MENTORS,

The last priority over the next two weeks before GSoC begins is for students to 
establish their commit privileges.  For that, they need your help.  Please 
review and provide feedback.  Thanks to those of you (Cliff, Erik, ??) that 
have been hitting them up hard and to others interacting with students on their 
project plans and wiki pages.

The standard practice employed since our open source project was established is 
that new contributors have to demonstrate competency with our contributor 
responsibilities.  Paraphrasing HACKING, those already with commit access have 
a responsibility to ensure new contributors are following the guidelines in the 
developer's guide.

In practice, that generally means new contributors need to provide at least one 
but ideally two "non-trivial" pristine patches that apply *flawlessly* without 
breaking anything.  The code should be demonstrably "better".

As you review patches [1], be critical.  It's okay to reject them when there is 
an issue.  The structure of the patch and change communicated by it is more 
important than the "content" of the patch itself.  If you have to make a change 
in order to apply the patch, any change, then that detail needs to be 
communicated to the contributor (even if it's simply incorrect formatting) and 
they need to try again.

[1] https://sourceforge.net/tracker/?group_id=105292&atid=640804&status=1

STUDENTS,

Please don't take criticism or a patch rejection personally! :)  It's all just 
about making sure everyone has the same healthy understanding so all developers 
can be productive.  Especially for a large project like BRL-CAD with an 
extensive source code history, code quality is vitally important.  If that 
change you spent a week working on injects a bug that isn't discovered until a 
year later -- but then takes three developers a couple days to diagnose, 
inspect, fix, test, cleanup, and document -- then that change was ultimately a 
net negative waste of time.  Nobody wants that.

Be sure you're read through the HACKING file, particularly the COMMIT ACCESS 
and CONTRIBUTOR RESPONSIBILITIES and CODING STYLE & STANDARDS sections.  Make 
sure you actually run the application/library/interface before your change and 
check it afterwards to make sure you didn't break anything!  Make sure you're 
following our conventions; make sure you test.  Ask questions.

Some patches are very quick to review and apply and others will take some time, 
but communicate with the developers to remind them of your patch if you're 
waiting on feedback.  Make your patches as small as possible (without making 
them trivial) so they are easier to review.

They don't have to be glorious big patches that change the world or fix 
impossible bugs.  On the contrary.  They need to demonstrate that you 
understand our code conventions, you know how to communicate a code change, you 
know how to properly document your work, you know how to test your changes, etc.

Everyone should have commit access before GSoC begins.  Several of you are 
nearly there (or already there): Wu, Christina, Ksenija, and Andrei.  Suryajith 
is the only one working in isolation so he gets held to the fire differently. 
;)  The rest of you are either pending review or were given feedback, your 
patch needs more work, or you need to submit additional patches. 

For everyone, continue to submit your work through the patches tracker until a 
developer lets you know that you've been granted commit access.  Once you have 
access, you will be expected to commit *daily* and throughout the day while you 
work on your project.

Cheers!
Sean


------------------------------------------------------------------------------
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/
_______________________________________________
BRL-CAD Developer mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/brlcad-devel

Reply via email to