Janulewicz, Matthew <MJanulewicz <at> westernasset.com> writes:
> If the original poster is doing this because his boss said to, and the project
goes awry, it's his head, not
> his boss'. Although Tommy maybe wasn't thinking along these lines, it's good
to help folks out in other
> ways beyond the original scope of the question.
...and you're suggesting that the project is more likely to 'go awry' on a
Windows server? Ergo SVN on win32 is less stable that on *nix????????? Can the
devs comment on this?
For every arguement 'Windows is slower per mhz' (may well be true) there is a
counter arguement 'Windows servers are cheaper to maintain, and cheaper to
recruit staff for'. In fact a similar discussion swept the development world
about 15 years ago, and now we all use 'fast-dev slow-running' RAD systems, with
much improved efficiency LOL. Another great example of this is managed code,
Java / .NET / Mono - much slower to run than native C, but so much safer it
works out cheaper in the end for many systems.
> My point to add is that with any application that is highly active on the
filesystem, Windows is a poor choice.
Oracle is deployed on more Windows servers that on all the others put together.
And we all know how light-weight Oracle is on the file system...
I'd just like to point out - I've not critizied *nix here, in any way. I just
get angry when Windows is made out to be a second-class platform in some
peoples' eyes. Like 'we don't need to implement core case-insensitivity in SVN'.
Why? Because all *real* os's are case sensitive!!!!! (no guys, perl / python
scripts don't count. It would be like having a perl script to cope with '/' in
directory names to make in Unix compatible!)
Lets cut the OS-facism, and try to be equal partners, yeah?
John
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Sat Jul 23 18:27:38 2005