[FFmpeg-devel] Forgejo entry threshold

[FFmpeg-devel] Forgejo entry threshold

[Nicolas George]

Dummy_test_user (HE12025-08-04):
> PR #20112 opened by Dummy_test_user
> URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/20112
> Patch URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/20112.patch
> 
> I did that pull request to see what is preventing me from submitting crappy code.
> Apparently nothing.

Hi.

That was me. Thank you for ensuring it went to the mailing-list.

But…

I could register an account and create a pull request without ever
reading the coding-style rules in doc/developer.texi, and the pull
requests has red patcheck:

mismatching doxy params
+ * @param erbut  useless
+ * @param ernut  does nothing

That is not good.

If nothing better comes to mind to exclude crappy code, could we at
least condition creating an account to answering a 10 questions quiz
about our coding conventions?

Regards,

-- 
  Nicolas George
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Hendrik Leppkes]

On Mon, Aug 4, 2025 at 4:45 PM Nicolas George <george@nsup.org> wrote:
>
> Dummy_test_user (HE12025-08-04):
> > PR #20112 opened by Dummy_test_user
> > URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/20112
> > Patch URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/20112.patch
> >
> > I did that pull request to see what is preventing me from submitting crappy code.
> > Apparently nothing.
>
> Hi.
>
> That was me. Thank you for ensuring it went to the mailing-list.
>
> But…
>
> I could register an account and create a pull request without ever
> reading the coding-style rules in doc/developer.texi,
>

Whats the barrier of entry to sending a patch to the ML right now?
Does it quiz you on the guidelines?

- Hendrik
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Nicolas George]

Hendrik Leppkes (HE12025-08-04):
> Whats the barrier of entry to sending a patch to the ML right now?

Being able to, which is strongly correlated to knowing good practices.

> Does it quiz you on the guidelines?

Do you have better to propose?

-- 
  Nicolas George
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Leon Grutters]

On 8/4/25 4:45 PM, Nicolas George wrote:
> Dummy_test_user (HE12025-08-04):
>> PR #20112 opened by Dummy_test_user
>> URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/20112
>> Patch URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/20112.patch
>>
>> I did that pull request to see what is preventing me from submitting crappy code.
>> Apparently nothing.
> Hi.
>
> That was me. Thank you for ensuring it went to the mailing-list.
>
> But…
>
> I could register an account and create a pull request without ever
> reading the coding-style rules in doc/developer.texi, and the pull
> requests has red patcheck:
>
> mismatching doxy params
> + * @param erbut  useless
> + * @param ernut  does nothing
>
> That is not good.
>
> If nothing better comes to mind to exclude crappy code, could we at
> least condition creating an account to answering a 10 questions quiz
> about our coding conventions?
>
> Regards,
>
Is it that much harder to submit a patch on the ML? All you do is sign 
up for the ML, run one git command and then attach it to an email and 
click send.

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Nicolas George]

Nicolas George (HE12025-08-04):
> That was me.

Whoever disabled the account: I went to just delete it, and now I
cannot, so the joke is on you, you have to do it yourself.

> Thank you for ensuring it went to the mailing-list.

It seems to work randomly.

Regards,

-- 
  Nicolas George
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Nicolas George]

Leon Grutters (HE12025-08-04):
> Is it that much harder to submit a patch on the ML? All you do is sign up
> for the ML, run one git command and then attach it to an email and click
> send.

No, it is not harder at all, but it has the reputation to be, which has
the same result.

Nowadays, any idiot wannabe coder already has a github account and knows
how to fumble on a similar webapp without having ever heard of
etiquette, coding style or good practices. At least, when learning
involves reading a doc instead of clicking randomly on buttons, they
would have an occasion to.

Regards,

-- 
  Nicolas George
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Kacper Michajlow]

On Tue, 5 Aug 2025 at 15:10, Nicolas George <george@nsup.org> wrote:
>
> Leon Grutters (HE12025-08-04):
> > Is it that much harder to submit a patch on the ML? All you do is sign up
> > for the ML, run one git command and then attach it to an email and click
> > send.
>
> No, it is not harder at all, but it has the reputation to be, which has
> the same result.
>
> Nowadays, any idiot wannabe coder already has a github account and knows
> how to fumble on a similar webapp without having ever heard of
> etiquette, coding style or good practices. At least, when learning
> involves reading a doc instead of clicking randomly on buttons, they
> would have an occasion to.

Those "idiot wannabe coders," as you call them, are simply not
interested in low-level C/ASM media processing libraries. They use
Playskool Plastic Tool Kit to develop web applications in javascript.

Seriously, I think elitism and gatekeeping are not the right approach.
Judging someone's contribution quality based on whether they use gmail
or other "non-approved" tools is unfair. It saddens me that this kind
of attitude comes from a senior member of the community with so much
experience in the project.

The whole point is to attract new contributors and make the workflow
easier for everyone to improve productivity. If you care about
quality, go review the patches.

- Kacper
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Rémi Denis-Courmont]

Le 4 août 2025 22:06:07 GMT+07:00, Leon Grutters <gruttersleonbot2@gmail.com> a écrit :
>Is it that much harder to submit a patch on the ML? All you do is sign up for the ML, run one git command and then attach it to an email and click send.

I don't know how much is that much. It is harder, at least the *first* time, though not hard enough to deter a "vibe coder". But it is exclusionary, as not everyone can get a working reputable outbound SMTP server.

That said, I prefer a forge to facilitate the review process, not the submission process. Judging by earlier discussions, I am far from the only one.
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Nicolas George]

Rémi Denis-Courmont (HE12025-08-06):
> But it is exclusionary, as not everyone can get a working reputable
> outbound SMTP server.

<sigh> It has already been said time and again that it is not a
requirement to participate. The only requirement is the ability to send
and receive mail.

-- 
  Nicolas George
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Nicolas George]

Kacper Michajlow (HE12025-08-05):
> Those "idiot wannabe coders," as you call them, are simply not
> interested in low-level C/ASM media processing libraries. They use
> Playskool Plastic Tool Kit to develop web applications in javascript.

Let us make sure they stay there.

> Seriously, I think elitism and gatekeeping are not the right approach.

I am pretty sure it is the only approach that can save a project that is
severely undercrewed of people who know the code well.

> Judging someone's contribution quality based on whether they use gmail
> or other "non-approved" tools is unfair. It saddens me that this kind
> of attitude comes from a senior member of the community with so much
> experience in the project.

Indeed, that would be unfair. But the only unfair and dishonest thing
here is you misrepresenting my words publicly.

The quality of your contribution can be seen with `git log -p`.

What your use of GMail's software discredits is not the quality of your
contribution but the relevance of your advice on using mail.

> The whole point is to attract new contributors and make the workflow
> easier for everyone

Still that idiotic idea that attracting more newbies will solve the
issue that we do not have enough senior developers.

> to improve productivity.

The current bullshit decreases the productivity of the most productive
member of the project.

-- 
  Nicolas George
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Rémi Denis-Courmont]

Le 6 août 2025 15:11:20 GMT+07:00, Nicolas George <george@nsup.org> a écrit :
>Rémi Denis-Courmont (HE12025-08-06):
>> But it is exclusionary, as not everyone can get a working reputable
>> outbound SMTP server.
>
><sigh> It has already been said time and again that it is not a
>requirement to participate. The only requirement is the ability to send
>and receive mail.

Sure. You can send a list of patch files... And make it even more difficult to review, thus making a forge even more pressing for reviewers.
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Nicolas George]

Rémi Denis-Courmont (HE12025-08-07):
> Sure. You can send a list of patch files... And make it even more difficult to review,

What? An attached file is not more difficult to review than a mail sent
directly by git, and both are more convenient than a web monstrosity?

Guys, have you never spend five minutes learning how to use mail?

-- 
  Nicolas George
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Rémi Denis-Courmont]

Le 7 août 2025 19:37:32 GMT+07:00, Nicolas George <george@nsup.org> a écrit :
>Rémi Denis-Courmont (HE12025-08-07):
>> Sure. You can send a list of patch files... And make it even more difficult to review,
>
>What? An attached file is not more difficult to review than a mail sent
>directly by git,

If your mail client treats attachments the same as inline text, your mail client is very badly broken as a mail client. And if not, then it is obviously much more cumbersome to review an attachment.

Also replying to multiple patches in a single mail makes everyone else's life miserable for tracking the review.

But more importantly, email has all the fundamental problems that I already raised several times in previous threads. It simply can't compete with something that's designed for code review and actually tracks and organises relevant metadata.

> and both are more convenient than a web monstrosity?

Unless you're using an extremely old or underpowered system that can't run a reasonably recent web browser, no.

And if you do, then good luck compiling FFmpeg with it.

>Guys, have you never spend five minutes learning how to use mail?

Yeah, we gave. And that definitely doesn't cover sending and reviewing patches because that's not what email was intended for. Git didn't even exist then.
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Nicolas George]

Rémi Denis-Courmont (HE12025-08-08):
> If your mail client treats attachments the same as inline text, your
> mail client is very badly broken as a mail client. And if not, then it
> is obviously much more cumbersome to review an attachment.

My mail client does not treat text attachments exactly the same way as
inline text, but it makes it extremely easy to view them and quote them
in the reply. It also makes it quite easy to fix the type of attachments
that were mistakenly flagged as octet/stream.

If your mail client does not let you do that easily… well, it proves my
point: a good mail client makes all the difference in productivity and
comfort to interact with the project.

> Also replying to multiple patches in a single mail makes everyone
> else's life miserable for tracking the review.

Then we delete all but the first patch and reply to it, the all but the
second patch and reply to it, and… And we ask the person who submitted
to next time have the courtesy to send the patches in separate mail.

By the way, we could also set up our SMTP server to accept authenticated
SMTP on port 465 with login/password guest/ffmpeg_is_great only for mail
to ffmpeg: it would fix most of the hypothetical “submitter cannot find
a SMTP server” issues.

> But more importantly, email has all the fundamental problems that I
> already raised several times in previous threads.

Sure, but we are in the process of establishing that these “fundamental”
problems are only the consequence of using mediocre mail clients.

> It simply can't compete with something that's designed for code review
> and actually tracks and organises relevant metadata.

What a naïve thing to say. There are plenty of cases where a master
craftsperson is more proficient using mainly one excellent universal
tool than with many specialized tools, even when such specialized tools
might be better for the dilettante. Cooking for example.

> Unless you're using an extremely old or underpowered system that can't
> run a reasonably recent web browser, no.

It only takes a few years of age and/or a cheap computer to make web
apps painfully slow. Not everybody lives in the First World, you seem to
be forgetting it.

I must say, I like my elitism about skills much better than your elitism
about material possessions.

Also, even with an extremely overpowered system, web applications are
still much worse in terms of productivity, because they have to work in
a browser, where ergonomics is very random.

Just consider the editor. Reviewing a patch involves writing text and
code snippets, which is best done with our respective favorite powerful
text editor. But in a web app, unless we engage in clumsy and annoying
copy-paste operations, we have to use the editor provided by the
browser, no choice at all.

Of course, if your mail client does not let you use your favorite
powerful text editor to reply to mail, then it is once again the proof
that the issue is using a mediocre mail client.

So, unless you come up with a trick to invoke Vim from the web monster
with minimal manipulations, you have to admit that switching to it would
make at least a few of us significantly less proficient.

> And if you do, then good luck compiling FFmpeg with it.

It is possible to compiler FFmpeg even on very underpowered computers,
it is one of the main qualities of the project. It will take a lot of
time, but that is not an obstacle since nobody has to stay and watch.
Rebuilding after a small change will take a little time too, during
which the contributor can read a doc or a blog article or do whatever.

Whereas the slowness of the web monsters happens when we are trying to
use them and make using them very uncomfortable, so that is a very
dishonest argument.

> Yeah, we gave. And that definitely doesn't cover sending and reviewing
> patches because that's not what email was intended for. Git didn't
> even exist then.

Any skill need to be updated.

-- 
  Nicolas George
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] Forgejo entry threshold

[Michael Niedermayer]

[-- Attachment #1.1: Type: text/plain, Size: 1481 bytes --]

Hi Remi

On Fri, Aug 08, 2025 at 10:14:52AM +0700, Rémi Denis-Courmont wrote:
> 
> 
> Le 7 août 2025 19:37:32 GMT+07:00, Nicolas George <george@nsup.org> a écrit :
> >Rémi Denis-Courmont (HE12025-08-07):
> >> Sure. You can send a list of patch files... And make it even more difficult to review,
> >
> >What? An attached file is not more difficult to review than a mail sent
> >directly by git,
> 

> If your mail client treats attachments the same as inline text, your mail client is very badly broken as a mail client. And if not, then it is obviously much more cumbersome to review an attachment.

mime_lookup application/octet-stream text/plain

auto_view text/x-patch text/x-diff

color body brightwhite default "^Index: .*"
color body red   default "^-.*"
color body green default "^[+].*"
color body white default "^ #.*"

First re-analyzes "application/octet-stream" and "text/plain" attachments and
provides them with the correct mime type according to what the file is

the second causes attachments that are patches to be shown

and third will colorize patches/diffs even when they are inline

I almost dont even notice if a patch is inline or an attachment, both
are shown automatically as i look at the mail

thx

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

I have never wished to cater to the crowd; for what I know they do not
approve, and what they approve I do not know. -- Epicurus

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

[-- Attachment #2: Type: text/plain, Size: 251 bytes --]

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".