Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Steven Liu]

James Almer 于2016年11月29日 周二上午5:46写道：

> On 11/28/2016 6:07 PM, James Almer wrote:
> > On 11/28/2016 5:40 PM, Andreas Cadhalpun wrote:
> >> On 28.11.2016 13:18, James Almer wrote:
> >>> [...] shit [...] shit [...] shit [...] shit. It's extremely
> disrespectful.
> >> Everyone reading your mail gets a clear picture of who is disrespectful.
> >>
> > Making my annoyance at the subject known using the above words is not the
> > same thing as being disrespectful to others.
> >
> >> Please re-read our code of conduct and follow it in the future.
>
> I should also point that you seem to be ignoring Nicolas called me a
> dictator,
> and indirectly an imbecile as well.
>
> You may want to reply to his emails and tell him to read our code of
> conduct.
>
> ___
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
I don't care about this, because I just use ffprobe ffplay ffmpeg, of
course, ffserver maybe important to some people for reference to
implementing a server use ffmpeg API

but I this the more and more important thing is the whole of ffmpeg
developing
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 6:07 PM, James Almer wrote:
> On 11/28/2016 5:40 PM, Andreas Cadhalpun wrote:
>> On 28.11.2016 13:18, James Almer wrote:
>>> [...] shit [...] shit [...] shit [...] shit. It's extremely disrespectful.
>> Everyone reading your mail gets a clear picture of who is disrespectful.
>>
> Making my annoyance at the subject known using the above words is not the
> same thing as being disrespectful to others.
> 
>> Please re-read our code of conduct and follow it in the future.

I should also point that you seem to be ignoring Nicolas called me a dictator,
and indirectly an imbecile as well.

You may want to reply to his emails and tell him to read our code of conduct.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 5:40 PM, Andreas Cadhalpun wrote:
> On 28.11.2016 13:18, James Almer wrote:
>> [...] shit [...] shit [...] shit [...] shit. It's extremely disrespectful.
> 
> Everyone reading your mail gets a clear picture of who is disrespectful.
> 

Making my annoyance at the subject known using the above words is not the
same thing as being disrespectful to others.

> Please re-read our code of conduct and follow it in the future.
> 
> I don't know why you are so enraged about this topic, but I assume that's
> because you are frustrated about how ffserver has been allowed to continue
> using private APIs for a very long time.

You should probably read my emails before asking me to re-read them myself.
You'll find out why I'm enraged about this topic. I made it very clear.

> 
> Believe it or not, but I share some of that frustration and thus am happy
> that it has been decided to no longer let ffserver block the library cleanup.
> So if it's still using those private APIs at the next major bump, it will
> sadly have to go.
> 
> However, there is no valid technical reason to remove ffserver earlier or
> if it gets fixed, which is why this should not be done, in particular if
> other developers are still working on it.
> 
> That's the long and short of it.

The long and short of it is: You didn't show up when this was up to discussion.
You're showing up now, late, at the very moment the months old final decision
was going to be made effective, and pretending to alter and override the agreed
course of action, using malicious arguments to favor of your position in the
matter.

Where were you and Nicolas last July? Where were you during the past years when
requests for developers were made?

> 
> I'm dismayed to see that this discussion has devolved into useless flaming.
> I have no desire nor intent to take part in it and urge everyone else to
> refrain from participating in that kind of behavior as well.

And I'm dismayed at how you and Nicolas think it's "rational" to abuse the
goodwill of people and use it against them as arguments to favor your own
stance and agenda in this subject.

Neither you or Nicolas are working on ffserver, or will be working on it if you
get your way. All you're doing is proving that anything can be done in this
project if you're persistent enough, and outlast the other side in arguments.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Reynaldo H. Verdejo Pinochet]

On 11/26/2016 01:00 PM, Rostislav Pehlivanov wrote:

[..]
Since a month has passed, reynaldo still hasn't responded, I think it's


This is not correct. I have been working on weeding
out it's private API usage problems. Last commit to
this effect is from 3 weeks ago.

Bests,

--
Reynaldo H. Verdejo Pinochet
Open Source Group - Samsung Research America
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 28.11.2016 13:18, James Almer wrote:
> [...] shit [...] shit [...] shit [...] shit. It's extremely disrespectful.

Everyone reading your mail gets a clear picture of who is disrespectful.

Please re-read our code of conduct and follow it in the future.

I don't know why you are so enraged about this topic, but I assume that's
because you are frustrated about how ffserver has been allowed to continue
using private APIs for a very long time.

Believe it or not, but I share some of that frustration and thus am happy
that it has been decided to no longer let ffserver block the library cleanup.
So if it's still using those private APIs at the next major bump, it will
sadly have to go.

However, there is no valid technical reason to remove ffserver earlier or
if it gets fixed, which is why this should not be done, in particular if
other developers are still working on it.

That's the long and short of it.

I'm dismayed to see that this discussion has devolved into useless flaming.
I have no desire nor intent to take part in it and urge everyone else to
refrain from participating in that kind of behavior as well.

Best regards,
Andreas

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Mon, Nov 28, 2016 at 09:14:15AM -0500, compn wrote:
[...]
> so michael, my advice to you is to just OK the patch and deal with
> ffserver later... if enough users come back to complain about its
> demise. i feel that this discussion is going no where if both sides are
> unable to come to a compromise.

I do not block the patch, i did not object to it. And iam not going to
kill anyone if its applied, it wont make me happy if its applied no
i would prefer a different path but thats my oppinion.

I did work on ffserver yes and i probably will continue as long as its
easy to work on it. I find the tool interresting

About approving or rejecting the patch, ffserver has a maintainer.
He is the one to make this hard decission, if a clear decission is
needed. Reynaldo is a reasonable person one can talk with

[...]
-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Never trust a computer, one day, it may think you are the virus. -- Compn


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Mon, Nov 28, 2016 at 05:03:53PM +0100, Nicolas George wrote:
> L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
> > How about the news entry on the website stating ffserver was meant to
> > go with 3.2? And the discussions that lead to it? That's a good rational
> > argument.
> 
> # November 29th, 2016, ffserver not removed
> #
> # Thanks to the efforts of dedicated developers, ffserver has been
> # updated and no longer needs to be removed immediately.
> 
> That's taken care of. And I assure you, ffserver users will prefer a
> change of mind like that than a follow-through.
> 
> > If you meant technical arguments, the time for those was months ago.
> 
> The time for technical arguments is always.
> 

> > If there's a vote, it will be to choose between ffserver being removed
> > tomorrow, or right before 3.3 is branched.
> > There's no "ffserver stays" option. That possibility was lost months
> > ago when neither you or anyone else showed up to back it up.
> 
> Hail James, our Great Dictator who decides what we are allowed to vote
> about.

can everyone please try to not escalate this and be polite to each
other

thx

[...]

-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Its not that you shouldnt use gotos but rather that you should write
readable code and code with gotos often but not always is less readable


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 1:52 PM, Michael Niedermayer wrote:
> On Mon, Nov 28, 2016 at 01:07:31PM -0300, James Almer wrote:
>> On 11/28/2016 12:59 PM, Nicolas George wrote:
>>> L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
 No Nicolas. It's a reminder that this patch, as discussed and announced, 
 will
 be pushed.
>>>
>>> And the Corleone give you reminders that they will break your legs.
>>>
 I don't care about ffserver.
>>>
>>> Then do not discuss ffserver.
>>
>> I discuss project management. This is a late attempt at overriding a decision
>> from parties that didn't participate in the real decision making discussions.
>>
> 
>> This crap wouldn't fly anywhere else, but somehow, and according to you, it's
>> rational on ffmpeg.
> 
> Why is it a problem that after announcing the removial of a feature
> we announce that the removial could be avoided as someone came forth
> and updated the code making the removial unneeded ?
> It should not be a unexpected course of events in a project driven by
> volunteers that a volunteer does some work that previously was
> believed noone would do

Any work making ffserver not depend on private API and ffm* will take a long
while. It's unrealistic to think it will be ready even in time for 3.3 or the
major bump.

As i said, once that work is done from within a separate repo after ffserver
has been removed from the main repository as announced, I'm ok if a patch
reintroducing it is sent to this list. We can then discuss if it's something
we want to have in the tree again, knowing well history as shown interest on
it has been minimal and development sparse.

> 
> (but before any more announcing really we should check and double check
>  that all issues that need fixing were fixed and i really want everyone
>  agreeing and being happy and ATM everyone seems in bloodrage berseker
>  mode, while that gives us a interresting suggestion for the next
>  release name, it doesnt make me happy at all to see everyone fight
>  and be angry)

Neither to me. But how this was handled, how the agreed efforts towards
a graceful removal of ffserver for the benefit of anyone that still wants
to use it with up to date git master were used, rubbed me the wrong way.

I don't care about ffserver, but i do care about the image of the project,
and when people abuse trust.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[wm4]

On Mon, 28 Nov 2016 16:59:54 +0100
Nicolas George  wrote:

> > Look it up, stop trying to rewrite history and stop being part of the
> > aforementioned malicious behavior.  
> 
> I see nothing malicious in trying to keep a useful program up to date.
> You have strange priorities.

I'm sorry to inform you that the way you do it (and your behavior
towards James Almer) is very malicious. If you can't see this, I'm
sorry for this project.

It's not my intention to participate in this discussion, though.
ffserver removal was decided long ago, and suddenly there is extensive
resistance against it. If you care so much, go fix it. And fix it
months ago, where we all agreed to delay ffserver removal a bit to
give you time to fix it.

These oddly timed recent patches don't fix anything about ffserver's
fundamental technical issues. They just appear to stall or sabotage the
removal for a little bit, again. This will happen over and over again.
Because you (e.g. michaelni) don't respect the decision, and just come
up with something to remove the current apparent reason for removal.
Not that it will actually help.

This just turns the FFmpeg project organization into a total joke.
Especially now that we seem to have shitstorm flamewars, where
the more persistent party gets to decide whether the decision really
goes through. This is certainly not rational behavior.

Anyway, not wasting my time on this any further. I'm sure you have a
sharp response, how about sending it to /dev/null. Spend your time
fixing ffserver out-of-tree instead, and we'll accept your patches to
readd ffserver once it's done.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 2:16 PM, Reto Kromer wrote:
> I'm also very strongly for keeping ffserver.
> 
> Best regards, Reto

Thanks, but this was up to discussion months ago, not now.

You're however welcome to help the efforts of making it work on
its own separate repository if you're interested in it.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 1:39 PM, Nicolas George wrote:
> L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
>> I discuss project management. This is a late attempt at overriding a decision
>> from parties that didn't participate in the real decision making discussions.
> 
> Maybe they did not participate because they were busy working on the
> actual code.

Keep thinking that if it helps you sleep or whatever.

> 
>> This crap wouldn't fly anywhere else
> 
> Untrue.

I'm eagerly waiting for your results about overriding Brexit, or making other
projects override their decisions of dropping their SPARC and SH5 ports.

> 
>> That nobody will take the project seriously anymore. That nobody will believe
>> in any announcement anymore. That anyone will be 100% sure that with a little
>> lobbying and trolling they will be able to get away with absolutely anything,
>> including overriding old decisions.
> 
> I see. Then what about people who will be disgusted that FFmpeg is a
> project where positive development (fixing bugs) is overridden by
> negative development (removing things) just because the negative people
> are more vocal?

Tell me of all the negative things that happened after we removed all those
library wrappers, when we replaced SDL with SDL2, and many other examples of
old, obsolete, broken or otherwise unmaintained removed modules.

> 
> My wet finger, which I consider more accurate than yours, tells me there
> are many more.

And this is why nobody tolerates talking with you. You only enter a discussion
to remind everyone they are wrong and you're right.

> 
>> Malicious was the attempt at turning efforts of making the program capable
>> of living on its own into an argument against the reason why it must go.
> 
> How is it malicious?

Do i have to say it for a third or fourth time? I think you're better re-reading
my previous emails. It's pretty clear there.

You haven't yet answered my question about how it isn't a lie, for that matter.
You're fast to make new questions while avoiding those aimed at you.

> 
>> You're aware that we could have told Reynaldo that no, we don't want to give
>> him time to make it work standalone, and this patch would have been pushed
>> a week or two ago, long before you even realized this all was even happening?
> 
> And that would have been incredibly rude. I would not have supported
> that decision.

Exactly, it would have been rude to reject a request of that kind, which is
why it was not.

What's rude is what happened after that was granted, and how.

> 
>> That the decision was made, and there's no going back.
> 
> There is a French saying that I already quoted in this discussion: only
> imbeciles never change their mind. For all it being a saying, it is
> still very true.

Call me an imbecile if you want, but i'll see this project decision being
made effective.

Push that ffprobe demuxer if you want and you find no more oposition. Unlike 
this
it's not bound or limited by a project decision, at least until someone calls
a vote about it, so both going in or not are still options, and i lost interest
on it.

You're also welcome to post a patch reintroducing ffserver after it doesn't
depend on private API and the ffm* modules.

> 
> Making one's mind involves taking into account the state of affairs at
> the time and making a conclusion. If the state of affairs changes, then
> the valid conclusion may change too. Someone who keeps the old
> conclusion in this case is indeed an imbecile.
> 
> Well, for ffserver, the state of affairs just changed. Maybe it was a
> consequence of the old decision, so what? The old decision was valid at

Finally! Looks like i'm talking with a rational being that doesn't ignore
facts after all.

> the time, given the state of affairs then. But the net result is still:
> ffserver is now maintained, it no longer blocks the development of the
> rest of the project, therefore the correct decision is no longer to
> remove it.
> 
>> You could also add
>>
>> # November 29th, 2016, From now on, announcements from this project are
>> # worth as much as a copy of ET for the Atari.
>> #
>> # Thanks to the efforts of people that couldn't get over the fact they
>> # showed up late and that abused the goodwill of some developers, nothing
>> # you read announced here from now on is to be trusted.
> 
> You seem to have in your values system the axiom "keeping promises for
> the sake of keeping promises", as if there was a superior being
> rewarding that kind of consistency. A more correct axiom would be
> "keeping promises for the sake of the person to whom the promise was
> made".

I'm pissed of at how this shit was handled. Of all the attempts at twisting
development efforts towards an specific goal in an unacceptable way.

I do not take the abuse of trust and goodwill kindly.

> 
> You can try a poll on the users: "Considering the reasons to remove
> ffserver are now void, would you have us keep our promise and remove it
> now, or change our mind at the 

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Reto Kromer]

I'm also very strongly for keeping ffserver.

