On Thu, Aug 11, 2011 at 5:21 PM, Richard Hipp <d...@sqlite.org> wrote:
> > > On Thu, Aug 11, 2011 at 4:53 PM, lists <li...@devilray.eu> wrote: > >> I'm a new user of fossil, having come grudgingly from CVS. Needless to >> say, my stubbornness was unfounded-life is immeasurably easier than it was >> on CVS for hundreds of little (and big) reasons. >> >> In trying to convert my workflow to fossil, I haven't been able to find >> any information in the wiki nor mail archives about creating a new >> repository, remotely, on the fossil server. > > > For a server you need two things: (1) The repository database file (the > *.fossil file) and (2) some mechanism to serve that file. > > Item (2) can be either CGI, or inetd/xinetd, or "fossil server". See the > documentation for details. In any of these cases you have to give it the > name of a repository database file to serve. > > But here is a cool feature: The name of the repo database file you give to > Fossil for (2) can actually be a directory rather than an individual file. > In that case, Fossil will serve all repos underneath that directory. > An amplification: In order for this to work, the repository files need to be named with a ".fossil" suffix. Other suffixes like ".fsl" or ".f" or anything else. If you are dealing with individual repositories, the suffix does not matter - it can be anything you want. But for this one server-every-file-in-a-subdirectory feature, all the repositories files have to end with ".fossil". > > Suppose you have (2) set up to serve files out of the /home/www/repos > directory on your server. Then in order to create a new repository on the > server you can do this: > > (a) Create the repository locally using "fossil init repo-name.fossil" > (b) Do whatever check-ins and configuration you want on the new repository, > including setting the administrator password. > (c) Test your setup locally using "fossil ui" > (d) Scp or ftp the repository file into the /home/www/repos directory on > the server. > > If you upload a repo file named /home/www/repos/abc.fossil" then you can > access it using http://domain/abc. If you upload the file to > /home/www/repos/dir1/dir2/xyz.fossil, then you access it using > http://domain/dir1/dir2/xyz. And so forth. > > So once you get (1) up and going, installing a new repository is just a > matter of uploading a new repository file. > > > >> I understand that in a large project and a tightly controlled server this >> may be undesirable, but in a home or small office environment, this is very >> useful without having to resort to log on to the server, issue the "fossil >> new <repo>" command and logging off, especially if you wish to restrict >> general log-in access to the server itself. >> >> One use for this is when creating static web sites; a new project comes >> along and whichever developer starts work on the project first simply >> creates the repository and commences work. Everyone else merely carries on >> as normal; "fossil clone <repo>" and "fossil open <repo>", etc. >> >> Am I missing something, or is this simply not a feature that exists yet? >> >> Sacha >> _______________________________________________ >> fossil-users mailing list >> fossil-users@lists.fossil-scm.org >> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users >> > > > > -- > D. Richard Hipp > d...@sqlite.org > -- D. Richard Hipp d...@sqlite.org
_______________________________________________ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
