Commentary for third public review of X3J11 C
David Hough
is nearing its close on 1 September 1988. This third review
is based upon a draft dated 13 May 1988 which is not greatly
changed from earlier drafts except that the controversial
"noalias" keyword was removed.
Consequently the Draft still leaves a good deal to be
desired from the numerical point of view.
I have two documents available for electronic
distribution.
I will be glad to send you
tbl/troff -ms source for these;
I'll send both unless you specify that you only want the newer one
described below. Unfortunately the
Draft ANSI Standard itself is not publicly available in
electronic form.
The first available document is my 29 March 1988 commentary prepared
for the second public review period (30 pages), with X3J11's
formal responses of 22 April interspersed. The following were
co-conspirators:
Greg Astfalk Larry Breed D. Burton
W. J. Cody Iain Johnstone W. Kahan
Zhishun Alex Liu David Mendel Jim Meyering
K-C Ng Gene Spafford Philippe Toint
Stein Wallace
The second available document is a draft,
subject to revision until submitted about 25 August, of my commentary
for the third public review. It's only about 10 pages
since I generally avoided directly repeating what was in
the earlier document. I'm looking for additional reviewers
and conspirators on this one. The abstract follows:
The proposed C standard suffers numerical
shortcomings - many inherited from its precursors
- in areas of interest to providers of portable
mathematical software. I comment in detail upon
the following aspects of the proposed standard:
Comment #1, Section 3.9: encourage sound practices
Comment #2, Section 3.9: disparage hazardous practices
Comment #3, Section 1.1: emphasize surprises in rationale
Comment #4, Section 1.1: anticipate supplemental standards
Comment #5, Section 2.2.4.2: use "significand"
Comment #6, Section 2.2.4.2: <float.h> has too many names, not enough information
Comment #7, Section 3.2.1.4: round conversions between floating types
Comment #8, Section 3.5.4.2: fix arrays
Comment #9, Section 4.5: exceptions in mathematical functions
Comment #10, Section 4.5: tell more in the rationale
Comment #11, Section 4.5: standardize hypot
Comment #12, Section 4.5.4.6: delete modf
Comment #13, Section 4.7: specify which signals can arise
David Hough
dho...@sun.com
na.h...@na-net.stanford.edu
{ucbvax,decvax,decwrl,seismo}!sun!dhough
Doug Gwyn
>I comment in detail upon the following aspects of the proposed standard...
One very important thing to be aware of is that X3J11 intends to get
the proposed standard wrapped up (formally adopted) as soon as possible.
When the third-round comments are reviewed at the September meeting, it
is highly likely that all suggestions for substantive changes to the
proposed standard will be rejected unless they remedy a proven serious
deficiency. The reason is that any non-editorial change would necessitate
yet another public review, therefore delay in publishing the official
standard. This process could go on forever, but there is a strong desire
to adopt a "good enough" standard in a timely fashion rather than working
toward a "perfect" standard that is too late to matter. Many of us feel
that the current draft is "good enough", perhaps modulo editorial nits.
I don't mean to discourage comments on the draft; however, you should be
advised that you'll need some extremely strong arguments for making any
substantive changes. Examples showing that the current draft is badly
broken would help.
Joseph Reger
>I don't mean to discourage comments on the draft; however, you should be
>advised that you'll need some extremely strong arguments for making any
>substantive changes. Examples showing that the current draft is badly
>broken would help.
The draft may not be 'badly broken' but is missing out on the opportunity
to make C a convenient language for numerical computing as well. It is a
pity that many of the 'real programmers' feel that any change that would
allow C to be the language of choice for 'non-real programmers'
(scientists) somehow would hurt their feeling/interests. I did not
participate in the debates about the power operator, noalias, conformant
arrays etc., because I was scared by some the vehemence of the 'defender
of the faith'. It is sad that never seemed to be enough time to discuss
some recommendations in detail. There are many scientist that I know
(mostly younger people) who really came to like C, and we are using it
despite its problems and deficiencies as far as numerical computing is
concerned.
I strongly feel that it is an unacceptable situation that many of us has
to program around these problems, although some of them could be easily
fixed. Much of today's (computational) science is done in a workstation
environment, mostly under Unix. In the future this is going to be even
more so, especially now that the supercomputer manufacturers are adopting
Unix, too. The best compilers in these environments are the C compilers,
period. Since the manufacturer often uses the same compilers for his own
development, the user can be fairly confident that most of the bugs have
already been eliminated. So there will be ever more scientist who program
in C. Why is it such a good idea to have a growing amount of code around
that contains ugly, difficult to understand "fixes"?
The power operator is a small issue, I agree. Noalias (no flames please, I
am afraid of you) is definitely going to come, since the vector machines
need it. Only that it will come in many (vendor specific) colors and
flavors. Conformant arrays? We (scientists) need them very much and I do
not see how they would mean any grand problem for C --and the end of the
western civilization-- in the simple version proposed by David Hough (see
his "Comments on Proposed ANSI C Standard").
All these problems could be solved, of course, by the inclusion of the
following statement into the Draft:
"Scientist and other non-real programmers are not allowed to use the
programming language C".
(The funny thing is that some scientist would actually like to see this
statement, not only in the Draft, but everywhere).
Joseph D. Reger, jos...@chromo.ucsc.edu
Jim Giles
Doug Gwyn
>The draft may not be 'badly broken' but is missing out on the opportunity
>to make C a convenient language for numerical computing as well.
I happen to use C for numerical programming, despite occasional flaws
such as those you mention, primarily because it offers much better
support for data structures than do other alternatives such as FORTRAN.
I agree that there are some changes that could make C more convenient
for such applications. Hough's suggestions are for the most part good
ones, but they haven't been receiving sufficient committee support.
The fundamental problem is that IT IS MUCH TOO LATE to be making
significant changes to the proposed standard. Look at all the trouble
the last-minute addition of "noalias" caused. The public review
period is intended as a REVIEW of work done by the committee, not as
an opportunity for language design. Where were all these scientific
users of C when the design work was being done? By leaving that up
to people who didn't think the flaws you perceive were significant,
you did not get those flaws addressed in the proposed C standard.
It's easy to complain about other people's work; much easier than
helping with the work. I suggest that you GET INVOLVED in drafting
the NEXT (revised) standard.
Obviously I am not speaking for X3J11 officially here..
Jim Valerio
Doug explains that X3J11's intention to get the proposed standard wrapped up
ASAP, and suggests that anything other than editorial changes are very likely
to be rejected.
Based on the replies I saw to the previous review cycle, I'm concerned that
even editorial comments are too likely to be rejected. Consider the
the Committee's choice not to replace "mantissa" and "value part" with
"significand". This editing change would bring the standard in conformance
with both IEEE 754/854 and ANSI/IEEE Std 1084-1986 (the IEEE Standard
Glossary of Mathematics of Computing Terminology).
I am also concerned that substantial mistakes are being made in the area of
floating-point support, despite the good critique provided by David Hough
in the previous review cycle.
