[phobos] [dmd-internals] Planning software?

[Brad Roberts]

(re-adding the list to keep the discussion public)

On 1/18/2012 11:19 AM, Andrei Alexandrescu wrote:
> On 1/18/12 1:11 PM, Brad Roberts wrote:
>> On 1/18/2012 8:35 AM, Andrei Alexandrescu wrote: Well, then there's
>> the problem.. your understanding of what bugzilla supports.  As I
>> said earlier, there's capabilites that we're not currently using that
>> we could be.  Please go read up on bugzilla's support of both
>> milestones and fully customizeable priorities and severities.  The
>> hierarchy part is sort of true.  Bugzilla does have depends on and
>> blocks relationships that can be established, and you can search
>> based on them.  It's not as good as it could be, but it does work.
>> For example, this search shows all open D bugs that block another
>> bug:
>>
>> http://d.puremagic.com/issues/buglist.cgi?query_format=advanced&field0-0-0=blocked&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&type0-0-0=notequals&value0-0-0=null&component=DMD&product=D
>>
>>
>>  Also of note, there's a number of major projects that do their
>> entire planning within bugzilla.  A classic example is the Mozilla
>> guys.. from which bugzilla was borne.  Another is GCC.
> 
> This echoes my thoughts: it's clear people who are well versed in, and
> committed to, bugzilla, can make it work for their planning.

So, because I listed projects I was sure you'd be familiar with you're suggesting it requires expert quality people to
use it?  Way to further preconceived notions.  Have you done any reading or research, or just ruled it out for some reason?

>> I'm not against use of tools (obviously, I've been _the_ person
>> installing and building tools for us so far), should we find one
>> that's ideal.  But I also don't feel like prioritization, scheduling,
>> and sticking to those are going to be solved by more tools.  They're
>> social issues.
> 
> They're social issues that can be helped by tools.

Yes, they can.  If the people want to be helped.  Our consistency with even the simplest, fixing regression bugs,
suggests we've got bigger issues than just tooling.