como@bnl-max (Andrew T. Como) (04/17/89)
Is there anyone out there that has a version of TeX that is running the the Encore Umax 4.2 (rev 3.2). I have a version written is Pascal but unfortunately the Encore compiler is terribly deficient. Appreciate any help from you multimax administrators. como@bnl.gov BITNET: como@bnl.BITNET UUCP: ....philabs!sbcs!bnl!como
art@buengc.BU.EDU (A. R. Thompson) (04/17/89)
In article <1101@bnlux0.bnl.gov> como@bnl-max (Andrew T. Como) writes: > > Is there anyone out there that has a version of TeX that >is running the the Encore Umax 4.2 (rev 3.2). > > I have a version written is Pascal but unfortunately >the Encore compiler is terribly deficient. > > Appreciate any help from you multimax administrators. It's not the Encore Pascal that's deficient, it's the Berkeley Pascal that is terribly deficient. Namely it passes file variables as value parameters which is a no no. The C versions of TeX work just fine. You should get them and everything should be peachy. If you have a problem please contact us and we will give you a hand. You should be very careful before you run around calling somebody's procuct deficient when you don't know what you're talking about. You owe Encore and Oregon Software and apology.
drs@bnlux0.bnl.gov (David R. Stampf) (04/18/89)
In article <2561@buengc.BU.EDU> art@buengc.bu.edu (A. R. Thompson) writes: >In article <1101@bnlux0.bnl.gov> como@bnl-max (Andrew T. Como) writes: >> >> Is there anyone out there that has a version of TeX that >>is running the the Encore Umax 4.2 (rev 3.2). >> >> I have a version written is Pascal but unfortunately >>the Encore compiler is terribly deficient. >> >> Appreciate any help from you multimax administrators. > >It's not the Encore Pascal that's deficient, it's the Berkeley Pascal that >is terribly deficient. Namely it passes file variables as value >parameters which is a no no. The C versions of TeX work just fine. You >should get them and everything should be peachy. If you have a problem >please contact us and we will give you a hand. > If I'm not mistaken, TeX compiles just fine on a VAX running BSD, and it doesn't on a max. If that makes the BSD version deficient ... Actually, I'm on the side that TeX leaves something to be desired in that it assumes that a compiler has an otherwise clause on case statements, which is not part of *vanilla* pascal and has been extended in many different ways, very few of which are even close to being compatable. >You should be very careful before you run around calling somebody's >procuct deficient when you don't know what you're talking about. You owe >Encore and Oregon Software and apology. If he apologizes, will you help him compile it, or will it all just be whistling in the wind? He did ask for help I believe. In fact, the people from Encore (sales people, but people all the same) admitted that the Pascal compiler would not compile TeX, without *lots* of hand massaging, and offered to help find a source for TeX. It didn't help much since there are some local mods, that relied upon a virgin source. So, something ain't right and *help* would be appreciated. Maybe from someone who knows what they are talking about... < dave
art@buengc.BU.EDU (A. R. Thompson) (04/18/89)
In article <1102@bnlux0.bnl.gov> drs@bnlux0.UUCP (David R. Stampf) writes: >In article <2561@buengc.BU.EDU> art@buengc.bu.edu (A. R. Thompson) writes: >>In article <1101@bnlux0.bnl.gov> como@bnl-max (Andrew T. Como) writes: >>> >>> Is there anyone out there that has a version of TeX that >>>is running the the Encore Umax 4.2 (rev 3.2). >>> >>> I have a version written is Pascal but unfortunately >>>the Encore compiler is terribly deficient. >>> >>> Appreciate any help from you multimax administrators. >> >>It's not the Encore Pascal that's deficient, it's the Berkeley Pascal that >>is terribly deficient. Namely it passes file variables as value >>parameters which is a no no. The C versions of TeX work just fine. You >>should get them and everything should be peachy. If you have a problem >>please contact us and we will give you a hand. >> > > If I'm not mistaken, TeX compiles just fine on a VAX running BSD, and >it doesn't on a max. If that makes the BSD version deficient ... Yes, but the VAX BSD Pascal is the Berkeley pc compiler which allows the illegal use of file parameters mentioned above. This handling of these parameters is contrary to the standard as well as long standing restrictions on the assignment to file variables. Quoting "Pascal User Manual and Report" 3rd edition by Jensen and Wirth revised by Mickel and Miner, p.110. "Note that file parameters or structured variables with files as components may not be specified as actual variable parameters, as this would constitute an assignment." So, any Pascal compiler that allows this is illegal. You can't fault Encore for requiring legal Pascal. > > Actually, I'm on the side that TeX leaves