[Buildbot-devel] Bug 87 -- custom build properties

Sun Mar 16 20:52:13 UTC 2008

Hi Dustin / Aaron,

Firstly, thankyou so much for your respective efforts on this ticket and BuildBot generally. I'm sorry I'm unresponsive at the moment - we're so Busy at BakBone that I have failed to get any more work done on this.

I think I can help with the situation regarding the multitude of available patches. It looks as if Aaron's customprops1.diff contains everything that was in customprops076.tgz. customprops076 was an update for 0.7.6 of all previous patches I had posted minus the functionality to display custom properties in the web interface. I believe Aaron has been good enough to perform the web interface work. So, I think the Aaron's customprops1.diff now contains all work for this ticket with the exception of customprops-fix-debugclient.patch which was posted by gward.

Aaron, does that add up to you?

-- Roch

-----Original Message-----
From: buildbot-devel-bounces at lists.sourceforge.net on behalf of Aaron Maxwell
Sent: Sat 15-Mar-08 8:17 PM
To: buildbot-devel at lists.sourceforge.net
Subject: Re: [Buildbot-devel] Bug 87 -- custom build properties

Hi Dustin,

I've attached a composite patch that I believe will help you with the 
development branch.

EXECUTIVE SUMMARY

Suggested recipe: Were I in your shoes, I would apply these patches in this 
order:

0) If you have not already, apply "trigger.patch" 
and "bb-item56-0.7.6fix.patch" from ticket 56.  Sounds like you have.
1) (requested) Apply commands.py_2.patch from ticket 140.
2) (requested) Apply OneLineMixin.make_line.patch from ticket 145.
3) Apply the attached customprops1.diff.
4) (optional) Integrate customprops-fix-debugclient.patch from ticket 87.

DETAILS

customprops1.diff includes "customprops076.tgz" and "customprops_web.diff" 
from ticket 87, and "customprops_trigger.diff" from ticket 56 (which adds 
custom properties support to the triggerable scheduler).

customprops1.diff assumes that the patches "trigger.patch" 
and "bb-item56-0.7.6fix.patch" from ticket 56 have been applied (in that 
order).  I believe they have already been integrated into the development 
branch, Dustin, but if not I remember them being straightforward to apply.

I have a couple of other patch requests while you're at it: to apply 
commands.py_2.patch from ticket 140, and OneLineMixin.make_line.patch from 
ticket 145.  (These are the two I mentioned in the last email, that I thought 
were mixed in - turns out I had them separated out nicely.)  The patch 
customprops1.diff was applied after these two; I *think* there is no 
dependency -- i.e., just customprops1.diff will merge in fine -- but I have 
not tested it.  Neither are related to custom props, but I believe both are 
useful additions to the codebase.  OneLineMixin.make_line.patch is, in my 
opinion, a small but highly useful refactoring that greatly aids in 
customizing buildbot reports.  My submission of commands.py_2.patch is a bit 
more selfish.  It may or may not affect many other people, because it's a bit 
obscure, but it makes my life a LOT easier.  Since it does not cause any 
issues and other people may well benefit from it, I request that it be added 
as well.

There is also customprops-fix-debugclient.patch (again still for ticket 87), 
which is contemporary but not included in my attached patch.  I'd recommend 
integrating it.  Unfortunately it does not quite cleanly merge in; I did not 
really investigate, but if you are able to make it fit I think it'd be 
worthwhile.

There are a few other patches in ticket 87 that are not included in 
customprops1.diff.  Some of the patches prior to customprops076.tgz actually 
should not be included: they date back to 0.7.5 or even 0.7.4, and are 
obsoleted by customprops076.tgz.  Some, however, it might make sense to 
apply.  Specifically, html.py.diff, runner.py.diff, tryclient.py.diff and 
scheduler.py.diff.  A potential problem: all of those predate 0.7.6.  Were I 
in your shoes, I'm not sure what I would do with these - it kind of depends 
on how much free time you have.  If your temporal resources are scarce, skip 
them for now.  Hopefully Roch is reading this, and he'll be able to comment 
(he originally wrote many of them).  They all provide (or did provide) useful 
functionality, so one thing you might do is just try to apply them one by one 
(*after* applying all patches 0-4 listed in the executive summary).  If they 
merge smoothly and do not cause any obvious problems, it's probably a good 
idea to keep them.  But then again, that approach is a little sketchy - 
preferably you'd test them out well, which is not easy.  Roch, any comments 
here?  (I don't have his email - if anyone knows, please ping him)

We're ramping up for a product release soon here at SnapLogic and coding like 
mad, so I won't have many free cycles to help more right now.  I'll answer 
questions, etc. as best I can.  Hopefully you have a lot of good information 
to work with here.

Thanks for maintaining the dev branch, Dustin.

-Aaron

-- 
Aaron Maxwell .:. amax at snaplogic.org .:. http://snaplogic.org
SnapLogic, Inc. - Data Integration for the Last Mile

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://buildbot.net/pipermail/devel/attachments/20080316/9290bcba/attachment.html>