Re: New fields for META.yml

2007-11-06 Thread Michael G Schwern
Andy Lester wrote:
 I'm moving my bug queue for WWW::Mechanize and Test::WWW::Mechanize over
 to Google Code, which is also where I host the source code.
 
 I think it'd be a good idea if META.yml had fields that say where the
 bug queue is, and where the source code lives.  That way
 search.cpan.org, and any others, could refer to the proper URLs, because
 very soon, the link at http://search.cpan.org/dist/WWW-Mechanize/ that
 points to the queue at
 http://rt.cpan.org/NoAuth/Bugs.html?Dist=WWW-Mechanize is going to be
 wrong.  Also, a project home page would be swell as well.
 
 Schwern: Is this your ball?

No, but I went back in my time machine and made sure it's there because I'm
that kind of guy.
http://module-build.sourceforge.net/META-spec.html#resources


-- 
Stabbing you in the face for your own good.


Re: New fields for META.yml

2007-10-30 Thread Eugene van der Pijll
A. Pagaltzis schreef:
 * Andy Lester [EMAIL PROTECTED] [2007-10-29 17:45]:
  I think it'd be a good idea if META.yml had fields that say
  where the bug queue is, and where the source code lives.  That
  way search.cpan.org, and any others, could refer to the proper
  URLs
 
 ??? The META.yml files in older versions of your distro can???t
   change even though the info in them would no longer be correct.

Surely that's no problem? Before reporting a bug, people will (or
should?) first install (or at least check) the latest version of the
ditribution, to check if it's already been fixed.

Eugene


New fields for META.yml

2007-10-29 Thread Andy Lester
I'm moving my bug queue for WWW::Mechanize and Test::WWW::Mechanize  
over to Google Code, which is also where I host the source code.


I think it'd be a good idea if META.yml had fields that say where the  
bug queue is, and where the source code lives.  That way  
search.cpan.org, and any others, could refer to the proper URLs,  
because very soon, the link at http://search.cpan.org/dist/WWW- 
Mechanize/ that points to the queue at http://rt.cpan.org/NoAuth/ 
Bugs.html?Dist=WWW-Mechanize is going to be wrong.  Also, a project  
home page would be swell as well.


Schwern: Is this your ball?

xoxo,
Andy

--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance






Re: New fields for META.yml

2007-10-29 Thread David Cantrell
On Mon, Oct 29, 2007 at 11:43:30AM -0500, Andy Lester wrote:

 I think it'd be a good idea if META.yml had fields that say where the  
 bug queue is ...

I'm far more likely to submit a bug report once I've got your stuff
installed and am using it (or trying to use it) than I am while trying
to install it.  By that time, CPAN.pm will have cleared up the build
directory and I won't have a copy of your META.yml handy.  Also,
automated tools are, umm, not very likely to submit bug reports.

Consequently, having it in a machine-readable META.yml file doesn't seem
very useful.  Put it in the docs instead.

-- 
David Cantrell | Godless Liberal Elitist

  For every vengeance, there is an equal and opposite revengeance.
-- Cartoon Law X


Re: New fields for META.yml

2007-10-29 Thread Eric Wilhelm
# from David Cantrell
# on Monday 29 October 2007 10:00:

By that time, CPAN.pm will have cleared up the build
directory and I won't have a copy of your META.yml handy.

Maybe CPAN.pm should keep them handy?  Of course, they can be 
systematically fetched from your friendly CPAN mirror.

Consequently, having it in a machine-readable META.yml file doesn't
 seem very useful.  Put it in the docs instead.

Maybe in the docs as well, but machine-readable would be quite 
convenient for e.g. a helper tool to launch your browser (or even 
(given some sort of discoverable web service API) fill-out the details 
of the report (starting with the module version, perl version, osname, 
and other Config.pm yummies which are so often missing from reports), 
plus possibly wizarding the user through the all-important search for 
bug step.)

