From weyus at att.net  Mon Jul  3 12:31:28 2006
From: weyus at att.net (Wes Gamble)
Date: Mon, 03 Jul 2006 11:31:28 -0500
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <20060630192647.GC26372@okcomputer.antiflux.org>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
Message-ID: <44A94660.1090100@att.net>

When I do gem update, and I see:

C:\eclipse\workspace>gem update
Updating installed gems...
Need to update 21 gems from http://gems.rubyforge.org
.....................
complete
Attempting remote update of rubyzip
Successfully installed rubyzip-0.9.1
Gems: [rubyzip] updated

what does the "Need to update 21 gems from ..." mean exactly? 

Does it mean that there are 21 gems out of date?
Does it mean that I have 21 gems installed and they're being inspected?

It's very confusing. 

Thanks,
Wes

From jim.weirich at gmail.com  Mon Jul  3 12:42:12 2006
From: jim.weirich at gmail.com (Jim Weirich)
Date: Mon, 3 Jul 2006 12:42:12 -0400
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <44A94660.1090100@att.net>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
	<44A94660.1090100@att.net>
Message-ID: <d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>

On 7/3/06, Wes Gamble <weyus at att.net> wrote:

> what does the "Need to update 21 gems from ..." mean exactly?
>
> Does it mean that there are 21 gems out of date?
> Does it mean that I have 21 gems installed and they're being inspected?



It means that there are 21 gems available on RubyForge that your local gem
repository knows nothing about.  It needs to update the local gem list with
information from those 21 gems before it can proceed with any requests.

-- 
-- Jim Weirich    jim at weirichhouse.org     http://onestepback.org
-----------------------------------------------------------------
"Beware of bugs in the above code; I have only proved it correct,
not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20060703/42673090/attachment-0001.html 

From weyus at att.net  Mon Jul  3 12:48:05 2006
From: weyus at att.net (Wes Gamble)
Date: Mon, 03 Jul 2006 11:48:05 -0500
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
References: <20060630192647.GC26372@okcomputer.antiflux.org>	<44A94660.1090100@att.net>
	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
Message-ID: <44A94A45.3040003@att.net>

So this is not a problem, correct?   It just means that I am not using 
every gem in creation right?

Jim Weirich wrote:

> On 7/3/06, *Wes Gamble* <weyus at att.net <mailto:weyus at att.net>> wrote:
>
>     what does the "Need to update 21 gems from ..." mean exactly?
>
>     Does it mean that there are 21 gems out of date?
>     Does it mean that I have 21 gems installed and they're being
>     inspected?
>
>
>
> It means that there are 21 gems available on RubyForge that your local 
> gem repository knows nothing about.  It needs to update the local gem 
> list with information from those 21 gems before it can proceed with 
> any requests.
>
> -- 
> -- Jim Weirich    jim at weirichhouse.org 
> <mailto:jim at weirichhouse.org>     http://onestepback.org
> -----------------------------------------------------------------
> "Beware of bugs in the above code; I have only proved it correct,
> not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas)
>
>------------------------------------------------------------------------
>
>_______________________________________________
>Rubygems-developers mailing list
>Rubygems-developers at rubyforge.org
>http://rubyforge.org/mailman/listinfo/rubygems-developers
>

From mfp at acm.org  Mon Jul  3 13:44:33 2006
From: mfp at acm.org (Mauricio Fernandez)
Date: Mon, 3 Jul 2006 19:44:33 +0200
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
	<44A94660.1090100@att.net>
	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
Message-ID: <20060703174433.GI12069@tux-chan>

On Mon, Jul 03, 2006 at 12:42:12PM -0400, Jim Weirich wrote:
> On 7/3/06, Wes Gamble <weyus at att.net> wrote:
> 
> >what does the "Need to update 21 gems from ..." mean exactly?
> >
> >Does it mean that there are 21 gems out of date?
> >Does it mean that I have 21 gems installed and they're being inspected?
> 
> It means that there are 21 gems available on RubyForge that your local gem
> repository knows nothing about.  It needs to update the local gem list with
> information from those 21 gems before it can proceed with any requests.

"Need to update 21 gemspecs from ..."

-- 
Mauricio Fernandez  -   http://eigenclass.org   -  singular Ruby

From weyus at att.net  Mon Jul  3 13:57:00 2006
From: weyus at att.net (Wes Gamble)
Date: Mon, 03 Jul 2006 12:57:00 -0500
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <20060703174433.GI12069@tux-chan>
References: <20060630192647.GC26372@okcomputer.antiflux.org>	<44A94660.1090100@att.net>	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
	<20060703174433.GI12069@tux-chan>
Message-ID: <44A95A6C.9050607@att.net>

Mauricio,

Can you explain this more?  Does this mean that there are 21 gems 
available on RubyForge that I am simply not using?

Why do I care about gems that I'm not using?

WG

Mauricio Fernandez wrote:

>On Mon, Jul 03, 2006 at 12:42:12PM -0400, Jim Weirich wrote:
>  
>
>>On 7/3/06, Wes Gamble <weyus at att.net> wrote:
>>
>>    
>>
>>>what does the "Need to update 21 gems from ..." mean exactly?
>>>
>>>Does it mean that there are 21 gems out of date?
>>>Does it mean that I have 21 gems installed and they're being inspected?
>>>      
>>>
>>It means that there are 21 gems available on RubyForge that your local gem
>>repository knows nothing about.  It needs to update the local gem list with
>>information from those 21 gems before it can proceed with any requests.
>>    
>>
>
>"Need to update 21 gemspecs from ..."
>
>  
>

From jmg3000 at gmail.com  Mon Jul  3 14:14:55 2006
From: jmg3000 at gmail.com (John Gabriele)
Date: Mon, 3 Jul 2006 14:14:55 -0400
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <44A95A6C.9050607@att.net>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
	<44A94660.1090100@att.net>
	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
	<20060703174433.GI12069@tux-chan> <44A95A6C.9050607@att.net>
Message-ID: <65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>

On 7/3/06, Wes Gamble <weyus at att.net> wrote:
> Mauricio,
>
> Can you explain this more?  Does this mean that there are 21 gems
> available on RubyForge that I am simply not using?
>
> Why do I care about gems that I'm not using?

The RubyForge gem repository gets updated regularly with new and updated gems.

When you run the gem command -- before doing anything else -- it has
to make sure it's got a current list of all available gems. With
previous versions of RubyGems (< 0.9.0), the gem command would have to
download the entire current list of gems (for a given repository -- by
default just the RubyForge one) whenever it noticed it was out of
date.

The current (0.9.0) gem command downloads this list incrementally.
When it tells you that there are 21 gems available, I think it's
partly supposed be a reassuring note to let you know that the
repository has been significantly updated since the last time you
checked. :)

---John

From weyus at att.net  Mon Jul  3 14:36:52 2006
From: weyus at att.net (Wes Gamble)
Date: Mon, 03 Jul 2006 13:36:52 -0500
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>
References: <20060630192647.GC26372@okcomputer.antiflux.org>	<44A94660.1090100@att.net>	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>	<20060703174433.GI12069@tux-chan>
	<44A95A6C.9050607@att.net>
	<65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>
Message-ID: <44A963C4.6050602@att.net>

John,

Thanks.  But the bottom line is - it's giving me information about the 
repository _in general_, not information that pertains to _my_ 
collection of gems.

Is this correct?

