Coda File System

Re: cvs coda?

From: Kragen Sitaker <>
Date: Mon, 29 Mar 1999 16:17:18 -0500 (EST)
Jan Harkes writes:
> | Nope, afraid not.  It wasn't nearly as big a loss as VNC would have
> | been; Coda isn't visually very spectacular even when it works, while
> | VNC is.  I just sat on the table and talked about Coda and what it
> | could do.
> Have you seen the logcmls/logreintegration/etc. tcl/tk/tix scripts?
> They are an attempt at presenting some of the information from codacon
> in a slightly more visually appealing way. They are still a far cry
> from a GUI ofcourse.

I haven't seen them, but they sound cool.

What I meant, though, was that Coda is a filesystem, and as such is
difficult to present visually in an exciting way.  That is certainly
not an indictment of Coda.  :)

> | > How's that? CVS gives every developer his own sandbox for development,
> | > so that people working on the same files will not influence others.
> | > However, in Coda, everyone is working on the exactly same files, with
> | > the small exception that they do not influence each other while working
> | > disconnected.
> | In Coda, when you open a file for write [in write-connected mode], how
> | soon can others see your changes?  Immediately?  After you close the
> | file?
> After you close the file. The file is stored on the server when the
> last writer closes it.

So if you did a 'cvs update' before each open for read, and a 'cvs
commit' after every close for write, working with CVS would be
indistinguishable from working with Coda.  (Except that it would be
much slower, and you'd still have to manually checkout and delete
sandboxes, and you'd have those CVS directories, etc.)

<>       Kragen Sitaker     <>
We are forming cells within a global brain and we are excited that we might
start to think collectively.  What becomes of us still hangs crucially on
how we think individually.  -- Tim Berners-Lee, inventor of the Web
Received on 1999-03-29 16:15:53