Best regards, Reto

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[wm4]

On Mon, 28 Nov 2016 16:05:29 +0100
Paul B Mahol  wrote:

> On 11/28/16, compn  wrote:
> > On Mon, 28 Nov 2016 02:22:48 +0100
> > Michael Niedermayer  wrote:
> >  
> >> I dont know if people want me and reynaldo to spend less time on
> >> FFmpeg, but time is a finite resource. If ffserver is maintained
> >> externally it would mean a noticable hit in maintaince man hours of
> >> FFmpeg. Now it might be that ffserver being pushed out would kill it.
> >> But really as dumb as i am, i dont belive theres a majority who wants
> >> to kill FFserver when there are people who actively work on it and
> >> care about it.  
> >
> > it seems like there are at least a few developers who would feel better
> > if ffserver was removed from master / moved somewhere else to die etc.
> >
> > these developers feel very strongly about removing ffserver.
> >
> > it might be wise to follow the opinions of these developers in order to
> > have a higher morale and continue the working environment here in the
> > project.
> >
> > otherwise some developers may feel betrayed, harmed, ignored, etc and
> > foster hatred because their opinions were not chosen.
> >
> > so michael, my advice to you is to just OK the patch and deal with
> > ffserver later... if enough users come back to complain about its
> > demise. i feel that this discussion is going no where if both sides are
> > unable to come to a compromise.
> >
> > sometimes you have to cut off the toe to save the patient. :)  
> 
> Exactly what happened with so called separate libpostprocess library.
> Is it still being maintained? - NOPE.

Because the only one who cared about it (michaelni) continued to
maintain the FFmpeg git copy of it, instead of the separate repository.

In other words, it was never given a chance.

But somehow it ends up as argument against splitting the repo better,
what a fucking joke is this?
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Mon, Nov 28, 2016 at 01:07:31PM -0300, James Almer wrote:
> On 11/28/2016 12:59 PM, Nicolas George wrote:
> > L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
> >> No Nicolas. It's a reminder that this patch, as discussed and announced, 
> >> will
> >> be pushed.
> > 
> > And the Corleone give you reminders that they will break your legs.
> > 
> >> I don't care about ffserver.
> > 
> > Then do not discuss ffserver.
> 
> I discuss project management. This is a late attempt at overriding a decision
> from parties that didn't participate in the real decision making discussions.
> 

> This crap wouldn't fly anywhere else, but somehow, and according to you, it's
> rational on ffmpeg.

Why is it a problem that after announcing the removial of a feature
we announce that the removial could be avoided as someone came forth
and updated the code making the removial unneeded ?
It should not be a unexpected course of events in a project driven by
volunteers that a volunteer does some work that previously was
believed noone would do

(but before any more announcing really we should check and double check
 that all issues that need fixing were fixed and i really want everyone
 agreeing and being happy and ATM everyone seems in bloodrage berseker
 mode, while that gives us a interresting suggestion for the next
 release name, it doesnt make me happy at all to see everyone fight
 and be angry)

[...]
-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

While the State exists there can be no freedom; when there is freedom there
will be no State. -- Vladimir Lenin


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
> The suggestion can be made if that happens in the future,

I can assure you, the suggestion will be made about five minutes about
the push, if it happens.

>   i guess, but i
> doubt it will be well received.

Well, the push without a valid reason will be very badly received too.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
> I discuss project management. This is a late attempt at overriding a decision
> from parties that didn't participate in the real decision making discussions.

Maybe they did not participate because they were busy working on the
actual code.

> This crap wouldn't fly anywhere else

Untrue.

> That nobody will take the project seriously anymore. That nobody will believe
> in any announcement anymore. That anyone will be 100% sure that with a little
> lobbying and trolling they will be able to get away with absolutely anything,
> including overriding old decisions.

I see. Then what about people who will be disgusted that FFmpeg is a
project where positive development (fixing bugs) is overridden by
negative development (removing things) just because the negative people
are more vocal?

My wet finger, which I consider more accurate than yours, tells me there
are many more.

> Malicious was the attempt at turning efforts of making the program capable
> of living on its own into an argument against the reason why it must go.

How is it malicious?

> You're aware that we could have told Reynaldo that no, we don't want to give
> him time to make it work standalone, and this patch would have been pushed
> a week or two ago, long before you even realized this all was even happening?

And that would have been incredibly rude. I would not have supported
that decision.

> That the decision was made, and there's no going back.

There is a French saying that I already quoted in this discussion: only
imbeciles never change their mind. For all it being a saying, it is
still very true.

Making one's mind involves taking into account the state of affairs at
the time and making a conclusion. If the state of affairs changes, then
the valid conclusion may change too. Someone who keeps the old
conclusion in this case is indeed an imbecile.

Well, for ffserver, the state of affairs just changed. Maybe it was a
consequence of the old decision, so what? The old decision was valid at
the time, given the state of affairs then. But the net result is still:
ffserver is now maintained, it no longer blocks the development of the
rest of the project, therefore the correct decision is no longer to
remove it.

> You could also add
> 
> # November 29th, 2016, From now on, announcements from this project are
> # worth as much as a copy of ET for the Atari.
> #
> # Thanks to the efforts of people that couldn't get over the fact they
> # showed up late and that abused the goodwill of some developers, nothing
> # you read announced here from now on is to be trusted.

You seem to have in your values system the axiom "keeping promises for
the sake of keeping promises", as if there was a superior being
rewarding that kind of consistency. A more correct axiom would be
"keeping promises for the sake of the person to whom the promise was
made".

You can try a poll on the users: "Considering the reasons to remove
ffserver are now void, would you have us keep our promise and remove it
now, or change our mind at the last minute and keep it?"

I have no doubt a huge majority of the users will answer: keep it.

> For things up to debate, sure. This is not the case.

What is this, if not a debate?

> No, those are the only options within the boundaries already established.

Changing the boundaries is always an option.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 1:31 PM, Paul B Mahol wrote:
> On 11/28/16, James Almer  wrote:
>>> What difference does it make?
>>
>> That the decision was made, and there's no going back. And much less after
>> the malicious attempts i already described and pointed you to, that you
>> seemingly intend to ignore.
> 
> It can be readded later, after its fixed?

The suggestion can be made if that happens in the future, i guess, but i
doubt it will be well received. Fears of it bitrotting again would be
justified if it were to be accepted in the tree again.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Paul B Mahol]

On 11/28/16, James Almer  wrote:
>> What difference does it make?
>
> That the decision was made, and there's no going back. And much less after
> the malicious attempts i already described and pointed you to, that you
> seemingly intend to ignore.

It can be readded later, after its fixed?
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
> No Nicolas. It's a reminder that this patch, as discussed and announced, will
> be pushed.

And the Corleone give you reminders that they will break your legs.

> I don't care about ffserver.

Then do not discuss ffserver.

>  I do care about the consequences of a bunch of
> people that showed up very late to the party getting away with trying to
> override an old project decision right when it was going to be made effective.

What are those consequences, pray?

> This is a lie

No.

> Look it up, stop trying to rewrite history and stop being part of the
> aforementioned malicious behavior.

I see nothing malicious in trying to keep a useful program up to date.
You have strange priorities.

> Too bad the decision was made and announced.

Decisions can be revised. If this one exists, then it should.

>   Had this happened months or
> years ago, when help was requested and promptly ignored, it wouldn't have
> ended up like this.

What difference does it make?

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 12:59 PM, Nicolas George wrote:
> L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
>> No Nicolas. It's a reminder that this patch, as discussed and announced, will
>> be pushed.
> 
> And the Corleone give you reminders that they will break your legs.
> 
>> I don't care about ffserver.
> 
> Then do not discuss ffserver.

I discuss project management. This is a late attempt at overriding a decision
from parties that didn't participate in the real decision making discussions.

This crap wouldn't fly anywhere else, but somehow, and according to you, it's
rational on ffmpeg.

> 
>> I do care about the consequences of a bunch of
>> people that showed up very late to the party getting away with trying to
>> override an old project decision right when it was going to be made 
>> effective.
> 
> What are those consequences, pray?

That nobody will take the project seriously anymore. That nobody will believe
in any announcement anymore. That anyone will be 100% sure that with a little
lobbying and trolling they will be able to get away with absolutely anything,
including overriding old decisions.

> 
>> This is a lie
> 
> No.

How it isn't?

> 
>> Look it up, stop trying to rewrite history and stop being part of the
>> aforementioned malicious behavior.
> 
> I see nothing malicious in trying to keep a useful program up to date.
> You have strange priorities.

Malicious was the attempt at turning efforts of making the program capable
of living on its own into an argument against the reason why it must go.

You're aware that we could have told Reynaldo that no, we don't want to give
him time to make it work standalone, and this patch would have been pushed
a week or two ago, long before you even realized this all was even happening?

> 
>> Too bad the decision was made and announced.
> 
> Decisions can be revised. If this one exists, then it should.
> 
>>  Had this happened months or
>> years ago, when help was requested and promptly ignored, it wouldn't have
>> ended up like this.
> 
> What difference does it make?

That the decision was made, and there's no going back. And much less after
the malicious attempts i already described and pointed you to, that you
seemingly intend to ignore.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 1:03 PM, Nicolas George wrote:
> L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
>> How about the news entry on the website stating ffserver was meant to
>> go with 3.2? And the discussions that lead to it? That's a good rational
>> argument.
> 
> # November 29th, 2016, ffserver not removed
> #
> # Thanks to the efforts of dedicated developers, ffserver has been
> # updated and no longer needs to be removed immediately.
> 
> That's taken care of. And I assure you, ffserver users will prefer a
> change of mind like that than a follow-through.

You could also add

# November 29th, 2016, From now on, announcements from this project are
# worth as much as a copy of ET for the Atari.
#
# Thanks to the efforts of people that couldn't get over the fact they
# showed up late and that abused the goodwill of some developers, nothing
# you read announced here from now on is to be trusted.


> 
>> If you meant technical arguments, the time for those was months ago.
> 
> The time for technical arguments is always.

For things up to debate, sure. This is not the case.

> 
>> If there's a vote, it will be to choose between ffserver being removed
>> tomorrow, or right before 3.3 is branched.
>> There's no "ffserver stays" option. That possibility was lost months
>> ago when neither you or anyone else showed up to back it up.
> 
> Hail James, our Great Dictator who decides what we are allowed to vote
> about.

No, those are the only options within the boundaries already established.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
> How about the news entry on the website stating ffserver was meant to
> go with 3.2? And the discussions that lead to it? That's a good rational
> argument.

# November 29th, 2016, ffserver not removed
#
# Thanks to the efforts of dedicated developers, ffserver has been
# updated and no longer needs to be removed immediately.

That's taken care of. And I assure you, ffserver users will prefer a
change of mind like that than a follow-through.

> If you meant technical arguments, the time for those was months ago.

The time for technical arguments is always.

> If there's a vote, it will be to choose between ffserver being removed
> tomorrow, or right before 3.3 is branched.
> There's no "ffserver stays" option. That possibility was lost months
> ago when neither you or anyone else showed up to back it up.

Hail James, our Great Dictator who decides what we are allowed to vote
about.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 12:25 PM, Nicolas George wrote:
> L'octidi 8 frimaire, an CCXXV, Ronald S. Bultje a écrit :
>> The majority.
> 
> Rational arguments first.

How about the news entry on the website stating ffserver was meant to
go with 3.2? And the discussions that lead to it? That's a good rational
argument.
You want rationality yet you fight to override an old project decision.

If you meant technical arguments, the time for those was months ago.

If there's a vote, it will be to choose between ffserver being removed
tomorrow, or right before 3.3 is branched.
There's no "ffserver stays" option. That possibility was lost months
ago when neither you or anyone else showed up to back it up.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 12:24 PM, Nicolas George wrote:
> L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
>> Threats? The only thing i see are reminders of what should have happened
>> earlier and will happen soon, as discussed an announced.
> 
> Reminding of unpleasant things over which you have total control, that
> is called threats.
> 
> "You will burn yourself." -> this is a warning.
> "I will burn you." -> that is a threat.
> 
> Since ffserver will not remove itself, it is indeed a threat.

No Nicolas. It's a reminder that this patch, as discussed and announced, will
be pushed.

> 
>> No, i want a project decision to finally go through. I don't give a shit
>> about the program. If you see me "emotional" is because i can't believe my
>> eyes when i see all this shit happening only after the decision was made
>> and not when actual development work was requested, or when the *final*
>> discussion was taking place.
>>
>> You all should have shown then. You didn't.
> 
> You act quite emotional over something you "don't give a shit" about.

I don't care about ffserver. I do care about the consequences of a bunch of
people that showed up very late to the party getting away with trying to
override an old project decision right when it was going to be made effective.

I'm here to make sure that doesn't happen.

> 
>> Every patch and change that's being done now are direct consequence and
>> effect of the decision to drop ffserver. It has remained broken,
>> unmaintained and bitrotting for years but now that we have reached the
>> fateful day, ffserver is suddenly more important than the h264 decoder.
> 
> This is a "post hoc ergo propter hoc" fallacy.
> 
> The recent patches and the proposal to remove ffserver are not a
> consequence of each other but both consequence of this single fact: the
> internal APIs that ffserver is (ab)using are being changed and it is
> blocking the merges.

This is a lie and you know it, regardless of how fancy you try to make it
sound like.
ffserver started getting all these little changes right after the removal
was going to be set in motion for the 3.2 release. Reynaldo even sent an
email stating as much. That these are meant to be efforts to make it not
depend on public api so it can live out of tree for whoever cares about it.
A little while after that, Andreas and such started to use said efforts as
an argument to "prove" ffserver was still being maintained.

Look it up, stop trying to rewrite history and stop being part of the
aforementioned malicious behavior.

> 
> The patches unblock the merges, therefore the removal is no longer
> needed.

Too bad the decision was made and announced. Had this happened months or
years ago, when help was requested and promptly ignored, it wouldn't have
ended up like this.

> 
>> This patch *should* have been committed weeks ago
> 
> I will remind you that most contributors work on the project on their
> free time. If you care so much about ffserver's code (apparently you do,
> or you would not be arguing so much about it), you could have sponsored
> them to work on it earlier.
> 
> Regards,

If you want to lose time trying to change a decision made by a group of
people, go to England and save them from Brexit. You'll probably have as
much success trying to change that as you'll have it here.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

L'octidi 8 frimaire, an CCXXV, Ronald S. Bultje a écrit :
> To what end? Aren't we dug in, don't we need a decision and follow through
> with the elected outcome?