I still don't know what it means.  Does it mean that 21 gems were 
somehow modified in the repository since the last time I ran the gem 
command?  If that's what it means, I don't think it makes a lot of sense 
to show this message "Need to update xxx gems..." unless it's qualified 
to be very clear that this is a message about the internal state of the 
gem command execution context and has nothing to do with the end user's 
local repository of gems.  I'm not sure I need to see detailed internal 
synchronization messages for the "gem" application.

Having said that, I won't worry any more when I see the message :).

Thanks,
WG

John Gabriele wrote:

>On 7/3/06, Wes Gamble <weyus at att.net> wrote:
>  
>
>>Mauricio,
>>
>>Can you explain this more?  Does this mean that there are 21 gems
>>available on RubyForge that I am simply not using?
>>
>>Why do I care about gems that I'm not using?
>>    
>>
>
>The RubyForge gem repository gets updated regularly with new and updated gems.
>
>When you run the gem command -- before doing anything else -- it has
>to make sure it's got a current list of all available gems. With
>previous versions of RubyGems (< 0.9.0), the gem command would have to
>download the entire current list of gems (for a given repository -- by
>default just the RubyForge one) whenever it noticed it was out of
>date.
>
>The current (0.9.0) gem command downloads this list incrementally.
>When it tells you that there are 21 gems available, I think it's
>partly supposed be a reassuring note to let you know that the
>repository has been significantly updated since the last time you
>checked. :)
>
>---John
>_______________________________________________
>Rubygems-developers mailing list
>Rubygems-developers at rubyforge.org
>http://rubyforge.org/mailman/listinfo/rubygems-developers
>
>  
>

From jim.weirich at gmail.com  Mon Jul  3 15:29:58 2006
From: jim.weirich at gmail.com (Jim Weirich)
Date: Mon, 3 Jul 2006 15:29:58 -0400
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <44A963C4.6050602@att.net>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
	<44A94660.1090100@att.net>
	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
	<20060703174433.GI12069@tux-chan> <44A95A6C.9050607@att.net>
	<65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>
	<44A963C4.6050602@att.net>
Message-ID: <d4dd9080607031229x29f7bc7dme29e19fd5a6fdf39@mail.gmail.com>

On 7/3/06, Wes Gamble <weyus at att.net> wrote:
>
>
> Thanks.  But the bottom line is - it's giving me information about the
> repository _in general_, not information that pertains to _my_
> collection of gems.


Ummm ....  Yes.

[...] I'm not sure I need to see detailed internal
> synchronization messages for the "gem" application.


The synchronization process takes time.  One of of the problems with the
0.8.x RubyGems series was that the synchronization could take a LONG time
without any feedback on progress.  The feedback provided by 0.9.0 is good,
but I agree that the message needs tweeking.

I'm thinking of something along the lines of:

   Synchronizing repository: xxx items out of date

That avoids the word "gem" so there won't be any confusion there.  It also
make it clear this action is related to the repository.

Good feedback.  Thanks.

-- 
-- 
-- Jim Weirich    jim at weirichhouse.org     http://onestepback.org
-----------------------------------------------------------------
"Beware of bugs in the above code; I have only proved it correct,
not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20060703/10adca69/attachment.html 

From glasser at mit.edu  Mon Jul  3 15:32:23 2006
From: glasser at mit.edu (David Glasser)
Date: Mon, 3 Jul 2006 15:32:23 -0400
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <d4dd9080607031229x29f7bc7dme29e19fd5a6fdf39@mail.gmail.com>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
	<44A94660.1090100@att.net>
	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
	<20060703174433.GI12069@tux-chan> <44A95A6C.9050607@att.net>
	<65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>
	<44A963C4.6050602@att.net>
	<d4dd9080607031229x29f7bc7dme29e19fd5a6fdf39@mail.gmail.com>
Message-ID: <1ea387f60607031232q316eb38bs5351472e404151a2@mail.gmail.com>

On 7/3/06, Jim Weirich <jim.weirich at gmail.com> wrote:
> I'm thinking of something along the lines of:
>
>    Synchronizing repository: xxx items out of date

How about "Synchronizing repository index" or just "Synchronizing index"?

--dave


-- 
David Glasser | glasser at mit.edu | http://www.davidglasser.net/

From jim.weirich at gmail.com  Mon Jul  3 15:34:38 2006
From: jim.weirich at gmail.com (Jim Weirich)
Date: Mon, 3 Jul 2006 15:34:38 -0400
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <1ea387f60607031232q316eb38bs5351472e404151a2@mail.gmail.com>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
	<44A94660.1090100@att.net>
	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
	<20060703174433.GI12069@tux-chan> <44A95A6C.9050607@att.net>
	<65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>
	<44A963C4.6050602@att.net>
	<d4dd9080607031229x29f7bc7dme29e19fd5a6fdf39@mail.gmail.com>
	<1ea387f60607031232q316eb38bs5351472e404151a2@mail.gmail.com>
Message-ID: <d4dd9080607031234l2fdb9988q479d6a3e98b4b12b@mail.gmail.com>

On 7/3/06, David Glasser <glasser at mit.edu> wrote:
>
> On 7/3/06, Jim Weirich <jim.weirich at gmail.com> wrote:
> >    Synchronizing repository: xxx items out of date
>
> How about "Synchronizing repository index" or just "Synchronizing index"?


Good.

-- 
-- 
-- Jim Weirich    jim at weirichhouse.org     http://onestepback.org
-----------------------------------------------------------------
"Beware of bugs in the above code; I have only proved it correct,
not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20060703/1abe9e7a/attachment.html 

From gil at fooplanet.com  Mon Jul  3 15:35:53 2006
From: gil at fooplanet.com (Gil)
Date: Mon, 03 Jul 2006 12:35:53 -0700
Subject: [Rubygems-developers] gem2rpm script
Message-ID: <44A97199.5080302@fooplanet.com>

Yesterday I hacked together a script that takes a gem file and
builds an rpm out of it.  I did it because I manage a pile of
servers and they already have rpm-based auto-update, freshness
checks, and a network-local repository.  I didn't want to
reimplement and monitor all of that for gems and to be honest I
don't feel comfortable with multiple package managers running around
lights-out systems.

I think my implementation is cheating a little bit because the RPM
folks go on and on about pristine sources and all and I'm just
starting from the gem.  However, the script seemed to build gems for
rake, rails etc and will make my life much easier in administering
said machines.

The script lives here:

    http://www.fooplanet.com/projects/gem2rpm/gem2rpm.gz

Some questions/comments for the list:

  - Would this approach yield a tool that's useful to anyone but me?

  - I couldn't find it anywhere, but does 'gem' scribble in a
    manifest or similar when it installs packages?

  - Is there a way I can wholesale disable remote upates/queries
    short of blowing away the sources gem?

  - In order to extract dependency requirement information I did
    this:

    class Gem::Version::Requirement
      attr_reader :requirements
    end

    I'm not happy with doing this but it seemed like the only way to
    go about the getting the version requirements from the gem
    metadata without actually parsing the result of as_list.

    Is there a better way to do this?

  - gem appeared to place the wrapper scripts in different
    directories when running with uid = 0 than under another uid.  I
    don't build rpms as root so this meant that I couldn't get the
    wrappers into $RPM_BUILD_ROOT/usr/bin without moving them
    myself.  Is there a better way to get gem to pretend that an
    unusual directory is the root of the filesystem (like make
    DESTDIR=foo install usually does)?

  - RPM has several fields that GEM doesn't appear to have including
    BuildArch, License, and Group.  Group appears to be an RPM thing
    so I expect it to be missing.  Did I overlook some way to
    get at the other two?

