Re: [RFC/WIP PATCH 04/11] upload-pack-2: Implement the version 2 of upload-pack

From: Stefan Beller <sbeller@google.com>
Date: May 27, 2015, 20:45 UTC
Message-ID: <CAGZ79kb6rsxpiXKg+eaVc+5bgUgPSaq+ocd-UHa=_K+rLxreTQ@mail.gmail.com>
In-Reply-To: <20150527203451.GD14309@peff.net>

On Wed, May 27, 2015 at 1:34 PM, Jeff King <peff@peff.net> wrote:

Show 34 quoted lines

> On Wed, May 27, 2015 at 10:40:37AM -0700, Stefan Beller wrote:
>
>> > If we are upload-pack-2, should we advertise that in the capabilities? I
>> > think it may make things easier later if we try to provide some
>> > opportunistic out-of-band data. E.g., if see tell git-daemon:
>> >
>> >   git-upload-pack repo\0host=whatever\0\0version=2
>> >
>> > how do we know whether version=2 was understood and kicked us into v2
>> > mode, versus an old server that ignored it?
>>
>> So in my vision we would call git-upload-pack-2 instead of having a "version=2".
>> and if git-upload-pack-2 doesn't exist, the whole conversation is
>> over, the client
>> it is left to make up some good error message or retry version 1.
>
> I'd like for that to be a starting point for us, and then to be able to
> add-on "hints" to ease the transition path in whatever way we want. We
> may even not do that in the long run, but I want to leave the door open
> if we can.
>
>> But I think advertising both which versions the server could deal
>> with, as well as
>> the currently expected version is a good thing.
>>
>> capability: can_speak=1,2
>> capability: speaking_now=2
>
> I was thinking just "speaking_now=2", but it probably makes sense to do
> both. I do not think using it to "downgrade" will ever be particularly
> useful (certainly not from v2 to v1, since to understand the flag both
> sides must be v2 in the first place). But advertising that via the v1
> conversation will be a good way to tell the other side that upgrade is
> possible.

If for some reason we discover a flaw in the current version, which makes it unusable (a buffer overflow?, some stupid abuse which makes the capability list huge), you may want to force downgrading (and in the very distant future when we are current on version 4 and have dropped version 1 already, you can only downgrade to 2 and 3, so I can see value in it.

Another idea to make it all more future proof: "capability: speaking_now=2" must be sent as the first line, so then you can adapt on the client side easily for which version you are listening.

Show 30 quoted lines

>
>> > Also, do we need the capability noise-word?
>>
>> I thought it opens up a new possible door in the future.
>> As we ignore anything not starting with "capability" for now, you
>> could introduce
>> your foo and bar ping pong easily and still be version 2 compatible.
>>
>> S: capability: thin
>> S: capability: another-capability
>> S: ping-pong foo
>> S: done
>> C: (not having understood ping-pong) just answering with capability: thin
>> C: done, let's proceed to refs advertisement
>>
>> The alternative client would do:
>>
>> C: ping-pong: foo-data1a
>> S: ping-pong: foo-data1b
>> C: ping-pong: foo-data2a
>> C: capability: thin
>> ...
>
> Right, but I think (and please correct me if there's a case I'm missing)
> that the behavior is the same whether it is spelled "ping-pong" or
> "capability:ping-pong". That is, the rule for "capability:" is "if you
> do not understand it, ignore it and do not mention it in your
> capabilities; the server is required to assume you were written before
> that capability was invented". But that is _also_ the rule for
> ping-pong, I think.

The rules are the same, right. But the allowed characters are limited (in theory) as the regular expressions given for the capabilities don't allow for binary data for example, but only well formed ASCII text, space separated. The "ping-pong" keyword could introduce a binary stream there including line feeds. (Today it sounds like a stupid idea though)

Show 10 quoted lines

>
>> > Eric mentioned the underflow problems here, but I wonder even more:
>> > what's wrong with the global ends_with() that we already provide?
>>
>> I was missing knowledge we have that, and apparently I was thinking it's
>> faster to come up with my own version than to look for it. :)
>
> Makes sense. :)
>
> -Peff