I'm moving this thread to team@.
On 4/13/2010 9:41 AM, Gale Andrews wrote:
> Summary -
> No flames, but the purpose of -quality is partly elitist, as Martyn
> confirms. I suggest banning posts on -devel unless they contain
> or refer to code, as the only logical/sensible outcome. Developers
> have to obey that rule too, which I believe they currently aren't.
> You raise many points, and I feel several are contradictory.
> * Release Manager *
> No "final say" is required on "release manager" without having the
> * I think it better a developer should be release manager.
> * Some known boundaries of "when a release should be aborted",
> and "what account should be taken of QA considerations" are IMO
> required, so we avoid a repeat of January and March this year. I
> discussed this off-list with James and we were not I think far apart.
> I thought 1.3.12 went really well. I had a little brief input, which was
> replied to with a decision.
> * -quality list *
> You/Martyn didn't like my content about "mixer issues on Windows 7
> and Mac", right? I can't accept it was hyper-verbose. There are *lots*
> of subtly different symptoms and I did summarise them. James explicitly
> asked me to post to -devel about those issues.
> What "conclusion" were you expecting me to come to? I was asking the
> developers to come to a conclusion about what action we should take
> (if any). If a developer had answered and suggested some code as a
> solution, what list should it go to?
> If that message belonged on -quality, I think the whole content did -
> dropping "provisional support" is even less to do with -devel. It's purely
> a policy decision. I've made my decision about it. I told you what it was.
> Where is the line between "just doing it" and "asking first"?
> Did you want my posts re: testing Extended Import on -quality too?
> I believe my replies were fairly terse, and most of my suggestions
> were implemented. LRN posted to -devel, not to -quality, so I replied
> to -devel.
> You know my view that -quality is largely pointless, with loads of
> confusing cross-over points with -devel, and that it creates more e-mails
> rather than less. As proposed, it makes bug tracking more difficult to
> have this split focus. Consider also if posts to -quality will actually be
> read by developers. If they are read, then the developers aren't saving
> themselves any time. Digest mode works for -devel too.
> Some of the initial posts to -quality were technical coding posts, but it
> was accepted that was incorrect. Is this then the distinction we want,
> so that unless the post contains/refers to C++/Nyquist code, it goes on
> -quality? I think that is pretty nonsensical, but it is an understandable
> distinction that I'd go along with.
> * Purpose of -devel*
> I quote again from:
> "This list is for Audacity developers and everyone else interested in
> following or ***contributing*** to the Audacity development process".
> I believe that text comes from Dominic (or some lead developer) from
> ages ago. That was attractive to me - a friendly, inclusive project
> where non-developers can have an influence (as opposed to the many
> geekish, user-unfriendly open source projects that you see around).
> I believe it's a good policy. I think it's helped us get user-facing
> contributions into a really nice Manual and into improving our quality.
> Martyn's view, that non-developers shouldn't contribute to this list, is
> one I find hard to sympathise with. I still think that view is the crux
> of the problem, and that quality discussions and testing worth their
> salt simply never can be short.
> I too was very sorry to lose Leland. He made/started loads of quality
> related posts on -devel. We corresponded a lot off list. If he thought I
> was driving him from the project, I would have hoped he would have
> said so to me.
> * Procrastination *
> I have been suggesting we release a slightly dirty 2.0 for ages. Others
> I believe are procrastinating over that. If we have no further plan or
> interest in addressing complex mixer and other moonphase issues, we
> should just release after replicable P2s are eliminated.
> | From Vaughan Johnson<vaughan@...>
> | Mon, 12 Apr 2010 15:57:53 -0700
> | Subject: [Audacity-devel] 1.3.12 Anyone?
>> I hope this doesn't escalate into flame war. I set up -quality for the
>> sake of split focus, as development and quality assurance are two
>> different things, and QA discussions were dominating -devel. As I
>> pointed out at the time, in most software organizations of any size,
>> they are separate departments. Gale, I know you felt it was on the order
>> of exile, but that was not my point. And clearly, anyone could subscribe
>> to either or both lists, so there was no exclusivity. There were lots of
>> conversations about it, but the resolution was to try it. I thought it
>> was actually working pretty well until it went into disuse recently. I
>> don't know why, but there are lots of purely QA discussions on -devel
>> Long-windedness is a completely different dimension, imo. To me, the
>> problem is a compulsion to discuss every tiny detail of any possible
>> decision in exhausting detail, rather than focus on making decisions.
>> Martyn is irate about it at this point. Flaming generally doesn't help,
>> but he's not the only one angry about it. I found it really alarming
>> that Leland left. I tried off-list to convince him otherwise.
>> Gale, I really have to attribute a lot of the reason he left to you.
>> Please understand that I know you're doing your best, you are sincere
>> and very dedicated to Audacity. You put in enormous amounts of time and
>> effort on Audacity. But several developers have expressed this same
>> message, about hyper-verbosity and not summarizing, not coming to
>> conclusions, to you over the years, and it gets better for a short
>> while, but then reappears. I really do not want to lose any more
>> developers because of it. This is really very serious, not just a matter
>> of personal style.
>> I too was outraged at your not being able to say yes/no to being release
>> manager. Hundreds of words about this-that-and-on-the-other-hand, then
>> finally it was clear the answer was "no" because of your workload
>> (totally valid reason), except that you wanted to have final say on how
>> it was done. (And clearly, you don't have that say-so if you're not
>> taking on the job, so there was no point in mentioning that.) So I too
>> think you could have just written "no" when James asked, saved yourself
>> writing hundreds of words, and saved most of us reading all of them.
>> Please, please, please remember James's helicopter analogy. Try to think
>> things through on your own, rather than posting every thought to
>> whatever list. Summarize. Post to the appropriate list.
>> On 4/11/2010 5:01 PM, Martyn Shaw wrote:
>>> On 21/03/2010 10:29, Gale Andrews wrote:
>>>>> I believe that procrastination is being the death of this project.
>>>>> Yours in particular is driving people (including me) away from it.
>>>> I could easily drop back to just doing a bit of docs, Wiki and tech.
>>>> support here and there, plus maintaining the mailing lists and
>>>> answering unsubscribed queries sent to them (i.e. the "paid for" bit).
>>> Thanks for that offer Gale, I would like to take you up on it. And
>>> I'm sure some others would too. Are you going to unsubscribe from
>>> audacity-devel before somebody else does that for you? It is a
>>> 'developers list' rather than a 'forum', after all.
>>> And wasn't audacity-quality set up specifically for your long-winded
>>> I appreciate that you have a connection with 'users' (you quote them
>>> often enough) but you really aren't a 'devel' are you? I'd rather see
>>> just 'devels' commenting on this list. Please unsubscribe.
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> audacity-devel mailing list