  - I'm unhappy with the contents of the changelog since the rpm
    changelog usually reflects the packaging history rather than the
    actual package history.  Implementing something else seemed like
    it would cost more than it was worth.  Anyone have any better
    ideas?

  - Some Linux distributions have an idea of prefixing or suffixing
    package names based on the language they come from, usually when
    the thing being packaged is a library.  For example, Fedora has
    names like perl-Compress-Zlib for the Compres::Zlib perl
    library.  Is anyone aware of such a convention for ruby packages
    other than just "ruby-"?

  - I couldn't find any easy way of extracting package metadata
    concerning rubygems itself prior to having rake installed.  Even
    then, the data mostly concerned the update package rather than
    rubygems proper.

Thanks for your time and any help,

--Gil

From weyus at att.net  Mon Jul  3 15:37:05 2006
From: weyus at att.net (Wes Gamble)
Date: Mon, 03 Jul 2006 14:37:05 -0500
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <d4dd9080607031229x29f7bc7dme29e19fd5a6fdf39@mail.gmail.com>
References: <20060630192647.GC26372@okcomputer.antiflux.org>	<44A94660.1090100@att.net>	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>	<20060703174433.GI12069@tux-chan>
	<44A95A6C.9050607@att.net>	<65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>	<44A963C4.6050602@att.net>
	<d4dd9080607031229x29f7bc7dme29e19fd5a6fdf39@mail.gmail.com>
Message-ID: <44A971E1.2010903@att.net>



Jim Weirich wrote:

> The synchronization process takes time.  One of of the problems with 
> the 0.8.x RubyGems series was that the synchronization could take a 
> LONG time without any feedback on progress.  The feedback provided by 
> 0.9.0 is good, but I agree that the message needs tweeking.
>
> I'm thinking of something along the lines of:
>
>    Synchronizing repository: xxx items out of date
>
> That avoids the word "gem" so there won't be any confusion there.  It 
> also make it clear this action is related to the repository.
>
> Good feedback.  Thanks.


Jim,

I have to be a pest, but I would suggest that the message be tweaked so 
that it's clear that the xxx items out of date are not necessarily the 
user's currently installed gems.  xxx items out of date means that the 
local "overall table of contents" is out of date with respect to the 
master "overall table of contents" (if I understand correctly).  But 
none of those xxx items that are "out of date" may be gems that the user 
has installed. 

I would also suggest that it be made clear that it's the table of 
contents that is being synchronized, not the gems themselves.  
"Synchronizing repository" is too generic a term in my view.

Sorry to be so picky, but people will interpret these messages without 
knowing anything about how 0.9.0 is somehow "smarter" than the versions 
before it.  In fact, I assumed that the updates were incremental to 
begin with, so there's no way that I would ever have known how to 
interpret that message, which is why I've written all these emails :).

Wes


From darix at web.de  Mon Jul  3 15:49:09 2006
From: darix at web.de (Marcus Rueckert)
Date: Mon, 3 Jul 2006 21:49:09 +0200
Subject: [Rubygems-developers] gem2rpm script
In-Reply-To: <44A97199.5080302@fooplanet.com>
References: <44A97199.5080302@fooplanet.com>
Message-ID: <20060703194909.GR9594@pixel.global-banlist.de>

On 2006-07-03 12:35:53 -0700, Gil wrote:
> Yesterday I hacked together a script that takes a gem file and
> builds an rpm out of it.  I did it because I manage a pile of
> servers and they already have rpm-based auto-update, freshness
> checks, and a network-local repository.  I didn't want to
> reimplement and monitor all of that for gems and to be honest I
> don't feel comfortable with multiple package managers running around
> lights-out systems.
> 
> I think my implementation is cheating a little bit because the RPM
> folks go on and on about pristine sources and all and I'm just
> starting from the gem.  However, the script seemed to build gems for
> rake, rails etc and will make my life much easier in administering
> said machines.

cool :)

i already have tons of rpms here: http://en.opensuse.org/Ruby
i just enabled sles9 support. next will be fedora 4/5.

> The script lives here:
> 
>     http://www.fooplanet.com/projects/gem2rpm/gem2rpm.gz

will look at it:)

>   - gem appeared to place the wrapper scripts in different
>     directories when running with uid = 0 than under another uid.  I
>     don't build rpms as root so this meant that I couldn't get the
>     wrappers into $RPM_BUILD_ROOT/usr/bin without moving them
>     myself.  Is there a better way to get gem to pretend that an
>     unusual directory is the root of the filesystem (like make
>     DESTDIR=foo install usually does)?

see my patch send to the mailinglist a few days ago. it adds a 
"gem install --build-root %{buildroot}"

>   - RPM has several fields that GEM doesn't appear to have including
>     BuildArch, License, and Group.  Group appears to be an RPM thing
>     so I expect it to be missing.  Did I overlook some way to
>     get at the other two?

buildarch doesnt need to be set from the spec file.
license could be passed as a commandline parameter.

>   - I'm unhappy with the contents of the changelog since the rpm
>     changelog usually reflects the packaging history rather than the
>     actual package history.  Implementing something else seemed like
>     it would cost more than it was worth.  Anyone have any better
>     ideas?

you could popup an editor so the packager could add some notes. i
usually combine the gem changelog + packager notes.

>   - Some Linux distributions have an idea of prefixing or suffixing
>     package names based on the language they come from, usually when
>     the thing being packaged is a library.  For example, Fedora has
>     names like perl-Compress-Zlib for the Compres::Zlib perl
>     library.  Is anyone aware of such a convention for ruby packages
>     other than just "ruby-"?

i use "rubygem-" so i can optionally provide a "ruby-" package.
rubygem- - packaged from gem
ruby-    - packaged from tarball

>   - I couldn't find any easy way of extracting package metadata
>     concerning rubygems itself prior to having rake installed.  Even
>     then, the data mostly concerned the update package rather than
>     rubygems proper.

i have a small gemunpack script. gem files are basically tarballs.
not too much magic involved.

you can catch me on irc (darix on freenode or quakenet).

so long ...

    darix

-- 
           openSUSE - SUSE Linux is my linux
               openSUSE is good for you
                   www.opensuse.org

From jim.weirich at gmail.com  Mon Jul  3 16:03:45 2006
From: jim.weirich at gmail.com (Jim Weirich)
Date: Mon, 3 Jul 2006 16:03:45 -0400
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <44A971E1.2010903@att.net>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
	<44A94660.1090100@att.net>
	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
	<20060703174433.GI12069@tux-chan> <44A95A6C.9050607@att.net>
	<65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>
	<44A963C4.6050602@att.net>
	<d4dd9080607031229x29f7bc7dme29e19fd5a6fdf39@mail.gmail.com>
	<44A971E1.2010903@att.net>
Message-ID: <d4dd9080607031303y5ff07b8yb661c675480b841b@mail.gmail.com>

On 7/3/06, Wes Gamble <weyus at att.net> wrote:
>
> > I'm thinking of something along the lines of:
> >
> >    Synchronizing repository: xxx items out of date
> >
> > That avoids the word "gem" so there won't be any confusion there.  It
> > also make it clear this action is related to the repository.
>
> I have to be a pest, but I would suggest that the message be tweaked so
> that it's clear that the xxx items out of date are not necessarily the
> user's currently installed gems.


Don't worry about being a pest.  This is good feedback.  I've been immersed
in this so much, it is difficult for me to tell what is not clear to new
users.

