SV: Re: SV: Re: Revision graph revisited

> Good idea, good description,

Thank you! Nice to hear that someone other than myself could follow. :-)

> but I don't think the method for identifying tags will work.

That's true. And it's why I wrote "tags" everywhere, as in
items-with-tag-like-appearance.

> 1. New branches are indistinguishable from tags.

Well, they will end up in the same column as tags. But is this such a
big deal? As soon as it gets its first commit it will get its own
column. (And the path is printed on every item, making them clearly
distinguishable.)

> 2. In TSVN at least, there are further commits to tags to update
> the language packs after the code is released.

I know some people who would argue that this is absolutely wrong to do.
(We are for example enforcing a zero modification policy on tags via
hook-scripts.) Anyway, the only thing that would happen is that the tag
would get it's own column. Much better than today anyway, and a (imho)
very small price to pay for those who stretch the tag-concept that way.

> Another way to do it would be to assume that every path containing
> "/tags/" (case-insensitive) is a tag, and should be included in the
> tags stack. Even better would be to allow the user to define what the
> tags area is called.

This is an inherent problem with subversions extremely flexible approach
to versioning... Maybe TSVN could create some way to identify folders
with special meanings (properties?) or maybe something is in the
pipeline over at the subversion project? If we want to solve that
problem I think we should look further than only at the revgraph.
 
However, I still think that my proposed way of organising the revgraph
is the right thing to do at present, and will generate reasonable
layouts for all graphs.

Hans-Emil

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tortoisesvn.tigris.org
For additional commands, e-mail: dev-help@tortoisesvn.tigris.org
Received on Fri Feb 10 09:22:54 2006