Democracy does not work that way. Before elections, you need
campaigning, arguments, debating. That way people can make an informed
decision about their vote rather than relying on which candidate has the
nicest face.

Without that, we get... basically all the major elections in 2016 in the
world. Please do not import that circus on ffmpeg-devel.

The arguments for keeping ffserver, IIRC, are:

- With the recent patches, it no longer does any harm.

- It has users, removing it would greatly annoy them, and just moving
  would just cause useless work for both the developers and users.

Let us see what are the arguments for removing. When we have them, if we
do not agree about which one is more important, then it becomes a
political decision and requires a vote.

Last note: "we decided it" is not a valid argument, because we can
always decide to change our mind; only idiots never do so.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Ronald S. Bultje]

Hi,

On Mon, Nov 28, 2016 at 10:25 AM, Nicolas George  wrote:

> L'octidi 8 frimaire, an CCXXV, Ronald S. Bultje a écrit :
> > The majority.
>
> Rational arguments first.


To what end? Aren't we dug in, don't we need a decision and follow through
with the elected outcome?

This has been going on for months. We need to move on, in some direction,
otherwise we'll keep talking about this for months to come.

Ronald
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

L'octidi 8 frimaire, an CCXXV, Ronald S. Bultje a écrit :
> The majority.

Rational arguments first.


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

L'octidi 8 frimaire, an CCXXV, James Almer a écrit :
> Threats? The only thing i see are reminders of what should have happened
> earlier and will happen soon, as discussed an announced.

Reminding of unpleasant things over which you have total control, that
is called threats.

"You will burn yourself." -> this is a warning.
"I will burn you." -> that is a threat.

Since ffserver will not remove itself, it is indeed a threat.

> No, i want a project decision to finally go through. I don't give a shit
> about the program. If you see me "emotional" is because i can't believe my
> eyes when i see all this shit happening only after the decision was made
> and not when actual development work was requested, or when the *final*
> discussion was taking place.
> 
> You all should have shown then. You didn't.

You act quite emotional over something you "don't give a shit" about.

> Every patch and change that's being done now are direct consequence and
> effect of the decision to drop ffserver. It has remained broken,
> unmaintained and bitrotting for years but now that we have reached the
> fateful day, ffserver is suddenly more important than the h264 decoder.

This is a "post hoc ergo propter hoc" fallacy.

The recent patches and the proposal to remove ffserver are not a
consequence of each other but both consequence of this single fact: the
internal APIs that ffserver is (ab)using are being changed and it is
blocking the merges.

The patches unblock the merges, therefore the removal is no longer
needed.

> This patch *should* have been committed weeks ago

I will remind you that most contributors work on the project on their
free time. If you care so much about ffserver's code (apparently you do,
or you would not be arguing so much about it), you could have sponsored
them to work on it earlier.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Ronald S. Bultje]

Hi,

On Mon, Nov 28, 2016 at 10:17 AM, Nicolas George  wrote:

> L'octidi 8 frimaire, an CCXXV, compn a écrit :
> > these developers feel very strongly about removing ffserver.
>
> I feel very strongly about keeping ffserver. Who is right?


The majority. OK, so this is going nowhere. Vote, everyone? We need to
settle this, it's getting ridiculous.

Ronald
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

L'octidi 8 frimaire, an CCXXV, compn a écrit :
> these developers feel very strongly about removing ffserver.

I feel very strongly about keeping ffserver. Who is right?


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Paul B Mahol]

On 11/28/16, compn  wrote:
> On Mon, 28 Nov 2016 02:22:48 +0100
> Michael Niedermayer  wrote:
>
>> I dont know if people want me and reynaldo to spend less time on
>> FFmpeg, but time is a finite resource. If ffserver is maintained
>> externally it would mean a noticable hit in maintaince man hours of
>> FFmpeg. Now it might be that ffserver being pushed out would kill it.
>> But really as dumb as i am, i dont belive theres a majority who wants
>> to kill FFserver when there are people who actively work on it and
>> care about it.
>
> it seems like there are at least a few developers who would feel better
> if ffserver was removed from master / moved somewhere else to die etc.
>
> these developers feel very strongly about removing ffserver.
>
> it might be wise to follow the opinions of these developers in order to
> have a higher morale and continue the working environment here in the
> project.
>
> otherwise some developers may feel betrayed, harmed, ignored, etc and
> foster hatred because their opinions were not chosen.
>
> so michael, my advice to you is to just OK the patch and deal with
> ffserver later... if enough users come back to complain about its
> demise. i feel that this discussion is going no where if both sides are
> unable to come to a compromise.
>
> sometimes you have to cut off the toe to save the patient. :)

Exactly what happened with so called separate libpostprocess library.
Is it still being maintained? - NOPE.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[compn]

On Mon, 28 Nov 2016 02:22:48 +0100
Michael Niedermayer  wrote:

> I dont know if people want me and reynaldo to spend less time on
> FFmpeg, but time is a finite resource. If ffserver is maintained
> externally it would mean a noticable hit in maintaince man hours of
> FFmpeg. Now it might be that ffserver being pushed out would kill it.
> But really as dumb as i am, i dont belive theres a majority who wants
> to kill FFserver when there are people who actively work on it and
> care about it.

it seems like there are at least a few developers who would feel better
if ffserver was removed from master / moved somewhere else to die etc.

these developers feel very strongly about removing ffserver.

it might be wise to follow the opinions of these developers in order to
have a higher morale and continue the working environment here in the
project.

otherwise some developers may feel betrayed, harmed, ignored, etc and
foster hatred because their opinions were not chosen.

so michael, my advice to you is to just OK the patch and deal with
ffserver later... if enough users come back to complain about its
demise. i feel that this discussion is going no where if both sides are
unable to come to a compromise.

sometimes you have to cut off the toe to save the patient. :)

-compn
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 11:04 AM, compn wrote:
> maybe when there is a conflict of opinion on a patch, we should agree
> to disagree on said patch and ignore it?
> 
> instead of arguing endlessly about it?

No. The one patch where that was an option was the news entry patch, the
last time this whole deal was effectively up to debate. Nobody participated
on it.
This patch is simply making effective the project's decision. It's not up
for discussion.

Maybe next time people will participate on things at the proper time and
on the proper threads.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[compn]

maybe when there is a conflict of opinion on a patch, we should agree
to disagree on said patch and ignore it?

instead of arguing endlessly about it?

not specifically targeting you, james, but a lot of people in this
thread.

-compn
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/28/2016 7:34 AM, Nicolas George wrote:
> Le septidi 7 frimaire, an CCXXV, James Almer a écrit :
>> We don't care about testing it with FATE.
> 
> Yes, we do.
> 
>>  It appears to me that you still 
>> don't
>> get that ffserver is being *dropped*. It is, as discussed and announced, no
>> longer part of the project.
> 
> Untrue. It is still there. There are threats of developers about to
> remove it, but they are in the future.

Threats? The only thing i see are reminders of what should have happened
earlier and will happen soon, as discussed an announced.

> 
> The gist of it is that YOU want ffserver dropped, and you do so for
> emotional reasons, not rational ones. Your emotional reasons are

No, i want a project decision to finally go through. I don't give a shit
about the program. If you see me "emotional" is because i can't believe my
eyes when i see all this shit happening only after the decision was made
and not when actual development work was requested, or when the *final*
discussion was taking place.

You all should have shown then. You didn't.

> inspired by rational ones, but only inspired, because you did not update
> your conclusions when the status of the rational reasons changed.
> 
> The rational behaviour is:

Follow through with a made and publicly announced project decision. Period.
Everything else are stalling attempts.

> 
> IF ffserver prevents other aspects of the development, for examples
> merges from the fork, THEN something must be done about it SOON, FOR
> EXAMPLE dropping it, but possibly something else, provided it happens
> soon.

Years, YEARS passed with ffserver in this very specific state of things.
"Something must be done", "Someone should fix it", "Someone should rewrite
it from scratch".

Nobody gave a shit. That period ended a couple months ago.

> 
> IF ffserver is unmaintained but does not block aspects of the
> development, then there is no rational reason to drop it in the short
> term. Maybe later. On the other hand, there are reasons NOT to drop it:
> some users use it.
> 
> IF ffserver is maintained, there is no reason to drop it whatsoever.
> 
> The way I read the discussions, the decision top drop was made when we
> were in the first case. But you can notice that "but possibly something
> else" just happened, and it happened soon enough.
> 
> Since Michael's patches ensure that ffserver no longer blocks merges,
> there is no longer a reason to drop ffserver urgently. Therefore, the
> decision to drop it tomorrow is void.

Every patch and change that's being done now are direct consequence and
effect of the decision to drop ffserver. It has remained broken,
unmaintained and bitrotting for years but now that we have reached the
fateful day, ffserver is suddenly more important than the h264 decoder.

Michael's and Reynaldo's patches were agreed to be efforts towards making
ffserver standalone for the only reason to not have a period of time where
whoever cares about this thing would not be able to use it with git head.
It was not our responsibility to do that, but Reynaldo wanted to do it as
a personal project of sorts. Thus ffserver was allowed to remain for a bit
to make his work easier.

Until then it was all fine. Now you people pretend using that agreement
and the development that came out of it as an argument to prove how
ffserver is in fact maintained after all.
This patch *should* have been committed weeks ago but with the excuse of
"think of the users" it was allowed to be delayed a bit. Now you come and
weaponize that. You do not do that shit. It's extremely disrespectful.

> 
> Regards,
> 

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Matthieu Bouron]

On Mon, Nov 28, 2016 at 02:22:48AM +0100, Michael Niedermayer wrote:
> On Sun, Nov 27, 2016 at 05:31:39PM -0500, Ronald S. Bultje wrote:
> > Hi,
> > 
> > On Sun, Nov 27, 2016 at 1:46 PM, Michael Niedermayer  > > wrote:
> > 
> > > On Sun, Nov 27, 2016 at 07:30:24PM +0100, Clément Bœsch wrote:
> > > > On Sun, Nov 27, 2016 at 06:49:35PM +0100, Michael Niedermayer wrote:
> > > > > On Sun, Nov 27, 2016 at 04:57:36PM +0100, Clément Bœsch wrote:
> > > > > > On Sun, Nov 27, 2016 at 04:25:18PM +0100, Michael Niedermayer wrote:
> > > > > > [...]
> > > > > > > ffserver had 14 commits to it in about the last month
> > > > > >
> > > > > > That's also pretty close to the number of commits in the last years.
> > > > > >
> > > > > > Good will in the last weeks is not enough of a technical
> > > > > > merit/justification to prevent the removal of junk code scheduled 
> > > > > > for
> > > > > > deletion for a long time now.
> > > > >
> > > > > why do you call it junk ?
> > > >
> > > > because it's highly dependent on internal stuff, very limited, 
> > > > completely
> > > > untested, unmaintained for several years but still contains a ton of
> > > code.
> > > >
> > > > > and the sheduling for deletion was conditional on it not being fixed
> > > > > IIRC.
> > > > > Why the hurry to remove it while people work on fixing it ?
> > > > > its not blocking anything ATM AFAIK
> > > >
> > > > There is no hurry, but piling up a bunch patches to fix small things 
> > > > just
> > > > to use it as an argument to say "hey look now it's maintained" in order
> > > to
> > > > save it from being killed is really annoying. The people interested in 
> > > > it
> > > > had years to act.
> > > >
> > >
> > > > You can fix a ton of little things in it, but unless the fundamental
> > > > problems are addressed (ZERO internal API usage + at least partial FATE
> > > > coverage) it's pointless
> > >
> > > of course the goal is ZERO internal API usage + at least partial FATE
> > > coverage.
> > > Well in fact the lack of fate tests have been the primary reason
> > > why i didnt fix some of the API issues years ago. I felt uneasy
> > > changing it without regression tests
> > >
> > >
> > > > and will be seen as yet another case of "KEEP
> > > > EVERYTHING FOREVER" toxic mentality.
> > >
> > > The opposit is toxic too
> > 
> > 
> > I'm perfectly fine with keeping the code, just not in the ffmpeg tree.
> > Please move it to its own tree.
> > 
> 
> > Everybody wants it out. Please follow majority.
> 
> Some people want it out yes, how many and what the majority want i do
> not know.
> Some people also wanted all tools treated equally and moved out (again
> i dont know how many support that)
> 
> Spliting just ffserver out means more work for whoever maintains it
> setting up seperate infrastructure, seperate coverity, coverage, fate,
> ...
> theres a lot of work in all this, its long term continuous effort
> and this is time that wont be spent on FFmpeg.
> 
> I dont know if people want me and reynaldo to spend less time on
> FFmpeg, but time is a finite resource. If ffserver is maintained
> externally it would mean a noticable hit in maintaince man hours of
> FFmpeg. Now it might be that ffserver being pushed out would kill it.
> But really as dumb as i am, i dont belive theres a majority who wants
> to kill FFserver when there are people who actively work on it and
> care about it.

Just a thought, maybe ffserver could live in a separate branch for some time ?

[...]
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Nicolas George]

Le septidi 7 frimaire, an CCXXV, James Almer a écrit :
> We don't care about testing it with FATE.

Yes, we do.

>   It appears to me that you still 
> don't
> get that ffserver is being *dropped*. It is, as discussed and announced, no
> longer part of the project.

Untrue. It is still there. There are threats of developers about to
remove it, but they are in the future.

The gist of it is that YOU want ffserver dropped, and you do so for
emotional reasons, not rational ones. Your emotional reasons are
inspired by rational ones, but only inspired, because you did not update
your conclusions when the status of the rational reasons changed.

The rational behaviour is:

IF ffserver prevents other aspects of the development, for examples
merges from the fork, THEN something must be done about it SOON, FOR
EXAMPLE dropping it, but possibly something else, provided it happens
soon.

IF ffserver is unmaintained but does not block aspects of the
development, then there is no rational reason to drop it in the short
term. Maybe later. On the other hand, there are reasons NOT to drop it:
some users use it.

IF ffserver is maintained, there is no reason to drop it whatsoever.

The way I read the discussions, the decision top drop was made when we
were in the first case. But you can notice that "but possibly something
else" just happened, and it happened soon enough.

Since Michael's patches ensure that ffserver no longer blocks merges,
there is no longer a reason to drop ffserver urgently. Therefore, the
decision to drop it tomorrow is void.