   Synchronizing xxx index entries from http://gems.rubyforge.com

-- 
-- 
-- Jim Weirich    jim at weirichhouse.org     http://onestepback.org
-----------------------------------------------------------------
"Beware of bugs in the above code; I have only proved it correct,
not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20060703/0939c20a/attachment.html 

From hgs at dmu.ac.uk  Tue Jul  4 05:49:24 2006
From: hgs at dmu.ac.uk (Hugh Sasse)
Date: Tue, 4 Jul 2006 10:49:24 +0100 (WEST)
Subject: [Rubygems-developers] Understanding new RubyGem 0.9 output
In-Reply-To: <44A971E1.2010903@att.net>
References: <20060630192647.GC26372@okcomputer.antiflux.org>
	<44A94660.1090100@att.net>
	<d4dd9080607030942g76fa3f80y1755c839db9d1ac8@mail.gmail.com>
	<20060703174433.GI12069@tux-chan> <44A95A6C.9050607@att.net>
	<65e0bb520607031114n1a7a30cav8257169699ea0f48@mail.gmail.com>
	<44A963C4.6050602@att.net>
	<d4dd9080607031229x29f7bc7dme29e19fd5a6fdf39@mail.gmail.com>
	<44A971E1.2010903@att.net>
Message-ID: <Pine.GSO.4.64.0607041039200.24813@brains.eng.cse.dmu.ac.uk>

On Mon, 3 Jul 2006, Wes Gamble wrote:

> master "overall table of contents" (if I understand correctly).  But 
> none of those xxx items that are "out of date" may be gems that the user 
> has installed. 

Actually, they are not really "out of date" -- they are probably 
perfectly useful.  Maybe "%d new items available"  might be less
like a warning.

<wafer-thin-mint>Should rubygems ever support warnings for gems that
really ought to be updated for security reasons?  Probably something
for another thread, if at all.</wafer-thin-mint>
> 
> I would also suggest that it be made clear that it's the table of 
> contents that is being synchronized, not the gems themselves.  
> "Synchronizing repository" is too generic a term in my view.

"Synchronizing index: %d new items available"

Having the number there tells people that the repository isn't stale,
dead, etc.
> 
        Hugh

From lists at kikobu.com  Wed Jul  5 16:34:21 2006
From: lists at kikobu.com (Morten)
Date: Wed, 05 Jul 2006 22:34:21 +0200
Subject: [Rubygems-developers] Ignoring mswin32 gems
In-Reply-To: <e5cbig$c54$1@sea.gmane.org>
References: <e5cbig$c54$1@sea.gmane.org>
Message-ID: <e8h7pb$e9c$1@sea.gmane.org>


Bump. Is this not a general problem?

Morten wrote:
> Hi.
> 
> I have the mysql 2.7 gem installed on my Linux box:
> 
> [root at machine ~]# gem list mysql
> 
> *** LOCAL GEMS ***
> 
> mysql (2.7)
>      MySQL/Ruby provides the same functions for Ruby programs that the
>      MySQL C API provides for C programs.
> 
> If I do a 'gem update' however, gem picks up updates for mswin32 as 
> being more recent that my current:
> 
> [root at machine ~]# gem update mysql
> Upgrading installed gems...
> Attempting remote upgrade of mysql
> Attempting remote installation of 'mysql'
> Select which gem to install for your platform (x86_64-linux)
>   1. mysql 2.7.2006.05.10 (mswin32)
>   2. mysql 2.7.2006.04.21 (mswin32)
>   3. mysql 2.7 (ruby)
>   4. mysql 2.6 (ruby)
>   5. mysql 2.5.1 (ruby)
>   6. Cancel installation
>  >
> 
> This is annoying as it means I have to update mysql each time I do a 
> "gem update" (which in turn means I need to remember to "gem update -- 
> --with-mysql-config").
> 
> Any way to tell gem that I want it to disregard mswin32 gems?
> 
> Thanks.
> 
> Morten


From chad at chadfowler.com  Wed Jul  5 16:43:42 2006
From: chad at chadfowler.com (Chad Fowler)
Date: Wed, 5 Jul 2006 14:43:42 -0600
Subject: [Rubygems-developers] Ignoring mswin32 gems
In-Reply-To: <e8h7pb$e9c$1@sea.gmane.org>
References: <e5cbig$c54$1@sea.gmane.org> <e8h7pb$e9c$1@sea.gmane.org>
Message-ID: <88AF7FBA-FAFA-4E7A-9D2C-B7B07C0BD57D@chadfowler.com>

On Jul 5, 2006, at 2:34 PM, Morten wrote:

>
> Bump. Is this not a general problem?
>
> Morten wrote:
>> Hi.
>>
>> I have the mysql 2.7 gem installed on my Linux box:
>>
>> [root at machine ~]# gem list mysql
>>
>> *** LOCAL GEMS ***
>>
>> mysql (2.7)
>>      MySQL/Ruby provides the same functions for Ruby programs that  
>> the
>>      MySQL C API provides for C programs.
>>
>> If I do a 'gem update' however, gem picks up updates for mswin32 as
>> being more recent that my current:


Sorry...I think the issue is that this problem is actually _more_  
general than the one example here.  Our whole notion of handling  
platforms needs a lot of work.  We did the simplest thing that worked  
at the time, but I think the current state of the gem repository has  
outgrown this approach.  Try installing mongrel for an example. :)

I think we need to bump "rework platform-specific gems" onto the  
pre-1.0 list.

Chad

From lyle.johnson at gmail.com  Wed Jul  5 16:50:16 2006
From: lyle.johnson at gmail.com (Lyle Johnson)
Date: Wed, 5 Jul 2006 15:50:16 -0500
Subject: [Rubygems-developers] RubyGems 0.9.0 introduces bug for installing
	source Gems
Message-ID: <d4cf71b00607051350n52ae90baw2edc404585810bdd@mail.gmail.com>

I filed a bug report about this this morning, but just wanted to
mention it here briefly. RubyGems 0.9.0 introduced a bug for
installing compiled extensions; in addition to doing the "make" and
"make install" steps of the build process, it *now* follows up with a
"make clean", which deletes the object files (*.o) and the extension
shared library (*.so). And, you know, that sort-of defeats the
purpose. ;)

From jim at weirichhouse.org  Thu Jul  6 13:26:05 2006
From: jim at weirichhouse.org (Jim Weirich)
Date: Thu, 06 Jul 2006 13:26:05 -0400
Subject: [Rubygems-developers] Ignoring mswin32 gems
In-Reply-To: <88AF7FBA-FAFA-4E7A-9D2C-B7B07C0BD57D@chadfowler.com>
References: <e5cbig$c54$1@sea.gmane.org> <e8h7pb$e9c$1@sea.gmane.org>
	<88AF7FBA-FAFA-4E7A-9D2C-B7B07C0BD57D@chadfowler.com>
Message-ID: <44AD47AD.4070209@weirichhouse.org>

Chad Fowler wrote:
> Sorry...I think the issue is that this problem is actually _more_  
> general than the one example here.  Our whole notion of handling  
> platforms needs a lot of work.  We did the simplest thing that worked  
> at the time, but I think the current state of the gem repository has  
> outgrown this approach.  Try installing mongrel for an example. :)
> 
> I think we need to bump "rework platform-specific gems" onto the  
> pre-1.0 list.

