Re: [Gnu-arch-users] What is arch? (Alternative answer)

[Robert Anderson]

Pierce T. Wetter III wrote:

> On Feb 24, 2004, at 10:18 AM, Robert Anderson wrote:
>
> > Pierce T. Wetter III wrote:
> >
> > > This is less true in arch, because arch assumes a model much closer 
> > > to a spiderweb then a tree.
> >
> >
> > Don't like it at all.  "Yuck.  I don't need a spiderweb.  That's 
> > messy and ugly."  I'm also not sure that there's anything true about it.
>
>
>  Do you have an alternative metaphor? It's true its not strictly a 
> spiderweb (since it can only merge from common ancestors), but its 
> much less a tree... 

I'm not sure that there's any real distinction here that's worth 
trumpeting about.  I don't think it is "much less a tree."  arch happens 
to be very good at merging in an N-tree branch structure, aka star topology.

>  I suppose that arch has a "detour" model instead of a "tree model", 
> in that it assumes that while code may wander off for a bit, it will 
> eventually come back.

Nah.  No such assumption.  Good tools for when it does, however.

> Most revision systems don't assume that, so their merging sucks. 

True conclusion, wrong reasons.

Bob