ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Steve Loughran" <stev...@iseran.com>
Subject Re: New Source Control Tasks
Date Wed, 26 Sep 2001 06:16:25 GMT

----- Original Message -----
From: "Jesse Stockall" <jesse@cryptocard.com>
To: "ant-dev" <ant-dev@jakarta.apache.org>
Sent: Tuesday, September 25, 2001 13:08
Subject: Re: New Source Control Tasks


> On Tue, 2001-09-25 at 12:08, Steve Loughran wrote:
> > Hi Jesse.
> >
> > Ignoring the amusing fact that you really do need a third party add on
to
> > make VSS work over long haul links, and the even more poignant fact that
> > sourcesafe has a well founded reputation for not keeping your source
safe,
> > your contribution is indubitably welcome.
> >
>
> SourceSafe is not my choice, The *nix guys here are outnumberd by the
> Windows folks so we are stuck with VSS. At least SOS gives us
> connectivity from Linux & Solaris. I've never tried the remote access
> part of SOS, but it does the job for internal use.
>

ok. you just need to be aware that if source safe has a failing, it doesnt
always keep your source safe. I had an incident checking in a document once
when someone else filled up the shared drive during the process. The check
in failed, the entire doc history was lost, and even the document took a
while to locate to where it had been stuck. That kind of put me off it, even
if has a reasonable UI -the authors dont have a full grasp of the concept of
atomic transaction. The fact that the manuals are so full of coverage of
database recovery process is another warning: if they waste paper on it, it
must be a serious and recurrent incident.

Which is better: not having an SCM system and knowing you dont, or thinking
you have an SCM system and finding out you dont?

-steve


Mime
View raw message