home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!dziuxsolim.rutgers.edu!pilot.njin.net!birchall
- From: birchall@pilot.njin.net (Shag)
- Newsgroups: comp.os.msdos.pcgeos
- Subject: The Adventures of Shag, in the GeoComm Dimension....
- Keywords: Debugging without a Debugger. Ack Ptooie.
- Message-ID: <Dec.26.01.42.57.1992.9076@pilot.njin.net>
- Date: 26 Dec 92 06:42:57 GMT
- Organization: Screaming in Digital, the Queensryche Digest
- Lines: 54
-
- This evening, I got bored, so I deliberately blew the living heck out of
- GeoComm something close to 3 dozen times, all for the sake of learning more
- trivial facts about its behaviour.
-
- The VT-100 "cs" (change scroll region) command is still the only documented
- culprit. If you are on a unix box, "echo '^[[1;24r'" can be a whole load of
- fun. (the ^[[ is gotten by hitting CTRL-V then ESC, by the way.) Typical
- results include random selections from the GeoCrash Jukebox.
-
- So far, nothing you didn't already know. Time to change that.
-
- First off: Only the first 5 lines of GeoComm's screen have this behaviour,
- at least on my system. I can connect to the unix, run Screen, run VI under
- screen, and make the system spit out as many "cs" codes as my heart desires,
- as long as the cursor isn't in the top 5 lines of the window when I do it.
- (Incidentally, GeoComm has lots of trouble redrawing EMACS windows when I'm
- working in the first few lines too, although it doesn't crash.)
- When I just echo the code itself from the command line, though, it doesn't
- matter where I am on the screen, it GeoCrashes.
-
- Secondly: VT100.TC lists code 25 as being:
- \E[%i%d;%i%dr
- That's esc-[-increment-integer-;-increment-integer-r
- /etc/termcap on this unix host lists the "cs" code as being:
- \E[%i%d;%dr
- That's esc-[-increment-integer-;-integer-r
- Note the lack of the second "increment" (%i) function. I modified VT100.TC
- by removing the second %i, and I _think_ I was actually able to insert lines
- in VI under Screen without GeoCrashing, but it was kinda difficult to tell,
- because it only put text in the top line of the window, ignoring the other
- 23 lines altogether. ;) [You try to fix one thing, and something else goes,
- this reminds me of my days as a DOS programmer ;)]
-
- So, once again, I do NOT save the day (foo, foo, foo!) but at least
- there's a little more food for thought. A brief recap:
-
- * The vt-100 "cs" code doesn't seem to bother GeoComm when the cursor
- is below the 5th line of the GeoComm window.
-
- * The stock unix vt-100 termcap lists the "cs" code as \E[%i%d;%dr and
- GeoComm's VT100.TC termcap lists it as \E[%i%d;%i%dr - not the same.
-
- As always, if anyone wants to apply these ideas and maybe find a fix,
- please feel free. Comments are also welcome (better here than in mail, someone
- else might actually find this interesting... ;).
-
- Have a happy holiday, or I'll put "cs" codes in my posts! *evil grin*
-
- -Shag
- --
- Shag | Operator, ShagNET | Editor of "Screaming in Digital"
- birchall@pilot.njin.net | Rutgers / NJIN | The Queensryche E-mail Digest
- birchall@njin.bitnet | dialup access for | queensryche-owner@pilot.njin.net
- shag@most.other.places | Burlington County | Anything Queensryche, every week
-