Regards,

-- 
  Nicolas George
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Sun, Nov 27, 2016 at 05:31:39PM -0500, Ronald S. Bultje wrote:
> Hi,
> 
> On Sun, Nov 27, 2016 at 1:46 PM, Michael Niedermayer  > wrote:
> 
> > On Sun, Nov 27, 2016 at 07:30:24PM +0100, Clément Bœsch wrote:
> > > On Sun, Nov 27, 2016 at 06:49:35PM +0100, Michael Niedermayer wrote:
> > > > On Sun, Nov 27, 2016 at 04:57:36PM +0100, Clément Bœsch wrote:
> > > > > On Sun, Nov 27, 2016 at 04:25:18PM +0100, Michael Niedermayer wrote:
> > > > > [...]
> > > > > > ffserver had 14 commits to it in about the last month
> > > > >
> > > > > That's also pretty close to the number of commits in the last years.
> > > > >
> > > > > Good will in the last weeks is not enough of a technical
> > > > > merit/justification to prevent the removal of junk code scheduled for
> > > > > deletion for a long time now.
> > > >
> > > > why do you call it junk ?
> > >
> > > because it's highly dependent on internal stuff, very limited, completely
> > > untested, unmaintained for several years but still contains a ton of
> > code.
> > >
> > > > and the sheduling for deletion was conditional on it not being fixed
> > > > IIRC.
> > > > Why the hurry to remove it while people work on fixing it ?
> > > > its not blocking anything ATM AFAIK
> > >
> > > There is no hurry, but piling up a bunch patches to fix small things just
> > > to use it as an argument to say "hey look now it's maintained" in order
> > to
> > > save it from being killed is really annoying. The people interested in it
> > > had years to act.
> > >
> >
> > > You can fix a ton of little things in it, but unless the fundamental
> > > problems are addressed (ZERO internal API usage + at least partial FATE
> > > coverage) it's pointless
> >
> > of course the goal is ZERO internal API usage + at least partial FATE
> > coverage.
> > Well in fact the lack of fate tests have been the primary reason
> > why i didnt fix some of the API issues years ago. I felt uneasy
> > changing it without regression tests
> >
> >
> > > and will be seen as yet another case of "KEEP
> > > EVERYTHING FOREVER" toxic mentality.
> >
> > The opposit is toxic too
> 
> 
> I'm perfectly fine with keeping the code, just not in the ffmpeg tree.
> Please move it to its own tree.
> 

> Everybody wants it out. Please follow majority.

Some people want it out yes, how many and what the majority want i do
not know.
Some people also wanted all tools treated equally and moved out (again
i dont know how many support that)

Spliting just ffserver out means more work for whoever maintains it
setting up seperate infrastructure, seperate coverity, coverage, fate,
...
theres a lot of work in all this, its long term continuous effort
and this is time that wont be spent on FFmpeg.

I dont know if people want me and reynaldo to spend less time on
FFmpeg, but time is a finite resource. If ffserver is maintained
externally it would mean a noticable hit in maintaince man hours of
FFmpeg. Now it might be that ffserver being pushed out would kill it.
But really as dumb as i am, i dont belive theres a majority who wants
to kill FFserver when there are people who actively work on it and
care about it.


[...]
-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

What does censorship reveal? It reveals fear. -- Julian Assange


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/27/2016 8:53 PM, Andreas Cadhalpun wrote:
> On 28.11.2016 00:37, James Almer wrote:
>> On 11/27/2016 8:23 PM, Andreas Cadhalpun wrote:
>>> This can't happen while ffserver still uses internal API, and when it 
>>> doesn't
>>> making it standalone isn't really useful, in particular because a standalone
>>> version can't be tested with FATE.
>>
>> We don't care about testing it with FATE.
> 
> Who is "we"?

The majority of developers that waited years for someone to try and make this
thing usable and got tired of doing it.

> 
> On 27.11.2016 19:30, Clément Bœsch wrote:
>> unless the fundamental problems are addressed (ZERO internal API usage +
>> at least partial FATE coverage)

Re-read his email. That paragraph is about work aimed at fixing ffserver if
it were to remain in the tree. Which it wont.

> 
> Apparently Clément cares about FATE coverage and so do I.
> 
>> It appears to me that you still don't get that ffserver is being *dropped*.
>> It is, as discussed and announced, no longer part of the project.
> 
> The main reason for deciding to drop ffserver was that it uses internal APIs
> and thus blocks the library development. Should that get fixed, the decision
> needs to be re-evaluated.

No. It will not be reviewed just because one or two people came out of the
woodworks years after requests for help and maintainers flew by unheard and
right when the final decision to drop the whole thing was made.

If you were interested, you had *years* to make that known. That time is now
past.

> 
>>> Why do you think removing it has any benefit, if it doesn't come together
>>> with the library cleanup?
>>
>> The "benefit" would be finally following through with a project decision for
>> once, even if late, but at least not embarrassingly late.
> 
> Do I understand you correctly that there is no technical benefit?

It goes the other way around. There are no technical reasons that block the
removal something that has been decided must go. Or rather, must have been
gone weeks ago.

> The project decision can be followed by removing ffserver at the next major
> bump, if it still uses any of the internal APIs that will be made private 
> then.

No "if".

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Josh de Kock]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA512

On 2016/11/27 23:56, Rostislav Pehlivanov wrote:
> On 27 November 2016 at 23:20, James Almer 
> wrote:
> 
>> On 11/26/2016 6:00 PM, Rostislav Pehlivanov wrote:
>>> On 26 October 2016 at 23:43, Rostislav Pehlivanov
>>> 
>>> 
>>> Since a month has passed, reynaldo still hasn't responded, I
>>> think it's long overdue to push this patch. I've attached a new
>>> version of the patch which only removes the ffserver program
>>> (without touching anything else), which had been OK'd by James
>>> Almer.
>>> 
>>> I'll push this on Tuesday next week, the 29th of November.
>> 
>> Since at least Michael is still working towards getting it
>> working as a standalone program it is acceptable to wait until
>> 3.3 is close to being branched from master.
>> 
>> It *must not* make it into that release, but until then being in
>> the tree or out of it will not hurt, and will make Michael's and
>> Reynaldo's efforts easier.
>> 
> 
> Yeah, let's not rush things, I wouldn't mind waiting until then
> either. And I agree that ffserver shouldn't be in the main repo in
> the 3.3 release.

I don't think anyone is rushing things, this has waited long enough.
We have an ultimatum, which is to push this on the 29th of November.
And it is important to stick to that now.

- -- 
Josh

PGP fingerprint: A93A602D7A6D3C5388D792BCD052D40DDEF9703D
-BEGIN PGP SIGNATURE-
Comment: GPGTools - https://gpgtools.org
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=RUC9
-END PGP SIGNATURE-
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Rostislav Pehlivanov]

On 27 November 2016 at 23:20, James Almer  wrote:

> On 11/26/2016 6:00 PM, Rostislav Pehlivanov wrote:
> > On 26 October 2016 at 23:43, Rostislav Pehlivanov 
> > wrote:
> >
> >>
> >>
> >> On 26 October 2016 at 23:33, Andreas Cadhalpun <
> >> andreas.cadhal...@googlemail.com> wrote:
> >>
> >>> On 27.10.2016 00:16, Rostislav Pehlivanov wrote:
>  On 26 October 2016 at 22:48, James Almer  wrote:
> 
> > On 10/26/2016 6:19 PM, Rostislav Pehlivanov wrote:
> >> Also removes url_feof from libavformat.v which should have been
> >> removed long ago and changed the multipart jpeg boundary tag to
> >> ffmpeg rather than ffserver (it's arbitrary).
> >
> > [...]
> >
> >> diff --git a/libavformat/libavformat.v b/libavformat/libavformat.v
> >> index c961cd8..47d5ddc 100644
> >> --- a/libavformat/libavformat.v
> >> +++ b/libavformat/libavformat.v
> >> @@ -1,19 +1,6 @@
> >>  LIBAVFORMAT_MAJOR {
> >>  global:
> >>  av*;
> >> -#FIXME those are for ffserver
> >> -ff_inet_aton;
> >> -ff_socket_nonblock;
> >> -ff_rtsp_parse_line;
> >> -ff_rtp_get_local_rtp_port;
> >> -ff_rtp_get_local_rtcp_port;
> >> -ffio_open_dyn_packet_buf;
> >> -ffio_set_buf_size;
> >> -ffurl_close;
> >> -ffurl_open;
> >> -ffurl_write;
> >> -#those are deprecated, remove on next bump
> >> -url_feof;
> >>  local:
> >>  *;
> >>  };
> >
> > No, this can't be done until the next major bump. url_feof is even
> >>> guarded
> > by an scheduled FF_API define.
> >
> > The rest should be ok, but anything that implies a library ABI break
> >>> can't
> > be done just yet. Wait for other comments and/or testing. This
> removes
> >>> a
> > lot of things after all.
> >
>  Fixed version attached without removing url_feof (apparently that's
> >>> meant
>  to go in the next bump, wasn't forgotten)
> >>>
> >>> No, none of the symbols can be removed without a major bump. The simple
> >>> reason
> >>> is that a ffserver built from the 3.1 branch still has to work with the
> >>> libavformat from the 3.2 branch.
> >>>
> >>> Best regards,
> >>> Andreas
> >>>
> >>> ___
> >>> ffmpeg-devel mailing list
> >>> ffmpeg-devel@ffmpeg.org
> >>> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> >>>
> >>
> >> Thanks for telling me,
> >> Attached patch will only remove the ffserver program and documentation
> >> relating to it.
> >> The ffm demuxer/muxer will be removed with the next bump like url_feof.
> >>
> >
> >
> > Since a month has passed, reynaldo still hasn't responded, I think it's
> > long overdue to push this patch. I've attached a new version of the patch
> > which only removes the ffserver program (without touching anything else),
> > which had been OK'd by James Almer.
> >
> > I'll push this on Tuesday next week, the 29th of November.
>
> Since at least Michael is still working towards getting it working as a
> standalone program it is acceptable to wait until 3.3 is close to being
> branched from master.
>
> It *must not* make it into that release, but until then being in the tree
> or out of it will not hurt, and will make Michael's and Reynaldo's efforts
> easier.
>
> ___
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>

Yeah, let's not rush things, I wouldn't mind waiting until then either. And
I agree that ffserver shouldn't be in the main repo in the 3.3 release.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 28.11.2016 00:37, James Almer wrote:
> On 11/27/2016 8:23 PM, Andreas Cadhalpun wrote:
>> This can't happen while ffserver still uses internal API, and when it doesn't
>> making it standalone isn't really useful, in particular because a standalone
>> version can't be tested with FATE.
> 
> We don't care about testing it with FATE.

Who is "we"?

On 27.11.2016 19:30, Clément Bœsch wrote:
> unless the fundamental problems are addressed (ZERO internal API usage +
> at least partial FATE coverage)

Apparently Clément cares about FATE coverage and so do I.

> It appears to me that you still don't get that ffserver is being *dropped*.
> It is, as discussed and announced, no longer part of the project.

The main reason for deciding to drop ffserver was that it uses internal APIs
and thus blocks the library development. Should that get fixed, the decision
needs to be re-evaluated.

>> Why do you think removing it has any benefit, if it doesn't come together
>> with the library cleanup?
> 
> The "benefit" would be finally following through with a project decision for
> once, even if late, but at least not embarrassingly late.

Do I understand you correctly that there is no technical benefit?
The project decision can be followed by removing ffserver at the next major
bump, if it still uses any of the internal APIs that will be made private then.

Best regards,
Andreas

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/27/2016 8:23 PM, Andreas Cadhalpun wrote:
> On 28.11.2016 00:14, James Almer wrote:
>> On 11/27/2016 7:32 PM, Andreas Cadhalpun wrote:
>>> On 26.11.2016 22:00, Rostislav Pehlivanov wrote:
 Since a month has passed, reynaldo still hasn't responded, I think it's
 long overdue to push this patch. I've attached a new version of the patch
 which only removes the ffserver program (without touching anything else),
 which had been OK'd by James Almer.
>>>
>>> I don't think this patch is a good idea.
>>> Removing ffserver now would be a regression for anyone using it, but on the
>>> other hand would have no benefit, since the library cleanup cannot happen
>>> before the next major bump.
>>> Thus this patch has a net-negative effect.
>>
>> With the efforts of making it standalone and thus available to anyone that
>> wishes to use it with an up to date git master, this point becomes null.
> 
> This can't happen while ffserver still uses internal API, and when it doesn't
> making it standalone isn't really useful, in particular because a standalone
> version can't be tested with FATE.

We don't care about testing it with FATE. It appears to me that you still don't
get that ffserver is being *dropped*. It is, as discussed and announced, no
longer part of the project.
This standalone attempt is something Reynaldo himself said he wanted to do, and
thus he was given time with ffserver still in the tree to make it easier.

> 
 I'll push this on Tuesday next week, the 29th of November.
>>>
>>> Why can't you wait for the next major bump?
>>
>> I'm still ok with waiting until 3.3 is close to be branched out, to give
>> Reynaldo or Michael more time if they are still working on making it
>> standalone, but nothing longer than that.
> 
> Why do you think removing it has any benefit, if it doesn't come together
> with the library cleanup?

The "benefit" would be finally following through with a project decision for
once, even if late, but at least not embarrassingly late.

> 
> Best regards,
> Andreas
> 
> 
> ___
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> 

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 28.11.2016 00:14, James Almer wrote:
> On 11/27/2016 7:32 PM, Andreas Cadhalpun wrote:
>> On 26.11.2016 22:00, Rostislav Pehlivanov wrote:
>>> Since a month has passed, reynaldo still hasn't responded, I think it's
>>> long overdue to push this patch. I've attached a new version of the patch
>>> which only removes the ffserver program (without touching anything else),
>>> which had been OK'd by James Almer.
>>
>> I don't think this patch is a good idea.
>> Removing ffserver now would be a regression for anyone using it, but on the
>> other hand would have no benefit, since the library cleanup cannot happen
>> before the next major bump.
>> Thus this patch has a net-negative effect.
> 
> With the efforts of making it standalone and thus available to anyone that
> wishes to use it with an up to date git master, this point becomes null.

This can't happen while ffserver still uses internal API, and when it doesn't
making it standalone isn't really useful, in particular because a standalone
version can't be tested with FATE.

>>> I'll push this on Tuesday next week, the 29th of November.
>>
>> Why can't you wait for the next major bump?
> 
> I'm still ok with waiting until 3.3 is close to be branched out, to give
> Reynaldo or Michael more time if they are still working on making it
> standalone, but nothing longer than that.

