[vlc-devel] [vlc-commits] JSON representation and Javascript unescaping

[Rémi Denis-Courmont]

Le maanantaina 28. syyskuuta 2020, 16.52.17 EEST Thomas Guillem a écrit :
> > Congratulations because I'm even more disgusted and even less likely to go
> > through code review onward, than I already was from other people's last
> > week's obstructionism.

> I don't see obstructionism, lot of people were interested, some of us didn't
> fully understand the implication of such patch (some us didn't even know
> that ytdl handled more than 100 sites and not only one).

Rejecting one log message per input is not misunderstanding, it's obstruction. 
And that's just the last example.

> I was personally afraid of the performance consequence (but I was still very
> enthusiastic by your patch set).

I never said that *you* did obstructionism.

> In the end, your patches were merged in around one week, it looks like a
> decent review process to me.

You can't have it both ways. You can't say that I shouldn't have cut short the 
review process, and then that a lapse of one week was acceptable.

If I'd sent version 4 for review, here's where we'd be at:
- Steve would still be complaining about logs and strings.
- Pierre would probably complain about user-agent.
- François and you would complain about code duplication and insists on using 
the first JSON API or updating the fingerprint code.
- In all likelihood someone would question using JSON.
- Likely somebody would complain about using Flex and Bison, because why not 
write a grammar directly in C when you have tools to generate the C code.
And who knows when this would actually be merged, if ever.

The reality is that I was in favor of getting the code reviewed at first. 
Otherwise, I wouldn't even have sent it for review in the first place. And then 
prerequisite inflation and filibuster disgusted me.

You can refuse to see those problems, or you can take a critical look at 
everyone's attitude (as opposed to just mine) and actually try to learn from 
that.

-- 
Реми Дёни-Курмон
http://www.remlab.net/