--Eric
-- 
Those who cannot remember the past are condemned to repeat it.
--George Santayana
---
http://scratchcomputing.com
---


Re: New fields for META.yml

2007-10-29 Thread Eric Wilhelm
# from Andy Lester
# on Monday 29 October 2007 09:43:

I think it'd be a good idea if META.yml had fields that say where the
   bug queue is, and where the source code lives.

It already does.

   Also, a project home page would be swell as well.

In the 'resources' hash, the 'bugtracker', 'repository', and 'homepage' 
keys are defined in the spec.

  http://module-build.sourceforge.net/META-spec-current.html#resources

In your Build.PL, just use the 'meta_merge' key.

  meta_merge = { resources = { repository = ... } },

See e.g.
  http://search.cpan.org/src/KWILLIAMS/Module-Build-0.2808/Build.PL

--Eric
-- 
Introducing change is like pulling off a bandage: the pain is a memory
almost as soon as you feel it.
--Paul Graham
---
http://scratchcomputing.com
---


Re: New fields for META.yml

2007-10-29 Thread Andy Lester

automated tools are, umm, not very likely to submit bug reports.

Consequently, having it in a machine-readable META.yml file doesn't  
seem

very useful.  Put it in the docs instead.


Please do me the courtesy of reading the entire message.

Repeating the part that it seems you missed: That way  
search.cpan.org, and any others, could refer to the proper URLs,  
because very soon, the link at http://search.cpan.org/dist/WWW- 
Mechanize/ that points to the queue at http://rt.cpan.org/NoAuth/ 
Bugs.html?Dist=WWW-Mechanize is going to be wrong.  Also, a project  
home page would be swell as well.  It has nothing to do with humans  
reading the META.yml in the build directory, and everything to do  
with how search.cpan.org sees it.


Thank you,
Andy

--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance






Re: New fields for META.yml

2007-10-29 Thread Andy Lester


On Oct 29, 2007, at 12:10 PM, Eric Wilhelm wrote:


It already does.


  Also, a project home page would be swell as well.


In the 'resources' hash, the 'bugtracker', 'repository', and  
'homepage'

keys are defined in the spec.


Ah, thank you.  I'd assumed they didn't exist.  Mea culpa.

Now to see if EU::MM supports them

--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance






Re: New fields for META.yml

2007-10-29 Thread Andy Lester


On Oct 29, 2007, at 5:11 PM, A. Pagaltzis wrote:


• The META.yml files in older versions of your distro can’t
 change even though the info in them would no longer be correct.

• Doing this via META.yml means the only way to update this info
 is to push out a new release.

Clearly this info should live somewhere and search.cpan should
use it, but META.yml is the wrong place. It belongs somewhere
unversioned.



Best is the enemy of the good.  Moving forward with META.yml for  
WWW::Mechanize and Test::WWW::Mechanize.


xoa

--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance






Re: New fields for META.yml

2007-10-29 Thread A. Pagaltzis
* Andy Lester [EMAIL PROTECTED] [2007-10-29 17:45]:
 I think it'd be a good idea if META.yml had fields that say
 where the bug queue is, and where the source code lives.  That
 way search.cpan.org, and any others, could refer to the proper
 URLs

• The META.yml files in older versions of your distro can’t
  change even though the info in them would no longer be correct.

• Doing this via META.yml means the only way to update this info
  is to push out a new release.

Clearly this info should live somewhere and search.cpan should
use it, but META.yml is the wrong place. It belongs somewhere
unversioned.

Unfortunately we don’t already have such a thing, and setting it
up just for the bugtracker etc. info involves much bigger changes
by more people than the incremental changes needed for doing that
via META.yml, so there’s virtually no point in me pointing out
that we could do better, but anyway.

Regards,
-- 
Aristotle Pagaltzis // http://plasmasturm.org/


Re: New fields for META.yml

2007-10-29 Thread A. Pagaltzis
* Andy Lester [EMAIL PROTECTED] [2007-10-29 23:15]:
 Best is the enemy of the good.

And next time this will be handy, it will also be too big a
change vs. an incremental suboptimal solution, and the overnext
time as well, and ever on.