Why do you think removing it has any benefit, if it doesn't come together
with the library cleanup?

Best regards,
Andreas


___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/26/2016 6:00 PM, Rostislav Pehlivanov wrote:
> On 26 October 2016 at 23:43, Rostislav Pehlivanov 
> wrote:
> 
>>
>>
>> On 26 October 2016 at 23:33, Andreas Cadhalpun <
>> andreas.cadhal...@googlemail.com> wrote:
>>
>>> On 27.10.2016 00:16, Rostislav Pehlivanov wrote:
 On 26 October 2016 at 22:48, James Almer  wrote:

> On 10/26/2016 6:19 PM, Rostislav Pehlivanov wrote:
>> Also removes url_feof from libavformat.v which should have been
>> removed long ago and changed the multipart jpeg boundary tag to
>> ffmpeg rather than ffserver (it's arbitrary).
>
> [...]
>
>> diff --git a/libavformat/libavformat.v b/libavformat/libavformat.v
>> index c961cd8..47d5ddc 100644
>> --- a/libavformat/libavformat.v
>> +++ b/libavformat/libavformat.v
>> @@ -1,19 +1,6 @@
>>  LIBAVFORMAT_MAJOR {
>>  global:
>>  av*;
>> -#FIXME those are for ffserver
>> -ff_inet_aton;
>> -ff_socket_nonblock;
>> -ff_rtsp_parse_line;
>> -ff_rtp_get_local_rtp_port;
>> -ff_rtp_get_local_rtcp_port;
>> -ffio_open_dyn_packet_buf;
>> -ffio_set_buf_size;
>> -ffurl_close;
>> -ffurl_open;
>> -ffurl_write;
>> -#those are deprecated, remove on next bump
>> -url_feof;
>>  local:
>>  *;
>>  };
>
> No, this can't be done until the next major bump. url_feof is even
>>> guarded
> by an scheduled FF_API define.
>
> The rest should be ok, but anything that implies a library ABI break
>>> can't
> be done just yet. Wait for other comments and/or testing. This removes
>>> a
> lot of things after all.
>
 Fixed version attached without removing url_feof (apparently that's
>>> meant
 to go in the next bump, wasn't forgotten)
>>>
>>> No, none of the symbols can be removed without a major bump. The simple
>>> reason
>>> is that a ffserver built from the 3.1 branch still has to work with the
>>> libavformat from the 3.2 branch.
>>>
>>> Best regards,
>>> Andreas
>>>
>>> ___
>>> ffmpeg-devel mailing list
>>> ffmpeg-devel@ffmpeg.org
>>> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>>>
>>
>> Thanks for telling me,
>> Attached patch will only remove the ffserver program and documentation
>> relating to it.
>> The ffm demuxer/muxer will be removed with the next bump like url_feof.
>>
> 
> 
> Since a month has passed, reynaldo still hasn't responded, I think it's
> long overdue to push this patch. I've attached a new version of the patch
> which only removes the ffserver program (without touching anything else),
> which had been OK'd by James Almer.
> 
> I'll push this on Tuesday next week, the 29th of November.

Since at least Michael is still working towards getting it working as a
standalone program it is acceptable to wait until 3.3 is close to being
branched from master.

It *must not* make it into that release, but until then being in the tree
or out of it will not hurt, and will make Michael's and Reynaldo's efforts
easier.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/27/2016 7:32 PM, Andreas Cadhalpun wrote:
> On 26.11.2016 22:00, Rostislav Pehlivanov wrote:
>> Since a month has passed, reynaldo still hasn't responded, I think it's
>> long overdue to push this patch. I've attached a new version of the patch
>> which only removes the ffserver program (without touching anything else),
>> which had been OK'd by James Almer.
> 
> I don't think this patch is a good idea.
> Removing ffserver now would be a regression for anyone using it, but on the
> other hand would have no benefit, since the library cleanup cannot happen
> before the next major bump.
> Thus this patch has a net-negative effect.

With the efforts of making it standalone and thus available to anyone that
wishes to use it with an up to date git master, this point becomes null.

> 
>> I'll push this on Tuesday next week, the 29th of November.
> 
> Why can't you wait for the next major bump?

I'm still ok with waiting until 3.3 is close to be branched out, to give
Reynaldo or Michael more time if they are still working on making it
standalone, but nothing longer than that.
And if we happen to bump major with it, then your request would have been
granted.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 26.11.2016 22:00, Rostislav Pehlivanov wrote:
> Since a month has passed, reynaldo still hasn't responded, I think it's
> long overdue to push this patch. I've attached a new version of the patch
> which only removes the ffserver program (without touching anything else),
> which had been OK'd by James Almer.

I don't think this patch is a good idea.
Removing ffserver now would be a regression for anyone using it, but on the
other hand would have no benefit, since the library cleanup cannot happen
before the next major bump.
Thus this patch has a net-negative effect.

> I'll push this on Tuesday next week, the 29th of November.

Why can't you wait for the next major bump?

Best regards,
Andreas
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 27.11.2016 19:36, Michael Niedermayer wrote:
> If somethig critical remains, please someone list exactly what that is.
> so reynaldo, i or others can go over the list and look into the issues

As far as I'm aware the following points are most problematic:
 * ffserver uses internal functions, see FIXME in libavformat/libavformat.v
 * ffserver uses the ffm format, which is incompatible with codecpar:
- It sets AVOptions of the stream->codec.
- It sets fields like qmin, which is not possible with codecpar.
 * There is no working FATE test, so developers not using ffserver
   can't really work on it for fear of breaking stuff.

Best regards,
Andreas
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Ronald S. Bultje]

Hi,

On Sun, Nov 27, 2016 at 1:46 PM, Michael Niedermayer  wrote:

> On Sun, Nov 27, 2016 at 07:30:24PM +0100, Clément Bœsch wrote:
> > On Sun, Nov 27, 2016 at 06:49:35PM +0100, Michael Niedermayer wrote:
> > > On Sun, Nov 27, 2016 at 04:57:36PM +0100, Clément Bœsch wrote:
> > > > On Sun, Nov 27, 2016 at 04:25:18PM +0100, Michael Niedermayer wrote:
> > > > [...]
> > > > > ffserver had 14 commits to it in about the last month
> > > >
> > > > That's also pretty close to the number of commits in the last years.
> > > >
> > > > Good will in the last weeks is not enough of a technical
> > > > merit/justification to prevent the removal of junk code scheduled for
> > > > deletion for a long time now.
> > >
> > > why do you call it junk ?
> >
> > because it's highly dependent on internal stuff, very limited, completely
> > untested, unmaintained for several years but still contains a ton of
> code.
> >
> > > and the sheduling for deletion was conditional on it not being fixed
> > > IIRC.
> > > Why the hurry to remove it while people work on fixing it ?
> > > its not blocking anything ATM AFAIK
> >
> > There is no hurry, but piling up a bunch patches to fix small things just
> > to use it as an argument to say "hey look now it's maintained" in order
> to
> > save it from being killed is really annoying. The people interested in it
> > had years to act.
> >
>
> > You can fix a ton of little things in it, but unless the fundamental
> > problems are addressed (ZERO internal API usage + at least partial FATE
> > coverage) it's pointless
>
> of course the goal is ZERO internal API usage + at least partial FATE
> coverage.
> Well in fact the lack of fate tests have been the primary reason
> why i didnt fix some of the API issues years ago. I felt uneasy
> changing it without regression tests
>
>
> > and will be seen as yet another case of "KEEP
> > EVERYTHING FOREVER" toxic mentality.
>
> The opposit is toxic too


I'm perfectly fine with keeping the code, just not in the ffmpeg tree.
Please move it to its own tree.

Everybody wants it out. Please follow majority.

Thanks,
Ronald
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Clément Bœsch]

On Sun, Nov 27, 2016 at 07:46:34PM +0100, Michael Niedermayer wrote:
[...]
> > You can fix a ton of little things in it, but unless the fundamental
> > problems are addressed (ZERO internal API usage + at least partial FATE
> > coverage) it's pointless
> 
> of course the goal is ZERO internal API usage + at least partial FATE
> coverage.

If you think you can manage this til the scheduled removal date, then
that's perfectly fine with me.

When is the due removal already? Next version?

Good luck.

-- 
Clément B.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Sun, Nov 27, 2016 at 07:30:24PM +0100, Clément Bœsch wrote:
> On Sun, Nov 27, 2016 at 06:49:35PM +0100, Michael Niedermayer wrote:
> > On Sun, Nov 27, 2016 at 04:57:36PM +0100, Clément Bœsch wrote:
> > > On Sun, Nov 27, 2016 at 04:25:18PM +0100, Michael Niedermayer wrote:
> > > [...]
> > > > ffserver had 14 commits to it in about the last month
> > > 
> > > That's also pretty close to the number of commits in the last years.
> > > 
> > > Good will in the last weeks is not enough of a technical
> > > merit/justification to prevent the removal of junk code scheduled for
> > > deletion for a long time now.
> > 
> > why do you call it junk ?
> 
> because it's highly dependent on internal stuff, very limited, completely
> untested, unmaintained for several years but still contains a ton of code.
> 
> > and the sheduling for deletion was conditional on it not being fixed
> > IIRC.
> > Why the hurry to remove it while people work on fixing it ?
> > its not blocking anything ATM AFAIK
> 
> There is no hurry, but piling up a bunch patches to fix small things just
> to use it as an argument to say "hey look now it's maintained" in order to
> save it from being killed is really annoying. The people interested in it
> had years to act.
> 

> You can fix a ton of little things in it, but unless the fundamental
> problems are addressed (ZERO internal API usage + at least partial FATE
> coverage) it's pointless

of course the goal is ZERO internal API usage + at least partial FATE
coverage.
Well in fact the lack of fate tests have been the primary reason
why i didnt fix some of the API issues years ago. I felt uneasy
changing it without regression tests


> and will be seen as yet another case of "KEEP
> EVERYTHING FOREVER" toxic mentality.

The opposit is toxic too

Thanks

[...]
-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

it is not once nor twice but times without number that the same ideas make
their appearance in the world. -- Aristotle


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Sun, Nov 27, 2016 at 02:05:44PM +0100, Michael Niedermayer wrote:
> On Sat, Nov 26, 2016 at 09:00:41PM +, Rostislav Pehlivanov wrote:
> > On 26 October 2016 at 23:43, Rostislav Pehlivanov 
> > wrote:
> > 
> > >
> > >
> > > On 26 October 2016 at 23:33, Andreas Cadhalpun <
> > > andreas.cadhal...@googlemail.com> wrote:
> > >
> > >> On 27.10.2016 00:16, Rostislav Pehlivanov wrote:
> > >> > On 26 October 2016 at 22:48, James Almer  wrote:
> > >> >
> > >> >> On 10/26/2016 6:19 PM, Rostislav Pehlivanov wrote:
> > >> >>> Also removes url_feof from libavformat.v which should have been
> > >> >>> removed long ago and changed the multipart jpeg boundary tag to
> > >> >>> ffmpeg rather than ffserver (it's arbitrary).
> > >> >>
> > >> >> [...]
> > >> >>
> > >> >>> diff --git a/libavformat/libavformat.v b/libavformat/libavformat.v
> > >> >>> index c961cd8..47d5ddc 100644
> > >> >>> --- a/libavformat/libavformat.v
> > >> >>> +++ b/libavformat/libavformat.v
> > >> >>> @@ -1,19 +1,6 @@
> > >> >>>  LIBAVFORMAT_MAJOR {
> > >> >>>  global:
> > >> >>>  av*;
> > >> >>> -#FIXME those are for ffserver
> > >> >>> -ff_inet_aton;
> > >> >>> -ff_socket_nonblock;
> > >> >>> -ff_rtsp_parse_line;
> > >> >>> -ff_rtp_get_local_rtp_port;
> > >> >>> -ff_rtp_get_local_rtcp_port;
> > >> >>> -ffio_open_dyn_packet_buf;
> > >> >>> -ffio_set_buf_size;
> > >> >>> -ffurl_close;
> > >> >>> -ffurl_open;
> > >> >>> -ffurl_write;
> > >> >>> -#those are deprecated, remove on next bump
> > >> >>> -url_feof;
> > >> >>>  local:
> > >> >>>  *;
> > >> >>>  };
> > >> >>
> > >> >> No, this can't be done until the next major bump. url_feof is even
> > >> guarded
> > >> >> by an scheduled FF_API define.
> > >> >>
> > >> >> The rest should be ok, but anything that implies a library ABI break
> > >> can't
> > >> >> be done just yet. Wait for other comments and/or testing. This removes
> > >> a
> > >> >> lot of things after all.
> > >> >>
> > >> > Fixed version attached without removing url_feof (apparently that's
> > >> meant
> > >> > to go in the next bump, wasn't forgotten)
> > >>
> > >> No, none of the symbols can be removed without a major bump. The simple
> > >> reason
> > >> is that a ffserver built from the 3.1 branch still has to work with the
> > >> libavformat from the 3.2 branch.
> > >>
> > >> Best regards,
> > >> Andreas
> > >>
> > >> ___
> > >> ffmpeg-devel mailing list
> > >> ffmpeg-devel@ffmpeg.org
> > >> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> > >>
> > >
> > > Thanks for telling me,
> > > Attached patch will only remove the ffserver program and documentation
> > > relating to it.
> > > The ffm demuxer/muxer will be removed with the next bump like url_feof.
> > >
> > 
> > 
> > Since a month has passed, reynaldo still hasn't responded, I think it's
> > long overdue to push this patch. I've attached a new version of the patch
> > which only removes the ffserver program (without touching anything else),
> > which had been OK'd by James Almer.
> 
> IIUC reynaldo is very busy.
> 
> have you looked at ffserver ? I think helping reynaldo and updating
> ffserver so it doesnt violate the API would make alot more sense for
> FFmpeg than removing it. Its a long existing feature and tool.
> And i dont think its really that hard to update it, iam not the
> author of it iam not an expert on that code so i might be wrong or
> miss things but the big mess in it seems just moving parameters around
> between demuxers and muxers. and a few uses of network and rtp
> functions which should be used through some public API or made
> available through a new one everyone likes ...

the 4 patches i just posted fix all the AVStream issues i am aware of.
codec/codecpar issues seem no major ones left, just some close and copy
that would probably go away if the field is removed.

I think this solves the bulk of the problem that i was aware of

If somethig critical remains, please someone list exactly what that is.
so reynaldo, i or others can go over the list and look into the issues

thanks

[...]

