't work for
>> me (ruby 1.8.6 (2007-09-24 patchlevel 111) [i486-linux], ubuntu)
>
> conflicts with a fix that's been in there on the 12th (commit 1cb0995b).

Fixed, rebased to avoid ugly merge. Sorry, anyone tracking me :)

-- 
Man is the only animal that laughs and weeps, for he is the only
animal that is struck with the difference between what things are and
what they ought to be.
 - William Hazlitt

Ohad Lutzky

From wmorgan-ditz at masanjin.net  Wed Aug 27 16:19:52 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Wed, 27 Aug 2008 13:19:52 -0700
Subject: [ditz-talk] Hoe dependency?
In-Reply-To: <8e50e92e0808271256x639a2fferdb2814614c9f2f8d@mail.gmail.com>
References: <8e50e92e0808220546l51c480aeve0dd371fcc75d901@mail.gmail.com>
	<1219624990-sup-5878@entry>
	<8e50e92e0808250123yc4086ebge12fd52317d4d5fe@mail.gmail.com>
	<1219861328-sup-1734@entry>
	<8e50e92e0808271256x639a2fferdb2814614c9f2f8d@mail.gmail.com>
Message-ID: <1219868189-sup-9759@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-27:
> ...so... should it be a dependency? Should it be in the gemspec?

No and no. It's only in there because it automatically adds itself to
all gems in creates (presumably to push it up in the list of most
frequently-downloaded gems), and my previous hack to remove it was
invalidated. Fixed now.
-- 
William <wmorgan-ditz at masanjin.net>

From dcalford at distributel.ca  Thu Aug 28 10:18:51 2008
From: dcalford at distributel.ca (Dalton Calford)
Date: Thu, 28 Aug 2008 10:18:51 -0400
Subject: [ditz-talk] command line ditz
In-Reply-To: <1219861183-sup-3519@entry>
References: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>
	<1219624940-sup-6459@entry>
	<1219668763.7205.1.camel@dcalford-laptop.corp.distributel.ca>
	<1219861183-sup-3519@entry>
Message-ID: <1219933131.7045.5.camel@dcalford-laptop.corp.distributel.ca>

I have not found a gem for sheila, but a search on the web came up with
something that looks like it.

Can someone please point me to the proper repository for sheila?

Also, the documentation speaks about the .ditz-plugins file but it is
not autocreated and no details about it's format is given.

Plus, as I am trying to learn this project and get it ready for a group
of non-ditz, non-git developers, I want to document the project as much
as possible.

Where can I find all existing documentation on the project (I am hoping
thier is more than what I have found to date) and where should I put my
changes so that everyone can benefit?

best regards

Dalton


On Wed, 2008-08-27 at 11:21 -0700, William Morgan wrote:
> Reformatted excerpts from Dalton Calford's message of 2008-08-25:
> > Is there a graphical interface for Ditz?
> 
> Not yet. I'm working on getting a webapp (called sheila) working with
> recent ditz versions. That's tentatively scheduled for 0.6 but no
> guarantees.


From wmorgan-ditz at masanjin.net  Thu Aug 28 14:19:24 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Thu, 28 Aug 2008 11:19:24 -0700
Subject: [ditz-talk] command line ditz
In-Reply-To: <1219933131.7045.5.camel@dcalford-laptop.corp.distributel.ca>
References: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>
	<1219624940-sup-6459@entry>
	<1219668763.7205.1.camel@dcalford-laptop.corp.distributel.ca>
	<1219861183-sup-3519@entry>
	<1219933131.7045.5.camel@dcalford-laptop.corp.distributel.ca>
Message-ID: <1219946903-sup-9253@entry>

Reformatted excerpts from Dalton Calford's message of 2008-08-28:
> Can someone please point me to the proper repository for sheila?

http://github.com/why/sheila/

Once my sheila changes are ready I'll publish them as a clone of that
repo.

> Also, the documentation speaks about the .ditz-plugins file but it is
> not autocreated and no details about it's format is given.

The plugins documentation tells you how to construct it, but I suppose
it doesn't say what the format official is: a YAML array of strings.

> Where can I find all existing documentation on the project (I am
> hoping thier is more than what I have found to date) and where should
> I put my changes so that everyone can benefit?

