On Mon, 28 Jan 2013 15:25:19 +0100 Gilles <[email protected]> wrote:
> On Mon, 28 Jan 2013 11:22:11 +0100, Lluís Batlle i Rossell > <[email protected]> wrote: > >> Is there a way to avoid writing all those files to disk, and just > >> output data to STDOUT and read this with grep? > > > >I don't know how to do that. > > Thanks. I'll see if there's a way to create some kind of temporary > disk in RAM. If you use FreeBSD, this is what i have in my central repository server: In /etc/fstab: md /mnt/fossiltmp mfs rw,-M-S32m 2 0 # 32 MB memory disk I use intensively this because in each sync i, or someone, make to my central repository server this script runs: #!/bin/sh cd /mnt/fossiltemp fossil open $1 makeheaders -f makeheaders.dat scan-build -o/mnt/fossiltemp/stats/ -vv make fossil addremove fossil commit -m $(date) fossil close rm -rf /mnt/fossiltmp/* This way i create the html pages from clang/llvm statical analyzer and commit them to the project when some one . It's fast, i can live without 32MB of ram on this server but not clean, because a new commit is done each sync. You can use a similar approach for your grep problem and if you are trying to find when you add something use fossil bisect --- --- Eduardo Morras <[email protected]> _______________________________________________ fossil-users mailing list [email protected] http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
