[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Gnu-arch-users] Re: designing how to kill pristine trees
From: |
Miles Bader |
Subject: |
[Gnu-arch-users] Re: designing how to kill pristine trees |
Date: |
07 Oct 2003 16:55:12 +0900 |
Denys Duchier <address@hidden> writes:
> > So I'm not exactly sure how tla could accomodate this, other than by
> > having a hardwired algorithm like the one I previously suggested.
>
> tla caching-method?
The issue in my mind was whether it was OK to have `special purpose'
algorithms like that at all (well I think my algorithm's generally
useful actually but at least some people disagreed).
I'd like to have a little more control than just a toggle though, if
possible, e.g., maybe my invlib path could look something like:
sibling#.pristines:sibling#.revlib:/usr/local/arch-pristines:/usr/local/arch-revlib
[Where `:' is the path separator, and `sibling#FOO' is some magic syntax
meaning `use a "sibling" dir called FOO']
However, I suspect there will be cries that the above is ugly, and it's
probably unnecessary if tla would call a script to construct the
revlib-path.
For instance, here's a script that implements the algorithm I described
earlier. A small twist being that it actually finds _two_ revision
libraries for the sources, one for `pristines' (automatically managed
cache), and one for `library entries' (user managed):
#!/bin/sh
# Construct a tla revision library path.
# CWD should be in the tree tla wants to operate on.
# The returned path is separated by `:'; the first entry is considered
# special in that it is used for automatically cached revisions.
# Default
REVLIB_PATH="/usr/local/arch-pristines:/usr/local/arch-revlib"
# Find sibling root
SIBS=""
DIR="`pwd`"
while :; do
test "$DIR" = / && break
cd ..
PARENT="`pwd`"
test -r "$DIR/{arch}/.arch-project-tree" && SIBS="$PARENT"
DIR="$PARENT"
done
if test "$SIBS"; then
REVLIB_PATH="$SIBS/.pristines:$SIBS/.revlib:$REVLIB_PATH"
fi
echo "$REVLIB_PATH"
-Miles
--
`To alcohol! The cause of, and solution to,
all of life's problems' --Homer J. Simpson
- [Gnu-arch-users] Re: designing how to kill pristine trees, Pau Aliagas, 2003/10/06
- [Gnu-arch-users] Re: designing how to kill pristine trees, Tom Lord, 2003/10/06
- [Gnu-arch-users] Re: designing how to kill pristine trees, Pau Aliagas, 2003/10/06
- [Gnu-arch-users] Re: designing how to kill pristine trees, Tom Lord, 2003/10/06
- Re: [Gnu-arch-users] Re: designing how to kill pristine trees, Florian Weimer, 2003/10/06
- [Gnu-arch-users] Re: designing how to kill pristine trees, Pau Aliagas, 2003/10/06
- Re: [Gnu-arch-users] Re: designing how to kill pristine trees, Miles Bader, 2003/10/06
- Re: [Gnu-arch-users] Re: designing how to kill pristine trees, Denys Duchier, 2003/10/07
- [Gnu-arch-users] Re: designing how to kill pristine trees,
Miles Bader <=
- [Gnu-arch-users] Re: designing how to kill pristine trees, Miles Bader, 2003/10/07
- Re: [Gnu-arch-users] Re: designing how to kill pristine trees, Andrew Suffield, 2003/10/06
- [Gnu-arch-users] Design proposal to kill pristine trees, Pau Aliagas, 2003/10/08
- [Gnu-arch-users] Re: Design proposal to kill pristine trees, Miles Bader, 2003/10/08
- [Gnu-arch-users] Re: Design proposal to kill pristine trees, Miles Bader, 2003/10/08
- [Gnu-arch-users] Re: Design proposal to kill pristine trees, Pau Aliagas, 2003/10/08
- [Gnu-arch-users] Re: Design proposal to kill pristine trees, Miles Bader, 2003/10/08
- [Gnu-arch-users] Re: Design proposal to kill pristine trees, Pau Aliagas, 2003/10/08
- [Gnu-arch-users] Re: Design proposal to kill pristine trees, Miles Bader, 2003/10/08
- [Gnu-arch-users] Re: Design proposal to kill pristine trees, Miles Bader, 2003/10/08