Everything that's been written is in the ditz repo.

As for where to put your changes, just clone the ditz git repo (on
Github, Gitorious, or your own hosting solution) and tell us about
it. :)
-- 
William <wmorgan-ditz at masanjin.net>

From dcalford at distributel.ca  Thu Aug 28 14:57:14 2008
From: dcalford at distributel.ca (Dalton Calford)
Date: Thu, 28 Aug 2008 14:57:14 -0400
Subject: [ditz-talk] command line ditz
In-Reply-To: <1219946903-sup-9253@entry>
References: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>
	<1219624940-sup-6459@entry>
	<1219668763.7205.1.camel@dcalford-laptop.corp.distributel.ca>
	<1219861183-sup-3519@entry>
	<1219933131.7045.5.camel@dcalford-laptop.corp.distributel.ca>
	<1219946903-sup-9253@entry>
Message-ID: <1219949834.5078.3.camel@dcalford-laptop.corp.distributel.ca>

Sorry,

I have just read through the ditz plugin documentation again, and
although it says how to modify an existing .ditz-plugins file, it has no
instruction on how to create a .ditz-plugins file.
Could you please include a .ditz-plugins file as an attachment so that I
can look at it and determine it's use in my projects.

Thanks

On Thu, 2008-08-28 at 11:19 -0700, William Morgan wrote:
> Reformatted excerpts from Dalton Calford's message of 2008-08-28:
> > Can someone please point me to the proper repository for sheila?
> 
> http://github.com/why/sheila/
> 
> Once my sheila changes are ready I'll publish them as a clone of that
> repo.
> 
> > Also, the documentation speaks about the .ditz-plugins file but it is
> > not autocreated and no details about it's format is given.
> 
> The plugins documentation tells you how to construct it, but I suppose
> it doesn't say what the format official is: a YAML array of strings.
> 
> > Where can I find all existing documentation on the project (I am
> > hoping thier is more than what I have found to date) and where should
> > I put my changes so that everyone can benefit?
> 
> Everything that's been written is in the ditz repo.
> 
> As for where to put your changes, just clone the ditz git repo (on
> Github, Gitorious, or your own hosting solution) and tell us about
> it. :)


From wmorgan-ditz at masanjin.net  Thu Aug 28 19:19:09 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Thu, 28 Aug 2008 16:19:09 -0700
Subject: [ditz-talk] command line ditz
In-Reply-To: <1219949834.5078.3.camel@dcalford-laptop.corp.distributel.ca>
References: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>
	<1219624940-sup-6459@entry>
	<1219668763.7205.1.camel@dcalford-laptop.corp.distributel.ca>
	<1219861183-sup-3519@entry>
	<1219933131.7045.5.camel@dcalford-laptop.corp.distributel.ca>
	<1219946903-sup-9253@entry>
	<1219949834.5078.3.camel@dcalford-laptop.corp.distributel.ca>
Message-ID: <1219965503-sup-5894@entry>

Reformatted excerpts from Dalton Calford's message of 2008-08-28:
> I have just read through the ditz plugin documentation again, and
> although it says how to modify an existing .ditz-plugins file, it has
> no instruction on how to create a .ditz-plugins file.

I've updated PLUGINS.txt.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Thu Aug 28 19:21:25 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Thu, 28 Aug 2008 16:21:25 -0700
Subject: [ditz-talk] command line ditz
In-Reply-To: <1219946903-sup-9253@entry>
References: <1219412567.7581.2.camel@dcalford-laptop.corp.distributel.ca>
	<1219624940-sup-6459@entry>
	<1219668763.7205.1.camel@dcalford-laptop.corp.distributel.ca>
	<1219861183-sup-3519@entry>
	<1219933131.7045.5.camel@dcalford-laptop.corp.distributel.ca>
	<1219946903-sup-9253@entry>
Message-ID: <1219965675-sup-6749@entry>

Reformatted excerpts from William Morgan's message of 2008-08-28:
> Once my sheila changes are ready I'll publish them as a clone of that
> repo.