YAGNI does not apply either: this *is already* the overnext time.
The gravatar spat would’ve been solved by something like that and
I remember another discussion where we said that this would
really be the right solution, though I no longer remember what it
was about. And I haven’t been on the relevant mailing lists for
that long, and I’m still not on all of them.

 Moving forward with META.yml for WWW::Mechanize and
 Test::WWW::Mechanize.

Yeah,

* A. Pagaltzis [EMAIL PROTECTED] [2007-10-29 23:15]:
 there’s virtually no point in me pointing out that we could do
 better, but anyway.

of course. It’s just not gonna happen. Oh well.

Regards,
-- 
Aristotle Pagaltzis // http://plasmasturm.org/


Re: New fields for META.yml

2007-10-29 Thread Andy Lester
On Mon, Oct 29, 2007 at 11:25:34PM +0100, A. Pagaltzis ([EMAIL PROTECTED]) 
wrote:
 * Andy Lester [EMAIL PROTECTED] [2007-10-29 23:15]:
  Best is the enemy of the good.
 
 And next time this will be handy, it will also be too big a
 change vs. an incremental suboptimal solution, and the overnext
 time as well, and ever on.

What do you propose we do?

 The gravatar spat would’ve been solved by something like that and

The Gravatar spat was entirely a non-problem.

-- 
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance


Re: New fields for META.yml

2007-10-29 Thread A. Pagaltzis
* Andy Lester [EMAIL PROTECTED] [2007-10-29 23:30]:
 On Mon, Oct 29, 2007 at 11:25:34PM +0100, A. Pagaltzis ([EMAIL PROTECTED]) 
 wrote:
  * Andy Lester [EMAIL PROTECTED] [2007-10-29 23:15]:
   Best is the enemy of the good.
  
  And next time this will be handy, it will also be too big a
  change vs. an incremental suboptimal solution, and the
  overnext time as well, and ever on.
 
 What do you propose we do?

I thought about it for a bit and found there are some faintly
tricky questions. Nothing too serious, but I’m not entirely sure
of all the answers, so I don’t have a proposal on hand.

Mostly the questions are things like whether the data should live
on the CPAN FTP itself or should search.cpan gain an account
system? Maybe have PAUSE would provide a bunch of pages with
settings and generate a file from them that would go on the FTP?
The concern is “distance of metadata” I guess – it shouldn’t be
too onerous for automatic tools working against the FTP, such as
CPAN.pm, to get at this data, even though it lives outside the
distribution. The 01mailrc.txt and 02packages.details.txt seem
like obvious candidates but I don’t know how extensible their
formats are?

I don’t see any insurmountable problems, it’s just a matter of
hashing out the choices and picking one. Is there honest
interest? I am up for it if the people who’d need to play along
are also. (Who maintains PAUSE f.ex.?)

Regards,
-- 
Aristotle Pagaltzis // http://plasmasturm.org/


Re: New fields for META.yml

2007-10-29 Thread Andy Armstrong

On 29 Oct 2007, at 22:25, A. Pagaltzis wrote:

YAGNI does not apply either: this *is already* the overnext time.
The gravatar spat would’ve been solved by something like that



I missed that... Got a link?

--
Andy Armstrong, Hexten






Re: New fields for META.yml

2007-10-29 Thread Andy Lester

I thought about it for a bit and found there are some faintly
tricky questions. Nothing too serious, but I’m not entirely sure
of all the answers, so I don’t have a proposal on hand.


Let me know when you get one then.  In the mean time, I'm moving  
forward.




Mostly the questions are things like whether the data should live
on the CPAN FTP itself or should search.cpan gain an account
system?


search.cpan.org is not the CPAN.  It is a view of the data.

--
Andy Lester = [EMAIL PROTECTED] = www.petdance.com = AIM:petdance






Re: New fields for META.yml

2007-10-29 Thread Andy Armstrong

On 29 Oct 2007, at 23:33, Andy Armstrong wrote:

I missed that... Got a link?



Have now - thanks.

--
Andy Armstrong, Hexten