jb10320@uxa.cso.uiuc.edu (Desdinova) (12/01/90)
Derek Taubert and I were just working on porting some 3-D object viewer software I wrote from an IBM RT to the GS. We were having some problems with the tool startup, so we figured since it was a desktop program we'd compile it under Prizm and take advantage of the source level debugger. Ha! 1) I loaded the file into memory (or so I thought). Random bits of the file were trashed when I scrolled through the file. First thing I figured was, "Well, maybe I have bad ram. My ramcard setup is pretty goofy." 2) So we looked thru all the blocks allocated to Prizm, to try and find the editor buffer so I could do some hand tests on the "bad" RAM. Much to our surprise, NONE of the blocks allocated to Prizm contained the file. 3) In a fit of impending anger towards what we imagined to be the worst programmer on the face of the earth, we did a bank by bank search of memory to locate the buffer. We found it. 4) THE BLEEDIN' BUFFER WAS NOT ALLOCATED TO *A_N_Y_T_H_I_N_G*. The file Prizm was supposed to be guarding was totally "free". We verified this by making some toolbox calls (allocated the unallocated block just to make it 'ours', then trashing it to see the results- file go boom!) and by using MemoryBar to Compact memory. Boom also. So, to make a long story short, DON'T DEALLOCATE A BUFFER BEFORE YOU'RE DONE WITH IT!!!! I thought they taught that in my first year CS course, but maybe Mike didn't attend an institution of higher learning (egads, even HIGH SCHOOL level.) Nothing personal, Mike. Anyone have an Amiga for sale :-) Or a shotgun? Grrr... Unknown! I hope US SNAIL hurries up and gets that PMMU info to me so we can finally do some REAL programming with REAL tools on the GS. Jawaid Bazyar Derek Taubert -- Jawaid Bazyar | Blondes in big black cars look better wearing Senior/Computer Engineering | their dark sunglasses at night. (unk. wierdo) jb10320@uxa.cso.uiuc.edu | The gin, the gin, glows in the Dark! Apple II Forever! | (B O'Cult) Comp.Sys.Apple2- Home of the Unofficial Apple II Developer Support Team (DST)
avery@netcom.UUCP (Avery Colter) (12/02/90)
jb10320@uxa.cso.uiuc.edu (Desdinova) writes: >1) I loaded the file into memory (or so I thought). Random bits of the > file were trashed when I scrolled through the file. First thing I > figured was, "Well, maybe I have bad ram. My ramcard setup is pretty > goofy." This, in my experience, happens when a file is loaded from a disk and the ORCA/C volume's files are not online anywhere. I'm not sure why this happens. I also seems to get window leaks, and when I do, I tend to just get the hell out. If I try to quit the normal way, there's a 50-50 chance the file I have active will get nailed. Hey, by the way, when are we gonna get together on a dark night and stuff some big black blondes in your shady car? -- Avery Ray Colter {apple|claris}!netcom!avery {decwrl|mips|sgi}!btr!elfcat (415) 839-4567 "I feel love has got to come on and I want it: Something big and lovely!" - The B-52s, "Channel Z"
gwyn@smoke.brl.mil (Doug Gwyn) (12/02/90)
In article <1990Dec1.082442.25195@ux1.cso.uiuc.edu> jb10320@uxa.cso.uiuc.edu (Desdinova) writes: >I thought they taught that in my first year CS course, but maybe Mike >didn't attend an institution of higher learning (egads, even HIGH SCHOOL >level.) Did it ever occur to you that perhaps that was a BUG in Prizm, not an intentional aspect of its implementation? Did you report this to ByteWorks? I believe Mike is working on the next release of the ORCA shell and Prizm. so there is a chance you could get this fixed fairly soon. In fact you might be able to volunteer as a prerelease test site. Ranting on the net accomplishes nothing worthwhile.
jh4o+@andrew.cmu.edu (Jeffrey T. Hutzelman) (12/02/90)
I, too have had problems with Prizm. My best suggestion is not to use
it. I never have since I started experiencing problems with it. It's a
real shame, because it could make a real nice debugger.
--------------------
Jeffrey Hutzelman America Online: JeffreyH11
Internet: jh4o+@andrew.cmu.edu BITNET: JHUTZ@DRYCAS
>> Apple // Forever!!! <<
jb10320@uxa.cso.uiuc.edu (Desdinova) (12/03/90)
In article <14631@smoke.brl.mil> gwyn@smoke.brl.mil (Doug Gwyn) writes: >In article <1990Dec1.082442.25195@ux1.cso.uiuc.edu> jb10320@uxa.cso.uiuc.edu (Desdinova) writes: >>I thought they taught that in my first year CS course, but maybe Mike >>didn't attend an institution of higher learning (egads, even HIGH SCHOOL >>level.) > >Did it ever occur to you that perhaps that was a BUG in Prizm, not an >intentional aspect of its implementation? Even so, it's a bug that should have been caught before it went beta, let alone commercial. THAT is what irks me. The reason for this is Mike is the only person doing major work on Pascal, C, the Assembler, and Prizm. The other people at Byteworks (his wife, for example) usually do incidental work like UI design or manuals. If he wants to have good products with realistic time frames he needs to HIRE MORE PEOPLE. I have more than a little experience with compilers- I'd be glad to work on Orca products for FREE! Just so I could have development tools that worked. > >Did you report this to ByteWorks? I believe Mike is working on the next >release of the ORCA shell and Prizm. so there is a chance you could get >this fixed fairly soon. In fact you might be able to volunteer as a >prerelease test site. I've done this many times (gotten betas to play with). In fact, I subscribed to GEnie just so I could send mail directly to Mike without paying long distance. I've personally discovered about three major bugs in Orca/C (a code generation bug which I reported long before 1.1 came out wasn't fixed, much to my dismay. He sent me a letter after 1.1 saying "the bug has been logged, and will be fixed in the new version". Whenever that is.) This is in addition to around 10 that other people had also found. I'd also like to know why *pc++ works but *pc-- causes the compiler to blow up. No, I haven't reported it yet, but I will as soon as I log in to GEnie again, I can assure you. >Ranting on the net accomplishes nothing worthwhile. Sure it does. It lets people know how screwed the entire orca environment (except the shell) is. BTW, Derek and I are going to try to patch that bug in the next day or two. -- Jawaid Bazyar | Blondes in big black cars look better wearing Senior/Computer Engineering | their dark sunglasses at night. (unk. wierdo) jb10320@uxa.cso.uiuc.edu | The gin, the gin, glows in the Dark! Apple II Forever! | (B O'Cult) Comp.Sys.Apple2- Home of the Unofficial Apple II Developer Support Team (DST)
lexter@pro-abilink.cts.com (Sam Robertson) (12/13/90)
In-Reply-To: message from jh4o+@andrew.cmu.edu I agree with the subject. Prizm, is such an unpredictable program, you don't even know when it gone the way of many of my programs! It gets lost in space and doesn't even alert me that some horrendous error has occurred. I quit using it soon after I got ORCA/Pascal and have since forgotten that it existed. Sam ------------------------------------------------------------------------------ Sam Robertson Pro-Abilink 300/1200/2400 1357 Santos Sysop (Saw) Abilene, Texas 79605 (915)673-6856 (915)673-4145 Proline: Lexter@Pro-Abilink INET: Lexter@Pro-Abilink.cts.com UUCP: Crash!pro-abilink!lexter ARPA: Crash!pro-abilink!lexter@nosc.mil BitNet: Lexter%pro-abilink.cts.com@nosc.mil ------------------------------------------------------------------------------ "You don't have to HIDEAWAY!" -Erasure