[haiku-development] Re: Banning Barrett
- From: waddlesplash <waddlesplash@xxxxxxxxx>
- To: Haiku Development ML <haiku-development@xxxxxxxxxxxxx>
- Date: Thu, 14 Mar 2019 22:26:39 -0400
On Thu, Mar 14, 2019 at 5:57 PM Dario Casalinuovo <b.vitruvio@xxxxxxxxx> wrote:
I removed myself temporarily from the ml as a sort of auto-banning, isn't it
clear?
I planned eventually to return when I felt I had adjusted everything.
Well ... that doesn't seem to make a lot of sense to me, to continue
committing code but not to be subscribed to the list where discussions
take place.
How nice is for you for considering me. No I was at work. I joined a company
4th of this month.
OK. I guess at this point I am inclined to agree with PulkoMandy that
we closed the vote a tad early... Still, I also agree with him that it
would not make any sense for you to have a vote here.
With that being said, though, despite all the new voices chiming in,
few of these have actually, formally, voted. Expressing opinions is
just fine, but this does not change the result of the vote.
Sure. I should have been subscribed. My fault. Just wondering of a bit of
fairness, I'd have done it for you.
Hmm, OK.
It seems you weren't here when the "big votes" happened.
I'm not sure what you mean by this...?
Let me clarify that I never spoke myself about conspiracy theories.
I don't have them handy right now, but if you are going to press this
point, I will go and locate emails where you essentially accuse the
Inc. of colluding or being biased against you, with no basis to accuse
them of this of course. I don't know what that would be besides a
conspiracy theory.
I never used that term.
Nobody ever said you did. Just because you did not use the term, does
not mean this is not what you are doing.
For instance, in this very thread, today:
May I ask excuses for being banned from #haiku-dev and removed from the Haiku
organization so fastly?
Who actually did it?
What does this mean? Are you accusing us of having some sort of
"shadow governance" that decides things behind the scenes? kallisti5
openly announced in this thread that he was removing your access. What
does this even mean?
Again we had two discussions in #haiku-dev about those commits. And the
discussion was an epical one that I doubt you didn't notice it. Anyway, I
also mentioned that some other times in IRC saying you that you didn't fix
what suggested and you ignored that another time.
Again, I remember very little of this. I sincerely doubt I did
nothing; more than likely I at least posted a reply to the
still-pending changes on Gerrit (or encouraged you to do so.) At any
rate, this eventually became a real discussion that was eventually
resolved; and the amount of strife it took to do so is what is mostly
being discussed here.
It isn't that you asked for the revert, but the manner in which it was
asked.
In which manner?
Stippi: "Huh? Please revert. Either the git-context here is misleading, and
this is actually not an error (please fix the tracing line then), or you need
to fix other code elsewhere. You never improve code by doing something
obviously wrong, just because it helps with one particular test you might be
focussing on."
Me: "What's the reason for adding this half-done API?
And even in the end we have the same TODO "we should not!!! make flat
copies of media_format".
Why you add "Clear" in place of memset, then replace memset with clear in
the code?
We are trying to avoid adding new media_kit API, for the reasons stated one
hundred times, please make this commit to fix something or keep the API
clean."
I may even see it's comment to be worse, by stating that my commit is
"obviously wrong".
Stephan's comment is short, clear, and to the point. Further,
Stephan's interactions with the community have been nothing but
polite, and he has worked extremely hard to defuse tensions whenever
they have arisen. If he says something is "obviously wrong", then,
well, I am willing to bet it indeed is.
Your comment on the other hand is extremely accusatory ("What's the
reason for this half-done API?", when the reason is stated *right in
the message*: "Fix GCC8 -Werror=class-memaccess", and so your question
comes off as just an accusation), and making a remark about having
"the same TODO" (as if we had to fix every TODO in every method or
function we change?) and then another question essentially equivalent
to the first one ("Why you add "Clear" in place of memset, then..." --
again, the reason is right in the commit message; if you hadn't heard
of -Wclass-memaccess, you could certainly have looked it up and seen
that this is one of the possible ways to mitigate it.)
