[vlc-devel] [vlc-commits] packetizer/startcode_helper: enhance with AVX2

[jnqnfe at gmail.com]

On Tue, 2019-03-12 at 18:54 +0200, Rémi Denis-Courmont wrote:
> I don't care.
> 
> I am very fed up with you and your persistent refusal to test your
> patches 
> before you submit them. I have warned you many times already.
> 
> You don't want to follow the rules that everybody else does?
> 
> Fine. Go away.
> 

Seriously?

Firstly, I have to begin by questioning why you have published your
grievance here publicly on the mailing list rather than speaking to me
privately? Your message may perhaps be perceived as somewhat tame, as
far as personal attacks go, but fundamentally that is what it is; I do
take some offence, and I very much do not appreciate your having chosen
to express this publicly.

Before I say anything further, let me be clear: I have absolutely no
desire to fall out with anyone here Remi, I'm here to contribute. I
respect that you do a lot of good work on the project and you seem very
competent at what you do; I welcome fair criticism and can accept
people wanting to sometimes vent frustration at something, but I do not
feel here that you are being fair to me or correct in your facts, and
especially since you've made this public, I'm going to have to take a
little time to address what you've said. I hope that you show me the
respect to take the time to read it and do so with an open mind.

The recent problem aside (to which I am sympathetic to any frustration
felt), I really do not understand how you can justify supposedly having
built up this frustration towards me over time.

You have misrepresented and/or misunderstood my position here on the
issue at hand; I believe that you are factually mistaken in how much
you seem to recall "warning" me previously, indeed in the idea that you
have even "warned" me at all in any regard rather than simply
expressing dissatisfaction and discouragement at best; and I feel that
your seemingly uncompromising attitude on this topic no matter it seems
how trivial the patch is very much unreasonable.

I do not even really understand in what form you actually mean when it
comes to "warning", which does not make it easy to respond to; do you
mean in terms of playing with fire - that ultimately a bad patch was
bound to result at some point - or warning me that if I kept it up that
you'd start refusing to accept my contributions? Either way I do not
accept that you have actually ever done either.

Furthermore a primary point of the message you've replied to was to
point you to the discussion regarding the patch set from which this
problem arose; if you'd bothered to pay attention to that, you might
have subsequently noticed the message I submitted last Friday the 8th,
where I stated perfectly clearly alongside the latest revision of those
patches that this revision **I had in fact compiled**, thus making some
of your comments here rather redundant!!! You might have also noticed
that this patch set revision included the fix for the packetizer issue,
and that you could have just picked out and merged that instead of
reverting the enhancement itself which will now need to be re-applied.

Let me expand on a few things briefly:

1) With respect to your portrayal of my supposed attitude to compiling
patches: You're making it out as though I have simply arrogantly been
refusing to do so. I am certain that the ONE time I recall us ever
having discussed this previously (many months ago) I not only made it
clear that I very much agreed with the principle of compiling first and
felt it not ideal that I was not doing so here, but also (as I recall)
I explained to some degree why I was not actually doing so, that I was
not in a position to at the time (nor have I been since); it has not in
any way ever amounted to arrogance/incompetence/recklessness or
whatever else you may want to imply, nor disrespect or disregard for
the time required on the part of VLC developers to review
contributions.

FYI in doing my bit to address this recent unfortunate mess I have had
to abruptly take some effort to sort out an at least temporary solution
to enable me to at last be in a position to compile VLC, and thus as
stated above, I have in fact compiled the latest revision of my recent
large contribution.

2) With respect to your having supposedly "warned" me "many times",
when and how? As far as I recall, there are only two occasions of it
ever having come up until now, once many months ago, and once about 6
weeks ago. In the former case it came up in discussion only because I
had felt the need to deliberately mention it with certain submissions.
I do not recall any sort of "warning" from you, just that we were in
agreement that it was not ideal; that on principle you did not like the
idea. In the latter case it was when I enquired about purging MMX and
you, seeming to have remembered me as someone who submits without
compiling, (though on a false basis of it supposedly being a deliberate
choice), made an assumption (as correct as it might be) that I was
still not doing so; you chose to add to the discussion suggesting that
I abstain if I could not even compile. I ignored you then considering
your stance on this to be unreasonable (not that if I had been in a
position to compile that I would not have done so, I certainly would).
(FYI, there were no compile issues at all with my purge patch).

3) Of all my odd contributions to VLC, most have been very small and
trivial, in part because I recognised and accepted that not compiling
patches was not ideal and out of not wanting to frustrate reviewers
with potentially broken ones. Many of them have been merged, often by
yourself, without any issue; at least one (init thread setup timing)
was rejected fairly because you disagreed with the change; some others,
similarly trivial, simply were never reviewed/merged (though JBK
randomly merged 3 recently I noticed), all of which I can tell you now
have no problems compiling, and ones which I see no just reason for
reject of. Of two old sets of patches that never made it in it, one
(log level value restrictions) I made several revisions of per your
requests, but I believe you lost interest in in the end; the other
(transform related) you failed to adequately review (I won't bother
detailing here for brevity), and again, both I can now tell you compile
without issue.

I have to ask, have any of those old patches been discarded due to any
such frustration regarding compiling? If they have, I rather take issue
with your not having made any attempt to make me understand that. I
hope this is not the case and they simply got overlooked or something.

So the point here is: from a history of small non-problematic changes,
and not being in a position to compile rather than arrogance, as I've
previously made clear, how is it reasonable for you to develop a
supposed build up of frustration towards me? With exception to this
recent issue, I very much think that you are getting yourself worked up
here without just cause.

I have to wonder, if in my earlier contributions I had never mentioned
them not having been compiled, considering they had no issues and thus
that you would be ignorant as far as this matter goes; whether you
would be feeling any such frustration at all now (except only with this
recent actual issue obviously). If not, as why would you not, then
again, how is is fair to have built up the frustration simply from
knowing they were not compiled when they were both trivial and also
good?

To be very, very clear, should you not be reading this thoroughly, it
is *not* the case that I was choosing to not compile, I was not in a
position to, as I am certain I had originally stated.

Again, Remi, I have no wish to fall out. Nor though do I have any
intentions of letting myself be unfairly pushed out from contributing
to this project when I see something to improve. I hope that you can
understand my point of view as to how I feel you are being unfair here,
and I hope we can ultimately move smoothly past this.