Actually, this is 100% in line with what I'm thinking.  As part of the 
local/remote integration (which, honestly, I haven't shared a lot of my 
ideas yet), I would like to define a standard way of referencing a gem, 
sort of a URI for gems (I'm calling it GRID, for Gem Resource 
IDentifier).  Part of the grid will be a better way of identifying 
platforms.  For example,  I think we need to differentiate between pure 
ruby gems (that will run anywhere) and gems with C extensions (that 
require a compile environment), in addition to the win32/darwin/whetever 
platforms.  The gem command, in resolving a GRID, will be able to select 
only those candidates that are appropriate for a user.

I have notes somewhere on how GRIDs are supposed to work, and they 
should be backwards compatible (more or less) with todays scheme.  Now, 
where did I put those notes?

- Jim Weirich

From lists at sourceillustrated.com  Sat Jul  8 12:32:33 2006
From: lists at sourceillustrated.com (John Wells)
Date: Sat, 08 Jul 2006 12:32:33 -0400
Subject: [Rubygems-developers] --ri not working on Windows?
Message-ID: <44AFDE21.2000000@sourceillustrated.com>

Guys,

Installed 0.9.0, which I believe includes the ri patch. However, after 
running:

gem rdoc --all --ri

on Windows (don't ask) the ri command still doesn't pick up any gem 
docs. Is there something special I need to do?

Thanks!
John

From dharple at generalconsumption.org  Sat Jul  8 20:35:36 2006
From: dharple at generalconsumption.org (Daniel Harple)
Date: Sat, 8 Jul 2006 20:35:36 -0400
Subject: [Rubygems-developers] --ri not working on Windows?
In-Reply-To: <44AFDE21.2000000@sourceillustrated.com>
References: <44AFDE21.2000000@sourceillustrated.com>
Message-ID: <9A1794DB-EA8A-4C52-9915-6A9127904651@generalconsumption.org>

On Jul 8, 2006, at 12:32 PM, John Wells wrote:

> Installed 0.9.0, which I believe includes the ri patch. However, after
> running:
>
> gem rdoc --all --ri
>
> on Windows (don't ask) the ri command still doesn't pick up any gem
> docs. Is there something special I need to do?

Try using the ?gemri? command.

-- Daniel


From lists at sourceillustrated.com  Sat Jul  8 21:09:08 2006
From: lists at sourceillustrated.com (John Wells)
Date: Sat, 08 Jul 2006 21:09:08 -0400
Subject: [Rubygems-developers] --ri not working on Windows?
In-Reply-To: <9A1794DB-EA8A-4C52-9915-6A9127904651@generalconsumption.org>
References: <44AFDE21.2000000@sourceillustrated.com>
	<9A1794DB-EA8A-4C52-9915-6A9127904651@generalconsumption.org>
Message-ID: <44B05734.6060309@sourceillustrated.com>

Daniel Harple wrote:

>Try using the ?gemri? command.
>
>-- Daniel
>  
>
That did the trick...thanks.

From rubygems-dev at whytheluckystiff.net  Sat Jul  8 22:12:38 2006
From: rubygems-dev at whytheluckystiff.net (why the lucky stiff)
Date: Sat, 8 Jul 2006 21:12:38 -0500
Subject: [Rubygems-developers] balloon needs...
Message-ID: <20060709021238.GC76585@lstsv-3264.layeredtech.com>

So, we didn't get much of a chance to talk about RubyGems at RailsConf, which makes
me oft dismayed.  I've done a bit of hacking which I think would benefit
RubyGems.  Most of it can be seen in this script:

  <http://balloon.hobix.com/lib/balloon.rb>

I know you're all extreeeeemely busy, so I'll keep this pointed.  Are you
intrested in any of these changes?  If so, I'm willing to check these in myself.
Or formulate a patch, if that's what you'd rather.
(Really, I'd also like to start conquering some of Mauricio's list.)

* rubygems/open-uri conflicts with open-uri.  i'd like to fix.
  (balloon.rb: 52-59)

* ExtExtConfBuilder would probably be better off overriding INSTALL,
  INSTALL_PROG, and INSTALL_DATA to use `ruby -run -e install` in situations
  where the user is installing within their home directory.  otherwise,
  extensions fail to install because `install` tries to chmod/chgrp to root.
  (balloon.rb: 61-72)

* would it be okay to flesh out the `loaded_specs` API so everyone can read
  from it?  50% of the stuff I work on has this is in it:

    class << Gem; attr_accessor :loaded_specs end

* After calling RemoteInstaller#install, I clear the source_index in order for
  it pick up the change.  I use Gem.use_paths again to do this.  Is this
  intentional?
  (balloon.rb: 78-80, 140)

* It would be nice if there were a blessed way to unload RubyGems, for libraries
  which happen to run under RUBYOPT="-rubygems".
  (balloon.rb: 3-12)

Beyond that, if I'm wrong about any of this (or being too hackywise,) please
correct me.  Ultimately, I just want to be able to manage Gem repositories aside
from the formally installed one.  (~/.balloon in this case.)

V. good, sweet amigos,

_why

From tilman at code-monkey.de  Sun Jul  9 04:59:36 2006
From: tilman at code-monkey.de (Tilman Sauerbeck)
Date: Sun, 9 Jul 2006 10:59:36 +0200
Subject: [Rubygems-developers] RubyGems 0.9.0 introduces bug for
	installing source Gems
In-Reply-To: <d4cf71b00607051350n52ae90baw2edc404585810bdd@mail.gmail.com>
References: <d4cf71b00607051350n52ae90baw2edc404585810bdd@mail.gmail.com>
Message-ID: <20060709085935.GA4674@code-monkey.de>

Lyle Johnson [2006-07-05 15:50]:
> I filed a bug report about this this morning, but just wanted to
> mention it here briefly. RubyGems 0.9.0 introduced a bug for
> installing compiled extensions; in addition to doing the "make" and
> "make install" steps of the build process, it *now* follows up with a
> "make clean", which deletes the object files (*.o) and the extension
> shared library (*.so). And, you know, that sort-of defeats the
> purpose. ;)

Blame me.

The purpose was to get rid of the object files in the gem directory, but
the clean target also removes the shared objects :(

Regards,
Tilman

-- 
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20060709/56feee0a/attachment.bin 

From lyle.johnson at gmail.com  Sun Jul  9 09:47:42 2006
From: lyle.johnson at gmail.com (Lyle Johnson)
Date: Sun, 9 Jul 2006 08:47:42 -0500
Subject: [Rubygems-developers] RubyGems 0.9.0 introduces bug for
	installing source Gems
In-Reply-To: <20060709085935.GA4674@code-monkey.de>
References: <d4cf71b00607051350n52ae90baw2edc404585810bdd@mail.gmail.com>
	<20060709085935.GA4674@code-monkey.de>
Message-ID: <d4cf71b00607090647n504f4fecw523e734d3c46791@mail.gmail.com>

On 7/9/06, Tilman Sauerbeck <tilman at code-monkey.de> wrote:

> Blame me.

I wasn't looking to assign blame. ;)

I just wanted to get it out there so that the problem can be addressed
in the next release of RubyGems. The stuff related to source gems
doesn't seem to get exercised as much as the stuff for "pure Ruby"
gems, and so problems like this one are easy to overlook.