-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

The real ebay dictionary, page 2
"100% positive feedback" - "All either got their money back or didnt complain"
"Best seller ever, very honest" - "Seller refunded buyer after failed scam"


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Clément Bœsch]

On Sun, Nov 27, 2016 at 06:49:35PM +0100, Michael Niedermayer wrote:
> On Sun, Nov 27, 2016 at 04:57:36PM +0100, Clément Bœsch wrote:
> > On Sun, Nov 27, 2016 at 04:25:18PM +0100, Michael Niedermayer wrote:
> > [...]
> > > ffserver had 14 commits to it in about the last month
> > 
> > That's also pretty close to the number of commits in the last years.
> > 
> > Good will in the last weeks is not enough of a technical
> > merit/justification to prevent the removal of junk code scheduled for
> > deletion for a long time now.
> 
> why do you call it junk ?

because it's highly dependent on internal stuff, very limited, completely
untested, unmaintained for several years but still contains a ton of code.

> and the sheduling for deletion was conditional on it not being fixed
> IIRC.
> Why the hurry to remove it while people work on fixing it ?
> its not blocking anything ATM AFAIK

There is no hurry, but piling up a bunch patches to fix small things just
to use it as an argument to say "hey look now it's maintained" in order to
save it from being killed is really annoying. The people interested in it
had years to act.

You can fix a ton of little things in it, but unless the fundamental
problems are addressed (ZERO internal API usage + at least partial FATE
coverage) it's pointless and will be seen as yet another case of "KEEP
EVERYTHING FOREVER" toxic mentality.

-- 
Clément B.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Sun, Nov 27, 2016 at 04:57:36PM +0100, Clément Bœsch wrote:
> On Sun, Nov 27, 2016 at 04:25:18PM +0100, Michael Niedermayer wrote:
> [...]
> > ffserver had 14 commits to it in about the last month
> 
> That's also pretty close to the number of commits in the last years.
> 
> Good will in the last weeks is not enough of a technical
> merit/justification to prevent the removal of junk code scheduled for
> deletion for a long time now.

why do you call it junk ?
and the sheduling for deletion was conditional on it not being fixed
IIRC.
Why the hurry to remove it while people work on fixing it ?
its not blocking anything ATM AFAIK

and why is this escalating in such a strange way, theres no need to
become aggressive.
Does the project really have too many developers?
If thats the case can you point me to someone to take over coverity
testing ? I cant find one ...

also reynaldo, has done alot for the project, not just through volunteer
work, but also by finding one of the very few sponsors who funded
students working in FFmpeg under the OPW as well as helping with GSoC
and OPW admin work.

For all i know reynaldo is not being paid to work on FFmpeg.
shouldnt you be a little nicer ...

[...]
-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Freedom in capitalist society always remains about the same as it was in
ancient Greek republics: Freedom for slave owners. -- Vladimir Lenin


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Clément Bœsch]

On Sun, Nov 27, 2016 at 04:25:18PM +0100, Michael Niedermayer wrote:
[...]
> ffserver had 14 commits to it in about the last month

That's also pretty close to the number of commits in the last years.

Good will in the last weeks is not enough of a technical
merit/justification to prevent the removal of junk code scheduled for
deletion for a long time now.

-- 
Clément B.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Sun, Nov 27, 2016 at 02:33:54PM +0100, wm4 wrote:
> On Sun, 27 Nov 2016 14:05:44 +0100
> Michael Niedermayer  wrote:
> 
> > On Sat, Nov 26, 2016 at 09:00:41PM +, Rostislav Pehlivanov wrote:
> > > On 26 October 2016 at 23:43, Rostislav Pehlivanov 
> > > wrote:
> > >   
> > > >
> > > >
> > > > On 26 October 2016 at 23:33, Andreas Cadhalpun <  
> > > > andreas.cadhal...@googlemail.com> wrote:  
> > > >  
> > > >> On 27.10.2016 00:16, Rostislav Pehlivanov wrote:  
> > > >> > On 26 October 2016 at 22:48, James Almer  wrote:
> > > >> >  
> > > >> >> On 10/26/2016 6:19 PM, Rostislav Pehlivanov wrote:  
> > > >> >>> Also removes url_feof from libavformat.v which should have been
> > > >> >>> removed long ago and changed the multipart jpeg boundary tag to
> > > >> >>> ffmpeg rather than ffserver (it's arbitrary).  
> > > >> >>
> > > >> >> [...]
> > > >> >>  
> > > >> >>> diff --git a/libavformat/libavformat.v b/libavformat/libavformat.v
> > > >> >>> index c961cd8..47d5ddc 100644
> > > >> >>> --- a/libavformat/libavformat.v
> > > >> >>> +++ b/libavformat/libavformat.v
> > > >> >>> @@ -1,19 +1,6 @@
> > > >> >>>  LIBAVFORMAT_MAJOR {
> > > >> >>>  global:
> > > >> >>>  av*;
> > > >> >>> -#FIXME those are for ffserver
> > > >> >>> -ff_inet_aton;
> > > >> >>> -ff_socket_nonblock;
> > > >> >>> -ff_rtsp_parse_line;
> > > >> >>> -ff_rtp_get_local_rtp_port;
> > > >> >>> -ff_rtp_get_local_rtcp_port;
> > > >> >>> -ffio_open_dyn_packet_buf;
> > > >> >>> -ffio_set_buf_size;
> > > >> >>> -ffurl_close;
> > > >> >>> -ffurl_open;
> > > >> >>> -ffurl_write;
> > > >> >>> -#those are deprecated, remove on next bump
> > > >> >>> -url_feof;
> > > >> >>>  local:
> > > >> >>>  *;
> > > >> >>>  };  
> > > >> >>
> > > >> >> No, this can't be done until the next major bump. url_feof is even  
> > > >> guarded  
> > > >> >> by an scheduled FF_API define.
> > > >> >>
> > > >> >> The rest should be ok, but anything that implies a library ABI 
> > > >> >> break  
> > > >> can't  
> > > >> >> be done just yet. Wait for other comments and/or testing. This 
> > > >> >> removes  
> > > >> a  
> > > >> >> lot of things after all.
> > > >> >>  
> > > >> > Fixed version attached without removing url_feof (apparently that's  
> > > >> meant  
> > > >> > to go in the next bump, wasn't forgotten)  
> > > >>
> > > >> No, none of the symbols can be removed without a major bump. The simple
> > > >> reason
> > > >> is that a ffserver built from the 3.1 branch still has to work with the
> > > >> libavformat from the 3.2 branch.
> > > >>
> > > >> Best regards,
> > > >> Andreas
> > > >>
> > > >> ___
> > > >> ffmpeg-devel mailing list
> > > >> ffmpeg-devel@ffmpeg.org
> > > >> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> > > >>  
> > > >
> > > > Thanks for telling me,
> > > > Attached patch will only remove the ffserver program and documentation
> > > > relating to it.
> > > > The ffm demuxer/muxer will be removed with the next bump like url_feof.
> > > >  
> > > 
> > > 
> > > Since a month has passed, reynaldo still hasn't responded, I think it's
> > > long overdue to push this patch. I've attached a new version of the patch
> > > which only removes the ffserver program (without touching anything else),
> > > which had been OK'd by James Almer.  
> > 
> > IIUC reynaldo is very busy.
> > 
> > have you looked at ffserver ? I think helping reynaldo and updating
> > ffserver so it doesnt violate the API would make alot more sense for
> > FFmpeg than removing it. Its a long existing feature and tool.
> 
> Where does this come from? It seems never ever it happened that someone
> wanted to actually update and maintain ffserver. That's why we want to
> remove it. The reason for removal isn't going to change apparently.
> Basically nobody wants to touch it, and those who tried apparently
> failed. It's over.

ffserver had 14 commits to it in about the last month, thats far from
unmaintained, especially considering the maintainer has little time
currently

[...]
-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

During times of universal deceit, telling the truth becomes a
revolutionary act. -- George Orwell


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[wm4]

On Sun, 27 Nov 2016 14:05:44 +0100
Michael Niedermayer  wrote:

> On Sat, Nov 26, 2016 at 09:00:41PM +, Rostislav Pehlivanov wrote:
> > On 26 October 2016 at 23:43, Rostislav Pehlivanov 
> > wrote:
> >   
> > >
> > >
> > > On 26 October 2016 at 23:33, Andreas Cadhalpun <  
> > > andreas.cadhal...@googlemail.com> wrote:  
> > >  
> > >> On 27.10.2016 00:16, Rostislav Pehlivanov wrote:  
> > >> > On 26 October 2016 at 22:48, James Almer  wrote:
> > >> >  
> > >> >> On 10/26/2016 6:19 PM, Rostislav Pehlivanov wrote:  
> > >> >>> Also removes url_feof from libavformat.v which should have been
> > >> >>> removed long ago and changed the multipart jpeg boundary tag to
> > >> >>> ffmpeg rather than ffserver (it's arbitrary).  
> > >> >>
> > >> >> [...]
> > >> >>  
> > >> >>> diff --git a/libavformat/libavformat.v b/libavformat/libavformat.v
> > >> >>> index c961cd8..47d5ddc 100644
> > >> >>> --- a/libavformat/libavformat.v
> > >> >>> +++ b/libavformat/libavformat.v
> > >> >>> @@ -1,19 +1,6 @@
> > >> >>>  LIBAVFORMAT_MAJOR {
> > >> >>>  global:
> > >> >>>  av*;
> > >> >>> -#FIXME those are for ffserver
> > >> >>> -ff_inet_aton;
> > >> >>> -ff_socket_nonblock;
> > >> >>> -ff_rtsp_parse_line;
> > >> >>> -ff_rtp_get_local_rtp_port;
> > >> >>> -ff_rtp_get_local_rtcp_port;
> > >> >>> -ffio_open_dyn_packet_buf;
> > >> >>> -ffio_set_buf_size;
> > >> >>> -ffurl_close;
> > >> >>> -ffurl_open;
> > >> >>> -ffurl_write;
> > >> >>> -#those are deprecated, remove on next bump
> > >> >>> -url_feof;
> > >> >>>  local:
> > >> >>>  *;
> > >> >>>  };  
> > >> >>
> > >> >> No, this can't be done until the next major bump. url_feof is even  
> > >> guarded  
> > >> >> by an scheduled FF_API define.
> > >> >>
> > >> >> The rest should be ok, but anything that implies a library ABI break  
> > >> can't  
> > >> >> be done just yet. Wait for other comments and/or testing. This 
> > >> >> removes  
> > >> a  
> > >> >> lot of things after all.
> > >> >>  
> > >> > Fixed version attached without removing url_feof (apparently that's  
> > >> meant  
> > >> > to go in the next bump, wasn't forgotten)  
> > >>
> > >> No, none of the symbols can be removed without a major bump. The simple
> > >> reason
> > >> is that a ffserver built from the 3.1 branch still has to work with the
> > >> libavformat from the 3.2 branch.
> > >>
> > >> Best regards,
> > >> Andreas
> > >>
> > >> ___
> > >> ffmpeg-devel mailing list
> > >> ffmpeg-devel@ffmpeg.org
> > >> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> > >>  
> > >
> > > Thanks for telling me,
> > > Attached patch will only remove the ffserver program and documentation
> > > relating to it.
> > > The ffm demuxer/muxer will be removed with the next bump like url_feof.
> > >  
> > 
> > 
> > Since a month has passed, reynaldo still hasn't responded, I think it's
> > long overdue to push this patch. I've attached a new version of the patch
> > which only removes the ffserver program (without touching anything else),
> > which had been OK'd by James Almer.  
> 
> IIUC reynaldo is very busy.
> 
> have you looked at ffserver ? I think helping reynaldo and updating
> ffserver so it doesnt violate the API would make alot more sense for
> FFmpeg than removing it. Its a long existing feature and tool.

Where does this come from? It seems never ever it happened that someone
wanted to actually update and maintain ffserver. That's why we want to
remove it. The reason for removal isn't going to change apparently.
Basically nobody wants to touch it, and those who tried apparently
failed. It's over.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Sat, Nov 26, 2016 at 09:00:41PM +, Rostislav Pehlivanov wrote:
> On 26 October 2016 at 23:43, Rostislav Pehlivanov 
> wrote:
> 
> >
> >
> > On 26 October 2016 at 23:33, Andreas Cadhalpun <
> > andreas.cadhal...@googlemail.com> wrote:
> >
> >> On 27.10.2016 00:16, Rostislav Pehlivanov wrote:
> >> > On 26 October 2016 at 22:48, James Almer  wrote:
> >> >
> >> >> On 10/26/2016 6:19 PM, Rostislav Pehlivanov wrote:
> >> >>> Also removes url_feof from libavformat.v which should have been
> >> >>> removed long ago and changed the multipart jpeg boundary tag to
> >> >>> ffmpeg rather than ffserver (it's arbitrary).
> >> >>
> >> >> [...]
> >> >>
> >> >>> diff --git a/libavformat/libavformat.v b/libavformat/libavformat.v
> >> >>> index c961cd8..47d5ddc 100644
> >> >>> --- a/libavformat/libavformat.v
> >> >>> +++ b/libavformat/libavformat.v
> >> >>> @@ -1,19 +1,6 @@
> >> >>>  LIBAVFORMAT_MAJOR {
> >> >>>  global:
> >> >>>  av*;
> >> >>> -#FIXME those are for ffserver
> >> >>> -ff_inet_aton;
> >> >>> -ff_socket_nonblock;
> >> >>> -ff_rtsp_parse_line;
> >> >>> -ff_rtp_get_local_rtp_port;
> >> >>> -ff_rtp_get_local_rtcp_port;
> >> >>> -ffio_open_dyn_packet_buf;
> >> >>> -ffio_set_buf_size;
> >> >>> -ffurl_close;
> >> >>> -ffurl_open;
> >> >>> -ffurl_write;
> >> >>> -#those are deprecated, remove on next bump
> >> >>> -url_feof;
> >> >>>  local:
> >> >>>  *;
> >> >>>  };
> >> >>
> >> >> No, this can't be done until the next major bump. url_feof is even
> >> guarded
> >> >> by an scheduled FF_API define.
> >> >>
> >> >> The rest should be ok, but anything that implies a library ABI break
> >> can't
> >> >> be done just yet. Wait for other comments and/or testing. This removes
> >> a
> >> >> lot of things after all.
> >> >>
> >> > Fixed version attached without removing url_feof (apparently that's
> >> meant
> >> > to go in the next bump, wasn't forgotten)
> >>
> >> No, none of the symbols can be removed without a major bump. The simple
> >> reason
> >> is that a ffserver built from the 3.1 branch still has to work with the
> >> libavformat from the 3.2 branch.
> >>
> >> Best regards,
> >> Andreas
> >>
> >> ___
> >> ffmpeg-devel mailing list
> >> ffmpeg-devel@ffmpeg.org
> >> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> >>
> >
> > Thanks for telling me,
> > Attached patch will only remove the ffserver program and documentation
> > relating to it.
> > The ffm demuxer/muxer will be removed with the next bump like url_feof.
> >
> 
> 
> Since a month has passed, reynaldo still hasn't responded, I think it's
> long overdue to push this patch. I've attached a new version of the patch
> which only removes the ffserver program (without touching anything else),
> which had been OK'd by James Almer.

