[MERGE] tags in log output

[Kent Gibson]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1



Aaron Bentley wrote:
> John Arbash Meinel wrote:
>
>> At one point we talked about making something like Revision or
>> RevisionTree have more functions on it (which would be lazily
>> evaluated). And then using something like that instead of passing
>>
> around
>> revision_ids.
>
>> However, as different branches have a different view about tags,
>> tags can't really be a property of a Revision, though it could be
>> a
> function
>> on Revision which is passed a Branch.
>
> Please make it a different kind of object, not a Revision.  Varying
>  public state of immutable objects is terribly icky.
>
> You can even call it a BranchRevision if you want.  Or there's
> precedent for RevisionInfo.
>
> (And yes, there seems to be great need for BranchRevisionTree or
> similar.)
>
>> I think it might be better to change the LogFormatter api itself,
>>
> so it
>> maintains a state of "this is what I'm logging", and that object
>> can gain more attributes when more information is available.
>> Which is, in essence, what Kent is talking about, only doing it
>> on a helper object rather than Revision.
>
> I don't follow how this is the same as what Kent is talking about.
>
And I'm sorry, I didn't distinguish between Revision and the
helper/wrapper class (which I'm calling LogRevision) in my initial
mail to prevent you guys going down this rat hole.
I was tempted to, but didn't want to get too far into implementation
detail.
Obviously the term "revision" is already pretty loaded so I should've
made it clear I wasn't talking about "the" revision.

I agree with Aaron - don't mess with Revision itself.

Cheers,
Kent.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGHta0goxTFTi1P8QRAv1DAJ0XPPHShAvOudhQd4JypbFxYnfh/QCfeZXO
YHI7vKgXtt685GvYx5TB0lg=
=wrDJ
-----END PGP SIGNATURE-----