[Nitro] Fwd: [NP.ORG] Nitro/Og Fora daily diggest

Mon Mar 19 09:48:27 EDT 2007

Dear devs,

I am trying to create a simple digest mail to represent the latest
activity of the forum to this list. At the end of this email you can
find the current format, I would like to hearmore suggestions.

-g.

---------- Forwarded message ----------
From: noreply at nitroproject.org <noreply at nitroproject.org>
Date: Mar 19, 2007 3:46 PM
Subject: [NP.ORG] Nitro/Og Fora daily diggest
To: george.moschovitis at gmail.com

>From gmosx.myopenid.com
Date: Mon Mar 19 13:13:37 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/33
Message-ID: http://www.nitroproject.org/fora/posts/view/39

Hello,

the article you quote is not that negative about UUIDS.

<blockquote>However, seeing what George said about single class
Comment for articles and whatever, appears to be
interesting.</blockquote>

at last, someone finds this interesting idea erhm... interesting ;-)

-g.

>From gmosx.myopenid.com
Date: Mon Mar 19 12:52:40 UTC 2007
Subject: Forum improvements.
Message-ID: http://www.nitroproject.org/fora/posts/view/38

Here are the latest forum improvements:

- no captcha for verified users (at the moment just me and judson).
Judson can you please verify this?
- uploads from windows should work (antoniolo can you verify this?)
- the topic hierarchy is kept, a thread view will be available in the
next update.
- initial work on forum digest.

keep the suggestions and bug reports coming,

-g.

>From gmosx.myopenid.com
Date: Mon Mar 19 09:53:02 UTC 2007
Subject: Ruby meta programming problem
Message-ID: http://www.nitroproject.org/fora/posts/view/37

I am using the following code to generate some utility methods:

<code><
class User::Proxy
  attr_accessor :oid
  attr_accessor :name
  attr_accessor :level
  for key, val in User::LEVEL_VALUES
    define_method("#{val}?") do
      @level >= key
    end
  end
</code>

My problem is that this doesn't work correctly. It seems that @level
referes to a value at the time of the method generation and not the
actual object value at run time. Any idea how to fix this?

-g.

>From gmosx.myopenid.com
Date: Mon Mar 19 08:35:33 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/3
Message-ID: http://www.nitroproject.org/fora/posts/view/36

thanks again.

>From gmosx.myopenid.com
Date: Mon Mar 19 08:34:07 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/18
Message-ID: http://www.nitroproject.org/fora/posts/view/35

I am not a big fan of test-first development. I always finalize my
design *after* I write some code. So writing tests beforehand, seems
like a waste of time. When the design is finalized it is the right
time to do tests.

-g.

>From judson.myopenid.com
Date: Mon Mar 19 08:09:45 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/3
Message-ID: http://www.nitroproject.org/fora/posts/view/34

This is some more fixes for the specs.  Specifically, and importantly,
the class definitions for the fixture classes have been moved into the
setup clauses for the contexts - the upshot of which is that they're
isolated from one another.

>From antoni0.myopenid.com
Date: Mon Mar 19 00:31:21 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/4
Message-ID: http://www.nitroproject.org/fora/posts/view/33

i read all the above articles and also this one:
http://joseph.randomnetworks.com/archives/2005/07/22/uuid-vs-sequences/

Based on above article, my opinion goes together with Judson. As far
as i could see, that type of id schema will be overkill for og at the
moment.

However, seeing what George said about single class Comment for
articles and whatever, appears to be interesting.

looking forward to see where this topic goes to!

>From judson.myopenid.com
Date: Mon Mar 19 00:07:43 UTC 2007
Subject: RE: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/18
Message-ID: http://www.nitroproject.org/fora/posts/view/32

<blockquote>I understand that, but I thought it would be a nice way of
someone else to verify my changes.</blockquote>

My shorter response (long one eaten by captchas) is that the specs are
a better indication of what you mean than the code is.  I'd rather fix
code to make specs pass than vice versa.