IIUC reynaldo is very busy.

have you looked at ffserver ? I think helping reynaldo and updating
ffserver so it doesnt violate the API would make alot more sense for
FFmpeg than removing it. Its a long existing feature and tool.
And i dont think its really that hard to update it, iam not the
author of it iam not an expert on that code so i might be wrong or
miss things but the big mess in it seems just moving parameters around
between demuxers and muxers. and a few uses of network and rtp
functions which should be used through some public API or made
available through a new one everyone likes ...

Dont you think its better if some people work together for a few days
to update ffserver and have a cool feature more in FFmpeg, a cleaned
up ffserver with some more modernized features. And the people liking
ffserver be happy and thankfull
vs.
Discussions and eventually a vote about removing ffserver and then
whatever the outcome some people "loosing" and being unhappy

I know i should update ffserver myself if i care so much, but my
todo is getting longer and longer (from new coverity issues that need
checking to all the little things here and there like updating the git
sync script earlier today so it deletes a mistakly added branch and
doesnt spread it to one of our git mirrors at least ...)

Above said, this mail is not intended to block or veto anyones patch

[...]

-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Dictatorship: All citizens are under surveillance, all their steps and
actions recorded, for the politicians to enforce control.
Democracy: All politicians are under surveillance, all their steps and
actions recorded, for the citizens to enforce control.


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 25.11.2016 22:22, wm4 wrote:
> On Fri, 25 Nov 2016 21:13:54 +0100
> Andreas Cadhalpun  wrote:
> 
>> On 25.11.2016 10:36, wm4 wrote:
>>> I'd like to add that removing the ffserver support from the libs, and
>>> all deprecated stuff it depends on, will make it actually easier to
>>> readd ffserver in a cleaner way. It reduces the risk that the new
>>> ffserver accidentally depends on mechanisms that were supposed to go
>>> away.  
>>
>> However, that can't be done before the next major bump.
> 
> I contest this too. The ffm muxer/demuxer does not affect ABI and has
> no use outside of ffserver.

Removing it would break ffserver completely, so it shouldn't be done
without a major bump, in particular if ffserver gets removed.
Also you don't know if it has uses outside of ffserver.

>>> Also, the promise was that developers would pick up ffserver again and
>>> update it. This hasn't really happened - instead it weirdly looks like
>>> the ffserver faction is stalling for time. Maybe the interest to keep
>>> it alive isn't that high after all.  
>>
>> To me it looks more like they simply lack the time to work on it.
> 
> Plenty of time was granted. Several years more than Libav did.

As far as I recall ffserver wasn't proposed for removal at the last
major bump.

Best regards,
Andreas

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[wm4]

On Fri, 25 Nov 2016 21:13:54 +0100
Andreas Cadhalpun  wrote:

> On 25.11.2016 10:36, wm4 wrote:
> > I'd like to add that removing the ffserver support from the libs, and
> > all deprecated stuff it depends on, will make it actually easier to
> > readd ffserver in a cleaner way. It reduces the risk that the new
> > ffserver accidentally depends on mechanisms that were supposed to go
> > away.  
> 
> However, that can't be done before the next major bump.

I contest this too. The ffm muxer/demuxer does not affect ABI and has
no use outside of ffserver.

> > Also, the promise was that developers would pick up ffserver again and
> > update it. This hasn't really happened - instead it weirdly looks like
> > the ffserver faction is stalling for time. Maybe the interest to keep
> > it alive isn't that high after all.  
> 
> To me it looks more like they simply lack the time to work on it.

Plenty of time was granted. Several years more than Libav did.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 25.11.2016 10:36, wm4 wrote:
> I'd like to add that removing the ffserver support from the libs, and
> all deprecated stuff it depends on, will make it actually easier to
> readd ffserver in a cleaner way. It reduces the risk that the new
> ffserver accidentally depends on mechanisms that were supposed to go
> away.

However, that can't be done before the next major bump.

> Also, the promise was that developers would pick up ffserver again and
> update it. This hasn't really happened - instead it weirdly looks like
> the ffserver faction is stalling for time. Maybe the interest to keep
> it alive isn't that high after all.

To me it looks more like they simply lack the time to work on it.

Best regards,
Andreas
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[wm4]

On Thu, 24 Nov 2016 21:40:14 -0300
James Almer  wrote:

> On 11/24/2016 8:48 PM, Andreas Cadhalpun wrote:
> > On 24.11.2016 10:53, Josh de Kock wrote:  
> >> There's no benefit in waiting for this to occur before ffserver is
> >> removed as it will always be in git history. Removing ffserver from
> >> master may even speed up the "resurrection of ffserver". I think this
> >> shouldn't be delayed any further.  
> > 
> > Similarly there is no benefit in removing ffserver before the next major
> > bump, because the libraries have to stay backwards compatible until then,
> > anyway.
> > However, at least Reynaldo and Michael seem to be still working on ffserver,
> > so gratuitously removing it now would be rather unfriendly.  
> 
> A sudden surge in interest product of the decision to kill it and *long after*
> calls for developers to prevent it from bitrotting in the first place that
> pretends to be blocking is not proper and far from nice.
> Reynaldo agreed to move ffserver to a separate repo in time for the next
> release as he stated in an email, and then work on making it not depend on
> the ffm de/muxers since those have to go with the next major bump, or with
> AVStream->codec at the latest. Pretending to turn that effort into an
> argument against the removal is a crappy thing to do.
> 
> The choice and the announcement were made months ago, and those who cared
> participated in the discussions and the relevant patches. Backing out at
> the very last second is extremely unprofessional and will kill any future
> trust people will have in decisions made and announced in public channels.
> 
> It's already embarrassing that ffserver made it to 3.2. Lets try to keep
> the image that this is a serious project.

I'd like to add that removing the ffserver support from the libs, and
all deprecated stuff it depends on, will make it actually easier to
readd ffserver in a cleaner way. It reduces the risk that the new
ffserver accidentally depends on mechanisms that were supposed to go
away.

Also, the promise was that developers would pick up ffserver again and
update it. This hasn't really happened - instead it weirdly looks like
the ffserver faction is stalling for time. Maybe the interest to keep
it alive isn't that high after all.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 11/24/2016 8:48 PM, Andreas Cadhalpun wrote:
> On 24.11.2016 10:53, Josh de Kock wrote:
>> There's no benefit in waiting for this to occur before ffserver is
>> removed as it will always be in git history. Removing ffserver from
>> master may even speed up the "resurrection of ffserver". I think this
>> shouldn't be delayed any further.
> 
> Similarly there is no benefit in removing ffserver before the next major
> bump, because the libraries have to stay backwards compatible until then,
> anyway.
> However, at least Reynaldo and Michael seem to be still working on ffserver,
> so gratuitously removing it now would be rather unfriendly.

A sudden surge in interest product of the decision to kill it and *long after*
calls for developers to prevent it from bitrotting in the first place that
pretends to be blocking is not proper and far from nice.
Reynaldo agreed to move ffserver to a separate repo in time for the next
release as he stated in an email, and then work on making it not depend on
the ffm de/muxers since those have to go with the next major bump, or with
AVStream->codec at the latest. Pretending to turn that effort into an
argument against the removal is a crappy thing to do.

The choice and the announcement were made months ago, and those who cared
participated in the discussions and the relevant patches. Backing out at
the very last second is extremely unprofessional and will kill any future
trust people will have in decisions made and announced in public channels.

It's already embarrassing that ffserver made it to 3.2. Lets try to keep
the image that this is a serious project.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 24.11.2016 10:53, Josh de Kock wrote:
> There's no benefit in waiting for this to occur before ffserver is
> removed as it will always be in git history. Removing ffserver from
> master may even speed up the "resurrection of ffserver". I think this
> shouldn't be delayed any further.

Similarly there is no benefit in removing ffserver before the next major
bump, because the libraries have to stay backwards compatible until then,
anyway.
However, at least Reynaldo and Michael seem to be still working on ffserver,
so gratuitously removing it now would be rather unfriendly.

Best regards,
Andreas

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Josh de Kock]

On 29/10/2016 16:53, Andreas Cadhalpun wrote:
> On 28.10.2016 19:31, compn wrote:
>> On Thu, 27 Oct 2016 15:42:22 -0300 James Almer wrote:
>>> On 10/27/2016 3:36 PM, Reynaldo H. Verdejo Pinochet wrote:
 On 10/27/2016 11:25 AM, James Almer wrote:
> This is not how things were agreed.
 I haven't agreed to this.
>>> You could have shown your displeasure in the relevant discussion and
>>> patch threads, and on IRC. Why didn't you?
>>>
>>> A quick search on the archive shows
>>> https://ffmpeg.org/pipermail/ffmpeg-devel/2016-April/192808.html
>>> https://ffmpeg.org/pipermail/ffmpeg-devel/2016-July/196500.html
>>
>> but then there was this in september, after the news entry was posted
>>
>> http://ffmpeg.org/pipermail/ffmpeg-devel/2016-September/199686.html
>>
>> i'd suggest talking to that person before applying said patch.
>>
>> we can talk before applying a patch still, right? :)
> 
> What's the current status of this "Resurrection of ffserver"?
> Is there a work-in-progress git repository available somewhere?
> 
> Best regards,
> Andreas

There's no benefit in waiting for this to occur before ffserver is
removed as it will always be in git history. Removing ffserver from
master may even speed up the "resurrection of ffserver". I think this
shouldn't be delayed any further.

-- 
Josh

PGP fingerprint: A93A602D7A6D3C5388D792BCD052D40DDEF9703D



signature.asc
Description: OpenPGP digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 10/30/2016 8:23 PM, Lukasz Marek wrote:
> On 27.10.2016 20:26, Michael Niedermayer wrote:
>> On Thu, Oct 27, 2016 at 11:03:07AM -0700, Reynaldo H. Verdejo Pinochet wrote:
>> I agree with moving the apps to a seperate repo hosted within
>> the same infra and keeping ffserver.
>> I will help with ffserver as my time & todo list permits
> 
> I don't follow ffmpeg list for long time, so please forgive me if I said 
> something already discussed, but the news said it is removed because of 
> cleanups. That is reasonable, but in such case moving it to separate repo is 
> nonsense. Also I'm not sure ffmenc/dec removal is good decision at the 
> moment. I don't want to suggest there is ohter app than ffserver that use it, 
> but with these removed you can forget ffserver will exists in other repo, 
> right?

ffmdec/enc will not be removed until the next major bump at the earliest
for backwards compat reasons. At worst, when AVStream->codec is removed.
And Reynaldo, who's moving ffserver to a separate repo, said he'll make
it work without them in the long run.

In any case, ffserver and ffm* will all be gone from the main tree sooner
or later.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Lukasz Marek]

On 27.10.2016 20:26, Michael Niedermayer wrote:

On Thu, Oct 27, 2016 at 11:03:07AM -0700, Reynaldo H. Verdejo Pinochet wrote:
I agree with moving the apps to a seperate repo hosted within
the same infra and keeping ffserver.
I will help with ffserver as my time & todo list permits


I don't follow ffmpeg list for long time, so please forgive me if I said 
something already discussed, but the news said it is removed because of 
cleanups. That is reasonable, but in such case moving it to separate 
repo is nonsense. Also I'm not sure ffmenc/dec removal is good decision 
at the moment. I don't want to suggest there is ohter app than ffserver 
that use it, but with these removed you can forget ffserver will exists 
in other repo, right?


___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

Hi Reynaldo,

On 28.10.2016 20:41, Reynaldo H. Verdejo Pinochet wrote:
> I'm already working on what I think it's a proper transition in
> line with the idea presented on my first reply to this thread.
> 
> My impression from today's discussion on IRC is that James
> and Rostislav are willing to give this process some time but
> nothing past next release.

I appreciate that you care about ffserver.
However, ffserver still uses symbols that were declared private
in 2011/2012, so there has been ample time to fix ffserver.
I'd really like libavformat not exporting them anymore as soon
as possible, i.e. at the next major bump.
I don't really mind keeping ffserver until then, as we have
to keep backwards-compatibility anyway, but if it still uses
private symbols when the next major bump happens, it will have
to go.

Best regards,
Andreas
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 28.10.2016 19:31, compn wrote:
> On Thu, 27 Oct 2016 15:42:22 -0300 James Almer wrote:
>> On 10/27/2016 3:36 PM, Reynaldo H. Verdejo Pinochet wrote:
>>> On 10/27/2016 11:25 AM, James Almer wrote:
 This is not how things were agreed.
>>> I haven't agreed to this.
>> You could have shown your displeasure in the relevant discussion and
>> patch threads, and on IRC. Why didn't you?
>>
>> A quick search on the archive shows
>> https://ffmpeg.org/pipermail/ffmpeg-devel/2016-April/192808.html
>> https://ffmpeg.org/pipermail/ffmpeg-devel/2016-July/196500.html
> 
> but then there was this in september, after the news entry was posted
> 
> http://ffmpeg.org/pipermail/ffmpeg-devel/2016-September/199686.html
> 
> i'd suggest talking to that person before applying said patch.
> 
> we can talk before applying a patch still, right? :)

What's the current status of this "Resurrection of ffserver"?
Is there a work-in-progress git repository available somewhere?

Best regards,
Andreas



___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Reynaldo H. Verdejo Pinochet]

Hi

I'm already working on what I think it's a proper transition in
line with the idea presented on my first reply to this thread.

My impression from today's discussion on IRC is that James
and Rostislav are willing to give this process some time but
nothing past next release.