I found the responses to David Hough's comments in the last review cycle were
often depressingly mechanical, unbalanced, and to my mind, unreasoned. I
don't understand how <float.h>, demonstably inadequate, arguably wrong, and
lacking prior art, can be accepted. Compare this to the decision not to
standardize hypot(), a function which exists on both BSD and SysV systems,
a function the Committee called an "invention of limited utility". Oddly
enough, the complementary atan2() function is standardized; the Committee
explains atan2() "can be used for purposes other than conversion to polar
coordinates", an argument that actually seems to apply more correctly to
hypot(). I could go on, but Hough's letter and the Committee's responses
say it all much better.
I hope that Doug, and the Committee as a whole, will very carefully read
and consider David Hough's comments in this coming review, and perhaps
supply better considered and self-consistent responses than those that
were provided in the previous review cycle.
--
Jim Valerio jimv%ra...@omepd.intel.com, {verdix,omepd}!radix!jimv
Joseph Reger
In article <83...@smoke.ARPA> gw...@brl.arpa (Doug Gwyn (VLD/VMB) <gwyn>) writes:
>>The draft may not be 'badly broken' but is missing out on the opportunity
>>to make C a convenient language for numerical computing as well.
>
>
>significant changes to the proposed standard.
It seemed to me - and I admittedly did not follow it from the very beginning -
that it was always MUCH TOO LATE.
>It's easy to complain about other people's work; much easier than
>helping with the work. I suggest that you GET INVOLVED in drafting
>the NEXT (revised) standard.
I certainly will.
Joseph Reger, jos...@chromo.ucsc.edu
Herman Rubin
I use C for numerical programming, and then have to edit the resulting .s
file. All of the languages, including C, are woefully deficient is letting
the user use the capacities of the machines. If C is to be a good flexible
language, the committee should widely advertise for complaints about the
deficiencies of the language before starting out.
I would have no trouble coming up with pages of these items. But the last
time I did something like this, in reply to the open invitation to attend
the meeting on the IEEE floating-point convention, was to receive an invi-
tation to attend! I do not have the time to attend meetings on software.
Another problem is that the language gurus are unsympathetic to ideas which
run counter to their perception of computing needs. They see integer
arithmetic as primarily for addressing and looping; I see integer arithmetic
as important for number-crunching. What about fixed-point (_not_ integer)
arithmetic? What about the use of overflow? What about division with
simultaneous quotient and remainder? What about an operation or function
returning a string of values? What about table-driven branches? What
about inserting new operators, using the processor syntax to specify the
argument structure of these operators? In fact, what about using the
easy-to-use hardware operators on most machines? A good example is &~,
which is more useful than &, and is hardware on many machines, including
the ones for which C was initially written. Many of those machines do not
even have a hardware &.
How many useful instructions have disappeared from hardware because they
do not occur in the HLLs? Multiprecision arithmetic needs unsigned
multiplication and division to be efficient, and not floating point
arithmetic. The presence of a single hardware instruction can be
essential to an algorithm being worthwhile; if the instruction is in
software, it is more likely to appear in hardware.
--
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907
Phone: (317)494-6054
hru...@l.cc.purdue.edu (Internet, bitnet, UUCP)
mcdo...@uxe.cso.uiuc.edu
>The fundamental problem is that IT IS MUCH TOO LATE to be making
>significant changes to the proposed standard. Look at all the trouble
>I suggest that you GET INVOLVED in drafting
>the NEXT (revised) standard.
The problem is, how does one do this? IF you are a regular reader of this
august information dispersal system, you might here about some such
effort not too late after it gets started. But, in the absence of that,
you are going to know about it only AFTER the standard gets approved,
when the next version of your compiler comes out and your programs
stop compiling. I never heard about Fortran 77 until my programs
refused to run because a new compiler didn't support Hollerith fields.
Henry Spencer
>I found the responses to David Hough's comments in the last review cycle were
Little birdies tell me that the review of the second-round public comments
was, in general, awfully rushed. It shows.
--
Intel CPUs are not defective, | Henry Spencer at U of Toronto Zoology
they just act that way. | uunet!attcan!utzoo!henry he...@zoo.toronto.edu
Henry Spencer
Well, remember two things. First, that there was opportunity for input
along these lines earlier, and little was received; it is now much too late
for major changes. Second, that X3J11's mission was to standardize an
existing language, not to invent a new one; they did make some small steps
toward making C friendlier for numerical work, and that is probably about
all one should expect from a standards committee.
If you really want to see C improved as a language for numerical computing,
the first thing to do is to scream at your compiler supplier until he/she/it
does some of the things you want. Then, when the time rolls around for the
next revision of the C standard, you can propose changes based on *actual
experience*. This will carry a lot more weight than untried inventions.
Given the time lags involved in all this, if you are serious about it, the
time to start haranguing your supplier is *now*.
Doug Gwyn
>often depressingly mechanical, unbalanced, and to my mind, unreasoned.
The "mechanical" aspect may be partly due to the existence of a list of
"stock response codes" that the Committee uses for its convenience in
recording answers to issues raised. In cases where the proposal was
rejected, there is also supposed to be additional explanation. As it
happens, sometimes the additional explanation is not provided, or the
one provided makes no sense to the response document editor and reviewers,
so we have to try to provide additional explanation that reflects the
Committee's position as best as we understand it. I will admit that
less-than-perfect responses are sometimes the result.
Also consider that a full rebuttal to a lengthy paper like Hough's
would take far more work than anyone was prepared to do.
>... Compare this to the decision not to
>standardize hypot(), a function which exists on both BSD and SysV systems,
>a function the Committee called an "invention of limited utility". Oddly
>enough, the complementary atan2() function is standardized; the Committee
>explains atan2() "can be used for purposes other than conversion to polar
>coordinates", an argument that actually seems to apply more correctly to
>hypot().
Although I favor standardizing hypot(), I have to say that I almost
never have found occasion to use it, unlike atan2() which is the main
inverse-trig function (if you find yourself using some other inverse-
trig function, odds are you made the wrong choice).
>I hope that Doug, and the Committee as a whole, will very carefully read
>and consider David Hough's comments in this coming review, and perhaps
>supply better considered and self-consistent responses than those that
>were provided in the previous review cycle.
I actually supported many of Hough's suggestions. Unfortunately, out
of perhaps 30 to 40 voting members of the Committee, fewer than a dozen
have significant experience using C for large-scale floating-point
applications. This makes it hard to "sell" improvements in this area.
Since the majority have little to gain (and their compiler/library work
would be increased) by making such changes, naturally such proposals
have a hard time obtaining the necessary 2/3 majority vote to be adopted.
In fact, only remedies for really glaring deficiencies have much chance
of gaining such a majority. And at this stage of the process, any
substantive change has very little chance no matter how nice it might
have been if it had been incorporated in the proposed Standard earlier.
Jerry Leichter (LEICHTER-JERRY@CS.YALE.EDU)leichter@venus.ycc.yale.edu (Jerry Leichter (LEICHTER-JERRY@CS.YALE.EDU)leichter@venus.ycc.yale.edu (Jerry Leichter (LEICHTER-JERRY@CS.YALE.EDU)leichter@venus.ycc.yale.edu (Jerry Leic)
>I use C for numerical programming, and then have to edit the resulting .s
>file. All of the languages, including C, are woefully deficient is letting
>the user use the capacities of the machines. If C is to be a good flexible
>language, the committee should widely advertise for complaints about the
>deficiencies of the language before starting out.
On the contrary: C is NOT woefully deficient for the vast majority of
applications to which the vast majority of "paying users" are interested in
applying it. As later comments make clear, the kinds of users Mr. Rubin has
in mind are rather different. The fact of the matter is, hardly anyone thinks
that "fixed-point arithmetic" (as opposed to integer) is important. It just
does not come up in the vast majority of uses to which computers are put.
Developing software is an expensive proposition. Everything added to a
language has to be implemented somewhere, by someone. Then it has to be
debugged, supported, and maintained. There are only two ways this will
happen: If someone is willing to pay for it; or when someone is willing
to do it out of their own love for the subject.
>I would have no trouble coming up with pages of these items. But the last
>time I did something like this, in reply to the open invitation to attend
>the meeting on the IEEE floating-point convention, was to receive an invi-
>tation to attend! I do not have the time to attend meetings on software.
Ah, so Mr. Rubin is willing to COMPLAIN, but he is NOT willing to do the work
out of his own love for the subject. He certainly gives no indication that
he is willing (or able) to pay to have it done either.
>Another problem is that the language gurus are unsympathetic to ideas which
>run counter to their perception of computing needs.
I am a "language guru", though my interests happen to be in parallel program-
ming languages. Again, why should I care what Mr. Rubin thinks "computing
needs" are when he can't provide money, isn't willing to invest his own time,
and can only provide the most specialized examples of what such features might
be used for?
> They see integer
>arithmetic as primarily for addressing and looping; I see integer arithmetic
>as important for number-crunching. What about fixed-point (_not_ integer)
>arithmetic? What about the use of overflow? What about division with
>simultaneous quotient and remainder? What about an operation or function
>returning a string of values? What about table-driven branches? What
>about inserting new operators, using the processor syntax to specify the
>argument structure of these operators? In fact, what about using the
>easy-to-use hardware operators on most machines? A good example is &~,
>which is more useful than &, and is hardware on many machines, including
>the ones for which C was initially written. Many of those machines do not
>even have a hardware &.
What about all these things? Being absolutely brutal about it: Why should
I (or other readers) care? What will it gain us to worry about this?
>How many useful instructions have disappeared from hardware because they
>do not occur in the HLLs?
Along the same brutal lines, my answer is: No USEFUL instructions have
disappeared at all. What has disappeared are a lot of non-essential ideas
that were tossed in back in the days when computer architecture was a new
field, with a large research component. No one really knew what would turn
out to be "useful".
Well, for better or for worse, computer architecture isn't like that any more.
Computer design is a multi-billion dollar industry. It is driven, not by what
people might WANT in some abstract sense, but by what they are willing and
able to pay for. THAT is the only workable definition of "useful", and on
that scale the things Mr. Rubin wants have long ago fallen to the bottom of
the list.
> Multiprecision arithmetic needs unsigned
>multiplication and division to be efficient, and not floating point
>arithmetic. The presence of a single hardware instruction can be
>essential to an algorithm being worthwhile; if the instruction is in
>software, it is more likely to appear in hardware.
>--
>Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907
>Phone: (317)494-6054
>hru...@l.cc.purdue.edu (Internet, bitnet, UUCP)
It's painful to see economics dominating a field one loves and pushing it in
directions one is not inclined to go. I'm not unsympathetic to Mr. Rubin's
position; my own background, way back when, is in mathematics (complex analy-
sis and a bit of analytic number theory). Even the work I do now is beyond
the current "commercial" leading edge, and I am sometimes frustrated by the
way hardware manufacturers put roadblocks in the way of doing "obviously
useful" things, because they are too busy heading in other directions. But
that's life.
The USEFUL thing for Mr. Rubin to do, if he really thinks these issues are
important, is to work at convincing others of it. Not by complaining in this
and other newsgroups about how he is being ignored. But exactly by spending
some time with those committees, by offering some real alternatives, by
showing how what he proposes is useful to people other than himself. Frankly,
I doubt anything he can do will ever get major commercial ventures interested.
But that doesn't mean he can't get other researchers interested. Many people
are able to design and build special-purpose hardware and software today; if
Mr. Rubin talked to some of them, he might discover that many good research
hardware hackers have the tools, but are lacking interesting problems. I
will say, however, that his chances of getting people interested would improve
markedly if he stopped complaining about how he didn't "have the time to
attend meetings on software". Very few computer scientists have the time to
attend meetings on statistics either.
-- Jerry
Rahul Dhesi
[wish list for HLLs]
I agree that many of the features wished for ought to be in higher-
level languages. But to put all of them in C would no longer leave
the relatively small, simple, low-level language that C was designed
to be.
Nearly all of the features that Herman Rubin wishes to see *are*
already in HLLs, only not all are in each HLL. C++ has some. Ada has
*many* of them, especially fixed point arithmetic and functions
returning structured values.
The real problem is not with C designers. The real problem is with
Fortran designers, who have always had an explicit mandate to design a
language for scientific computing, and have continued to fail miserably
to achieve this. In a way the C users who do numerical computing want
to put on C the burden that Fortran was supposedly designed to carry.
The trouble with doing so is that other users will lose. Each new
feature added to a language increases the complexity of the language
translator, and *all* users, even those who don't need to use these
features, will pay in money, disk space, and CPU time.
--
Rahul Dhesi UUCP: <backbones>!{iuvax,pur-ee,uunet}!bsu-cs!dhesi
Paul R. Haas
1. ANSI C (as proposed) does not support numerical computing adequately.
2. There is not enough time to fix it for the current standard.
There are several ways of coping:
1. Use Fortran.
2. Hope your compiler vendor comes up with reasonable extensions.
3. Do something to encourage your compiler vendor to come up with
something reasonable.
I would favor writing a standard for "correct" math extensions to the C
language. If the standard is adopted by at least one compiler vendor
(or producer so as to include the FSF) then you can show prior art for
the next round of X3J11. If the "correct" math extensions are simple
enough to implement then many manufacturers will put it into their
compilers.
A proposal for a standard to be produced by an individual, an
independent committee or a committee from one of the user groups (ACM,
IEEE, /usr/group, Usenix, etc...). A committee could meet in person
or electronically, etc...
Unfortunately, I lack the skills to produce such a document.
----
Paul Haas, uunet!actnyc!prh (if that doesn't work: ha...@msudoc.egr.msu.edu)
Doug Gwyn
>you are going to know about it only AFTER the standard gets approved,
I'm pretty sure the formation of X3J11 was announced in CACM, and it
has been well known in the C community for years (e.g. "The C Advisor"
and other regular columns). I don't think it made the TV network news.
Charles Marslett
> On the contrary: C is NOT woefully deficient for the vast majority of
> applications to which the vast majority of "paying users" are interested in
> applying it.
