CompuServe Thread

Lattice Flame

10 messages in this thread
#114497From: Scott LitvinoffJul 26, 1990 12:16 AM
I didn't use GO, and the K+R and ANSI points were just sidepoints. Either way the compiler was in the wrong because the problem was in the handling of case statements and if statements. It just can't handle it. Too bad for Lattice but its simple, the Amiga C compiler should hit the road and don't come back! Scott Litvinoff
#114843From: Mike Roth/LatticeJul 28, 1990 11:56 PM
Scott, I know of NO reported problems in the area you are refering to, and I'm sure Lattice would have been avalanched with reports if there were really major problems in standard C statement code generation. — Mike Roth
#114932From: Scott LitvinoffJul 29, 1990 8:28 PM
Mike, Just because you have no knowledge of such a problem does not mean one does not exist. Out of the 50,000 or so Amiga programmers out there, how many do you think there are that have tried using such deeply nested logic. And then how many do you think decide to spend the $5 in phone calls to find out what they already know. Scott Litvinoff
#114956From: Mike Roth/LatticeJul 29, 1990 10:58 PM
Scott, And what makes you think that with such deeply nested constructs that you don't possibly have a bug in your code. With the number of people using the Lattice compiler for complex work (including compiling the Lattice compiler) I'd be VERY surprised if the type of bug you mention is in there, or is a very special case situation if it is. Instead of complaining about all of the code generation problems, perhaps you should send sample code to SAS so that they can try to pinpoint the problem and fix it if it really is a code generation bug. — Mike Roth
#114960From: Scott LitvinoffJul 29, 1990 11:48 PM
Mike, The way I know that it's not a bug in my code is that that part of the program works fine on the IBM platform when compiled by Turbo C. But thank you, I think I will send it to SAS along with a nice big nastygram. Scott
#114998From: Betty Clay/SYSOPJul 30, 1990 2:37 PM
Scott, it might make you feel good to send a nastygram – but I seriously doubt that it will get the results you desire! As time goes by, you will learn that it always pays to be nice to people, and especially so if you need them to do something. 🙂 ……Betty
#115088From: Tom EshelmanJul 30, 1990 9:56 PM
Scott: As the oldest, living member of the Forum, 😎 I would counsel you to follow the "wiles of the woman." hee-hee. You have every right to assume the other guy, in fact, has screwed up. But, put yourself in the shoes of this vile and miserable wretch. Permit him to atone for his misdeeds without too much loss of face, and he will scramble to do so. It's human nature. Yell loudly, and he will surely slither away into the night. This is the time for diplomacy and guile. 😎 Oooo. I like that…. guile. Yeah. Lotsa guile. 😎 Regards Tom
#115238From: Scott LitvinoffAug 1, 1990 12:16 AM
Tom Thanx for the advice. Think I can get the vile and miserable wretch to give me my money back? Friendlily (is that a word <grin> ), Scott
#115440From: Tom EshelmanAug 2, 1990 9:10 PM
Scott: Only then would he *literally* be a vile and miserable wretch. <grin>. The fact you caught some typo "deserves" another anecdote. Last year, I audited a course in algorithms at Albright College, here in town. One evening there was a message from the instructor in my VAX mailbox. He advised me that I was the first person he ever chanced upon who properly spelled "Dijkstra", as in the Dijkstra algorithm, in the same sentence that had flubbed the word, "bubble", as in bubble sort. He thought that was funny as h…. Very much the wise guy, obviously. 😎 BTW and IMHO, they will get this taken care of to your satisfaction. I have owned both brands of compiler from the year one until versions 5.0, when it became a little salty to have both around just for reference. I normally muck around with the Aztec for reasons of convenience and because most available code *I* am interested in studying was written with the Aztec 3.6. Also, it served me as an assembly language study aid. Both outfits were slow but solid supporters. I must admit, I'm not blind to the number of revisions Lattice put out since their 5.0 was released, and I wouldn't like that either were I in your mocassins. Regards Tom
#115466From: Scott LitvinoffAug 2, 1990 10:38 PM
Tom Thanks for being sympathetic about my situation. I had been beginning to think that everybody thought the Lattice compiler was the end-all be-all compiler. So, which do you think is better, the latest revision of Aztec or Lattice 5.0.4? THANX Scott