Then you continue with "We are trying to avoid adding new media_kit
API...", which IIRC was disputed (who is "we", and was it ever really
decided that we are not introducing new APIs? BMediaClient is a new
API from not so very long ago, so when was the bar added for "no new
APIs"?) and then add a term of exasperation ("for the reasons stated
one hundred times"), which implies you are extremely frustrated, and
this sets the tone for anyone replying, as they will feel they are
entering a high-tension situation, no matter what they thought before
reading your message.
I found that often people comment on my code without actually reading it.
I remember at least a discussion with Jua, where I forgot that under 64
bit adds are not atomic and then fixed it. Then he replied that the code
was still broken, and I replied "Did you read the code?".
I don't know the context, but that sounds not so polite to me ... "How
so?" or "In what way?" would be a much less aggressive response to a
statement that code one wrote is broken.
Should I add that in this discussion he admitted having missed the thing and
this wasn't actually an aggressive discussion?
Well, jua has since stated that he has essentially decided to ignore
you, even before PulkoMandy made the decision to avoid replying to
your commits. So ... whatever happened there, it probably was not so
simple as "he admitted having missed the thing"; or if it was, then
there is some future scenario which is the cause of that.
I had another similar discussion recently with waddlesplash where he
challenged BMetaData, he actually didn't read the code and I had a debate
with him for 20 mins for nothing.
I seem to recall that at the end of that time I just essentially gave
up, because you were not seeing my points and I was failing to
understand your perspective.
No you gave up because you disputed the implementation of it, that it wasn't
IPC-able and you didn't even know it actually had a Message() method.
That was my base accusation, yes. However I still then, and now,
dispute the class's necessity at all; it's just a shallow wrapper
around BMessage with not much point to it; in fact I had more than a
few reasons why it was a bad wrapper. But you just failed to see
these; and so in the end I just was making sure IPC was still
possible. So I gave up.
The point is that it's hard to communicate. Especially when you develop a
relatively obscure part of the OS. Again I never had a grudge with PK for
that. I knew he was tired and reports are always a pain in the ass to write,
so I just didn't want to worsen the situation.
"Relatively obscure"? PulkoMandy has not recently, but he has
contributed to the Media Kit in the past. Same for Axel, jua, etc.
etc. It's not that "obscure", just that nobody is working on it
presently. Certainly they could review code.
I'm fixing it to increase support for streaming, correctly handling metadata,
and make way for subtitles support.
I'm not just refactoring. And I actually fixed a few things in ffmpeg while
doing that.
The fixes in FFmpeg recently were mostly PulkoMandy's, yes? All the
rest of those items are not "fixes", they are "new features". So,
again, how was what I said incorrect?
there are others from the past few years, and as a result very
few developers want to spend the time to (1) learn about the kit and
then fix whatever the bug is, and then
Always was this way.
(2) get yelled at on the
mailing list for doing it "incorrectly", and have it take 5-10 emails
before the real problem is finally understood.
It takes time to understand each other. I find various examples of past
discussions where this already happened.
As said, I don't refuse that I may have behaved incorrectly, but please don't
make it more long than what actually it is.
What do you mean by that...?
This seems like admitting you were provoking me, isn't it?
Not necessarily. It is me apologizing to attempt to mend fences, and
take responsibility for something for the sake of mending fences. If
you had accepted my apology, that would have been the end of it; I
mean what I say and I do what I promise. This would have been no
exception.
I tried all those months and in various instances to solve the thing with
you. I was unsuccessfull, but it seems that there you are stating that you
are going to hold the grudge anyway.
How did you attempt to "solve the thing with me," other than informing
me you had "turned over a new leaf"? That really is not enough. Either
you are admitting wrongdoing or you are not. Simply stating that you
are now different does not resolve anything; if there was truly
wrongs, then one of us has to apologize for things to be properly
resolved.
But with that being said: I don't see how we could bounce back from
this. Even if kallisti5's closure of the discussion and removal of
your access was premature, I don't see how we can repair things now. I
was the last to reply here with a +1 to PulkoMandy's email before
that; there are plenty of others who feel very strongly that you are
not suited for this project. I don't know when or how this could have
been rectified, and these relationships repaired; but I think we are
past that point, wherever it was, by now.
-waddlesplash
Other related posts:
