RE: [Gnu-arch-users] Re: recent changes

[Derek Zhou]

I can make some patches so that tla will complain loud if a greedy revlib is 
not found. Also how about a default greedy revlib in either ~/my-tla-revlib or 
/tmp/uid-tla-revlib? 
With a revlib tla is quite usable for project with a medium-large size (<10k 
files) and that can cover 99% of all free software projects.  
Those changes are probably tla 1.4 things.
Derek

> -----Original Message-----
> From: Andy Tai [mailto:address@hidden 
> Sent: Sunday, November 20, 2005 1:12 AM
> To: Derek Zhou; address@hidden; Matthieu Moy
> Cc: address@hidden
> Subject: RE: [Gnu-arch-users] Re: recent changes
> 
> Well, Arch 1 may always suck in some way in this regard... I 
> remember people 
> complaining about the need to manually specify the use of the 
> revision libraries... and every 50th
> is just a heuristic.  Anyway, maybe after the next release we 
> may then adapt the baz behavior,
> with the number of 50 as default but adjustable, and then we 
> can further explore if even better
> algorithms can be applied on top of the Arch 1.x archive formats... 
> 
> Based on my understanding, I am afraid performance on large 
> trees will be a continuous headache
> for tla... why, say, tla was not a good choice for the Linux kernel...
> 
> 
> --- Derek Zhou <address@hidden> wrote:
> 
> > One thing that have frustrate lots of newbies (including 
> me) is the fact that to get reasonably
> > good performance in any realistic sized tree, you have to 
> have a revision library. So why do we
> > just enforce this policy in the software so nobody 
> accidentally say arch sucks?
> > Derek  
> > 
> > > -----Original Message-----
> > > From: Andy Tai [mailto:address@hidden 
> > > Sent: Friday, November 18, 2005 1:56 PM
> > > To: Matthieu Moy; Derek Zhou
> > > Cc: address@hidden
> > > Subject: Re: [Gnu-arch-users] Re: recent changes
> > > 
> > > 
> > > Yes,  cache reversions every 50th one may be useful but 
> > > should be done in a smarter manner...  if
> > > the baz algorithms can be applied in a less disruptive way 
> > > that would be great...
> > > 
> > > --- Matthieu Moy <address@hidden> wrote:
> > > 
> > > > address@hidden (Ludovic Court�s) writes:
> > > 
> > > > >> * cacherev every 50 revisions and every tag even 
> within the same
> > > > >> archive. Disk is cheap
> > > > >
> > > > > While I agree this should be the default, I think it 
> should not be
> > > > > hard-wired.
> > > > 
> > > > In particular, cachedrevs for all tags are a bad choice if you
> > > > microbranch a lot. It does not only cost disk space, it 
> also costs
> > > > bandwidth: if you have a close ancestor in your revision 
> > > library, it's
> > > > cheaper to apply a few changesets to it than to get the cached
> > > > revision. Bazaar has clever algorithms to chose which full tree
> > > > revision to start with (a cachedrev, the initial 
> import, or in your
> > > > revision library), but that's relatively deep changes, 
> I don't think
> > > > this will ever be merged into tla. 
> > > > -- 
> > > > Matthieu
> > > 
> > > 
> > 
> > 
>