Mail Archives: djgpp-workers/1996/09/13/05:41:06
On Thu, 12 Sep 1996, John M. Aldrich wrote:
> 1) Would it be better to design the program as a plain text program or
> using some sort of windowing interface?
I vote for plain text, at least at first. What I had in mind was a
program that produces simple text report, that should be easily
redirected into a file (to be posted later together with a problem
report).
> is a good idea, I would welcome pointers to sample code to implement a
> text-based windowing system (using conio functions or whatever).
If you do decide to go fancy, use PDCurses: why reinvent the wheel?
> 2) Should I even write the program with djgpp? A significant number of
> users fail to install a DPMI host when they install their files, yet the
> program would require DPMI to run.
As DJ pointed out, you will have to make sure that DPMI-related errors
detected early on by the stub should print human-understandable error
messages. Maybe the easiest way to do so is to change the stub slightly
so that it returns different exit codes for different problems. Then you
could write a batch file that will print the actual messages. It's ugly,
but it will work for everybody, and doesn't require a special stub.
*Then* you can bind the program with PMODE to get a stand-alone
executable.
> - DJGPP files present
> - DJGPP files installed in the correct directory structure
> - Environment variables set
> - DJGPP version
> - Available memory
> - Evaluate PATH to look for other compilers' progs
> - Trivial program compilation
Seems enough for the first cut.
> Recommendations on the ordering of these tests?
Env vars (DJGPP and DJDIR), then the directory structure and required
files existence, then compile a test proggy (you can make it return a
code that tells you what version of DJGPP is installed: just look at
__DJGPP__ and __DJGPP_MINOR__), then look at memory, then PATH.
> 5) Here are some additional ideas for things to look at, but I'm not sure
> how workable/necessary they would be:
I'd say the above is enough for the first version. Additions may be done
later.
> - Debug DJGPP.ENV
Add this to the list above. I suggest to look for these problems:
- Changes from the original contents distributed with DJGPP (you
can encode a CRC of it inside the program)
- Trailing ^Z or whitespace
- Butchered DJDIR line
> - Implement recommended changes to system files
Never. Just tell them what do you suggest and let them do it.
> 6) Any other ideas for features to include?
Later, as the program gains experience.
> 7) What should I call it? ;) "DJDIAG" comes to mind, but I understand
> that only DJ himself is allowed to write programs starting with his initials.
He can grant you a one-time license to do that ;-).
I vote for `djverify' or `djtest'.
- Raw text -