Sub: Minutes of FIGGY BAR RT Conference. Date: 06/06/91 Time: 22:00EDT Notice on the door reads: Welcome to the GEnie/FIG RoundTable's answer to Monty Python - the Thursday night open 'FIGGY BAR'. All Forth and computer related subjects are fair game. Items discussed include Unix difficultoes and Forth problems with a linked list/sort in FPC. Attendees: [GARY-S] [[Cliff] C.MORAVETZ1] [[Len] NMORGENSTERN] Minutes: <[Cliff] C.MORAVETZ1> Howdy! <[Cliff] C.MORAVETZ1> Hi Len! You're the first victim - customer Hi <[Len] NMORGENSTERN> I just had a frustrating week! Leonard - welcome <[Cliff] C.MORAVETZ1> My frustrating week isn't over yet. R-e-a-l-l-y ? a man of liesure ? <[Len] NMORGENSTERN> You tell me your frustration and I will tell you mine! <[Len] NMORGENSTERN> You tell me your frustration and I will tell you mine! <[Cliff] C.MORAVETZ1> Unix stinks! What Unix platform and why do you think it stinks <[Len] NMORGENSTERN> ga Cliff! <[Cliff] C.MORAVETZ1> It's the same old story about how the terminal I/O is done... <[Len] NMORGENSTERN> My headache will (has) responded to baby aspirin. <[Len] NMORGENSTERN> To cure yours may require decapitation! <[Cliff] C.MORAVETZ1> On the PC, when I hit XOFF, the display stops... Are you using SysV or BSD ? <[Cliff] C.MORAVETZ1> On Unix, who knows when It'll stop. Inevitably, what... <[Cliff] C.MORAVETZ1> I wanted to look at has disappeared. <[Cliff] C.MORAVETZ1> SysV <[Cliff] C.MORAVETZ1> Aborting a program... Same disgusting stuff. have you never heard of more or less ? that's why those paging routines were invented <[Cliff] C.MORAVETZ1> I've used Ctrl-Q/Ctrl-S for years. Why can't flow control work right? shift/delete will break _any_ Unix process <[Cliff] C.MORAVETZ1> Shift-delete? what does that send? Because you are using a multi-uuser, multi-taskker and it may be in mid process when you hit cntl/Q <[Cliff] C.MORAVETZ1> There's no upper case DEL character. <[Cliff] C.MORAVETZ1> But the I/O routines should know how to keep barfing at me! You should use the tools Unix provides and you will avoid those aggrivations <[Cliff] C.MORAVETZ1> Oh well, It could be worse and never pay attention. next time pipe your text, program, whatever through either more or preferably less ie: cat | more <[Cliff] C.MORAVETZ1> I know how to use those things... I just think things could have been done "right" less is more automagic ... less

Why on earth would you expect a multi-tasking environment to operate like brain dead messy-dos ? <[Cliff] C.MORAVETZ1> Because the last multi-tasking/multiuser OS I used handled this just fine! which was ?... <[Cliff] C.MORAVETZ1> CP-V <[Len] NMORGENSTERN> What is that? <[Cliff] C.MORAVETZ1> Hah! The operating system for Sigma mainframes. <[Len] NMORGENSTERN> Thanx CP-V, really ? On what ? This may get very interesting. I fully expected you to say OS/9 <[Cliff] C.MORAVETZ1> Xerox used to make mainframes, and I've worked on the Sigma 6 and Sigma 9 models. <[Cliff] C.MORAVETZ1> They are from the days of discrete components. What sort of shop Cliff ? <[Cliff] C.MORAVETZ1> A University. Did most of my undergrad assignments on it. There was one where I helped maintain a CDC 3300 - it was used in a medical lab, Sigma ??? (model unknown) <[Cliff] C.MORAVETZ1> Later I worked for a company that made a clone for the Sigma 9 called a T-85. I thinkk most of the Sigmas must have ended up in University type environments <[Cliff] C.MORAVETZ1> That, and real time. There's quite a few still running RT applications. Lot of GE and Westing House TTL boxes went that route, too <[Cliff] C.MORAVETZ1> Well, enough ancient history. We mustn't forget len's frustrations - go for it, Len <[Cliff] C.MORAVETZ1> I just think it's funny that that clunky old box outdoes Unix. <[Len] NMORGENSTERN> My problem is trivial compared to yours. <[Len] NMORGENSTERN> Much of the frustration is the feeling that I should <[Len] NMORGENSTERN> have figured it out after 3 minutes, not 3 