Status of speed regressions...

[Adrian Buehlmann]

On 2010-11-20 17:16, Jason Harris wrote:
> 
> On Nov 19, 2010, at 11:03 PM, Matt Mackall wrote:
> 
>> On Fri, 2010-11-19 at 18:40 +0100, Jason Harris wrote:
>>> Hi All,
>>>
>>> I was just wondering what the status of fixing the speed regressions is? Are there patches I can apply?
>>
>>> I have a MacHg with Mercurial 1.5.4 (baseline) with 1.6.4 (10-15%
>>> slower) and with 1.7.1 (20% slower). I can publish timings but I think
>>> others have noticed this as well...
>>
>> Seems a refresher is in order.
>>
>> Here is what I expect of someone who is basing an ostensibly serious
>> product on our software:
>>
>> - proactively test BEFORE every release
>> - report issues IMMEDIATELY
>> - and with SPECIFICS
>> - work to fix issues UPSTREAM, don't patch our code in your releases
> 
> <sigh> do we really have to go here... yet again???

It seems. Unfortunately. Since it looks like you still have some serious
misunderstandings.

What I still don't get is, why do you think there would be some patches
flying around somewhere that you could somehow pick up and apply to some
fork of yours?

You were told already that we don't want and can't care about forks of
Mercurial.

There is one official repo and it is at

   http://selenic.com/repo/hg

it contains all official changes that have been reviewed and accepted by
the Mercurial project so far. The stable branch currently contains all
bugfixes since 1.7, and the default branch currently contains
development that will be released in 1.8.

Asking "Are there patches I can apply?" *could* be seen as abrasive and
arrogant, as this *might* show your disrespect for what this project
here is doing in that you implicitly say with that "I don't care what's
in the official repo, I keep track of my own and pick whatever patches I
like". Which would be unprofessional indeed. In case you indeed do that,
then even asking here for such patches -- "valuable" patches that have
been somehow "ignored" by the project -- would be quite strange to say
the least.

Or another conclusion might just be that you haven't understood what the
official repo really is and how things get in there and how it is
supposed to be used.

I hope it is just the latter.