I find this comment and the attitude of the author woefully parochial -- I do
not program in COBOL and I might not even recognize a either a data entry
language or a data base language if it hit me in the face, but I do know that
more money (real dollars, payroll hours or however you want to look at it) is
spent on programs that are much more difficult to write in C than in the
language they are written in (and in some cases -- heresy -- that language
is even 8086 assembly language!). I am quite certain that spreadsheets
garner more user dollars than C compilers for any computers other than
Crays and Suns (and Fortran compilers are probable ahead of C compilers on
at least the Crays).
C is rapidly catching up with Pascal as the second most well known language
but it has a long way to go before it becomes as well know (and perhaps as
useful)as BASIC (more heresy?).
For my purposes, C is the language of choice most of the time (by a fair
margin -- I have no second choice, except maybe Modula were C to vanish
from the face of the earth). But C is not a universal language and she
does not appear to be expanding into other areas of applicability any more
rapidly than her elder brother and sister, FORTRAN and LISP. And I think this
is both A GOOD THING, and the reason that it is unlikely to be a major language
20 years from now. I have plenty of spare time in 20 years to learn several
new small languages and I have no real need to program in Ada or PL/I.
(How do you like my personification of programming language? Shall we create
a few mythic tales to describe her birth?)
Charles Marslett
ch...@killer.dallas.tx.us
Steve C. Simmons
this is largely a library issue, it is possible to cure the problem
even with implementations that are seriously munged. This is not
enough of an issue to make further delays in the standard.
There's no reason why a revision of the standard cannot begin almost
immediately. Clearly the folks who need the numeric changes should
get together, formalize their needs, educate the rest of us on their
necessity, and propose it for the revision.
If we insist on making the changes in the *present* proposed standard,
I think we're looking at major delays. Let's get the standard out
NOW, and tighten down the libraries later.
--
Steve Simmons ...!umix!itivax!vax3!scs
Industrial Technology Institute, Ann Arbor, MI.
"You can't get here from here."
Henry Spencer
>It seemed to me - and I admittedly did not follow it from the very beginning -
>that it was always MUCH TOO LATE.
X3J11 has been trying to get the %#@%$% thing out the door for quite some
time now. The combination of lengthy public-review cycles and official
meetings held only quarterly means that a standard which needs *three*
public-review cycles will be in "almost finished, no substantive changes
without a damn good reason" status for a long time. Sounds like you came
on the scene after that phase started.
Ian L. Kaplan
>
>The real problem is not with C designers. The real problem is with
>Fortran designers, who have always had an explicit mandate to design a
>language for scientific computing, and have continued to fail miserably
>to achieve this. In a way the C users who do numerical computing want
>to put on C the burden that Fortran was supposedly designed to carry.
The Fortran 8x committee has its problems, but lack of features is
not one of them. The April '87 Fortran draft standard includes a
number of "modern programming language" features, including something
like structures (referred to as derived types) and modules, with
imports and exports. The real problem with the Fortran
standardization process is the in ability of the Fortran community to
arrive at a standard. The decade is almost over. Soon it will be
Fortran 9x.
Ian Kaplan
"I don't know what the most popular numeric programming language will
look like in the year 2000, but it will be named Fortran."
These opinions are my own.
Chris Torek
(Charles Marslett) writes:
>In article <36...@yale-celray.yale.UUCP> leic...@venus.ycc.yale.edu
>(Jerry Leichter (LEICHTE...@CS.YALE.EDU)leic...@venus.ycc.yale.edu
>(Jerry Leichter (LEICHTE...@CS.YALE.EDU)leic...@venus.ycc.yale.edu
>(Jerry Leichter (LEICHTE...@CS.YALE.EDU)leic...@venus.ycc.yale.edu
>(Jerr writes:
[A rather unusual name :-) .]
>>On the contrary: C is NOT woefully deficient for the vast majority of
>>applications to which the vast majority of "paying users" are interested in
>>applying it.
[back to chasm@killer:]
>I find this comment and the attitude of the author woefully parochial
>-- I do not program in COBOL and I might not even recognize a either a
>data entry language or a data base language if it hit me in the face,
>but I do know that more money ... is spent on programs that are [done
>in other languages] .... C is not a universal language and she does
>not appear to be expanding into other areas of applicability any more
>rapidly than her elder brother and sister, FORTRAN and LISP. And I
>think this is both A GOOD THING, and the reason that it is unlikely to
>be a major language 20 years from now.
This is curious, because I see Jerry Leichter and Charles Marslett as
basically in agreement---so why should this attitude be `woefully
parochial'? That C does not make a good functional programming
language is no surprise; that people who pay for programs written in C
are not paying for such code should also be no surprise; and hence that
there is no great push for C to be augmented with everything out of
Miranda and FP combined should likewise be no surprise.
To return somewhat to the original subject: If you believe that, with
a few tweaks that would either improve, or at least not damage, the
language, C could become an ideal language for numerical software, it
is then your job to demonstrate it. Make the changes---write yourself
a compiler, or have someone else write it---and show that the new
language is better than the old. If it is sufficiently better,
programmers will beat a path to your mailbox, and the new language will
become popular in the same way that C became popular. And if *you* are
not willing to put in the effort, why then should *we* be?
--
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7163)
Domain: ch...@mimsy.umd.edu Path: uunet!mimsy!chris
Hank Dietz
I've been using C for most programs since 1978. I've taught and am
currently teaching a C programming course at Purdue University. However, C
isn't supposed to be all things to everyone: it is a systems programming
language and has little real competition as such (Ada? Modula 2?).
Making C a numerical applications language has never been a priority,
nor should it be. For example, fixed-point arithmetic would never be used
by most of the originally-intended C user community; it would simply clutter
the language definition and impede the development of good quality compilers.
I personally feel that X3J11 has done an outstanding job of resisting the
"kitchen sink" syndrome, keeping the language reasonably clean and
implementable, while resolving more than a few ambiguous/omitted details.
Propose a new language if you're not happy with any existing one.
As for the language standardization process, if you're not willing
to attend the meetings nor to correspond in a reasonably formal way, I don't
think you've got much of a reason to complain. Now, I'm a bit unhappy in
that I wasn't invited to be on X3J11 and would like to have had more input,
but even so I have had no trouble in getting X3J11 folk to listen to me. My
number one remaining beef with X3J11 is that they changed the function
declaration syntax in an incompatible way without simultaneously providing
public-domain software to automatically convert old C programs to the new
notation... but this is a problem I personally intend to remedy.
So, let's not flame on about X3J11. It isn't perfect, but it is C
and it is a better definition than we had before. Enough said.
-hankd
Rob Carriere
Doubtless. But we were talking about the numerical community, they
generally don't read CACM. So, was it announced in, say publications
of the AAAS or the IEEE? (NOTE: I'm not saying it wasn't, I don't
know, and I'm curious)
Rob Carriere
Steven Ryan
Are you crazy?
Extensibilty implies the gods are mortal and a rational mode system exists.
Shame for mentioning this is comp.lang.c.
c.severijns
We have been using C for scientific computing for some time now and so far we
only feel the need for a very few changes to the language ( we use a non-ANSI
C compiler). One of these changes is already made in the ANSI standard, the
possibility to pass a float as an argument to a function. The second change
we would like to be made is the possibility to compile C with "intrinsic"
function to be able to use a floating point processor like the MC68881 more
efficiently. This requires only an extra option for the compiler.
