small changes - larger aggravation --

asmboy · Joined: 20 Nov 2007 Posts: 2128 Location: albany ny

post install of 4.128 - command line version-
i don't know if i should laugh or cry.

suddenly in this version - the extension ".c" is explicit .....
as in the full source file name tagged with .C is required to compile.
( it didna used to be that way - but now - an improvement ??)

Another weird thing that is inscrutable and a real PITA

WHY would they change the FUSES in an older device file?

the 4.085 device files for the 18f4525
( of which i just ran afoul) define BORV28 as a valid fuse

Now in the "new improved" device header - ? only BORV27
is acceptable.

am i simply not in tune with the reason why header file content should suddenly need to be different?

I am getting a very bad feeling about how fuses are handled --
as altering long stable, defined - text descriptors does NOT seem like a clever way to advance usability, especially if i have to edit old files for bloody FUSES!!!

i suddenly realize the "misery pill" i have taken in order to compile for some new 18xxK parts and the feeling is a tad less than wonderful.

temtronic · Posted: Tue Jan 03, 2012 8:16 pm

There's something to be said about 'NOT" upgrading ! Whether it's hardware or software, I've found it's best to work with what you KNOW is good and solid and NOT take the 'Upgrade train'.
Just because it's 'new', 'more features','faster','better',etc. doesn't mean it really is, in the long run.
Having to debug 'this worked' before version xxx or 'that pin' used to do this...is a huge loss of time and money.
I feel your pain. Yeesh, I lost all of today(14 hrs) trying to get a Vinculum to be both 'host' and 'device'. Yes, technically possible, but I ain't got microscopes for eyes to tie a resistor to one of the 44 pins that's thinner than a human hair!
Yup, guess I'm getting old and losing patience on this microSMD stuff.

asmboy · Joined: 20 Nov 2007 Posts: 2128 Location: albany ny

more of what i mean about fuses

FvM · Joined: 27 Aug 2008 Posts: 2337 Location: Germany

Are you saying that you changed from V4.078 to V4.128? A number of compiler changes has been accumulated through the years...

asmboy · Joined: 20 Nov 2007 Posts: 2128 Location: albany ny

4.085 actually
but the changes i NEED were new chip support -
finding a count-zero bug in array indices is not fun -
and what i was bitching about was not liking
changed params in old header files - that
break how fuses are interpreted in teh newer version ;-((

i think #fuses is one of the WEAKEST - most poorly
documented parts of the CCS compiler -

what i would LOVE to see is an expert setting like this:

i would prefer to leave out #fuses statements
and spec this as valid

#FUSEREG 0x30001 0x3A
#FUSEREG 0x30002 0x03

etc etc

i bet it can be done w/o a special #directive - SO
if the wiser heads here can pont the way - i would be grateful

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

The CCS manual says this: