#SAS dumps Amiga
51 messages in this thread
I've been reading messages the last couple of days on Internet that say
SAS will dump the Amiga totally by the end of this year. They will be
releasing one last patch (it won't be 6.52 because of a pirate version
floating around), and then Fini. Not even tech support.
I didn't see the original message, so the above is gleaned from the
various questions and answers I've seen. If someone else knows any details
I'm sure everyone would like to know …
Harry
As I understand it, support for the Amiga version of SAS C has stopped
already. The programmers at SAS are working in their own time on a final
update. They hope to have this available early next year.
Peter Wade
Autopiloting from London, England
This is correct. Technical support has been terminated for the SAS/C
Development System for AmigaDOS (as of 12/23/94). We announced this in
September on the net. Development was halted last summer, and this was also
announced on the net.
I regret that this decision was necessary, but the business case was very
weak for continuing the product. If it can't make money, it's got to be
canceled.
–Doug
SH*T. One more very big nail in the coffin.
Robb
Nah, don't look at it like that…. maybe SAS will sell it to someone who can
or will run with it.
Highly unlikely. You couldn't seperate the source code from that for other
platforms, one of which they still sell compilers for…
Licensing agreement? Development and distribution rights in return for
royalties or something?
The risk of letting out imporant source code has to be viewed against potential
returns.
Unfortunately, I don't think that's likely and I'm getting (this is the first
time I've said this in these fora) real pessimistic about anything ever
happening with the Amiga again.
I've still got my female friend and am expanding her capabilities weekly, but
I'm investing in hardware that can be used on other platforms, lately.
Robb
Robert,
One thing to remember: The SAS C compiler works reasonably well, as is. The
changes they provided in the last "big update" seemed entirely useless to us,
and we use SAS _every day_ to compile a huge application (Imagemaster R/t) as
well as some other not-so-well-known applications such as my amatuer radio
software, which is a demanding real-time application. It all works fine. It's
not big deal, mainly because the product isn't a bad one now – SAS did a
credible job. Who knows what the next version would offer, anyway.
Yeah, it could be faster, and it darn sure could make better code – but it
works ok now. Compare that to the Aztec compiler when it went nipples north,
wasn't worth donkey spit in comparison, IMHO. You can use SAS as it stands to
develop just about anything you can imagine (or have resources to expend on).
This isn't a serious problem – the compiler's usable for the forseeable future,
as is.
–Ben
…via AutoPilot
SAS's current C++ implementation is based on cfront 2.1. The next version
can offer more C++ support like templates and exception handling.
Also, how about a graphical Intuition designer with some code generation
built in. I know there are some shareware (or freeware) programs out
there but it would be nice to have a complete integration with the
development environement.
How about more built in class libraries. Something like RogueWave Tools++
for SAS…
There is a big list of features that could have been implemented by SAS
for the developers that use C++ as there main language.
Hani
I have absolutely NO use for C++, frankly – I think it is one of the
biggest boondoggles and wasters of system resources ever committed
upon programmers. So I'm not likely to worry about development in
that direction.
I actually would prefer to write in asm, if I had the time. I don't,
so I've compromised and moved to C, mostly. C++, however, creates
incredibly slow, bloated code – to the point where I can't tolerate
the performance at all. Too big a step for me, I'm afraid, at least
until code generators and C++ precompilers and/or compilers improve
about 3 orders of magnitude.
I completely agree that a nice intuition interface designer would be
wonderful, but I always thought is was CBM's job – one they only sort
of did with a halfhearted attempt (that gadtools thing). I'm sure SAS
could do a good job, if they thought it was worth the time…
obviously not now, of course.
–Ben
…via AutoPilot
Try as I might, I'm still no fan of C++, either. And I recently saw a news
item that computer guru Ed Yourdon has been more vocal against it, too. Demand
for it has leveled off.
I think one big drain in C++ is the malloc/free library in most compilers.
They're so darn slow and degrade in performance over time (the first call to
malloc() takes N time, the 10,000 call takes 100xN time or worse). With all
those automatic "new" operators in C++ code, programmer forget that there's a
malloc() in there someplace.
>Try as I might, I'm still no fan of C++, either. And I recently saw a
>news item that computer guru Ed Yourdon has been more vocal against it,
>too. Demand for it has leveled off.
>
>I think one big drain in C++ is the malloc/free library in most
>compilers. They're so darn slow and degrade in performance over time (the
>first call to malloc() takes N time, the 10,000 call takes 100xN time or
>worse). With all those automatic "new" operators in C++ code, programmer
>forget that there's a malloc() in there someplace.
I can see how C++ in general will result in larger and slower code, but I
don't see the need to write all my code in assembly (at least in my shop)
to gain performance improvement. I did not respond to Ben's earlier
message since he seemed to describe his needs, and was not really stating
that assembly or C is better than C++. But since this thread is still on
(and C++ is becoming equivalent to BASIC) I'll put in my 2 cents:
– Even though my code is bigger and generally slower, I think the benefits
of using C++ (or Object Oriented Languages) outway the performance
degradation.
– After writting your code, you can always find the areas that are
performance sensitive and write pieces of that code in assembly. If you
find yourself rewritting everything, then I think there is a problem in
the design.
– You can right class specific 'new' and 'delete' operators to increase
performance in the creation and destruction of classes that do not need
the extras that come with 'new' and 'delete'.
🙂 Hani
There's a lot of hidden cost in C++; that's where the power comes from. IMHO,
computers aren't fast enough for that kind of power, yet. 🙂
–Ben
…via AutoPilot
I agree with Ben that C++ does offer a lot of hidden cost, and at first I
didn't really like the language because of that. After getting used to it,
there are features of C++ that don't invoke much, if any, extra cost at
run-time and yet do improve the programming environment when compared to C.
If you restrict yourself to a subset of the C++ features, you can write very
efficient C++ code.
I avoid passing or returning classes by value, for example, so I sidestep the
whole issue of created and destroyed temporaries. I do use the concept of
member functions and even virtual member functions; these don't add much
overhead compared to well-written C code, but are easier to deal with. I use
function overloading and default parameters, but again these features don't
cost anything at run-time. I use templates very carefully to ensure that I
don't end up with 50 gazillion automatically generated template functions. I
don't use iostreams. I do use new/delete. Anyway, the point is that the C++
language doesn't make it impossible to write high-performance code like some
other OO languages do (with their run-time class binding, etc.) but some of
its features do exact a run-time cost – and we must be aware of what each
language feature is actually DOING underneath, or we risk writing really slow
code.
–Doug
And you also, at least with the present tools, buy yourself a slower compile…
so if you're comfortable writing in straight c, you may be doing yourself a
disservice – and that goes triple for designed with multiple compiles or pseudo
support via preprocessing stages.
I'm sure if you think that way, it's wonderful. I still think like an asm
programmer, I'm afraid. 🙂
–Ben
…via AutoPilot
>If you restrict yourself to a subset of the C++ features, you can write very
>efficient C++ code.
There can be truth to this for a certain app, or combo of things, but
this is somewhat pushing it to have to restrict yourself to the
subset in most cases.
>I avoid passing or returning classes by value, for example, so I sidestep the
>whole issue of created and destroyed temporaries. I do use the concept of
>member functions and even virtual member functions; these don't add much
>overhead compared to well-written C code, but are easier to deal with. I use
>function overloading and default parameters, but again these features don't
>cost anything at run-time. I use templates very carefully to ensure that I
>don't end up with 50 gazillion automatically generated template functions. I
>don't use iostreams. I do use new/delete. Anyway, the point is that the C++
>language doesn't make it impossible to write high-performance code like some
>other OO languages do (with their run-time class binding, etc.) but some of
>its features do exact a run-time cost – and we must be aware of what each
>language feature is actually DOING underneath, or we risk writing really slow
>code.
No problem with this list. The summary of it is that most of these do not
have speed hit but that some will have a size hit, which I believe was my
first response on this. The three "real" issues are:
– the size of the iostream library (and of course in most apps, this is
transparently absorbable with little problem)
– temporaries: this is definitely surprising at best and so one
definitely has to know about this
– template functions: the problem is not the generation of the functions
so much as the laying down of them. But yes, this can indeed be bloat.
