Rahul Dave said:
> Yes, infact there is no reason why 3-or 4 way sync cannot be achieved..
> by which I mean
> (a) home machine
> (b) office machine, both permanentlty or semi-perm connected
> (c) laptop
> (d) upstreamer, cacher, backuper.
>
> The benefir of the backuper here is that I can write something at home, when
> I login from office or laptop, I get the latest stuff.
It's there and it's called CVS ;)
Most of my development work, I use a CVS server on a permanently-connected
machine (the upstreamer in that terminology), and check in any changes.
Then, the (a) (b) and (c) machines just have to do "cvs update" to get the
latest version. Very handy for multi-site edits and synchronisation.
CVS is creaking at this stage a little though, but Perforce (commercial)
or Apache's Subversion (DAV with a working V!), or possibly Jonathan
Shapiro's DCMS, works/will work a lot better.
Having said that, these CMSes (in the older terminology, "config mgmt
system" IIRC?) only deal well with data that changes in relatively small
jumps. Very changeable data, like my mail spool, is a bit harder to
sync. :(
I'd like to see a generalised multi-site synchronisation algorithm
implementation for use in a range of stuff, e.g. mail syncers, etc. It's
a big problem for decentralised working, and crops up with all kinds of
systems...
I'm sure there's lots of research into this one I'm just not aware of.
I know there's a 2-way sync tool available called Unison, which I've
been meaning to look at; can't find the URL right now though. :(
One note: CVS doesn't do any automatic downgrading though. IMHO I'm not
convinced it should be built into the multi-way sync technology however,
IMO it should be a separate service or app.
--j.
This archive was generated by hypermail 2b29 : Fri Apr 27 2001 - 23:18:28 PDT