Problem with cogito and Linux tree tags

Problem with cogito and Linux tree tags

[Radoslaw Szkodzinski]

My cg-version is:
cogito-0.10 (20e473c9afd8b5d2d549b0e7881473600beb9c37)

cg-tag-ls output is:
v2.6.11    5dc01c595e6c6ec9ccda4f6f69c131c0dd945f8c
v2.6.11-tree    5dc01c595e6c6ec9ccda4f6f69c131c0dd945f8c
v2.6.12-rc2    9e734775f7c22d2f89943ad6c745571f1930105f
v2.6.12-rc3    0397236d43e48e821cce5bbe6a80a1a56bb7cc3a
v2.6.12-rc4    ebb5573ea8beaf000d4833735f3e53acb9af844c
v2.6.12-rc5    06f6d9e2f140466eeb41e494e14167f90210f89d

The ID of the commit that says "Linux 2.6.12-rc5" is:
137318b273db26b889675101fbd02d2e84cae5e3

The v2.6.12-rc5 tag hash isn't a tree or commit hash, it is unusable.
I don't know what causes that...

AstralStorm

Re: Problem with cogito and Linux tree tags

[Dan Holmsand]

Radoslaw Szkodzinski wrote:
  The v2.6.12-rc5 tag hash isn't a tree or commit hash, it is unusable.
> I don't know what causes that...

It's a tag object. If you cg-pull via http, you won't get any of those, 
though.

Try rsync instead:

$ cg-branch-add kernel 
rsync://rsync.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
$ cg-pull kernel

/dan

Re: Problem with cogito and Linux tree tags

[Radoslaw Szkodzinski]

Dan Holmsand wrote:

>
> Radoslaw Szkodzinski wrote:
>  The v2.6.12-rc5 tag hash isn't a tree or commit hash, it is unusable.
>
>> I don't know what causes that...
>
>
> It's a tag object. If you cg-pull via http, you won't get any of
> those, though.
>
I thought it was one of these. Seems cogito can't handle them yet.

> Try rsync instead:
>
Guess what, I did it by rsync.

AstralStorm

Re: Problem with cogito and Linux tree tags

[Dan Holmsand]

Radoslaw Szkodzinski wrote:
> Dan Holmsand wrote:
>>It's a tag object. If you cg-pull via http, you won't get any of
>>those, though.
> I thought it was one of these. Seems cogito can't handle them yet.

Ah. I forgot just how old cogito-0.10 has gotten. Current cogito handles 
these just fine. You should be able to read the tag object slightly more 
manually, though:

$ git-cat-file tag 06f6d9e2f140466eeb41e494e14167f90210f89d

which tells you that v2.6.12-rc5 is commit 
2a24ab628aa7b190be32f63dfb6d96f3fb61580a

/dan

Re: Problem with cogito and Linux tree tags

[Radoslaw Szkodzinski]

Dan Holmsand wrote:

>
> $ git-cat-file tag 06f6d9e2f140466eeb41e494e14167f90210f89d
>
> which tells you that v2.6.12-rc5 is commit
> 2a24ab628aa7b190be32f63dfb6d96f3fb61580a
>
Ok.  This doesn't work too. That cogito version really works fine, but I
had that branch cloned from a local repository gotten from rsync and
probably that's the problem.
The tags weren't propagated, but they should be.

AstralStorm

Re: Problem with cogito and Linux tree tags

[Radoslaw Szkodzinski]

Radoslaw Szkodzinski wrote:

>Dan Holmsand wrote:
>
>  
>
>>$ git-cat-file tag 06f6d9e2f140466eeb41e494e14167f90210f89d
>>
>>which tells you that v2.6.12-rc5 is commit
>>2a24ab628aa7b190be32f63dfb6d96f3fb61580a
>>
>>    
>>
>Ok.  This doesn't work too. That cogito version really works fine, but I
>had that branch cloned from a local repository gotten from rsync and
>probably that's the problem.
>The tags weren't propagated, but they should be.
>
>AstralStorm
>  
>
Well, new cogito-0.11.1 (300ab153620d2492e824cb3561c32debb5e80bf8) has
this fixed and picked up missing objects automatically.

AstralStorm

Re: Problem with cogito and Linux tree tags

[Petr Baudis]

Dear diary, on Fri, Jun 03, 2005 at 01:14:18AM CEST, I got a letter
where Radoslaw Szkodzinski <astralstorm@gorzow.mm.pl> told me that...
> Radoslaw Szkodzinski wrote:
> 
> >Dan Holmsand wrote:
> >
> >  
> >
> >>$ git-cat-file tag 06f6d9e2f140466eeb41e494e14167f90210f89d
> >>
> >>which tells you that v2.6.12-rc5 is commit
> >>2a24ab628aa7b190be32f63dfb6d96f3fb61580a
> >>
> >>    
> >>
> >Ok.  This doesn't work too. That cogito version really works fine, but I
> >had that branch cloned from a local repository gotten from rsync and
> >probably that's the problem.
> >The tags weren't propagated, but they should be.
> >
> >AstralStorm
> >  
> >
> Well, new cogito-0.11.1 (300ab153620d2492e824cb3561c32debb5e80bf8) has
> this fixed and picked up missing objects automatically.

Oh yes, I wanted to mention it in this thread but forgot.

However note that what it does now is awful and it's just a
quick'n'dirty fix to make it mostly work. I guess a better way would be
to teach the pull tools to download refs/ files too and do the proper
dependency things from them too (in addition to the passed commit id).

I will need to go back and look again at the set of Daniel's patches
which implemented something like that.

-- 
				Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
C++: an octopus made by nailing extra legs onto a dog. -- Steve Taylor