[lcdproc] TCP or .so

Matt madmatt@bits.bris.ac.uk
Tue, 27 Mar 2001 20:17:44 +0100 (BST)


William W. Ferrell mentioned the following:

| --- andreb@rd.bbc.co.uk's mailer spewed these beefy chunks ---
| > 
| > Yes, right. But here are platforms who don't support this. On these it's
| > the same as 0.4
| 
| Time to reveal more ignorance on my part. Which platforms don't support
| dynamic loading?

AFAIK, none of our currently supported platforms would suffer. They all
have an appropriate interface for us to use.

One point I think I should clear up, is that the drivers aren't libs that
are linked at compile-time. They are loaded at run-time, and therefore we
can't use the static compiling approach without some modifications. This
is because LCDd won't have any dependencies on the driver code such that
the linker will "stitch in" the drivers on compiling. Even though you
would pass -lMtxOrb, ld will disregard it.

I think it would be possible to solve this, but I'll have to have a think
about how to best solve it nicely. I'm afraid a couple of #ifdef's may be
required...this is all dependent on whether we need to solve this problem
immediately.

We can make the names case-insensitive too, well parsing anyway. I like
MtxOrb and CFontz... :)

Matt



-----------------------------------------------------------
To unsubscribe from this list send a blank message to
lcdproc-unsubscribe@lists.omnipotent.net