mirror of
https://github.com/GnoConsortium/gno.git
synced 2024-12-23 05:30:18 +00:00
72 lines
2.7 KiB
Plaintext
72 lines
2.7 KiB
Plaintext
|
Well, first off:
|
||
|
|
|
||
|
|
we are: Mike Horwath
|
||
|
|
James Brookes
|
||
|
|
Dave Huang
|
||
|
|
|
||
|
|
5/12/92
|
||
|
|
Jawaid bitched about some stack overflowing, so, between us both, we searched
|
||
|
|
diligently for the error. Inside one of the functions, there was a stack
|
||
|
|
allocated variable over 2K in size. So, the fix was to make it a static
|
||
|
|
variable, so that it is gloabally allocated from normal memory, instead of the
|
||
|
|
stack. That error is now fixed.
|
||
|
|
|
||
|
|
Jawaid also has fixed the getenv() library call, and that was the problem with
|
||
|
|
using a $editor variable in LESS. That is now fixed.
|
||
|
|
|
||
|
|
Less now also works with either files terminated with lf's (unix normal) or
|
||
|
|
cr's (Apple IIgs normal). Works GREAT.
|
||
|
|
|
||
|
|
Fixed a bug where pipes would cause a problem if you shelled out '!' (alone),
|
||
|
|
added GS/OS specific code to take care of which stdin to use, either the pipe
|
||
|
|
or normal stuff...
|
||
|
|
|
||
|
|
5/11/92
|
||
|
|
Well, with big help from Jawaid "The God" Bazyar, piping and redirection are
|
||
|
|
now working correctly. Problems stemmed from file descriptors being set to
|
||
|
|
numbers that were not correct, meaning that, it was trying to write to the
|
||
|
|
input of a pipe or stdin.
|
||
|
|
|
||
|
|
JB put assembly into the putchr() and putstr() calls, noticable speed increase
|
||
|
|
and some savings on disk.
|
||
|
|
|
||
|
|
It prints the control codes by default, so that when you use man, it will do
|
||
|
|
things like inverse and the like correctly.
|
||
|
|
|
||
|
|
did a #pragma noroot so that it cuts off about 1K of exec file. Kind of nice
|
||
|
|
of jawaid to tell me this. Nice little savings I must say.
|
||
|
|
|
||
|
|
Reset the stacksize from 4.5K to 1K, as I was wrong about the stack usage.
|
||
|
|
Thanks go to jawaid for this bit of news also.
|
||
|
|
|
||
|
|
5/5/92
|
||
|
|
Well, another change this time to prchr(), so that you can pass terrible
|
||
|
|
command lines, and have them done right. Problem came from not outputing the
|
||
|
|
right escape type character for end of line stuff. So, that now works. NOW, I
|
||
|
|
only wish I had a terminal to test all this on, hope everythign is fine, if not
|
||
|
|
we are screwed for alot of work here.
|
||
|
|
|
||
|
|
Made the default as -e (let you hit space or return to quit when at eof) and
|
||
|
|
-m (medium prompt line, shows you the file you are viewing, nothing more).
|
||
|
|
|
||
|
|
5/4/92
|
||
|
|
changed most of the open()'s to use the correct modes of the file. solves
|
||
|
|
errors with coming back from vi and such.
|
||
|
|
|
||
|
|
changed the stacksize from default of 8K to 4.5K, seems to work like a dream.
|
||
|
|
|
||
|
|
put in an optimize -1 into the code.
|
||
|
|
|
||
|
|
Worked on the lsystem() call, as it is really messed in the head. Hopefully,
|
||
|
|
by the time you have read this, it will be written from scratch, and working
|
||
|
|
as it should.
|
||
|
|
|
||
|
|
5/3/92-5/4/92
|
||
|
|
Dave fixed all the problems with \n and \r's being wrong in places.
|
||
|
|
|
||
|
|
Dave also fixed the gno termcap entry, as sr and sl (am I right?) were not
|
||
|
|
correct for less to work right.
|
||
|
|
|
||
|
|
5/2/92-5/4/92
|
||
|
|
James and Dave did all the prototyping into the sourcecode.
|