something to be desired in >that it assumes that a compiler has an otherwise clause on case statements, >which is not part of *vanilla* pascal and has been extended in many different >ways, very few of which are even close to being compatable. Yes, but the file problem is even worse. > >>You should be very careful before you run around calling somebody's >>procuct deficient when you don't know what you're talking about. You owe >>Encore and Oregon Software and apology. > > If he apologizes, will you help him compile it, or will it all just >be whistling in the wind? He did ask for help I believe. In fact, the people >from Encore (sales people, but people all the same) admitted that the Pascal >compiler would not compile TeX, without *lots* of hand massaging, and offered >to help find a source for TeX. It didn't help much since there are some >local mods, that relied upon a virgin source. Yes, but when you ask for help you should not go about claiming somebody's product is "terribly deficient" when it's not. Of course the Encore compiler will not compile that version of TeX because TeX is not written in Pascal, but rather a restricted dialect of Pascal. Secondarily you mention that there are local mods which require a "virgin source". Here we have a second departure from the use of "standard" software. I don't think it's reasonable to make non-standard modifications and then complain about the product failing to adhere to local custom. > > So, something ain't right and *help* would be appreciated. Maybe from >someone who knows what they are talking about... I did, I suggested he use the C version which compiles and runs on the Encore without the slightest problem. We put that version up here without untoward difficulty. I also suggested he contact me for additional help. Again, the original poster stated that the product was deficient and thereby made false statements about the compiler. Such statements can cause harm to the sellers when they are not warranted. If the compiler wouldn't compile the if statement or something like that then he would have a point but here he was wrong and said something that was potentially libelous. If the poster really wanted help he should have asked for it and not trashed the product. It's not conducive to good relations to solicit help by saying something is bad when it works just fine. All that does is get people who know better upset and cast the poster in a very bad light. I am not associated with Encore in any way except as a very satisfied customer and one who has made extensive use of the Pascal compiler and find it to be hands down the best one I have ever used (including the one I wrote in 1980).
pierson@mist.encore.com (Dan L. Pierson) (04/18/89)
I suppose that it's time to say something. All of the following is my
opinion, not Encore policy.
Date: 17 Apr 89 22:20:52 GMT
From: bnlux0!drs@sbcs.sunysb.edu (David R. Stampf)
In article <2561@buengc.BU.EDU> art@buengc.bu.edu (A. R. Thompson) writes:
>It's not the Encore Pascal that's deficient, it's the Berkeley Pascal that
>is terribly deficient. Namely it passes file variables as value
>parameters which is a no no. The C versions of TeX work just fine. You
>should get them and everything should be peachy. If you have a problem
>please contact us and we will give you a hand.
>
If I'm not mistaken, TeX compiles just fine on a VAX running BSD, and
it doesn't on a max. If that makes the BSD version deficient ...
I'm afraid it's another case of the bad driving out the good. Oregon
Pascal, is a real high quality optimizing Pascal compiler. BSD Pascal
is an obsolecent instructional toy with too much non-Pascal wierdness.
We thought we were doing our users a service by providing a superior
product, IMHO we made a mistake by not sticking to the inferior, but
omnipresent BSD Pascal.
>You should be very careful before you run around calling
>somebody's procuct deficient when you don't know what you're
>talking about. You owe Encore and Oregon Software and apology.
If he apologizes, will you help him compile it, or will it all
just be whistling in the wind? He did ask for help I believe. In
fact, the people from Encore (sales people, but people all the
same) admitted that the Pascal compiler would not compile TeX,
without *lots* of hand massaging, and offered to help find a
source for TeX. It didn't help much since there are some local
mods, that relied upon a virgin source.
Depends on what you mean by virgin source. I brought up the latest
web2c based TeX on both Mach and UMAX4.2 with no changes about three
weeks ago. The problem is that I did it by ftping the web2c stuff
from one place, sightly different versions of the web sources from
another place, reconciling the differences, etc. It works just fine
but is not something I'd want to distribute. Based on this, I believe
that the latest Unix TeX tape from the University of Washington should
compile just fine.