For the rest we consider C a good language for scientific computing that
generates code that is not much slower than FORTRAN and has the advantage of
structures. In one case were we needed complex data structures our C version
turned out to be even more than twice as fast as a similar code in FORTRAN.
Camiel Severijns UUCP: mcvax!eutrc3!eutnv1!camiel
Surface Physics Group, Dept. of Physics
Eindhoven Universtiy of Technology
The Netherlands
Herman Rubin
It is important not just that it appear in a journal, but prominently. If
you want input, go out and loudly proclaim it. As a researcher, I find it
necessary to glance at more than 200 journals. I certainly missed the
announcement in CACM (one of my lower priority journals).
I do not believe it appeared in _Science_, the journal of AAAS. Now most
mathematicians and statisticians do not read any of the above named journals.
How about including the _Notices_ of the AMS, the _Bulletin_ of the IMS,
and the appropriate information journals of SIAM and ASA? How about asking
the physicists and chemists and astronomers and geologists and biologists?
(Apologies to the groups left out are in order.)
Barry Margolin
>In article <83...@smoke.ARPA> gw...@brl.arpa (Doug Gwyn (VLD/VMB) <gwyn>) writes:
>>I'm pretty sure the formation of X3J11 was announced in CACM, and it
>generally don't read CACM. So, was it announced in, say publications
>of the AAAS or the IEEE?
I believe that IEEE Computer (and maybe also IEEE Software) has a
regular column containing standards-related notices.
As for announcing such things in non-computer journals, that would
take quite a bit of foresight. I'm sure that X3 simply has a list of
publications they regularly announce things in, rather than trying to
figure out all the possible journals that might be interested in a
particular standard. It wouldn't seem obvious that journals of the
AAAS or AMS would be interested in a standard for a systems
programming language (that's all C has ever been intended to be, no
matter how many statisticians and scientists use it). If the
scientific/numeric communities are interested, I think it should be
the responsibility of the editors of their journals to gather the
information, rather than relying on us CS people to know that they
care. ANSI publishes a regular newsletter on all standards-related
activity; while I would not expect most people to read this, I WOULD
expect at least one journalist for each magazine to keep an eye on it.
Barry Margolin
Thinking Machines Corp.
bar...@think.com
{uunet,harvard}!think!barmar
Henry Spencer
In other words, you want it fixed, but you can't be bothered investing your
own time and effort in getting it fixed? Don't expect much sympathy.
Standards are hard work; if you can't be bothered helping with it, those
who do put in long hours on them are likely to feel that you don't really
care all that much.
> What about fixed-point (_not_ integer) arithmetic?
What about it? Last time I did something along those lines, there wasn't
any formidable difficulty in implementing it on top of integer arithmetic.
That was a long time ago, and the stuff I was doing was specialized and
simple, mind you.
> What about the use of overflow?
A nice idea, but it's hard to make it portable.
> What about division with simultaneous quotient and remainder?
Already in X3J11 C; see div() and ldiv() in section 4.10.6. If your
compiler supplier doesn't implement them or implements them inefficiently,
complain to him, not to X3J11 or to the net.
>What about an operation or function returning a string of values?
What about it? Can be done right now, although a bit clumsily, using
pointers; see scanf for an example. It's not at all clear that adding it
as an explicit construct would improve efficiency; in fact it could well
reduce it.
>What about table-driven branches?
See the "switch" construct, which has been in C all along. If your
compiler doesn't do this well, again, complain to the supplier.
>What
>about inserting new operators, using the processor syntax to specify the
>argument structure of these operators?
Again, perfectly possible now if you're willing to live with distasteful
syntax (function calls). The past experiments with user control of syntax
have mostly been limited successes at best.
>In fact, what about using the
>easy-to-use hardware operators on most machines? A good example is &~,
>which is more useful than &, and is hardware on many machines, including
>the ones for which C was initially written...
And which any sensible compiler on those machines will use if you write
x & ~y, just as you'd expect. See above comments on compiler defects.
>How many useful instructions have disappeared from hardware because they
>do not occur in the HLLs?
How many useless instructions have appeared in hardware because some clot
had the mistaken idea that they could be useful to HLLs? Exacting a speed
and cost penalty from the customers as a result of the extra complexity,
too. Such things are always compromises.
Henry Spencer
>>the NEXT (revised) standard.
>
>you are going to know about it only AFTER the standard gets approved,
>when the next version of your compiler comes out and your programs
>stop compiling. I never heard about Fortran 77 until my programs
>refused to run because a new compiler didn't support Hollerith fields.
If you wish to be involved in drafting standards, you are going to have
to sit up and pay attention so you know when work is in progress. X3J11
was fairly well publicized as such things go; anyone who was seriously
monitoring language-standards activity heard about it. Again, I'm afraid
the answer is that the only way to get involved in such things is to make
an effort to do so. This will generally involve spending both time and
money on it. A good first step is to join ACM's SIGPLAN (Special Interest
Group on Programming Languages); its monthly journal, SIGPLAN Notices,
publishes (a) quite a bit of drivel, and (b) a certain amount of news on
things like impending standards work. For example, subscribers to it
were not caught unprepared by Fortran 77, since an entire (preliminary)
draft of the F77 standard appeared there. That was kind of an extreme
case, which hasn't been repeated, but in general, if you subscribe to
the major publications of the programming-languages community, you will
not be caught by surprise by standards efforts.
News system
> The Fortran 8x committee has its problems, but lack of features is
>not one of them. The April '87 Fortran draft standard includes a
>number of "modern programming language" features, including something
>like structures ...
>"I don't know what the most popular numeric programming language will
> look like in the year 2000, but it will be named Fortran."
Deja vu. In 1963 a commitee (Share) got together to produce FORTRAN V. It
had structures, if-then-else, switch statements (spelled 'GOTO
<expression>'), eight types of numeric data, and a whole bunch more. After
the commitee saw what they had wrought, they decided that it was good. But
FORTRAN V was a bad name. So they called the language NPL (New programming
Language). When the Naval Physics Lab complained, the commitee changed the
name again. And Voila! PL/I was born.
Marv Rubinstein
Andrew Klossner
"I'm a bit unhappy in that I wasn't invited to be on X3J11 ..."
The X3 technical committees are open to anyone with $200 per year (less
if you don't want to vote) and the willingness to attend meetings.
They are not formed by inviting selected people; the notice of
committee formation is published (in an obscure ANSI rag) and then it's
come all who are interested.
-=- Andrew Klossner (decvax!tektronix!tekecs!andrew) [UUCP]
(andrew%tekecs....@relay.cs.net) [ARPA]
(Vice chair, X3J2)
Michael Burgett
floating point ops, and the *failure* of X3J11 to solicit input and rectify
these things are getting _old_....
1) C is not now, has not been in the past, and (hopefully) will not be in
the future, a lanugage designed for writing scientific applications
2) C was designed and implemented to remove the onus of using assembly language
to write operating systems, utilities, device drivers and the ilk. In this
regard, it has no equal.
In light of 1 & 2... where's the beef? C is doing what it is designed to do,
and from what I've seen of the ANSI standard, will continue to do so. My hat
off to the committee for not bowing to public pressure to try to make C all
things to all people (can you say PL/1... I knew you could.)
If you want to write an application demanding scientific functions, write the
damn things in fortran and then write all the stuff that makes sense to, in C.
(How would you like it if you hired a carpenter and he showed up with one tool
to try and add a room on your house?) This seems to me the essence of why we
have different languages to begin with, and all the whining, sniveling and
crying *shouldn't* change that... just face it, to program effectively
you just might have to learn more than one language.... (shock! disbelief!!)
awww well..... i guess i've flamed enuff for one letter....
mike burgett burgett!ad...@decwrl.dec.com
"my intellectual work belongs to my employer, but my flames are my own..."
Henry Spencer
>Sounds like somebody wants an extensible C.
It's been done, it works well, and it's readily available: C++.
Robert Lupton
so it can't have been that hard to find out. I probably first saw it on
net.lang.c, which we all read.
Robert
Joseph Reger
>These discussions about the flaws of the C language in dealing with complex
>floating point ops, and the *failure* of X3J11 to solicit input and rectify
>these things are getting _old_....
>
>1) C is not now, has not been in the past, and (hopefully) will not be in
>the future, a lanugage designed for writing scientific applications
>In light of 1 & 2... where's the beef? C is doing what it is designed to do,
>and from what I've seen of the ANSI standard, will continue to do so. My hat
>off to the committee for not bowing to public pressure to try to make C all
>things to all people (can you say PL/1... I knew you could.)
>
>
>"my intellectual work belongs to my employer, but my flames are my own..."
And they are nothing to be proud of! Mr. burgett sounds like if
he had had invented the language C, and as if he were the only
authority to decide just who is permitted to use it. The humble
proposition was to make a _few_ changes that would _not_ make the
language more complex, or bigger or more difficult to implement
or whatever the usual "arguments" against these are. I am pulling
out from this debate now and just would like to comment that it
is ending yet another time where it sadly usually does: "Scien-
tist go home, you buggers program in YOUR language not in OURS."
I thank all of you who sent me e-mail on this topic (none of
which was like Mr. burgett's above piece). I will continue to use
C as long as the standard does not require the implementors to
code special "Scientific Application Detectors" (SAD) into it,
which would produce erroneous code if the probability of scien-
tific use exceeds a certain value.
Joseph D. Reger, jos...@chromo.ucsc.edu
Doug Gwyn
The essential point is, C's ability to support numerical applications could
be significantly improved without adverse effect on its use for systems work.
In fact, X3J11 has adopted some of these improvements, e.g. allowing single-
precision intermediate computation instead of forcing it to be done in double.
Rob Carriere
>just face it, to program effectively
>you just might have to learn more than one language.... (shock! disbelief!!)
And having done so, you might then find that there is a language that
does almost everything you want, could do everything you want with
*only small changes*, and is already better than anything else around.
Can you blame people for then trying to get these minor changes done?
I am not talking about anything like a PL/1 syndrome, because I like C
for its simplicity, and I'd much rather do some work than have the
language bloated, but there are a couple of minor changes that would
greatly improve the utility of C in the numerical field.
Rob Carriere
Face it, C is just to damn *_GOOD_* for you systems guys to keep it
all to yourselves... :-)
Herman Rubin
> These discussions about the flaws of the C language in dealing with complex
> floating point ops, and the *failure* of X3J11 to solicit input and rectify
> these things are getting _old_....
>
> 1) C is not now, has not been in the past, and (hopefully) will not be in
> the future, a lanugage designed for writing scientific applications
>
> 2) C was designed and implemented to remove the onus of using assembly language
> to write operating systems, utilities, device drivers and the ilk. In this
> regard, it has no equal.
Whether C was designed for writing scientific applications is absolutely
irrelevant. English was not designed for discussing computer issues.
It was at least somewhat recognized that it might not be possible or
desirable to eliminate the use of assembler in C.
FORTRAN was designed specifically for the IBM704, and was not intended for
subroutine libraries. Unfortunately, this language, whose inadequacies should
have been obvious to anyone with any understanding of computer hardware and
numerical mathematics, has become so common that many of its devotees cannot
understand that they could profitably use other languages.
Many people have posted that they can do a better job of programming numerical
applications in C than in FORTRAN. How can _you_ flame them for that?
> In light of 1 & 2... where's the beef? C is doing what it is designed to do,
> and from what I've seen of the ANSI standard, will continue to do so. My hat
> off to the committee for not bowing to public pressure to try to make C all
> things to all people (can you say PL/1... I knew you could.)
That a badly designed language was rejected is irrelevant.
> If you want to write an application demanding scientific functions, write the
> damn things in fortran and then write all the stuff that makes sense to, in C.
> (How would you like it if you hired a carpenter and he showed up with one tool
> to try and add a room on your house?) This seems to me the essence of why we
> have different languages to begin with, and all the whining, sniveling and
> crying *shouldn't* change that... just face it, to program effectively
> you just might have to learn more than one language.... (shock! disbelief!!)
Can you provide me with a good way to program where one _line_ is in C and
another in FORTRAN? Of course not. Subroutine calls, cheap when FORTRAN and
ALGOL were produced, range from expensive to very expensive. I do not exag-
gerate when I say one line.
I consider an instruction a tool, and a programming language a tool box.
It is useful to have electric drills, power saws, etc. But a competent tool-
user knows when to use a given tool. I expect a carpenter to know when not
to use a high-level power saw and use a low-level hand saw instead. I expect
a programmer to know when to use an assembler instruction instead of clumsily
using C.
Also, a programmer is more like a constructor than a carpenter. It is
sometimes even necessary for the same person to combine the tasks of a
carpenter, plumber, and electrician simultaneously. Thus, the tool box
must contain all of the relevant tools.
BTW, I find the instruction set of any computer far simpler than any HLL.
Now the obfuscated assembler directives are another matter.
Herman Rubin
> In article <12...@garth.UUCP> smr...@garth.UUCP (Steven Ryan) writes:
> >Sounds like somebody wants an extensible C.
>
> It's been done, it works well, and it's readily available: C++.
There are gross weaknesses in C++. It does not allow the introduction of
new operators, for example. It does not address the problem of multiword
hardware types, using machine dependencies where they can profitably be
used (see the discussion about short x short -> long), and other such
goodies. I have used one type when C would assume another type; C++ would
complain.
Fortunately, the newer C++ compilers do not reduce to C; that gave such
atrocious code that if there was another way it would be preferable.
C++ addresses a few of the weaknesses of C. However, it ignores the worst
of the problems.
T. William Wells
: Make the changes---write yourself
: a compiler, or have someone else write it---and show that the new
: language is better than the old.
Anticipating at least one possible complaint: compiler writing is
*hard* work. Agreed. But you don't have to write the whole
thing. If you are going to make what are essentially minor
changes, you can do them in available compilers. For example,
the Gnu compiler which is more-or-less ANSI compatible and which
does not cost money (this is not an endorsement of Stallman et
al. just recognizing that they exist), or the Minix C compiler
which does cost (but only ~$100), or the Amsterdam Compiler Kit
(which costs a whopping $10,000). No doubt there are others as
well.
