[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnu-arch-users] multi-committer functionality revisited

From: Tom Lord
Subject: Re: [Gnu-arch-users] multi-committer functionality revisited
Date: Mon, 17 Nov 2003 16:12:16 -0800 (PST)

    > From: Robert Anderson <address@hidden>

    > > Why don't you use some non-local transport for that archive, then?

    > Because I want the local performance (but it's a negotiable point). 
    > However, I don't think it solves the problem, does it?  The only
    > transport available to me is sftp, and I wouldn't know how to solve the
    > problem there, either.

Because the user that the sftp daemon runs as owns all the files and
directories in the archive.

Oh... I see... you're thinking that to use SFTP you have to give each
member of your group a distinct login.   No, that isn't what I had in
mind.   Perhaps I'm violating your "no new accounts" constraint.

    > > Alternatively, looking (at least briefly) back at the Xouvert thread,
    > > it seems we overlooked a sane solution at that time (blush.  But then
    > > that's why I proposed the "make a new savannah account" solution after
    > > all - to buy time to think):

    > > While I would be against the "copy permissions" hack and against a
    > > "tla umask" command, I would not be against support for archive 
    > > URLs of the forms (suitably adjusted if I've violated uri syntax):

    > >         file%umask=XXX://path/to/your/archive

    > >         sftp%umask=XXX:/address@hidden/path/to/your/archive

    > >         sftp:/%umask=XXX,address@hidden/path/to/your/archive

    > >         etc.

    > That's how you'd register the archive, I presume?


"morally speaking", these would just be extensions to the capability
of a transport rather than changes to the "dumb server" model of arch.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]