home *** CD-ROM | disk | FTP | other *** search
Text File | 1991-03-06 | 1.6 KB | 42 lines | [TEXT/GEOL] |
- Item 5550398 30-June-88 06:31
-
- From: D0619 Palomar Software, Dev, Joel West
-
- To: MACAPP$ MacApp Interest List
-
- Sub: Limited vision
-
- Greg is of course right about arbitrary limits. Sure, FORTRAN programmers
- have a good excuse for arbitrary limits, but everybody else doesn't.
-
- As a software design psychiatrist, one of the most common ailments I see
- is limitosis--a needless limiting of one's horizons and futures. Remember
- 640K? We laugh now, but some dweeb six years ago thought it was a
- reasonable limit for MS-DOS. Remember 32K resources?
-
- Limits would be fine if you knew with 100% certainty what your software would
- be used for throughout its lifetime. OS and compiler writers never do. Ever.
-
- Suppose you chose an approach that has a limit, and you were wrong. Be
- creative. For example, Greg suggests we could go from 4096 to 8192 entries
- in the JT using a 16-bit signed (or unsigned) displacement.
-
- As for this neo-fascist sanctimoniousness, "Our limits are good for you, so
- you should thank us for them," it seems like someone has their role confused.
- Is it as a preacher, or as a supplier of useful software? Based on my years
- of experience, the correct answer by anyone who works for me to the complaint
-
- I keep bumping up against your limits...
-
- is
-
- You want them bigger? Sure, no problem. How big?
-
- This is particulary true for something like MPW, where you can easily add
- compiler and linker options to let the user specify his own preferred table
- size, without increasing the resources required by the ordinary guys.
-
- Joel
-
-
-