* 2004-12-09, Geoffrey Brown:
> I downloaded code composer essentials. It appears to be eclipse + gcc
> +gdb. Anybody know if the
> gcc and gdb ports are related to those on the mspgcc sourceforge page ?
> Also, it doesn't appear
> that TI has made the gcc and gdb sources available; the downloadable
> sources appear to be
> for eclipse.
http://focus.ti.com/docs/toolsw/folders/print/msp-cce430.html
Even if CCE is based on the GNU toolchain without GNU GPL's "make source
available in case of binary distribution", i think, after many years of
GNU philosophy brainwashing and FOSS development experience, that it is
better than nothing.
Better in sense, that it is TI's supported product. The no-charge
version is good suitable start option for users and MSPGCC active
developers:
* as it seems to me, there are almost no pure-non-Windows(R) people here
* 8k is quite better, than IAR's 4k
* without fsck-ups form compiler vendors
** TI can control quality for both uCs and Ccs/asms
** same for mortals uC users -- emb. developers
- bad thing is, that generally it's not fair, if codebase was started
from the original msp430 GCC port and nothing goes back.
Well, from the historical perspective, GCC was holy cow for Cygnus for
many years. They had many private GCC branches for many `customers'. I'm
sure (see wikipedia last external link) something wasn't published as
now is restrictively (sue FUD) required as per GPL.
Not fair, *if* msp430 port developers require themselves more than 8k
binary and much quicker uC support/errata patches. *But*, it is fair to
allow to link many binaries in more than 8k final. After all many of
FOSS developers deserve some sweets and cakes sometimes. If asm output
is encrypted, then it isn't possible technically, if license is
restrictive, then...
Publishing source is low hanging fruit for any kind of proprietary
compiler vendor's piracy. Making output binary linkink format easy to
reverse engineer, is one for TRUE-developers :).
OTOH, "sharing" and "maintaining" codebase is a very deep FOSS
problem.
But all this is not the case for Atmel's brand new AVR32, as it seems...
http://atmel.com/dyn/products/tools_card.asp?tool_id=4118
--
-o--=O`C
#oo'L O
<___=E M