Bests,


--
Reynaldo H. Verdejo Pinochet
Open Source Group - Samsung Research America

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 10/28/2016 2:31 PM, compn wrote:
> On Thu, 27 Oct 2016 15:42:22 -0300 James Almer wrote:
>> On 10/27/2016 3:36 PM, Reynaldo H. Verdejo Pinochet wrote:
>>> On 10/27/2016 11:25 AM, James Almer wrote:
 This is not how things were agreed.
>>> I haven't agreed to this.
>> You could have shown your displeasure in the relevant discussion and
>> patch threads, and on IRC. Why didn't you?
>>
>> A quick search on the archive shows
>> https://ffmpeg.org/pipermail/ffmpeg-devel/2016-April/192808.html
>> https://ffmpeg.org/pipermail/ffmpeg-devel/2016-July/196500.html
> 
> but then there was this in september, after the news entry was posted
> 
> http://ffmpeg.org/pipermail/ffmpeg-devel/2016-September/199686.html

This is like someone showing up the day a collapsing building is scheduled
to be demolished suggesting to instead add some extra cement to keep it
in place. That simply would fly.
One at most can suggest what could be built in its place afterwards.

> 
> i'd suggest talking to that person before applying said patch.

As i said, nobody is against a working ffserver replacement made from scratch,
out of tree or otherwise, but the decision was already made and announced in
more than one channel.

> we can talk before applying a patch still, right? :)

Sure, we're doing it right now. But the time to have an effect on the
decision is long past. It was at the latest back with the news entry patch.
A thread where for that matter you didn't talk.

> 
> -compn

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[compn]

On Thu, 27 Oct 2016 15:42:22 -0300 James Almer wrote:
> On 10/27/2016 3:36 PM, Reynaldo H. Verdejo Pinochet wrote:
> > On 10/27/2016 11:25 AM, James Almer wrote:
> >> This is not how things were agreed.
> > I haven't agreed to this.
> You could have shown your displeasure in the relevant discussion and
> patch threads, and on IRC. Why didn't you?
> 
> A quick search on the archive shows
> https://ffmpeg.org/pipermail/ffmpeg-devel/2016-April/192808.html
> https://ffmpeg.org/pipermail/ffmpeg-devel/2016-July/196500.html

but then there was this in september, after the news entry was posted

http://ffmpeg.org/pipermail/ffmpeg-devel/2016-September/199686.html

i'd suggest talking to that person before applying said patch.

we can talk before applying a patch still, right? :)

-compn
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Ronald S. Bultje]

Hi,

On Thu, Oct 27, 2016 at 2:36 PM, Reynaldo H. Verdejo Pinochet <
reyna...@osg.samsung.com> wrote:

> Hello
>
> On 10/27/2016 11:25 AM, James Almer wrote:
>
>> [..]
>>
>> This is not how things were agreed.
>>
>>
> I haven't agreed to this.


Agreement that does not have to be unanimous.

Ronald
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Reynaldo H. Verdejo Pinochet]

Hello

On 10/27/2016 11:25 AM, James Almer wrote:

[..]

This is not how things were agreed.



I haven't agreed to this.

The patch fixes nothing while breaking master
for ffserver users. There's a way forward, but
the patch can't be applied before it happens.

Bests,

--
Reynaldo H. Verdejo Pinochet
Open Source Group - Samsung Research America
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 10/27/2016 3:36 PM, Reynaldo H. Verdejo Pinochet wrote:
> Hello
> 
> On 10/27/2016 11:25 AM, James Almer wrote:
>> [..]
>>
>> This is not how things were agreed.
>>
> 
> I haven't agreed to this.

You could have shown your displeasure in the relevant discussion and patch
threads, and on IRC. Why didn't you?

A quick search on the archive shows
https://ffmpeg.org/pipermail/ffmpeg-devel/2016-April/192808.html
https://ffmpeg.org/pipermail/ffmpeg-devel/2016-July/196500.html

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 10/27/2016 3:22 PM, Reynaldo H. Verdejo Pinochet wrote:
> I understand you might be upset but please refrain from
> using bad language.
> 
> The way I see it your removal will happen once there's a
> solution in place just not before. This is why the patch
> is not rejected.

This is not how things were agreed.

I'm all for a working ffserver replacement, but if a decision is
made and announced, it can't be changed just like that out of
nowhere.
This is not how a project should work.

> 
> Bests,
> 

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Michael Niedermayer]

On Thu, Oct 27, 2016 at 11:03:07AM -0700, Reynaldo H. Verdejo Pinochet wrote:
> I'm not OK with this patch. My reasons are:
> 
> * There are users relying on it
> * We gain nothing by removing ffserver without a
>   replacement or alternate solution
> 
> One solution I would support is a transition made by
> moving all applications to a separate repo hosted within
> the same infra. With a commitment not break our own apps.
> The server will be fixed to work using public API in the
> process. I can work on this with whoever wants to help.

I agree with moving the apps to a seperate repo hosted within
the same infra and keeping ffserver.
I will help with ffserver as my time & todo list permits

[...]
-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Why not whip the teacher when the pupil misbehaves? -- Diogenes of Sinope


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Reynaldo H. Verdejo Pinochet]

I understand you might be upset but please refrain from
using bad language.

The way I see it your removal will happen once there's a
solution in place just not before. This is why the patch
is not rejected.

Bests,

--
Reynaldo H. Verdejo Pinochet
Open Source Group - Samsung Research America
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Rostislav Pehlivanov]

On 27 October 2016 at 19:14, James Almer  wrote:

> On 10/27/2016 3:03 PM, Reynaldo H. Verdejo Pinochet wrote:
> > I'm not OK with this patch. My reasons are:
> >
> > * There are users relying on it
> > * We gain nothing by removing ffserver without a
> >   replacement or alternate solution
> >
>
> We've discussed the removal of ffserver and even /announced/ it on news
> section of the website. It would be really nice if for once this project
> gets its shit together and stops looking like a joke from an outsider
> PoV.
>
> > One solution I would support is a transition made by
> > moving all applications to a separate repo hosted within
> > the same infra. With a commitment not break our own apps.
> > The server will be fixed to work using public API in the
> > process. I can work on this with whoever wants to help.
>
> Starting an ffserver replacement outside the ffmpeg tree was suggested
> as a potential todo if people were willing to write it, but it didn't
> affect the removal decision, something that should have happened today,
> as scheduled, on the release that was just made, but that ultimately
> didn't happen because who knows why.
>
> Might as well write a news entry telling people to stop trusting
> news entries.
>
> >
> > Bests,
> >
> >
>
> ___
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>

Exactly. This was discussed to death, and unless there are actual technical
objections to the patch (e.g. breaks something or does something else it
shouldn't) I'll push the patch in a few days after the release has been
tagged, with James' suggestions.

Also the plan is to move all programs to a separate repository leaving all
the libraries into one repo. So this is about the best chance of a chance
at starting this good practice as any. So technically if you improve it not
to use public APIs it won't get removed - just eventually relocated. But
definitely removed from the main repo.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 10/27/2016 3:03 PM, Reynaldo H. Verdejo Pinochet wrote:
> I'm not OK with this patch. My reasons are:
> 
> * There are users relying on it
> * We gain nothing by removing ffserver without a
>   replacement or alternate solution
> 

We've discussed the removal of ffserver and even /announced/ it on news
section of the website. It would be really nice if for once this project
gets its shit together and stops looking like a joke from an outsider
PoV.

> One solution I would support is a transition made by
> moving all applications to a separate repo hosted within
> the same infra. With a commitment not break our own apps.
> The server will be fixed to work using public API in the
> process. I can work on this with whoever wants to help.

Starting an ffserver replacement outside the ffmpeg tree was suggested
as a potential todo if people were willing to write it, but it didn't
affect the removal decision, something that should have happened today,
as scheduled, on the release that was just made, but that ultimately
didn't happen because who knows why.

Might as well write a news entry telling people to stop trusting
news entries.

> 
> Bests,
> 
> 

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Reynaldo H. Verdejo Pinochet]

I'm not OK with this patch. My reasons are:

* There are users relying on it
* We gain nothing by removing ffserver without a
  replacement or alternate solution

One solution I would support is a transition made by
moving all applications to a separate repo hosted within
the same infra. With a commitment not break our own apps.
The server will be fixed to work using public API in the
process. I can work on this with whoever wants to help.

Bests,


--
Reynaldo H. Verdejo Pinochet
Open Source Group - Samsung Research America
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 10/26/2016 7:43 PM, Rostislav Pehlivanov wrote:
> From 48d1ed653e7841f49f50559510ca06f524cbbf1f Mon Sep 17 00:00:00 2001
> From: Rostislav Pehlivanov 
> Date: Wed, 26 Oct 2016 23:38:46 +0100
> Subject: [PATCH] Remove the ffserver program
> 
> Also changes the mpjpeg boundary tag to "ffmpeg" (it's arbitrary).

This is unrelated, so IMO should be in a separate patch.

> 
> Signed-off-by: Rostislav Pehlivanov 
> ---
>  .gitignore|1 -
>  Changelog |1 +
>  MAINTAINERS   |3 -
>  Makefile  |4 +-
>  README.md |2 -
>  configure |5 -
>  doc/ffmpeg-bitstream-filters.texi |4 +-
>  doc/ffmpeg-codecs.texi|4 +-
>  doc/ffmpeg-devices.texi   |4 +-
>  doc/ffmpeg-filters.texi   |4 +-
>  doc/ffmpeg-formats.texi   |4 +-
>  doc/ffmpeg-protocols.texi |4 +-
>  doc/ffmpeg-resampler.texi |4 +-
>  doc/ffmpeg-scaler.texi|4 +-
>  doc/ffmpeg-utils.texi |4 +-
>  doc/ffmpeg.texi   |4 +-
>  doc/ffplay.texi   |4 +-
>  doc/ffprobe.texi  |4 +-
>  doc/ffserver.conf |  372 
>  doc/ffserver.texi |  923 -
>  doc/issue_tracker.txt |3 -
>  doc/libavcodec.texi   |4 +-
>  doc/libavdevice.texi  |4 +-
>  doc/libavfilter.texi  |4 +-
>  doc/libavformat.texi  |4 +-
>  doc/libavutil.texi|4 +-
>  doc/libswresample.texi|4 +-
>  doc/libswscale.texi   |4 +-
>  doc/protocols.texi|2 +-
>  ffserver.c| 4043 
> -
>  ffserver_config.c | 1320 
>  ffserver_config.h |  134 --
>  libavformat/mpjpeg.c  |2 +-
>  tests/Makefile|   10 -
>  tests/ffserver-regression.sh  |   39 -
>  tests/ffserver.conf   |  311 ---
>  tools/bisect-create   |2 +-
>  37 files changed, 43 insertions(+), 7210 deletions(-)
>  delete mode 100644 doc/ffserver.conf
>  delete mode 100644 doc/ffserver.texi
>  delete mode 100644 ffserver.c
>  delete mode 100644 ffserver_config.c
>  delete mode 100644 ffserver_config.h
>  delete mode 100755 tests/ffserver-regression.sh
>  delete mode 100644 tests/ffserver.conf

"make ffserver" fails as unknown target as expected, "make all" and
other targets i could think about like testprogs and examples still
work, and fate passes.

LGTM, but ask Michael before backporting to 3.2 since he's making
the release and this is a big change.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[Andreas Cadhalpun]

On 27.10.2016 00:16, Rostislav Pehlivanov wrote:
> On 26 October 2016 at 22:48, James Almer  wrote:
> 
>> On 10/26/2016 6:19 PM, Rostislav Pehlivanov wrote:
>>> Also removes url_feof from libavformat.v which should have been
>>> removed long ago and changed the multipart jpeg boundary tag to
>>> ffmpeg rather than ffserver (it's arbitrary).
>>
>> [...]
>>
>>> diff --git a/libavformat/libavformat.v b/libavformat/libavformat.v
>>> index c961cd8..47d5ddc 100644
>>> --- a/libavformat/libavformat.v
>>> +++ b/libavformat/libavformat.v
>>> @@ -1,19 +1,6 @@
>>>  LIBAVFORMAT_MAJOR {
>>>  global:
>>>  av*;
>>> -#FIXME those are for ffserver
>>> -ff_inet_aton;
>>> -ff_socket_nonblock;
>>> -ff_rtsp_parse_line;
>>> -ff_rtp_get_local_rtp_port;
>>> -ff_rtp_get_local_rtcp_port;
>>> -ffio_open_dyn_packet_buf;
>>> -ffio_set_buf_size;
>>> -ffurl_close;
>>> -ffurl_open;
>>> -ffurl_write;
>>> -#those are deprecated, remove on next bump
>>> -url_feof;
>>>  local:
>>>  *;
>>>  };
>>
>> No, this can't be done until the next major bump. url_feof is even guarded
>> by an scheduled FF_API define.
>>
>> The rest should be ok, but anything that implies a library ABI break can't
>> be done just yet. Wait for other comments and/or testing. This removes a
>> lot of things after all.
>>
> Fixed version attached without removing url_feof (apparently that's meant
> to go in the next bump, wasn't forgotten)

No, none of the symbols can be removed without a major bump. The simple reason
is that a ffserver built from the 3.1 branch still has to work with the
libavformat from the 3.2 branch.

Best regards,
Andreas

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] Remove the ffserver program and the ffm muxer/demuxer

[James Almer]

On 10/26/2016 6:19 PM, Rostislav Pehlivanov wrote:
> Also removes url_feof from libavformat.v which should have been
> removed long ago and changed the multipart jpeg boundary tag to
> ffmpeg rather than ffserver (it's arbitrary).

[...]

> diff --git a/libavformat/libavformat.v b/libavformat/libavformat.v
> index c961cd8..47d5ddc 100644
> --- a/libavformat/libavformat.v
> +++ b/libavformat/libavformat.v
> @@ -1,19 +1,6 @@
>  LIBAVFORMAT_MAJOR {
>  global:
>  av*;
> -#FIXME those are for ffserver
> -ff_inet_aton;
> -ff_socket_nonblock;
> -ff_rtsp_parse_line;
> -ff_rtp_get_local_rtp_port;
> -ff_rtp_get_local_rtcp_port;
> -ffio_open_dyn_packet_buf;
> -ffio_set_buf_size;
> -ffurl_close;
> -ffurl_open;
> -ffurl_write;
> -#those are deprecated, remove on next bump
> -url_feof;
>  local:
>  *;
>  };

No, this can't be done until the next major bump. url_feof is even guarded
by an scheduled FF_API define.

The rest should be ok, but anything that implies a library ABI break can't
be done just yet. Wait for other comments and/or testing. This removes a
lot of things after all.

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel