[FFmpeg-devel] [PATCH 1/2] avcodec/snowdec: Check intra block dc differences.

Michael Niedermayer michael at niedermayer.cc
Fri Nov 17 21:06:41 EET 2017

On Fri, Nov 17, 2017 at 10:28:49AM -0500, Ronald S. Bultje wrote:
> Hi,
> On Fri, Nov 17, 2017 at 10:14 AM, Michael Niedermayer <
> michael at niedermayer.cc> wrote:
> > As a maintainer If a file fails with an error i want to know
> > what error it is. Why is this so hard or controversal ?
> Because *it doesn't happen* for real files.
> But you don't want to listen.

So you know exactly which error conditions will never happen ?
And thats considering all the odd files we saw over the years,
hacked and modified encoders and hacked and modified specs and drafts
on which some encoders are based which we have seen files from.

And you expect that every time we consider adding an error message
someone should do this analysis ?

And after we have ommitted hundreads of error messages this way with
all the time that costs. the 10kb of space saved is worth more than
the hundread hours spend on this analysis ?

and none of these predictions of non occurance in real world files
will be wrong ?
In no significant number of cases will bugs be missed because they
are just vissible through one of these ommited error messages ?

The above sounds unreasonable if that is your position. But please
correct me if that is not your position. It seemed to follow from
a selective ommision of error messages

> You still -after all these years- think that
> if you believe something, then it must be true. How can this project be
> collaborative if any time we disagree, you just dig in your heels in the
> sand and start this same old argument where "you are maintainer" and
> therefore whatever you believe is true, and whatever you say is final?

Something in your reasoning is a bit unreasonable here.
I do not remember that you ever backed down from an argument.  And
you say i am diging my heels in ?
and that after i agreed to remove the messages yesterday so this
can move forward and this bug can be fixed.

About "collaborative" i must repeat what i said multiple times now
"Iam happy to follow what the community prefers." How can this be
anti-collaborative !?!?!?!

We can do what i suggested many times already, poll the community and
follow the communities preferrance on error messages.


Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

No snowflake in an avalanche ever feels responsible. -- Voltaire
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20171117/993f740a/attachment.sig>

More information about the ffmpeg-devel mailing list