[asterisk-dev] Bugs/patches 16033 and 16590 ignored forever

[Leif Madsen]

Kirill 'Big K' Katsnelson wrote:
> On 100420 0548, Leif Madsen wrote:
>> I'm not sure how to resolve the paradox you're speaking of. I'm not even sure I
>> agree a paradox exists. Separate pools of resources do not exist for testing and
>> development.
> 
> But they do:
> 
>> The Digium development team does both testing and development
> 
> Yes, but not only. Both developer and test pools are larger than the dev 
> team.

Right, but we're talking about the Digium team. Beyond that no one has any 
control other than the developer themselves.

> Again, suppose I file a bug, then developer X writes a fix, *I* confirm 
> it. In this case, the pool of testers includes at least one who is 
> particularly willing to test the fix.
> 
> If, however, I file a bug *and* supply the fix, I am excluded from the 
> testing pool.

Why are you excluded from the testing pool? Presumably you have tested the patch 
you've submitted and it works.

> It is true that in the first case I am likewise excluded from the dev 
> pool: I can be only one resource in relation to one bug report. 

Why? Being a developer doesn't preclude you from being a tester.

> So here 
> I am weighing the cost -- essentially the game is to choose from which 
> resource pool I should get myself excluded to let the fix go through.

So far I'm unconvinced of the arguments presented thus far to agree with this.

> So ultimately the question is, what is more likely: that someone other 
> than myself fixes a bug given a bug report, or that someone other than 
> me tests my own fix. From my observation, the former happens, and the 
> latter does not. So far, empirically, to win the game I must choose the 
> former solution: do not fix the bug myself.

Can you provide examples of this? We've stated that issues by the Digium team 
are fixed when it's determined an issue affects a large number of users, and 
that issues with patches require less engineering effort, so those are the 
issues most likely to be resolved faster.

 From the viewpoint of the community, issues resolved are those which affect the 
developer in question, or those which interest the developer to an extent to 
take an issue to completion.

Nothing I've stated leads me to the resolution that an issue without a patch is 
resolved sooner. Your observations simply state that issues you've submitted 
with patches have not garnered the attention of developers that can take an 
issue to completion. From the Digium teams viewpoint this must mean the issue 
was obscure enough to not have made it into a sprint (yet), and from the 
viewpoint of the community developer, it has not been an issue of enough 
interest to take to completion; neither of which I have an answer to how to resolve.

I don't see how taking the exact same issue, where the only difference is a 
patch or not a patch, leads me to believe the same issue without code attached 
to it is resolved sooner.

> This is a paradox: If I am willing to contribute my time to fix the 
> problem, I am punished by having to maintain the patch, well, for a very 
> long time.

Submitting a patch does not require anything further from you other than 
feedback from a developer who is attempting to move an issue to completion; and 
even then, there is nothing binding you to that. By creating a patch and 
submitting it to the issue tracker you are not legally bound to maintain anything.

If you didn't submit the patch to the Asterisk issue tracker, then you'd be in 
the same position of maintaining any changes to the code base locally indefinitely.

Leif.