For those who don't know about web2c, this is the new base for Unix
TeX distributions. The web source is tangled (with a special change
file) to produce the usual Pascal brick. The brick is then run
though a custom Pascal->C translator and compiled with the host C
compiler. This has two major advantages even for a system with BSD
Pascal: the C executables are smaller and about twice as fast, and
with the web2c version you can escape from TeX's use of 16-bit
internal memory indexes. The second advantage can be very helpful to
heavy LaTeX+BibTeX+PicTeX, etc. users.
dan
In real life: Dan Pierson, Encore Computer Corporation, Research
UUCP: {talcott,linus,necis,decvax}!encore!pierson
Internet: pierson@encore.comdrs@bnlux0.bnl.gov (David R. Stampf) (04/19/89)
There is a lot more wind than light here - I'll edit a bit. >Yes, but the VAX BSD Pascal is the Berkeley pc compiler which allows the >illegal use of file parameters mentioned above. This handling of these >parameters is contrary to the standard as well as long standing >restrictions on the assignment to file variables. > >Quoting "Pascal User Manual and Report" 3rd edition by Jensen and Wirth >revised by Mickel and Miner, p.110. > >"Note that file parameters or structured variables with files as >components may not be specified as actual variable parameters, as this >would constitute an assignment." So, any Pascal compiler that allows this >is illegal. You can't fault Encore for requiring legal Pascal. > That's wonderful, but Pascal has long suffered from creeping featuritis. There is simply no credible pascal standard. Look at the PC and Mac markets which are arguably the largest pascal markets, and try to talk get/put with these folks. It doesn't really matter on some obsure points, when there are major differences. >> >> Actually, I'm on the side that TeX leaves something to be desired in >>that it assumes that a compiler has an otherwise clause on case statements, >>which is not part of *vanilla* pascal and has been extended in many different >>ways, very few of which are even close to being compatable. > >Yes, but the file problem is even worse. We all have our pet peeves, for the work I've had to do, the lack of a reasonable switch statement, lack of static variables, lack of function i pointers and the lack of a standard way to modularize are the worst parts of Pascal. Otherwise it is one of my favorite languages. As Knuth said, he used Pascal because it is everyone's second favorite language. >Yes, but when you ask for help you should not go about claiming somebody's >product is "terribly deficient" when it's not. Of course the Encore >compiler will not compile that version of TeX because TeX is not written >in Pascal, but rather a restricted dialect of Pascal. > deficient (adj) - lacking in some quality, faculty or characteristic necessary for completeness. The *only* reason we have Pascal other than to support some Programming 101 types of effort is to run TeX. In that context, it is deficient. That isn't necessarily a prejoritive remark. Don't be so touchy. The fact is, that Unix systems are (within the various subspecies of BSD and System V) rather nice platforms for portability. I don't see why Encore cannot provide both Pascal compilers since there are obvious compatability problems. >Secondarily you mention that there are local mods which require a "virgin >source". Here we have a second departure from the use of "standard" >software. I don't think it's reasonable to make non-standard >modifications and then complain about the product failing to adhere to >local custom. > So instead I get people telling me to compile parts of sources on one machine and run them through special pascal to C converters on another. I guess one man's resonable modifications is another man's departure from standards. >I did, I suggested he use the C version which compiles and runs on the >Encore without the slightest problem. We put that version up here without >untoward difficulty. I also suggested he contact me for additional help. > I'm taking you up on the offer. Please send information on the C version of TeX. I assume it passes the trip test etc. >Again, the original poster stated that the product was deficient and >thereby made false statements about the compiler. Such statements can >cause harm to the sellers when they are not warranted. If the compiler >wouldn't compile the if statement or something like that then he would >have a point but here he was wrong and said something that was potentially >libelous. If the poster really wanted help he should have asked for it >and not trashed the product. It's not conducive to good relations to >solicit help by saying something is bad when it works just fine. All that >does is get people who know better upset and cast the poster in a very bad >light. > And you libeled him back by saying "he didn't know what he was talking about". Loosen up a bit and look at what everyone is saying. I think he asked a reasonable question and got back a nasty reply. That's the only reason I followed up at all. I'm sure that the Encore compiler is a fine tool, and I will be sure to inform our TeX users here of that fact. Since they are TeX users, I'm sure their esthetics will be high enough to appreciate that. >I am not associated with Encore in any way except as a very satisfied >customer and one who has made extensive use of the Pascal compiler and >find it to be hands down the best one I have ever used (including the one >I wrote in 1980). I find it hard to believe that it was better than the one you wrote. At least we have that it common - it is probably also better than the one I wrote too. ;-) Anyway, so there is no mistake, I think Encore is an OK company, and I am willing to accept on your recomendation that their compiler is the best darn Pascal compiler in the world which does not compile TeX. < dave (P.S. Back to the original question - can anyone help us get TeX running on an Encore.)