days. <[Len] NMORGENSTERN> <[Len] NMORGENSTERN> I was creating a sorted linked list and could <[Len] NMORGENSTERN> not get it to work right (elementary problem in <[Len] NMORGENSTERN> C.Sci. 101 ) It took me over 3 days to discover <[Len] NMORGENSTERN> that FPC's sort algorithm treats bytes between <[Len] NMORGENSTERN> 128 and 255 as negative!!!! <[Len] NMORGENSTERN> <[Len] NMORGENSTERN> The fix took me about 10 minutes, once I figured <[Len] NMORGENSTERN> it out! <[Len] NMORGENSTERN> <[Len] NMORGENSTERN> <[Cliff] C.MORAVETZ1> Oh, negative characters! <[Cliff] C.MORAVETZ1> How neat! Is this not documented somewhere ? If not- or even if so, it deserves mention in the BB area <[Len] NMORGENSTERN> It's written in assembler using byte cmpare <[Len] NMORGENSTERN> The Intel CMP function works that way for bytes <[Len] NMORGENSTERN> Since most of don't use 128-255 for characters, <[Len] NMORGENSTERN> there is usually no problem. <[Cliff] C.MORAVETZ1> And it does signed compares? <[Len] NMORGENSTERN> The word is COMP ( a1 a2 l -- f) <[Cliff] C.MORAVETZ1> Well, I wouldn't say you're problem is trivial compared to mine. 3 days? <[Cliff] C.MORAVETZ1> What a booger! <[Len] NMORGENSTERN> You are right, Gary. I will post a message to that effect. <[Len] NMORGENSTERN> One of the disadvantages of programming alone is That's one of the things we should do - serve as bug fix repositories <[Len] NMORGENSTERN> that your mind gets in a rut. Someone else can frequently <[Len] NMORGENSTERN> spot your problem for you. <[Cliff] C.MORAVETZ1> I know - I'm working alone too. <[Len] NMORGENSTERN> I kept fiddling with the link algorithm. I was sure it was <[Len] NMORGENSTERN> right but it wouldn't work right, so I kept changing <[Len] NMORGENSTERN> and rewiriting it!!! How true ! Once you accept a 'lie' you are in serious trouble. <[Len] NMORGENSTERN> (I must have a dozen ways to create a linked list!!!) <[Cliff] C.MORAVETZ1> If only a little bell would go *ding* inside your head when you're looking in the wrong place. hahahahahahaha <[Len] NMORGENSTERN> Speaking of looking in the wrong place <[Len] NMORGENSTERN> I have a rule: <[Len] NMORGENSTERN> When your test suite uncovers a problem carefully <[Len] NMORGENSTERN> examine the test suite before you examine your program! You know - your recipe bookklet of linked lists would be a nice library posting, too, Len <[Cliff] C.MORAVETZ1> If I'd been following that advice... <[Len] NMORGENSTERN> Actually the thing is so elementary, I don't think it would <[Cliff] C.MORAVETZ1> I could have saved myself 3 days of work. <[Cliff] C.MORAVETZ1> once. <[Len] NMORGENSTERN> be worth while except as a tutuorial <[Cliff] C.MORAVETZ1> Well, I think it's time to buzz off. I totally disagree. people coming to Forth from a C or Fortran environment often (always) bitch about the lack of libraries - your linked list routines would be a nice addition. <[Cliff] C.MORAVETZ1> I've been spending too much on GEnie. <[Cliff] C.MORAVETZ1> has left. Hey Cliff, nice chattiing <[Len] NMORGENSTERN> See you again, soon. === End of Steno notes. === =END= Item 0696001 91/06/07 13:20 From: ELLIOTT.C Elliott Chapin To: GARY-S Gary L. Smith cc: FIGSYSOPS$ FIG RT SysOps Sub: topic police Reply: Item #7395630 from GARY-S on 91/06/07 at 08:30 Go ahead & add a topic. If you haven't done MMOs to it from 10-15 by the time I see it I'll do 'em - or leave a message saying where we'd like the discussion to continue. I've been wondering what to do about the fuzzy boundary between cat 6 & 7. And there's also the question of how to strike some meaningful division between hardware & software oriented discussions. Perhaps it won't be very long before another category appears. There are probably some more that should be in 7 rather than 1 or 6. Maybe the new one should be called something like "Environmental Issues" and include OS and Quarterdeck/Windows/etc. oriented discussions. I'll air a plan in time to do something before Rochester. =END= P 200?