From chad at chadfowler.com  Sun Jul  9 12:23:35 2006
From: chad at chadfowler.com (Chad Fowler)
Date: Sun, 9 Jul 2006 10:23:35 -0600
Subject: [Rubygems-developers] balloon needs...
In-Reply-To: <20060709021238.GC76585@lstsv-3264.layeredtech.com>
References: <20060709021238.GC76585@lstsv-3264.layeredtech.com>
Message-ID: <057AE294-6E2A-42ED-BB70-56F311832706@chadfowler.com>


On Jul 8, 2006, at 8:12 PM, why the lucky stiff wrote:

> So, we didn't get much of a chance to talk about RubyGems at  
> RailsConf, which makes
> me oft dismayed.  I've done a bit of hacking which I think would  
> benefit
> RubyGems.  Most of it can be seen in this script:
>
>   <http://balloon.hobix.com/lib/balloon.rb>
>
> I know you're all extreeeeemely busy, so I'll keep this pointed.   
> Are you
> intrested in any of these changes?  If so, I'm willing to check  
> these in myself.
> Or formulate a patch, if that's what you'd rather.
> (Really, I'd also like to start conquering some of Mauricio's list.)
>
> * rubygems/open-uri conflicts with open-uri.  i'd like to fix.
>   (balloon.rb: 52-59)
>
> * ExtExtConfBuilder would probably be better off overriding INSTALL,
>   INSTALL_PROG, and INSTALL_DATA to use `ruby -run -e install` in  
> situations
>   where the user is installing within their home directory.   
> otherwise,
>   extensions fail to install because `install` tries to chmod/chgrp  
> to root.
>   (balloon.rb: 61-72)
>
> * would it be okay to flesh out the `loaded_specs` API so everyone  
> can read
>   from it?  50% of the stuff I work on has this is in it:
>
>     class << Gem; attr_accessor :loaded_specs end
>
> * After calling RemoteInstaller#install, I clear the source_index  
> in order for
>   it pick up the change.  I use Gem.use_paths again to do this.  Is  
> this
>   intentional?
>   (balloon.rb: 78-80, 140)
>
> * It would be nice if there were a blessed way to unload RubyGems,  
> for libraries
>   which happen to run under RUBYOPT="-rubygems".
>   (balloon.rb: 3-12)
>
> Beyond that, if I'm wrong about any of this (or being too  
> hackywise,) please
> correct me.  Ultimately, I just want to be able to manage Gem  
> repositories aside
> from the formally installed one.  (~/.balloon in this case.)
>
> V. good, sweet amigos,
>


Personally I'm cool with all of this, though I'll wait for Jim to  
respond since he's been putting all of the time into RubyGems  
lately.  The one area that might be good to get your input on is the  
new, not-yet-started, remote/local installer unification that Jim has  
brewing.  That may affect or be affected by the stuff you're doing.

Over and out for now.

Chad

From tilman at code-monkey.de  Sun Jul  9 13:53:00 2006
From: tilman at code-monkey.de (Tilman Sauerbeck)
Date: Sun, 9 Jul 2006 19:53:00 +0200
Subject: [Rubygems-developers] RubyGems 0.9.0 introduces bug for
	installing source Gems
In-Reply-To: <d4cf71b00607090647n504f4fecw523e734d3c46791@mail.gmail.com>
References: <d4cf71b00607051350n52ae90baw2edc404585810bdd@mail.gmail.com>
	<20060709085935.GA4674@code-monkey.de>
	<d4cf71b00607090647n504f4fecw523e734d3c46791@mail.gmail.com>
Message-ID: <20060709175258.GA29076@code-monkey.de>

Lyle Johnson [2006-07-09 08:47]:
> On 7/9/06, Tilman Sauerbeck <tilman at code-monkey.de> wrote:
> 
> > Blame me.
> 
> I wasn't looking to assign blame. ;)
> 
> I just wanted to get it out there so that the problem can be addressed
> in the next release of RubyGems. The stuff related to source gems
> doesn't seem to get exercised as much as the stuff for "pure Ruby"
> gems, and so problems like this one are easy to overlook.

Damn! I posted my final extension-stuff patch for Rubygems in
20050925081121.GA5808 at code-monkey.de, but it seems that Chad applied a
patch that I posted earlier.
I assumed he committed the right one, but didn't verify :(

Here's a patch that contains the bits that have been missed back then.
I also backed out the "clean" chunk for ExtConfBuilder.
I only tested it briefly, I haven't looked at this code in months and
I'm not sure it's 100% correct.

See the mail I referenced above for comments on these changes.

I'm sorry for the inconvience :(

Regards,
Tilman

-- 
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
-------------- next part --------------
Only in rubygems-0.9.0/lib/rubygems: .installer.rb.swp
diff -aur rubygems-0.9.0.orig/lib/rubygems/installer.rb rubygems-0.9.0/lib/rubygems/installer.rb
--- rubygems-0.9.0.orig/lib/rubygems/installer.rb	2006-06-07 05:39:54.000000000 +0200
+++ rubygems-0.9.0/lib/rubygems/installer.rb	2006-07-09 19:48:23.000000000 +0200
@@ -292,9 +292,12 @@
       say "Building native extensions.  This could take a while..."
       start_dir = Dir.pwd
       dest_path = File.join(directory, spec.require_paths[0])
+      ran_rake = false # only run rake once
 
-      results = []
       spec.extensions.each do |extension|
+        break if ran_rake
+        results = []
+
         case extension
         when /extconf/ then
           builder = ExtExtConfBuilder
@@ -302,6 +305,7 @@
           builder = ExtConfigureBuilder
         when /rakefile/i then
           builder = ExtRakeBuilder
+          ran_rake = true
         else
           builder = nil
           results = ["No builder for extension '#{extension}'"]
@@ -310,7 +314,7 @@
         begin
           err = false
           Dir.chdir File.join(directory, File.dirname(extension))
-          results = builder.build(extension, directory, dest_path)
+          results = builder.build(extension, directory, dest_path, results)
         rescue => ex
           err = true
         end
@@ -551,29 +555,27 @@
   end  # class Uninstaller
 
   class ExtConfigureBuilder
-    def self.build(extension, directory, dest_path)
-      results = []
+    def self.build(extension, directory, dest_path, results)
       unless File.exist?('Makefile') then
         cmd = "sh ./configure --prefix=#{dest_path}"
         results << cmd
         results << `#{cmd}`
       end
 
-      results.push(*ExtExtConfBuilder.make(dest_path))
+      ExtExtConfBuilder.make(dest_path, results)
       results
     end
   end
 
   class ExtExtConfBuilder
-    def self.build(extension, directory, dest_path)
-      results = ["#{Gem.ruby} #{File.basename(extension)} #{ARGV.join(" ")}"]
+    def self.build(extension, directory, dest_path, results)
+      results << "#{Gem.ruby} #{File.basename(extension)} #{ARGV.join(" ")}"
       results << `#{Gem.ruby} #{File.basename(extension)} #{ARGV.join(" ")}`
-      results.push(*make(dest_path))
+      make(dest_path, results)
       results
     end
 
-    def self.make(dest_path)
-      results = []
+    def self.make(dest_path, results)
       raise unless File.exist?('Makefile')
       mf = File.read('Makefile')
       mf = mf.gsub(/^RUBYARCHDIR\s*=\s*\$[^$]*/, "RUBYARCHDIR = #{dest_path}")
@@ -585,27 +587,25 @@
         make_program = (/mswin/ =~ RUBY_PLATFORM) ? 'nmake' : 'make'
       end
 
-      ['', 'install', 'clean'].each do |target|
+      ['', 'install'].each do |target|
         results << "#{make_program} #{target}".strip
         results << `#{make_program} #{target}`
-      end
 
