darcs

Patch 547 Add inferred XML Schema for outputs. (and 1 more)

Title Add inferred XML Schema for outputs. (and 1 more)
Superseder Nosy List dixiecko, guillaume.hoffmann
Related Issues
Status accepted Assigned To guillaume.hoffmann
Milestone

Created on 2011-02-10.21:01:04 by dixiecko, last changed 2011-04-01.14:44:49 by gh.

Files
File name Status Uploaded Type Edit Remove
add-inferred-xml-schema-for-outputs_.dpatch dixiecko, 2011-02-10.21:01:04 application/x-darcs-patch
add-inferred-xml-schema-for-outputs_.dpatch dixiecko, 2011-02-19.13:31:28 application/x-darcs-patch
unnamed dixiecko, 2011-02-10.21:01:04 text/x-darcs-patch
unnamed dixiecko, 2011-02-10.21:01:04
unnamed dixiecko, 2011-02-19.13:31:28 text/x-darcs-patch
unnamed dixiecko, 2011-02-19.13:31:28
See mailing list archives for discussion on individual patches.
Messages
msg13668 (view) Author: dixiecko Date: 2011-02-10.21:01:04
2 patches for repository http://darcs.net/screened/:

Thu Feb 10 21:08:31 CET 2011  dixiecko@gmail.com
  * Add inferred XML Schema for outputs.

Thu Feb 10 21:46:41 CET 2011  dixiecko@gmail.com
  * Test darcs changes XML with XSD/xmllint.
Attachments
msg13700 (view) Author: kowey Date: 2011-02-17.15:23:09
On Thu, Feb 10, 2011 at 21:01:04 +0000, Radoslav Dorcik wrote:
> Thu Feb 10 21:08:31 CET 2011  dixiecko@gmail.com
>   * Add inferred XML Schema for outputs.
> 
> Thu Feb 10 21:46:41 CET 2011  dixiecko@gmail.com
>   * Test darcs changes XML with XSD/xmllint.

I gave these a cursory look as I think of the schema as
contrib.

Seems nice to have one. Is anybody worried about the prospect
of committing to some sort of formal documentation of our XML
format?  Do we need to worry about evolving things? (Presumably
such an evolution would be involved anyway since we'd have to
think about backwards compatibility).  I also take it Rado is
volunteering to maintain the schema file if we ever decide we
need to change the format ;-)

The test sounds like a good accompaniment.  Rado, if you'd be
kind enough to amend with the licence boiler plate that'd save
somebody asking you for permission to add it in later.

I'd be a bit concerned about blurring the boundaries between
contrib and the rest of darcs a bit (on the one hand, we're
sort of treating this as a contrib, on the other hand we have
a test for it... which is it?), but I don't think it's a big
deal.  Comments welcome.

-- 
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
For a faster response, try +44 (0)1273 64 2905 or
xmpp:kowey@jabber.fr (Jabber or Google Talk only)
msg13703 (view) Author: dixiecko Date: 2011-02-19.13:31:28
3 patches for repository http://darcs.net/screened/:

Thu Feb 10 21:08:31 CET 2011  dixiecko@gmail.com
  * Add inferred XML Schema for outputs.

Thu Feb 10 21:46:41 CET 2011  dixiecko@gmail.com
  * Test darcs changes XML with XSD/xmllint.

Sat Feb 19 14:17:04 CET 2011  dixiecko@gmail.com
  * Update test for XML Schema with MIT header.
  Review comment from Eric.
Attachments
msg13704 (view) Author: dixiecko Date: 2011-02-19.13:55:35
Hello,

On Thu, 2011-02-17 at 15:12 +0000, Eric Kow wrote:
> Seems nice to have one. Is anybody worried about the prospect
> of committing to some sort of formal documentation of our XML
> format?  Do we need to worry about evolving things? (Presumably
> such an evolution would be involved anyway since we'd have to
> think about backwards compatibility).  

[rado] Backward compatibility is really good topic to think about
generally, at least in case of the software with users :)
The XML Schema itself does not help with backward compatibility nor it
makes it worse. But how many mission critical projects use Darcs XML
outputs? For the Haskell projects the Darcs library is much better
choice.

> I also take it Rado is
> volunteering to maintain the schema file if we ever decide we
> need to change the format ;-)

[rado] no problem. 

> The test sounds like a good accompaniment.  Rado, if you'd be
> kind enough to amend with the licence boiler plate that'd save
> somebody asking you for permission to add it in later

[rado] added.

> I'd be a bit concerned about blurring the boundaries between
> contrib and the rest of darcs a bit (on the one hand, we're
> sort of treating this as a contrib, on the other hand we have
> a test for it... which is it?), but I don't think it's a big
> deal.  Comments welcome.
> 

[rado] you are right. Maybe the better section would be directory with
manual. On the other hand I think the XML Schema and XML-like outputs
from Darcs deserves few cleanups like well formed XML and dedicated
namespace before make it official supported by Darcs Team. 
Anyway I think Darcs Team should spend the effort on more demanded
features. Let keep this XSD in contrib and when XML interface become
demanded, the focus can be allocated on that.

Rado
msg13856 (view) Author: gh Date: 2011-04-01.14:44:49
Already reviewed and amended, and looks ok, it's in.
History
Date User Action Args
2011-02-10 21:01:04dixieckocreate
2011-02-17 15:23:09koweysetmessages: + msg13700
2011-02-19 13:31:28dixieckosetfiles: + unnamed, add-inferred-xml-schema-for-outputs_.dpatch, unnamed
messages: + msg13703
2011-02-19 13:55:36dixieckosetmessages: + msg13704
2011-04-01 13:48:02koweysetassignedto: guillaume.hoffmann
nosy: + guillaume.hoffmann
2011-04-01 14:44:49ghsetstatus: needs-review -> accepted
messages: + msg13856