clyde@ut-emx.UUCP (Head UNIX Hacquer) (04/19/89)
> > I have a version written is Pascal but unfortunately > >the Encore compiler is terribly deficient. > > It's not the Encore Pascal that's deficient, it's the Berkeley Pascal that > is terribly deficient. Namely it passes file variables as value > (...) > You should be very careful before you run around calling somebody's > procuct deficient when you don't know what you're talking about. You owe > Encore and Oregon Software and apology. Sorry, but the Encore Pascal is deficient in comparsion to the Berkley Pascal (I know that's like comparing a running junk heap to a broken down junk heap). >>I<< know because we wasted weeks trying to get it (Encore Pascal) to build TeX for us. In fact, we put in our bid specification that the Pascal HAD to be compatable with UCB Pascal in order to build TeX. One of the factors that led us to buy our Encore was that we could have TeX. Encore never delivered a Pascal that could build the old TeX. One can blame this upon the folks who wedded TeX to UCB Pascal, or those who munged up UCB Pascal or the Encore people who claimed their Pascal was compatable with UCBs. I choose all of the above, but have special contempt for the latter who outright lied to us (out of ignorance, I hope). Don't get me started about how limp the Encore Fortran is... On the other hand, the newest TeX release (which uses C instead of Pascal), builds and runs just fine on our Mulitmax (though none of the provided undumps work with COFF). Not that the Encore C compiler hasn't given us fit also... -Clyde Hoover -- Shouter-To-Dead-Parrots @ Univ. of Texas Computation Center; Austin, Texas clyde@emx.utexas.edu; ...!cs.utexas.edu!ut-emx!clyde Tip #268: Don't feel insecure or inferior! Remember, you're ORGANIC!! You could win an argument with almost any rock!
marco@buengc.BU.EDU (Marco Zelada) (04/19/89)
In article <1101@bnlux0.bnl.gov> como@bnl-max (Andrew T. Como) writes: > > Is there anyone out there that has a version of TeX that >is running the the Encore Umax 4.2 (rev 3.2). > > I have a version written is Pascal but unfortunately >the Encore compiler is terribly deficient. > > Appreciate any help from you multimax administrators. > > > como@bnl.gov >BITNET: como@bnl.BITNET >UUCP: ....philabs!sbcs!bnl!como I obtained the TeX source from Stanford that is setup to compile all in C. After a few modifications I got it to compile and installed it on our machine and even made a lot of fonts for our lab. I am more than willing to make the stuff available through ftp if you should be interested. Just drop me a line or give me a call. -- ________________________________________________________________________________ Name: Marco Zelada Phone: 617 353 9882 VLSI CAD Engineer E-mail: marco@buengc.bu.edu Dept: Electrical & Computer Eng. Mail: 44 Cummington St. Room 236 Org: Boston University Boston MA, 02215 ________________________________________________________________________________
bph@buengc.BU.EDU (Blair P. Houghton) (04/21/89)
In article <6217@xenna.Encore.COM> soper@xenna.UUCP (Pete Soper,,,) writes: >From article <8904181517.AA00510@mist.>, by pierson@mist.encore.com (Dan L. Pierson): >> I suppose that it's time to say something. All of the following is my >> opinion, not Encore policy. > > This goes for me too. But I'm the Encore Pascal product manager and > [...] you'll find it harder to believe these are just my opinions. I'm just a grad student, who happens to be excessively fond of these Multimax machines... it's getting to be alt.sex time every time I find another way to speed it up (like speeding up the Blue Flame, for you Death Valley buffs :-) and make it float off the floor. And I haven't even tried out the C extensions in the Parallel Threads handbook...(yow) These are my opinions, if not God's word, tho' you never know... >> We thought we were doing our users a service by providing a superior >> product, IMHO we made a mistake by not sticking to the inferior, but >> omnipresent BSD Pascal. > > When Encore was very young a survey was done to examine all the compilers > either available or soon to be available for 32000 targets. With others I > looked at a few dozen compilers, visited vendors, made charts and graphs > and finally took part in some agonizing choices based on a number of > criteria. Begs the question: Where went the option to ship several of each? In general, I imagine there are licensing and packaging and other logistical concerns to be added into the economic determinations, but is that decision, the one that a potentially constraining partition will be erected in the programming space, an easy one made early with arguments of "two compilers is twice the cost, pick one;" or is it left until a few more bits of data come in to confirm that some other need is preventing the minor luxury of multiple dialects of a language? > As for compilation speed, we solved this by speeding up the hardware :-) A-men! (Plug Alert!) These babies hummmmm. (Okay, so it was a short plug... well, maybe just one more :) And we haven't even seen the newfangled XPC processors, yet. --Blair "RRRrrrrrrramming Speeeeed!" - Captain Nemo.