I have something that works here:
  http://github.com/wmorgan/sheila/tree/master

It requires the very latest git versions of both ditz and git. It also
currently doesn't have any user verification, or do any git committing.
But it does allow you to add issues!
-- 
William <wmorgan-ditz at masanjin.net>

From mitch at cgarbs.de  Fri Aug 29 16:51:47 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Fri, 29 Aug 2008 22:51:47 +0200
Subject: [ditz-talk] [PATCH] simplify bash completion
In-Reply-To: <1220043108-23896-1-git-send-email-mitch@cgarbs.de>
References: <1220043108-23896-1-git-send-email-mitch@cgarbs.de>
Message-ID: <1220043108-23896-2-git-send-email-mitch@cgarbs.de>

---
 contrib/completion/ditz.bash |   11 ++++-------
 1 files changed, 4 insertions(+), 7 deletions(-)

diff --git a/contrib/completion/ditz.bash b/contrib/completion/ditz.bash
index 4a17c72..89f047f 100644
--- a/contrib/completion/ditz.bash
+++ b/contrib/completion/ditz.bash
@@ -9,13 +9,10 @@ _ditz()
     local cur=${COMP_WORDS[COMP_CWORD]}
     if [ $COMP_CWORD -eq 1 ]; then
 	COMPREPLY=( $( compgen -W "$(ditz --commands)" -- $cur ) )
-    elif [ $COMP_CWORD -eq 2 ]; then
-	local cmd=${COMP_WORDS[1]}
-	COMPREPLY=( $( compgen -W "$(ditz "$cmd" '<options>' 2>/dev/null)" -- $cur ) )
-    elif [ $COMP_CWORD -eq 3 ]; then
-	local cmd=${COMP_WORDS[1]}
-	local parm1=${COMP_WORDS[2]}
-	COMPREPLY=( $( compgen -W "$(ditz "$cmd" "$parm1" '<options>' 2>/dev/null)" -- $cur ) )
+    else
+	unset COMP_WORDS[COMP_CWORD]  # remove last
+	unset COMP_WORDS[0]           # remove first
+	COMPREPLY=( $( compgen -W "$(ditz "${COMP_WORDS[@]}" '<options>' 2>/dev/null)" -- $cur ) )
     fi 
 }
 
-- 
1.5.6.3



From mitch at cgarbs.de  Fri Aug 29 16:51:48 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Fri, 29 Aug 2008 22:51:48 +0200
Subject: [ditz-talk] [PATCH] bash completion: support commandline options
In-Reply-To: <1220043108-23896-2-git-send-email-mitch@cgarbs.de>
References: <1220043108-23896-1-git-send-email-mitch@cgarbs.de>
	<1220043108-23896-2-git-send-email-mitch@cgarbs.de>
Message-ID: <1220043108-23896-3-git-send-email-mitch@cgarbs.de>

add support for --comment, --no-comment and --silent where applicable
---
 ...e-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml |    8 +++++-
 contrib/completion/ditz.bash                       |   21 +++++++++++++++++++-
 2 files changed, 26 insertions(+), 3 deletions(-)

diff --git a/bugs/issue-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml b/bugs/issue-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml
index 6989b8a..8015766 100644
--- a/bugs/issue-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml
+++ b/bugs/issue-b3468b77d26514af3c7667f0d644b3f38f9e29f0.yaml
@@ -5,8 +5,8 @@ type: :task
 component: ditz
 release: "0.6"
 reporter: Christian Garbs <mitch at cgarbs.de>
-status: :in_progress
-disposition: 
+status: :closed
+disposition: :fixed
 creation_time: 2008-08-22 12:19:45.501116 Z
 references: []
 