-      results
+        raise unless $?.exitstatus.zero?
+      end
     end
 
   end
 
   class ExtRakeBuilder
-    def ExtRakeBuilder.build(ext, directory, dest_path)
+    def ExtRakeBuilder.build(ext, directory, dest_path, results)
       make_program = ENV['rake'] || 'rake'
       make_program += " RUBYARCHDIR=#{dest_path} RUBYLIBDIR=#{dest_path}"
 
-      results = []
+      results << "#{make_program} extension".strip
+      results << `#{make_program} extension`
 
-      ['', 'install', 'clean'].each do |target|
-        results << "#{make_program} #{target}".strip
-        results << `#{make_program} #{target}`
-      end
+      raise unless $?.exitstatus.zero?
 
       results
     end
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20060709/12559513/attachment.bin 

From drbrain at segment7.net  Tue Jul 11 16:56:41 2006
From: drbrain at segment7.net (Eric Hodel)
Date: Tue, 11 Jul 2006 13:56:41 -0700
Subject: [Rubygems-developers] --ri not working on Windows?
In-Reply-To: <9A1794DB-EA8A-4C52-9915-6A9127904651@generalconsumption.org>
References: <44AFDE21.2000000@sourceillustrated.com>
	<9A1794DB-EA8A-4C52-9915-6A9127904651@generalconsumption.org>
Message-ID: <53929254-C73F-430B-8C0D-ADE43A4617A0@segment7.net>

On Jul 8, 2006, at 5:35 PM, Daniel Harple wrote:

> On Jul 8, 2006, at 12:32 PM, John Wells wrote:
>
>> Installed 0.9.0, which I believe includes the ri patch. However,  
>> after
>> running:
>>
>> gem rdoc --all --ri
>>
>> on Windows (don't ask) the ri command still doesn't pick up any gem
>> docs. Is there something special I need to do?
>
> Try using the ?gemri? command.

You will need to install ruby 1.8.5p1 or newer for ri to work with gems.

-- 
Eric Hodel - drbrain at segment7.net - http://blog.segment7.net
This implementation is HODEL-HASH-9600 compliant

http://trackmap.robotcoop.com



From lists at sourceillustrated.com  Tue Jul 11 19:09:34 2006
From: lists at sourceillustrated.com (John Wells)
Date: Tue, 11 Jul 2006 19:09:34 -0400 (EDT)
Subject: [Rubygems-developers] --ri not working on Windows?
In-Reply-To: <53929254-C73F-430B-8C0D-ADE43A4617A0@segment7.net>
References: <44AFDE21.2000000@sourceillustrated.com><9A1794DB-EA8A-4C52-9915-6A9127904651@generalconsumption.org>
	<53929254-C73F-430B-8C0D-ADE43A4617A0@segment7.net>
Message-ID: <52097.172.16.100.0.1152659374.squirrel@192.168.2.4>

Eric Hodel said:
> You will need to install ruby 1.8.5p1 or newer for ri to work with gems.

Interesting...it's working fine for me (so far) with 1.8.4...


From drbrain at segment7.net  Wed Jul 12 00:12:37 2006
From: drbrain at segment7.net (Eric Hodel)
Date: Tue, 11 Jul 2006 21:12:37 -0700
Subject: [Rubygems-developers] --ri not working on Windows?
In-Reply-To: <52097.172.16.100.0.1152659374.squirrel@192.168.2.4>
References: <44AFDE21.2000000@sourceillustrated.com><9A1794DB-EA8A-4C52-9915-6A9127904651@generalconsumption.org>
	<53929254-C73F-430B-8C0D-ADE43A4617A0@segment7.net>
	<52097.172.16.100.0.1152659374.squirrel@192.168.2.4>
Message-ID: <11374DE1-953D-4B8A-A3BF-8813951356C3@segment7.net>

On Jul 11, 2006, at 4:09 PM, John Wells wrote:

> Eric Hodel said:
>> You will need to install ruby 1.8.5p1 or newer for ri to work with  
>> gems.
>
> Interesting...it's working fine for me (so far) with 1.8.4...

I mean natively (without gemri).

-- 
Eric Hodel - drbrain at segment7.net - http://blog.segment7.net
This implementation is HODEL-HASH-9600 compliant

http://trackmap.robotcoop.com



From ehs at pobox.com  Thu Jul 13 16:14:49 2006
From: ehs at pobox.com (Edward Summers)
Date: Thu, 13 Jul 2006 16:14:49 -0400
Subject: [Rubygems-developers] using a .gem in situ
Message-ID: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>

My apologies if this has been asked before. I'm a newbie to gems and  
was wondering if the actual .gem file can be used by the Ruby  
interpreter without having to 'install' it into a particular  
location. Something like how .jar files can be used?

Also, is there a search interface to these archives anywhere?

//Ed

From grant at antiflux.org  Thu Jul 13 16:30:18 2006
From: grant at antiflux.org (Grant Hollingworth)
Date: Thu, 13 Jul 2006 14:30:18 -0600
Subject: [Rubygems-developers] using a .gem in situ
In-Reply-To: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>
References: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>
Message-ID: <20060713203018.GA9554@okcomputer.antiflux.org>

* Edward Summers <ehs at pobox.com> [2006-07-13 14:15]:
> Also, is there a search interface to these archives anywhere?

google site:http://rubyforge.org/pipermail/rubygems-developers/ -ext:txt

From me at misto.ch  Thu Jul 13 16:35:53 2006
From: me at misto.ch (Mirko Stocker)
Date: Thu, 13 Jul 2006 22:35:53 +0200
Subject: [Rubygems-developers] using a .gem in situ
In-Reply-To: <20060713203018.GA9554@okcomputer.antiflux.org>
References: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>
	<20060713203018.GA9554@okcomputer.antiflux.org>
Message-ID: <200607132235.53287.me@misto.ch>

On Thursday 13 July 2006 22:30, Grant Hollingworth wrote:
> > Also, is there a search interface to these archives anywhere?
>
> google site:http://rubyforge.org/pipermail/rubygems-developers/ -ext:txt

Or you could use gmane: http://dir.gmane.org/gmane.comp.lang.ruby.gems.devel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20060713/251867af/attachment.bin 

From drbrain at segment7.net  Thu Jul 13 16:46:08 2006
From: drbrain at segment7.net (Eric Hodel)
Date: Thu, 13 Jul 2006 13:46:08 -0700
Subject: [Rubygems-developers] using a .gem in situ
In-Reply-To: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>
References: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>
Message-ID: <394D3979-EAE3-4D28-9081-D96B3C0814F3@segment7.net>

On Jul 13, 2006, at 1:14 PM, Edward Summers wrote:

> My apologies if this has been asked before. I'm a newbie to gems and
> was wondering if the actual .gem file can be used by the Ruby
> interpreter without having to 'install' it into a particular
> location. Something like how .jar files can be used?

No.

-- 
Eric Hodel - drbrain at segment7.net - http://blog.segment7.net
This implementation is HODEL-HASH-9600 compliant

http://trackmap.robotcoop.com



From jim.weirich at gmail.com  Thu Jul 13 17:46:24 2006
From: jim.weirich at gmail.com (Jim Weirich)
Date: Thu, 13 Jul 2006 17:46:24 -0400
Subject: [Rubygems-developers] using a .gem in situ
In-Reply-To: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>
References: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>
Message-ID: <d4dd9080607131446w751d7ccdr57c50bb3dcaf8bad@mail.gmail.com>

On 7/13/06, Edward Summers <ehs at pobox.com> wrote:
>
> My apologies if this has been asked before. I'm a newbie to gems and
> was wondering if the actual .gem file can be used by the Ruby
> interpreter without having to 'install' it into a particular
> location. Something like how .jar files can be used?