However, I suspect that the essential work would have to be done
in the libraries, but, given that the existing libraries are not
adequate (mostly the point of the complaints, I think), and that
numerical computing is your field, that should be, rather than a
problem, the heart of your activity. (Urk! The structure of
that sentence!)
---
Bill
novavax!proxftl!bill
T. William Wells
: you want input, go out and loudly proclaim it. As a researcher, I find it
: necessary to glance at more than 200 journals. I certainly missed the
: announcement in CACM (one of my lower priority journals).
:
: I do not believe it appeared in _Science_, the journal of AAAS. Now most
: mathematicians and statisticians do not read any of the above named journals.
: How about including the _Notices_ of the AMS, the _Bulletin_ of the IMS,
: and the appropriate information journals of SIAM and ASA? How about asking
: the physicists and chemists and astronomers and geologists and biologists?
: (Apologies to the groups left out are in order.)
: --
: Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907
: Phone: (317)494-6054
: hru...@l.cc.purdue.edu (Internet, bitnet, UUCP)
Hind-sight is wonderful. You shoulda looked in those journals.
If only you knew. Now you do. This is not intended as
condemnation, but rather just observing that humans are not
omnicient. And to emphasize what you already know: that if you
want to keep on top of things, you must read the relevant
publications.
I do not think that it is the business of standard committees
(and, by extension, lots of other groups) to make exceptional
effort to be known about by people outside their fields. To put
it bluntly, it is damn near impossible to do this even half right
and certainly a waste of effort; it is the job of those who have
an interest in the field to look in the right places and not the
job of the standards committees to shout their business from the
rooftops. (Think of the information pollution we already have!)
---
Bill
novavax!proxftl!bill
Brian Glendenning
>
>It's been done, it works well, and it's readily available: C++.
vectorizing C++ compilers available on "crunching" machines, e.g. Cray, Convex
and Alliant? (In fact, are vectorizing _C_ compilers available for the latter
two)?
Do C and C++ compilers generally give about the same level of optimization,
i.e. are C compilers much more mature than C++ compilers.
--
Brian Glendenning INTERNET - br...@radio.astro.toronto.edu
Radio Astronomy, U. Toronto UUCP - {uunet,pyramid}!utai!radio!brian
+1 (416) 978-5558 BITNET - glen...@utorphys.bitnet
Henry Spencer
>... Clearly the folks who need the numeric changes should
>get together, formalize their needs, educate the rest of us on their
>necessity, and propose it for the revision.
You forgot a couple of major intermediate steps: convince some compiler
suppliers to implement the changes, and use them for a while to find out
whether they really do the job. Many people, including me, will remain
unconvinced that your proposals are reasonable unless you've actually tried
them. Anyone who's actually tried language design (or library-function
design, which is a specialized subcase of language design) can tell you
that intuition is no substitute for experience.
Yes, this is more work. There ain't no such thing as a free lunch.
Henry Spencer
However, if you want input primarily from people who are competent in C and
interested in its future development, it suffices to mention it quietly in
the places such people frequent. Which is what was done.
>As a researcher, I find it necessary to glance at more than 200 journals...
Then I'd say you can't possibly have time to read endless X3J11 drafts, and
related documents, with care and attention. People who have never been
involved in standards work have *NO CONCEPT* of the tonnage of paper one
has to read if one wants to do a proper job of it. This requires real
motivation, not a dilettante's casual interest. People with that level
of motivation are going to see a quiet announcement in selected journals,
because they'll be reading those journals already.
Henry Spencer
>Does C++ solve the oft-mentioned problems with C for numerical work?
Probably not completely, although its extensibility makes it better than
C (for example, defining new kinds of numbers is simple).
>Are
>vectorizing C++ compilers available on "crunching" machines, e.g. Cray, Convex
>and Alliant? (In fact, are vectorizing _C_ compilers available for the latter
>two)?
The answer is probably "not yet". However, the same comment would apply
to any other proposed solution to the problems. The language itself is
pretty much right; getting the implementations right is important, but
is a separate problem.
>Do C and C++ compilers generally give about the same level of optimization,
>i.e. are C compilers much more mature than C++ compilers.
Most existing C++ implementations are based on C compilers to some degree,
so they're pretty much comparable.
Steven Ryan
>
>It's been done, it works well, and it's readily available: C++.
Some people have been asking for access to machine specific features. C is
good at getting at machine features for one particular machine whether they
exist or not.
Query: Does C++ do the same or does it define its machine independent operators
in terms of specific machine features and give programmers access to the same
mechanism?
(Why bother buying an unavailable book if I can con someone else in to doing
my research for me?)
Henry Spencer
>> >Sounds like somebody wants an extensible C.
>>
>> It's been done, it works well, and it's readily available: C++.
>
I didn't say it was perfect, I said it worked well. There is a difference.
Nobody expects a language to keep everybody happy. (Personally I doubt
that any language would keep Herman Rubin happy.) C++ is a fairly well-done
and highly usable extensible C.
>It does not allow the introduction of new operators, for example.
There is room for debate about whether dynamic alteration of language
syntax is a good idea. C++ does provide for new operators, provided that
you are willing to use function-call syntax for them. Call syntax is
admittedly clumsy for anything complicated, but user-defined syntax is
a real minefield for both users and implementors.
>It does not address the problem of multiword
>hardware types, using machine dependencies where they can profitably be
>used (see the discussion about short x short -> long)...
You mean, the current *implementations* do not provide for this. There
is no reason why the implementation of a C++ type can't use hardware-
specific extensions when they exist. The client interface can remain
machine-independent, as it generally should be.
Chris Torek
>Cray, Convex and Alliant? (In fact, are vectorizing _C_ compilers
>available for the latter two)?
Convex has (and has had for years) a vectorising C compiler. It is the
same as their FORTRAN back-end. And if you use a Pascal-to-C translator,
you even get vectorised Pascal!...
--
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7163)
Domain: ch...@mimsy.umd.edu Path: uunet!mimsy!chris
Henry Spencer
>Some people have been asking for access to machine specific features. C is
>good at getting at machine features for one particular machine whether they
>exist or not.
>in terms of specific machine features and give programmers access to the same
>mechanism?
C++ is essentially a superset of C, so it takes the same approach as C.
In both, there is no reason why a perceptive implementor can't provide
machine-specific hooks for users to use to implement packages which have
machine-independent interfaces. This works rather better in C++, mind
you, because package interfaces are much nicer in C++.
T. William Wells
: And having done so, you might then find that there is a language that
: does almost everything you want, could do everything you want with
: *only small changes*, and is already better than anything else around.
: Can you blame people for then trying to get these minor changes done?
: I am not talking about anything like a PL/1 syndrome, because I like C
: for its simplicity, and I'd much rather do some work than have the
: language bloated, but there are a couple of minor changes that would
: greatly improve the utility of C in the numerical field.