@@ -20,5 +20,9 @@ log_events:
   - Christian Garbs <mitch at cgarbs.de>
   - changed status from unstarted to in_progress
   - ""
+- - 2008-08-29 20:40:26.189706 Z
+  - Christian Garbs <mitch at cgarbs.de>
+  - closed with disposition fixed
+  - fixed, also enabled --silent for the edit command
 git_branch: 
 claimer: 
diff --git a/contrib/completion/ditz.bash b/contrib/completion/ditz.bash
index 89f047f..1bbdf43 100644
--- a/contrib/completion/ditz.bash
+++ b/contrib/completion/ditz.bash
@@ -7,12 +7,31 @@
 _ditz() 
 {
     local cur=${COMP_WORDS[COMP_CWORD]}
+
     if [ $COMP_CWORD -eq 1 ]; then
+	# no command yet, show all commands
 	COMPREPLY=( $( compgen -W "$(ditz --commands)" -- $cur ) )
+
     else
 	unset COMP_WORDS[COMP_CWORD]  # remove last
 	unset COMP_WORDS[0]           # remove first
-	COMPREPLY=( $( compgen -W "$(ditz "${COMP_WORDS[@]}" '<options>' 2>/dev/null)" -- $cur ) )
+	
+	# add options if applicable...
+	local options
+	if [ "${cur:0:1}" = '-' ]; then
+	    # ...but only if at least a dash is given
+	    case "${COMP_WORDS[1]}" in
+		add|add_reference|add_release|assign|close|comment|release|set_component|start|stop|unassign)
+		    options="--comment --no-comment"
+		    ;;
+		edit)
+		    options="--comment --no-comment --silent"
+		    ;;
+	    esac
+	fi
+	
+	# let ditz parse the commandline and print available completions, then append the options form above
+	COMPREPLY=( $( compgen -W "$(ditz "${COMP_WORDS[@]}" '<options>' 2>/dev/null) $options" -- $cur ) )
     fi 
 }
 
-- 
1.5.6.3



From mitch at cgarbs.de  Fri Aug 29 16:51:46 2008
From: mitch at cgarbs.de (Christian Garbs)
Date: Fri, 29 Aug 2008 22:51:46 +0200
Subject: [ditz-talk] bash completion updates
Message-ID: <1220043108-23896-1-git-send-email-mitch@cgarbs.de>

Hi there,

I've updated the bash completion.  It has been simpliefied and
afterwards expanded to support the options --comment, --no-comment and
--silent where applicable.  The second patch depends on the first.

The options will only be expanded when you at least type a single dash
and press tab.  I did this on purpose so that "ditz edit <tab>" will
still show all issues instead of a completion list like this:

--comment
ditz_1_bar
ditz_2_baz
ditz_3_grok_the_foo
ditz_4_frobnetic
--no-comment
--silent

I this case you'd have to type "d" to get your issue list, I don't
find that very useful.

Regards,
Christian
-- 
....Christian.Garbs.....................................http://www.cgarbs.de

Quark!  Quark!  Beware the quantum duck!


From wmorgan-ditz at masanjin.net  Sat Aug 30 19:01:27 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sat, 30 Aug 2008 16:01:27 -0700
Subject: [ditz-talk] bash completion updates
In-Reply-To: <1220043108-23896-1-git-send-email-mitch@cgarbs.de>
References: <1220043108-23896-1-git-send-email-mitch@cgarbs.de>
Message-ID: <1220137273-sup-4471@entry>

Reformatted excerpts from Christian Garbs's message of 2008-08-29:
> I've updated the bash completion.  It has been simpliefied and
> afterwards expanded to support the options --comment, --no-comment and
> --silent where applicable.  The second patch depends on the first.

Applied, thanks.
-- 
William <wmorgan-ditz at masanjin.net>

From wmorgan-ditz at masanjin.net  Sat Aug 30 19:15:17 2008
From: wmorgan-ditz at masanjin.net (William Morgan)
Date: Sat, 30 Aug 2008 16:15:17 -0700
Subject: [ditz-talk] [merge] lutzky/master Easier running of
	non-installed version, fix for unavailable rubygems
In-Reply-To: <8e50e92e0808271305ufb7b556of6a5c87ea4765ef0@mail.gmail.com>
References: <8e50e92e0808250147yb0d25a2h63011486bba25310@mail.gmail.com>
	<1219860044-sup-1020@entry>
	<8e50e92e0808271305ufb7b556of6a5c87ea4765ef0@mail.gmail.com>
Message-ID: <1220138096-sup-8088@entry>

Reformatted excerpts from Ohad Lutzky's message of 2008-08-27:
> Fixed, rebased to avoid ugly merge. Sorry, anyone tracking me :)

Merged, thanks.
-- 
William <wmorgan-ditz at masanjin.net>