Re: [Audacity-devel] [Audacity-quality] Fwd: Audacity-quality post from edgarmusgrove@wavecable.com requires approval

[Vaughan J. <va...@au...>]

Gale, when I said I inferred Ed's unsubscribe action was political, I 
thought it was a move he made on his own, as a statement against 
-quality. I did not say or imply there was any conspiracy going on, and 
specifically, it did not even cross my mind that you had anything to do 
with it. I assumed Ed makes his own decisions. I don't assume conspiracy 
and don't see why you thought I would.

Anyway, Ed clarified to me off-list that he unsubscribed from -devel 
too, that he feels his level of contribution is such that feedback@ is 
sufficient. Certainly should save him some time!

The rest of this message I think should be on team@, and I'll respond 
there.

- Vaughan


On 4/15/2010 12:55 AM, Gale Andrews wrote:
>
> Summary:  No backdoor political manoeuvres about -quality are going
> on. If QA and developers have to be much more distinct than now, and
> we are aware of the dangers, so be it (IMO).
>
> Tinkering at the margins by sending a proportion of -devel posts to
> -quality using way too vague a criterion, meaning more e-mails spread
> around in total, isn't going to work. Can we be more radical and make
> -devel a much more demarcated list that QA don't even need to be
> concerned with?
>
>
> | From Vaughan Johnson<va...@au...>
> | Wed, 14 Apr 2010 14:28:25 -0700
> | Subject: [Audacity-quality] Fwd: Audacity-quality post from edg...@wa... requires approval
>> Ed, it's disappointing that you've unsubscribed from -quality. I guess
>> it's a political move.
>
> Vaughan, when I had time I was going to make what I hoped was a
> constructive responsive to what you sent to Team. But to clarify, I had
> no knowledge or direct part in Ed's decision, whatever his reasons were.
> I encourage him to re-subscribe. I think the list is likely to remain in
> some form or other, even though I think the content on it now could be
> handled just as well in other ways.
>
> Since -quality started, I've made clear what I think it should be used for:
>
> * discussions to flesh out possible Proposals to go on Wiki
>
> * discussing issues for possible inclusion on Bugzilla, where it is
>     totally unclear if it's a bug or not
>
> ...and what it should *not* be used for:
>
> * posts about quality of code
>
> * discussing issues that are clear bugs - they are to go straight on
>     Bugzilla in order to keep e-mails down. To keep abreast of all bug
>     changes, wherever they emanate from, please look at the Bugzilla
>     Recent Changes.
>
> The list has never been used much. I can only assume that people (and
> definitely *not* only me) have been posting to -devel out of preference -
> maybe because of the confusions. Or maybe because -devel traffic had
> dropped off, they thought they didn't need the inconvenience (which it is)
> of posting/cross-posting to -quality.
>
> We also have the contradiction that at release time, no-one seems to mind
> lots of posts even though clearly most of them have quality content.
>
>
>> But you didn't see my response to Gale's characterization of -quality as
>> elitist. I think he and Martyn are totally wrong about that.
>
> I completely accept that you (Vaughan) did not set it up to *be* elitist.
> Obviously I've read what Martyn has said a few times about "-devel for
> devels". But although I understand we don't want heavy-handed rules,
> the fact remains that some concrete, unequivocal guidance is needed
> about where to post. We don't have that now, IMHO. And -quality may
> still be *seen* as elitist unless everyone follows that guidance.
>
>
>> Martyn was wrong in characterizing -quality as a place to exile
>> non-developers. From my perspective, he's frustrated to the point of
>> fury at Gale alone, not non-developers as a class. Frankly he's not
>> alone in that frustration with Gale, which is the real problem here. I
>> hope that in discussions on team@ we can resolve that. This whole
>> revisiting of -quality as elitist is a pointless distraction from that.
>> But Martyn's harsh expression of his frustration has caused more
>> problems than clarity, which is what I think he actually intended. But
>> he can speak for himself.
>
> I wish someone would tell me what these latest "frustrations" are that
> you perceive. I'm certainly frustrated by the bucketloads of my time
> these discussions waste. I accept it would have been better to give .
> James a straighter answer first time about Release Manager. I said why
> I didn't, but I'm glad no apparent offence was taken when I made clear
> my concern over the two aborted releases. Thanks for that.
>
> But since then, I thought I was doing well! I worked very hard testing
> LRN's Extended Import, including on making the posts as concise and
> accurate as I could. Those posts went to -devel because (1) everyone
> else's responses did too  2) unless I'm mistaken, the deal was that new
> features should be discussed on -devel. Clearly "test results" e-mails
> are a problem and I do have some ideas what to do about that.
>
> Plus, I posted about Windows 7 support and moonphase problems (to
> -devel, for the reasons I explained). The issues were summarised. I've
> fallen over backwards to "suggest" things (even if I've actually "decided"
> as far as I'm concerned). So please explain off-list - what went wrong
> since 1.3.12? That I posted to -devel when  everyone else did too?
>
> ---
>
> Obviously I (and I think unless they want to contradict me) everyone in
> QA would "prefer" a more inclusive -devel list like it always used to be,
> with -non devels feeling able to fully contribute. We'd prefer we are all
> properly aware of each others' concerns and aspirations, so we make a
> product that's good for users, as well as merely fun to develop on.
>
> Equally, to whatever extent an inclusive -devel list with many
> non-technical, quality-driven posts drives current and new developers
> away - we have to address that. I can live with a clearer split between QA
> and developers if that's what it takes. I do strongly object to doing this
> while our -devel info page says the direct opposite... so let's change the
> web page.
>
> Just as Vaughan made the -quality list for functional purposes, I don't like
> it as it is now, for functional reasons. The rationale of when you post/
> cross-post just is not clear enough. Look at that Extended Import thread
> alone - the posts go from part QA/part code to all code (the libmad issue)
> and all QA. To effectively bug track I think we need things "going on" in
> as few places as possible. I certainly don't need any duplicate e-mails
> and expansion of the total number of e-mails - I actually do have to read
> them all, cover to cover.
>
> At present, my thoughts are going along the lines Martyn suggested when
> -quality was first proposed. Most posts on -devel are to some extent or
> other about quality (to me, that's expected, but still...). So, move almost
> all -devel content to -quality. Everything that the QA side needs is there;
> my objection is gone. Devel list is for things like code reviews and
> refactoring, discussion of improving SVN, possibly technical discussion
> of how to implement a Proposal or fix a bug (with the result of discussions
> posted to Wiki or Bugzilla).
>
> "Developing new features" may be a crossover case and always will be.
> Clearly, testing and interface sugestions have to go to -quality. Maybe
> there is some way to make it less to and fro - let it accumulate comments
> on -quality, maybe on a temporary wiki page, and then send all the feedback
> back to -devel at once.
>
> This is much more far-reaching than the "user-facing" definition of -quality.
> This is what I had in mind when I suggested about -devel containing nothing
> except posts containing code. Release discusssions are not on there. The
> assumption is most posts will not be on there. Can -devels get what they
> need by reading only -devel, Wiki, Bugzilla and skimming -quality in digest
> mode? No idea.
>
> And if we fall back to the old less rigorous bug-tracking, maybe decisions on
> "future" of -devel could be postponed, for a time anyway.
>
> [As an aside, has anyone considered using Nabble to work on -devel?  If I
> haven't read the list for a few days, I find that much less intimidating than
> an inbox-full, even if threaded]. Also the posts really are chronological
> there, which I don't find always works in e-mail clients.]
>
> I've left discussion of "release procrastination" from Vaughan's post
> alone. I'll reply to Team about that.
>
>
>
>
>
>
> Gale
>