>From gmosx.myopenid.com
Date: Sun Mar 18 16:54:17 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/18
Message-ID: http://www.nitroproject.org/fora/posts/view/31

Thanks antonio,

I will check your patch and integrate it if everything is ok ;-)

-g.

>From antoni0.myopenid.com
Date: Sun Mar 18 14:58:32 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/18
Message-ID: http://www.nitroproject.org/fora/posts/view/30

Hello folks! now np.org is amazing! :D

There are 4 patches in this file.

Today morning i pulled the latest version from repo and I've added a
new patch also. This one is a 'revision' for modelmixin renaming.

Take a special care on many_to_many_rich_associations. This one is
neither a new patch or code-revision, it's a bug fix. George, please
run the tests to see what i'm talking about.

>From gmosx.myopenid.com
Date: Sun Mar 18 10:41:25 UTC 2007
Subject: Minor np.org update
Message-ID: http://www.nitroproject.org/fora/posts/view/29

I just added some  minor fixes to this site:

- improved topic list view
- you can use a subset of html
- you can use the blockquote to quote earlier posts
- you can use the code tag to markup source code (under construction,
I will add syntax highlighting support and some fixes)
- I fixed the stupid 'page-jump' css bug. Many thanks to the CSS
master George Paparakis (http://www.qupy.com/) for his help.

keep the suggestions and bug reports coming...

regards,
George.

PS: Also added links to oxywtf.de and the repository to the homepage.

>From gmosx.myopenid.com
Date: Sun Mar 18 08:15:18 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/4
Message-ID: http://www.nitroproject.org/fora/posts/view/28

Btw,

Tom emailed me this interesting link:

http://codesnipers.com/?q=node/143&amp;title=Using-UUID/GUID-as-Primary-Key-in-Rails

I kind of like the aproach.

-g.

>From gmosx.myopenid.com
Date: Sun Mar 18 08:10:27 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/4
Message-ID: http://www.nitroproject.org/fora/posts/view/27

 I could see it as an option for the Manager.
 Making it mandatory would make upgrades problematic as well.

I will implement this as an option. I am still investigating the best
way to tackle this problem.

-g.

>From gmosx.myopenid.com
Date: Sun Mar 18 08:09:10 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/18
Message-ID: http://www.nitroproject.org/fora/posts/view/26

 but it's probably better to keep specs in line with the code, you know?

I understand that, but I thought it would be a nice way of someone
else to verify my changes.

-g.

>From judson.myopenid.com
Date: Sun Mar 18 03:44:58 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/4
Message-ID: http://www.nitroproject.org/fora/posts/view/25

I could see it as an option for the Manager.  Making it mandatory
would make upgrades problematic as well.

Regarding the old thread: the best solution still seems to me to be
PostgreSQL style sequences.  In RDMSes that don't provide sequences,
it seems like they could be emulated with autoincrementing tables.

>From judson.myopenid.com
Date: Sun Mar 18 03:35:13 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/18
Message-ID: http://www.nitroproject.org/fora/posts/view/24

No really: here they are.

Another quick note: I've started moving some of the contexts from
store into other spec files - since they really relate to relations
and models, I put them in those spec files.

>From judson.myopenid.com
Date: Sun Mar 18 03:31:13 UTC 2007
Subject: RE:
In-Reply-To: http://www.nitroproject.org/fora/posts/view/18
Message-ID: http://www.nitroproject.org/fora/posts/view/23

Here's fixes against the new API stuff.

There are 8 specs still failing.   All of them were failing before, I think.

I've started adopting RSpec's new UI on "should" clauses - instead of
should_thing they look like "should thing."  So specs might require
rspec 0.8

--
This mail is automatically generated from the http://nitroproject.org/fora
digest robot. It presents the discussions in the fora during the last 24 hours.
Do not reply to this email.

-- 
http://blog.gmosx.com
http://cull.gr
http://www.joy.gr
http://nitroproject.org