My thoughts.

First when I think blocker I think blocker for users, not developers. So
the first question i would ask when filing an issue is "is it likely a user
will actually run into this issue?".

The more build bots the better in my opinion but it should be communicated
somewhere what that build bot does beyond the official one. And in part
using that to determine if the issue should be a blocker. For instance,
 issues that arise from file/uri conversions for file paths with spaces is
potentially a major show stopper for windows users which is a large
contingent of users, so i would classify it as a blocker.

However a build that is breaking because it is testing a non-standard
module, or even a non-standard (ie non-oracle) jdk i would not classify as
blocker since it represents a smaller subset of geoserver users. Often I
run into test failures on OSX because who knows why... but I realize i am
using a platform most devs don't so i just ignore it.

Long story short I think the person has to use there better judgement when
filing a build breaking issue that doesn't show up on the main build server.

$0.02

-Justin

On Wed, May 9, 2012 at 3:56 AM, Andrea Aime <[email protected]>wrote:

> On Wed, May 9, 2012 at 10:57 AM, Ben Caradoc-Davies <
> [email protected]> wrote:
>
>> Blocker is, or should be, "end of the world, attend to immediately" type
>>> of issue, e.g., a build breakage
>>> on the official build server, something that blocks every developer, but
>>> I see it used
>>> as "it's a blocker _for me_" instead, which I believe is misuse.
>>>
>>
>> One word you have added here is "official". While I appreciate the work
>> that OpenGeo have done in providing Hudson coverage, it is not the only
>> build coverage for this project.
>>
>> Despite several requests from me, the OpenGeo Hudson no longer builds in
>> a path with spaces, so does not catch the File/URL conversion build
>> failures I often report. These break the build on some Windows development
>> boxes, and can break production deployments on Windows and other boxes in
>> paths with spaces or internationalised paths ("Program Files", "Documents
>> and Settings").
>>
>> There are other reasons to run an "unofficial" build bot, including
>> coverage of obscure modules like webservice; these are not OpenGeo's
>> problem. Others have run build bots to cover other JDKs. If we set one up
>> for Jody for OpenJDK would it be unofficial? Are only OpenGeo Hudson
>> failures blockers?
>>
>
> Imho only the official build server counts as a blocker, because nobody
> should commit while the build is broken there.
> Other build bots are treating cases that the community has not agreed to
> maintain at all, or at least at the same
> level of urgency.
>
> If you want your build bots to be "official" in my opinion you should
> address the community and get them
> to agree that the case handled by such build bot deserves everybody
> immediate attention (since that's what
> a blocker issue is, a "stop the world and fix immediately" one).
> The build bot should obviously made be available to developers and allow
> people to start a build
> manually when they are trying to check if a certain fix worked.
> The bot should also be described in detail so that people know how to
> reproduce the specific case it's handling locally.
>
> The OpenJDK case you're mentioning is telling, nobody has agreed to give
> OpenJDK any urgency
> since it's so far not relevant business wise and nobody stepped up to be a
> "spare time OpenJDK maintainer",
> if someone sets up a OpenJDK build bot it's appreciated, but it won't be
> given particular priority over
> any other random issue in jira (at least not by me).
>
> Cheers
> Andrea
>
> --
> Ing. Andrea Aime
> GeoSolutions S.A.S.
> Tech lead
>
> Via Poggio alle Viti 1187
> 55054  Massarosa (LU)
> Italy
>
> phone: +39 0584 962313
> fax:      +39 0584 962313
> mob:    +39 339 8844549
>
> http://www.geo-solutions.it
> http://geo-solutions.blogspot.com/
> http://www.youtube.com/user/GeoSolutionsIT
> http://www.linkedin.com/in/andreaaime
> http://twitter.com/geowolf
>
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Geoserver-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/geoserver-devel
>
>


-- 
Justin Deoliveira
OpenGeo - http://opengeo.org
Enterprise support for open source geospatial.
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
GeoTools-Devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-devel

Reply via email to