As Eric said, you cannot use gems without installing them first.  Very early
on, the gems team discussed hacking the require command to load ruby source
from a zipped up gem, but that approach would work with .so or .dll files,
so we didn't even try to got that route.

-- 
-- 
-- Jim Weirich    jim at weirichhouse.org     http://onestepback.org
-----------------------------------------------------------------
"Beware of bugs in the above code; I have only proved it correct,
not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20060713/68f116a0/attachment.html 

From ehs at pobox.com  Fri Jul 14 10:18:38 2006
From: ehs at pobox.com (Edward Summers)
Date: Fri, 14 Jul 2006 10:18:38 -0400
Subject: [Rubygems-developers] using a .gem in situ
In-Reply-To: <d4dd9080607131446w751d7ccdr57c50bb3dcaf8bad@mail.gmail.com>
References: <38A09A1B-51F2-42EC-B21E-ECD646032EF2@pobox.com>
	<d4dd9080607131446w751d7ccdr57c50bb3dcaf8bad@mail.gmail.com>
Message-ID: <1DF86E2B-7678-4D65-BAA8-FE981C253A7C@pobox.com>

On Jul 13, 2006, at 5:46 PM, Jim Weirich wrote:
> As Eric said, you cannot use gems without installing them first.   
> Very early on, the gems team discussed hacking the require command  
> to load ruby source from a zipped up gem, but that approach would  
> work with .so or .dll files, so we didn't even try to got that route.

Thanks much, I hadn't thought of the compiled code libraries.

//Ed

From judson-rubygems at redfivellc.com  Thu Jul 20 16:14:42 2006
From: judson-rubygems at redfivellc.com (Judson Lester)
Date: Thu, 20 Jul 2006 13:14:42 -0700
Subject: [Rubygems-developers] Fwd: Kernel::require
In-Reply-To: <8905c87a0607201158y5e9a1606i7cb54ed7e40a798a@mail.gmail.com>
References: <8905c87a0607191350k45239029u9d59e221e3fdff3d@mail.gmail.com>
	<8905c87a0607201158y5e9a1606i7cb54ed7e40a798a@mail.gmail.com>
Message-ID: <8905c87a0607201314h1436b378rac6d578bb03dd749@mail.gmail.com>

Pardon me if this is uninformed, but I haven't been able to find the
answer anywhere.

Is there a reason that custom_require doesn't call module_function
:require on Kernel?  As a result, once rubygems is loaded


require 'a-library'

is different from

Kernel.require 'a-library'

Notably because the Kernel.require form ignores gems.

A specific consequence is that in Needle, Needle::Registry#require

'service-library', ServiceModule fails if the service-library is a gem.
Which is a shame, since DI service libraries and rubygem package
management seem like a peanut butter and jelly combination.

Judson
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20060720/ffd2beae/attachment.html 

From weyus at att.net  Fri Jul 21 12:10:05 2006
From: weyus at att.net (Wes Gamble)
Date: Fri, 21 Jul 2006 11:10:05 -0500
Subject: [Rubygems-developers] Fwd: Kernel::require
In-Reply-To: <8905c87a0607201314h1436b378rac6d578bb03dd749@mail.gmail.com>
References: <8905c87a0607191350k45239029u9d59e221e3fdff3d@mail.gmail.com>	<8905c87a0607201158y5e9a1606i7cb54ed7e40a798a@mail.gmail.com>
	<8905c87a0607201314h1436b378rac6d578bb03dd749@mail.gmail.com>
Message-ID: <44C0FC5D.4040406@att.net>

Is there any way to get visibility into the state of the rubygems 
repository via some Web based interface?

So that when my "gem" command tells me "updated 3 gems" I could go 
somewhere and try and see which 3 gems were updated?

Or is "gem" the only interface to the RubyGems repository?

Thanks,
Wes

From rubygems at freeze.org  Fri Jul 21 12:25:09 2006
From: rubygems at freeze.org (Jim Freeze)
Date: Fri, 21 Jul 2006 11:25:09 -0500
Subject: [Rubygems-developers] Install path
Message-ID: <8B0F691C-6AA8-423B-B287-08127042F5AE@freeze.org>

Hello

At work we have created a common area to install gems. These
gems are used by Sun, HP and Linux machines.

I recently installed rails into this common area with something like

    gem instal rails -i /path/to/common/gem/area --include-dependencies

When some of my apps didn't run, I found that in

    /path/to/common/gem/area/bin/rails

the shebang had:

  #!/big/long/path/to/ruby/linux/install

What would be better for our purposes is it to read

   #!/usr/bin/env ruby

This will let any architecture machine use the gem (at least those  
with executables).

Is there a way to customize the shebang without having to edit the  
bin files after install?

Thanks

Jim Freeze



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20060721/7034179e/attachment.html 

From anne at wjh.harvard.edu  Sun Jul 30 08:43:26 2006
From: anne at wjh.harvard.edu (Anne G)
Date: Sun, 30 Jul 2006 08:43:26 -0400 (EDT)
Subject: [Rubygems-developers] installing ruby gems on Mac OS X with ruby
	1.8.4
Message-ID: <Pine.SOL.4.30.0607300822240.25651-100000@wjh2.wjh.harvard.edu>

Mac OS X upgraded to 10.4.7 on Thursday
(so it uses gcc4...), ruby 1.8.4

I also changed profile to try to install gnome
/etc/profile
PKG_CONFIG_PATH=/usr/local/lib/pkgconfig
DYLD_LIBRARY_PATH=/usr/local/lib:/opt/local/lib
export PKG_CONFIG_PATH DYLD_LIBRARY_PATH

don't know if that is contributing.

-------------------------------->

I tried to install the gettext gem,

got the message that it could not find gem. gem install
used to work fine, what happened?.

Went about uninstalling the darwin port gems then
reinstalled darwinport gems.

sudo gem install gettext
/opt/local/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:21:in
`require__': no such file to load -- zlib (LoadError)

I installed zlib 1.2.3 from source
then I saw the darwinport install, but says zlib is
installed.

dowloaded ruby 1.8.4 source
followed instructions at
http://rubyforge.org/tracker/?func=detail&group_id=126&aid=4520&atid=576
anne-g:~/Desktop/ruby-1.8.4/ext/zlib anne$ ruby extconf.rb
checking for deflateReset() in -lz... no
checking for deflateReset() in -llibz... no
checking for deflateReset() in -lzlib... no
*** extconf.rb failed ***

I have seen pages about this, but they talk about debian to
load zlibc, zlib1g, zlib1g-dev but I don't think debian
works with OS X.

easy find did not find deflateReset on my mac.

what next?



From anne at wjh.harvard.edu  Sun Jul 30 11:05:55 2006
From: anne at wjh.harvard.edu (Anne G)
Date: Sun, 30 Jul 2006 11:05:55 -0400 (EDT)
Subject: [Rubygems-developers] installing ruby gems on Mac OS X with
	ruby 1.8.4
In-Reply-To: <Pine.SOL.4.30.0607300822240.25651-100000@wjh2.wjh.harvard.edu>
Message-ID: <Pine.SOL.4.30.0607301104460.11582-100000@wjh1.wjh.harvard.edu>

Figured out a fix. zlib should come with ruby 1.8.4.
So I uninstalled all my ports and reinstalled, and now gems
are fine.

somehow ruby had be corrupted.
now irb, require 'zlib' returns true.