Re: [glob2-devel] New Net Code Is Priority

[Kai Antweiler]

> I think that deleting the current code, far from perfect but working, in the 
> hope it will motivate anyone making the new one is an *extremly bad idea*. I 
> strongly oppose it.
>
> If the problem is the difficulty to do branch with CVS ; we should think 
> instead at dropping CVS and going towards a more distributed version/patch 
> system. But I do not think destroying features are a good way to go forward.

Might work though.  Finding all bugs in scarcely documented code can
be harder than writing new code from scatch.

But please look for an existing library first.  Even if it has
not all the features we are used to.  If everything necessary
works and is stable and sufficient:  take it!

I'm glad that someone takes care of the network code, but I doubt
that Bradley is the right one.  You should have access to a network
to write/debug network code.  


-- 
Kai Antweiler