madd@bucsb.bu.edu.UUCP (02/03/87)
People have already sent in lots of suggestions on MORE v1.2 that I
recently posted. First, there will be at least one more posting.
This posting has several planned fixes:
* ^H will really backspace. This fixes some hilighting problems
that were reported
* The problem with tabs will be fixed (it's already fixed on
v1.3 on my machine, but with other fixes coming I'll do them
before posting)
* Carriage returns will function properly
These are the only fixes currently in the works. Because of the time
available to me when I posted, I'm afraid I may have left several
things out of my posting that might be important to people.
* The "faster" character io that I had been using has been
removed in favor of Turbo Pascal's built in char io, with
the optional larger buffer. Performance tests indicated
little performace loss and there was lots less code.
* Carriage returns are completely ignored. This was for
speed, but I found a neat way to keep it fast and still
support them, so they'll work fine next version. This
wasn't usually a problem, but can be if you're dumping
formatted text to MORE.
* Newline is interpreted as CR/LF. It will be properly
handled in the next version, though this could give some
problems when listing UNIX files.
* It can store 3x as many lines as before. I cut down the
size of the structure used to store lines to a minimum.
* The bug that cropped up on machines w/ > 512K has been
fixed. This resulted from negative return values from
Turbo Pascal's maxavail function. Caveat programmer!
* A couple of minor screen errors were fixed (mostly having
to do with scrolling single lines near the top of the file).
They work fine now.
* If possible, MORE.COM shows the percentage through the file.
It took considerable coaxing to make TPascal QUICKLY main-
tain counts of characters when you can have a file up to
32Mb in length. On short files, however, the percentage can
be quite wrong. If a file is reported by DOS to be 512
bytes long and it actually is 2 bytes long, this throws my
percentages all off. Generally, this isn't too bad and
should be hard to notice because it prints "end" when you
are at the end of the file.
* There is a definite bug in the tab function. It refuses
to tab if the tab was received while the cursor was on
a tabstop. This has been fixed in all subsequent versions.
* The screen output routines completely ignore the BIOS, so
there should be no difference in speed between systems
with advaced scrren utilities such as fansi console.
Thanx to everyone who replied. If you find a bug, mail it to me fast.
Also suggestions. These sometimes prove as helpful as bug reports.
I'll be posting the newer version in about a week, depending on when
I get hold of a modem. Also, thanx to everyone who sent me Xmodem
sources and especially whoever that was who sent me procomm. Life is
much easier when you don't have to crossload via text!
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
- Jim Frost * The Madd Hacker -
UUCP: ..!harvard!bu-cs!bucsb!madd | ARPANET: madd@bucsb.bu.edu
CSNET: madd%bucsb@bu-cs | BITNET: cscc71c@bostonu
-------------------------------+---+------------------------------------
"Oh beer, oh beer." -- Me | [=(BEER) <- Bud the Beer (cheers!)