Definitely need smarter compilers and programmers wrt this.
What about CBM's ToolMaker? For me, ToolMaker (an Inutition builder- code
generator) runs faster and is easier to use on my stock A2000 than Borland's
resource editor is on my 66MHz 486DX2.
..Andy
Toolmaker is very limited. I've written several applications with it, and found
that it wasn't up to big jobs.
Screens with many gadgets are slow as mud.
There is little flexibility (because gadtools, like many of the good ideas in
the OS, was not supported well).
–Ben
…via AutoPilot
Well, maybe if your app didn't have eleventy-seven gadgets on its screen…
🙂
Yeah.. it did, too. 🙂 It was the Talkboard speech thing for those with speech
problems… had a LOT of gadgets on screen. My intuition based thing had NO
delay… the gadtools one just slogged on a 2000. It was unusable, and I was
p*ssed.
–Ben
…via AutoPilot
Hello Hani,
On the German market, you can buy the Maxon's C++ Compiler V3.0 which
is supporting the latest AT&T standard. This Compiler includes
Classes for Intuition in the way MUI does. And the most important
thing is that the C++ code is translated directly without generating
C-source, which is finally translated into executable code. This
means, the generated code is small and fast. This package costs
about 430.- DM ( SAS C V6.51 costs 650.- DM)
bye …
Theo,
Do they have English Docs available?
._ _ _ INTERNET: mshea@math.arizona.edu
| | |ichael _>hea COMPUSERVE: 76665,3360
Hi Michael,
I'm not sure, but I think so, because Maxon does normaly pretty good software.
I have not bought the software yet, due to not having the money, but I red an
article about this compiler and it seems to be good stuff.
Don't you know the company Maxon in the US ?
bye …
Thanks for the info Theo, I wonder how I can get more information on
Maxon. Do you have a phone number or email address I can use?
Thanks,
Hani
Maxon…
Maxon Computer
Industriestrasse 26
D-65760 Eschborn
Tel. +49 6196 48 18 11
Fax: +49 6196 4 18 85
They make 3 programming languages: MaxonPASCAL 3, MaxonC++ 3
and MaxonBASIC 3 (compiler). Also a raytracing package called
Cinema 4D.
>F<riso
Friso P. Gosliga | Auto
CIS: 100341,464 – Internet: Friso.Gosliga@kub.nl | Pilot
Thanks!.
Ben, But what about when those rumored '060 accelerators come? Are there
optimizations for the '060 that might never occurr as there is no supported
compiler for the Amiga anymore?
wmc – via Autopilot!
Yes, almost certainly so, given that the 060 is trending towards supscalar
architecture.
At least the compiler will product working code, and you can count on the 060
driving it faster to some degree just because of its nature.
This of course asssumes that there will _be_ 060 accellerators. I hope so.
Things look awfully bleak – I never thought I'd say this, but I was _very_
distressed to see ASDG (now ER) reduce technical support hours and try to sell
off all that Amiga gear.
At least Nova's still there and appears to be going strong. We do NOT want to
be the only software developer in our market remaining. Not hardly.
–Ben
…via AutoPilot
Ben,
The thing that puzzles me (as well as distresses me) is the way
support for A2000 based machines disappeared virtually overnight.
Everytime I visit a facility that uses Toasters professionally I see
a large number of 2000's, not 4000s. (This is particularly true of
Amiga animation based houses since the accelerated A2000 is faster
than the stock 4000 and also because of the slot-poor 4000 chassis.
Like me, they were all waiting for decent priced A4000Ts to come
along :^{…) For us its really bad because the only talk of a '060
for a2000 is an adapter kit from RCS that will be able to provide
plug-in replacement for socketted '040s. There are a lot of GVP
users with surface mount '040s that are SOL. A4000 users are not
quite so bad off as at least 2 firms have promised '060s when the
chip ships.
