#STRTRK broken (!?)
10 messages in this thread
Gosh, Toto, I don't think we're in Kansas!
O.K., I _know_ I played STARTREK several times the other evening, but when
I powered-up today, I got the same behavior that Paul reported last week.
I haven't changed my CONFIG.SYS or my AUTOEXEC.BAT for a long time, so
nothing has changed in that regard. I even tried such wild&crazy stuff as
making sure that the printer was on first (??)
Now I don't know what's going on here, but I'm going to try replacing
COMMAND.COM from the original SS disk, and pray that it's not the boot
loader that's flakey. Yup, today is the day after Friday the 13th, and I'm
starting to think "virus"!!
No, I'm not saying that STRTRK is infected, but rather that the failure of
the Microsoft run-time (compiled into STARTREK) is tripping over a "benign"
viral infection, which is why it worked on Thursday, but not today. (I've
already changed the MS-DOS clock to try other dates… no help there.)
Several people have reported that it works fine, and it worked for me until
today. Could the others with "working" copies _please_ try theirs and see
if they've gone south as well? This smacks of Foul Necromancy, and I'm not
very handy with Bell, Book, and Candle!
Dennette
Dennette,
I did run a not-entirely up-to-date version of ViruScan over both
STRTRK.EXE (the self-extracting file) and STARTREK.EXE (the program itself)
and got no viruses reported. The version of ViruScan that I used was a
couple months old. I'll go get the latest version and report back if there
are any problems.
— Carl
Thnx, Carl … this is a real "puzzlement".
Oops (vapor lock!) I'm convinced that STARTREK is virus-free, but I'm
concerned that my version of MS-DOS may be corrupted because it's the MHFU
interrupt (I think) that's activating. I mean, even hitting the SHIFT key
kills it, so it's the keyboard interrupt, and that makes me suspect the OS.
(Does this make sense to you?)
Dennette
Dennette,
That would require someone to have written a virus for the Rainbow. It's
obviously possible for someone to write generic MS-DOS viruses as well as
IBM-PC viruses, as well as Rainbow viruses. Most of what I have seen
described appears to be IBM-specific, though some are at least partially
generic MS-DOS.
But if there's a virus that attacks the Rainbow's keyboard interrupt,
we're talking about a hacker who's strange in more ways than one.
— Carl
AAAARRRRGGGGGGGGHHH!!!
O.K., 'Bow Babies, sit down for this one …
There is a _bug_ in the VT100 emulation that invol\@ves "RESTORE CURSOR",
or "<ESC>8" – if you have _not_ done a "SAVE CURSOR"{first, then it goes
south as described in this message thread. THERE IS NOT VIRUS!!!
After having used LC-TERM to call my VAX and use EDT on some files, i]I was
able to run STARTREK. Reboot, and STARTREK bombs the first time. Called
the VAX again, used EDT, came back, STARTREK works! Checked the BASIC
code, pondered, added a "SAVE CURSOR" to the start-up code, rebooted, and
it works like a charm!!{/paranoia) and the expense of the download of a bad version. I'll rebuild
the executable and upload a new version this evening. (Boy, is _my_ face
red!)
I suspect that KEDT and other utilities that generate a "<ESC>7" (SAVE) can
be run first as well, and that's what I must have been doing while
developing this, otherwise I would have run into it sooner. I think that
{thSAVE/RESTORE uses a buffer rather than a stack, but it's all the same,
really – bad programming (on my part!) to try to "restore" an uninitialized
value. Ah, well, just shows you're never too old to learn a new way to
make a mess!
Dennette
Dennette: I discovered that ESC 8 would crash my 'Bow last March (I made a
little file with ESC 8 in it so I would remember)
and I'm glad to know "why". (And why it only happens sometimes…)
The infamous "Rainbow Owner's Manual" Appendix H, under "Restore Cursor/ESC
8",
states clearly (and apparently wrongly):
"If none [i.e. no cursor] were [i.e. has been] saved, the cursor moves to
home position."
BFT
Thnx, Brian – I'm glad I'm not the only one who's been misled by
documentation!
BTW, just got creamed by the C run-time routine "atof()", which takes a
string and returns a double… if the string is in double-precision
notation (1.234D-12, as opposed to 1.234E-12), then the exponent is
_ignored_!! The Microsoft C version accepts "E", "e", "D", and "d", but
"standard" C only supports "E" and "e", so it stops parsing when it hits
the "D" or "d". Needless to say, the code I ported from the 'Bow to my VAX
and SUN did not give correct results, but it's been that way for the last
TWO YEARS and I just noticed it this week!
Share & Enjoy! Dennette
Dennette: After I collected a few howlers from the 'Bow "Owners Manual" I
filled out the "Manual Evaluation Card" in the back and sent it in. That
was last week. Do you suppose Digital will prepare a new edition
incorporating my corrections?! (The ESC 8 section wasn't one of them.)
I keep hearing that I ought to learn C, but then I keep hearing about these
problems with floating-point data types…
BFT
Carl – RE: your comment about a virus specific to the 'Bow…
My thoughts at the time were that it was a virus specific to the IBM, so it
was only _partially_ effective on the 'Bow! I dunno, but that does make a
crazy kind of sense… there are enough differences in the BIOS that it may
not have been able to attach itself completely, and it only manifested
itself when certain applications (like this game) were run.
Turns out that it was sloppy programming on my part, and once I was able to
make it run or fail "on demand", I took a closer look at the code, with
special attention to the cursor control commands. Please replace your
version with the new one (when it's available), and have fun!
Dennette