BTW, I saw three interesting articles in the Bangkok Nation newspaper
while I was away. One was about the Draco ("First Amiga Clone
Announced") and the other tw dealt separately with Motorola's record
sales an profits, and the other mentioned that DEC sold the factory
that produces Alhpa chips (among others) to Motorola who will
continue to produce Alpha's under contrec to DEC. (Dec ain't doing
so good which means another peice of advanced tech may disappear
because of a company going belly-up.).
wmc – via Autopilot!
Well, there are a lot of 2000's out there. We'll see who makes what
when the chips become available. The Amiga market is likely to suffer
from a great deal of vapor, well intentioned or not, as the market
continues it's downward spiral. I'd bet that someone makes a 2000
version if any version at all appears from anyone.
Being's how I have 2 A2000s and 0 A4000s, I hope that in future you will be
referred to as a true visionary… :^}
wmc – via Autopilot!
I just got ImageFX from Nova, and it's HOT!
-sja
Yep, it's a very nice product, has been all along.
–Ben
…via AutoPilot
Doug –
Even though support has been terminated, is SAS/C still for sale?
Jeff Schweiger
Yes, as long as they still have boxes and manuals in the warehouse (or
at dealers and distributors).
Since they no longer provide technical support for the program, have they
reduced the price?
Alex
I don't know if the price has been reduced but I payed A$350 (about $275 us)
about 2 weeks ago. I think that the price was about that before they droped
support.
Stew
Sounds about par for the course. Drop development, drop tech support, and
then still charge list price cause you figure those poor dumb Amiga slobs
won't know any better.
SAS made quite a point of telling everyone last spring that they'd be dropping
support effective December something. I think they've been quite honerable.
They dropped their prices a lot last year, I don't know whether they dropped
them again since. Full list might still be on the high side, but they have
competitive upgrade prices and other discounts that are very reasonbale for
what you get in the package.
Somebody had it for around $130 in the latest Amiga World.
DJ
On AP and on the road to recovery from the Road to Mandalay,
From Queens, where the ice cream truck plays Helter Skelter …
I don't think that's a fair characterization. We notified people at least
six months in advance that support was being dropped, and 9 months in advance
that development was being dropped. We did drop our prices, and we were
telling people that called for a long time what the situation was.
Our prices were dropped in August; current retail price is $199 in the US.
Price for academic discount, competitive upgrades, and V5 upgrades is $99.50.
Price for a V6 upgrade is $59. These are NOT gouging prices for a product
with as much development effort and production cost as the Amiga development
system.
It's unfortunate that the decision had to be made to withdraw from the Amiga
market, but given that it was made, I think SAS Institute handled the
situation about as well as it could be handled.
–Doug
Doug,
> …given that the decision had to be made, I think SAS Institute handled >
the situation about as well as it could be handled.
I completely agree. And left Amiga developers with a _very_ usable product,
too.
–Ben
…via AutoPilot
No complaints here, Doug.
Just wondering: any breaks for users with your compiler who would like to
change platforms? (e.g., an Amiga type who wanted to do work in MS-DOS)?
–Jim
Sorry, to say this, but … get Borland or Microsoft … both much more capable
than SAS/C.
-sja
Well, we don't really have a product on MS-DOS or any other "reasonable"
subsitute platform any more. Lattice C for DOS is pretty creaky. Most of
our compiler work is for in-house use now…
Of course, there are TWO platforms for which SAS Institute does still offer a
development system. I'll put in a good word for you if you want to try to
get a discount on either our MVS or CMS compiler system 8^)
–Doug
Doug,
I too believe that SAS was very upfront with the Amiga Community about their
decision to drop support for us. The current product is very useable and will
serve well for many years.
Maybe in the future if the Amiga is revived in some manner SAS will consider
supporting it again.
Thanks to you and others who provided a excellent product.
Doug,
I tried to get SAS/C 6 from a couple of mail order houses, but no one
had any to sell. Can I order it from SAS? I would like to upgrade from
version 5 before it all goes away…..
James Du Bois via AutoPilot
Member of the Bermuda Triangle Exploratory Expedition 1992 – 1951
Hi James –
Yes, you can still get SAS/C V6.51 from SAS Institute. Email them for
more info: saspubx@vm.sas.com ; or just call them (919)677-8000 and ask for
Book Sales.
–Doug