So, keeping within the Spirit of C :-), what are the *small*
changes that one could make to the dpANS that would make it an
ideal :-) language for numerical computing? Perhaps if we could
boil down these into a coherent recommendation, we could get them
fixed in some later standard. Anyone who wants to bat that
around should probably start posting in a new series of messages,
to separate it from these anti-ANSI flames.
: Rob Carriere
: Face it, C is just to damn *_GOOD_* for you systems guys to keep it
: all to yourselves... :-)
Amen to that.
---
Bill
novavax!proxftl!bill
Michael Burgett
#he had had invented the language C, and as if he were the only
#authority to decide just who is permitted to use it. The humble
#proposition was to make a _few_ changes that would _not_ make the
#language more complex, or bigger or more difficult to implement
#or whatever the usual "arguments" against these are. I am pulling
#out from this debate now and just would like to comment that it
#is ending yet another time where it sadly usually does: "Scien-
#tist go home, you buggers program in YOUR language not in OURS."
[...]
#Joseph D. Reger, jos...@chromo.ucsc.edu
not so Mr. Reger, you misrepresent what I said. I lay no special *claim* to
the C language nor pretend to make decisions on who may or may not use it.
I do contend that :
a) The ANSI C Committee seems to have done their job in standardizing
current practice (as opposed to implementing anyone's wish list.)
b) EVERYONE is welcome to use C and I greatly ENCOURAGE this. I have
been evangelizing C for some time now and will continue to do so.
what I don't approve of is attempts to make C the best language for
all applications at the expense of it beautiful simplicity and
compactness... This is akin to taking a set of brushes and
paints after the Mona Lisa because you don't like her smile... :-)
have you "scientists" considered using an extensible language (like C++ :-))
to solve some of your woes??
Mike Burgett
"my intellectual work belongs to my employer, but my flames are my own."
(and yes I'mm proud of them!)
John Sambrook
months now. He has argued his points with great vigor and seems
genuinely interested in proving his case. However, it seems to me
that few people share his concern.
It seems likely that this debate will continue for a very long time.
While that may not be a bad thing, in and of itself, it isn't as
satisfying as, say, an implementation of the ideas that Mr. Rubin
has advanced.
I would like to ask Mr. Rubin what he is doing, outside of posting to
this and other USENET newsgroups, to bolster his position. Is there
any research and/or design work in progress, or is it just talk. It is
my feeling that such work would be useful, and that everyone would
benefit.
Perhaps a good first start would be a carefully considered paper that
presents the fundamental issues Mr. Rubin would like to see addressed.
John Sambrook Internet: jo...@nsr.bioeng.washington.edu
University of Washington RC-05 UUCP: uw-nsr!john
Seattle, Washington 98195 Dial: (206) 548-4386
--
John Sambrook Internet: jo...@nsr.bioeng.washington.edu
University of Washington RC-05 UUCP: uw-nsr!john
Seattle, Washington 98195 Dial: (206) 548-4386
Karl Heuer
>We have been using C for scientific computing for some time now and so far we
>only feel the need for a very few changes to the language. [One is passing
>float by value, which is already in ANSI C.] The second change we would like
>to be made is the possibility to compile C with "intrinsic" function
This also is already in ANSI C.
Karl W. Z. Heuer (ima!haddock!karl or ka...@haddock.isc.com), The Walking Lint
David Keppel
>have you "scientists" considered using an extensible language (like
>C++ :-)) to solve some of your woes??
Some have. The School of Oceanography here has a major project trying
to put to gether a cannonical library of useful tools (e.g.,
statistics, curve fitting, data disply, ...) using C++. To quote from
the chief programmer's .project:
Project: Evangelizing the true C++ to the heathen Fortran Oceanographers
Want to know more about C++? Try:
Bjarne Stroustroup "The C++ Programming Language"
comp.lang.c++
gnu.g++
My personal feeling: Since Sun/AT&T has announced (some ammount of
commitment) that they will be using C++ in future Un*xs and since some
major groups (such as Cray) have a committment to high-performance
compilers and C++ (appears to be) a good way to go about doing
scientific programming, I think that the quality if C++ compilers will
get good quite soon.
Want to know more about the Oceanography project? Send me e-mail,
I'll try to get a project summary out to you.
;-D on ( Not involved with the project in any way ) Pardo
--
pa...@cs.washington.edu
{rutgers,cornell,ucsd,ubc-cs,tektronix}!uw-beaver!june!pardo
Michael Meissner
| mathematicians and statisticians do not read any of the above named journals.
| How about including the _Notices_ of the AMS, the _Bulletin_ of the IMS,
| and the appropriate information journals of SIAM and ASA? How about asking
| the physicists and chemists and astronomers and geologists and biologists?
| (Apologies to the groups left out are in order.)
Complain to the X3 parent body then. They are the ones that are
responsible for publishing when Ansi committees are formed, when the
public reviews are, etc. They have a list of journals that they send
such annoucements to -- maybe the journals you read didn't wish to
include it, or their backlog is too large to be able to print such
annoucements. Several announcements were made on Usenet stating a new
C standard was started four years ago. I participated in an early
USENIX BOF on the C standard, and others have done so in more recent
times. I recall that somebody made the observation that if you really
wanted to reach the mass of C programmers, Ansi should have made the
announcement in the funny papers the last time this discussion came
up. :-)
In case you wish to complain, the address for the X3 secretariat is:
X3 Secretariat
Computer and Business Equipment Manufacturers Association
311 First Street, N.W. Suite 500
Washington, DC 20001-2178
--
Michael Meissner, Data General.
Uucp: ...!mcnc!rti!xyzzy!meissner
Arpa: meis...@dg-rtp.DG.COM (or) meissner%dg-rtp...@relay.cs.net
Steven Ryan
>Nobody expects a language to keep everybody happy. (Personally I doubt
>that any language would keep Herman Rubin happy.) C++ is a fairly well-done
>and highly usable extensible C.
That's a very nice compliment. Complacency is a sign of death.
Peter W. Brewer
I have ported the AT&T version to the Convex C-1 and with alterations to allow
the Convex long long 64-bit type. It works with the Convex vectorizing C compiler
and with the Convex generic C compiler. Both of these have been extended to allow
the 64-bit integer type long long. The Convex vectorizing C-compiler is also a
very good optimizing scalar compiler.. however there are, as usual, some problems
when it compiles things like YACC generated stuff even when all optimizations are
turned off. And so for various reasons my rendition of cfront on the Convex has
to use both versions of Convex C compiler technology. And so yes there is a vectorizing
C++ on the convex... soon I hope to port it to a Cray 2... as soon as Cray gives
us a working C compiler {It is still too buggy and it is going to be difficult to
port to a machine which is not byte addressable.. is not really System V even though
it claims to be and yet is not Berkeley.. :) shades of SUN OS 4.0! (: The version of
the Convex C++ is still undergoing testing. I can compile a working cfront with it.
But it seems to me someone should develop a test suite in order to find out if things
like C++ or even vectorizing/optimizing C's do the right thing. Any takers?
Peter Brewer
ler...@super.org
--
Peter Brewer |_____________| THE
ler...@super.org |___|____|/ SUPERCOMPUTING
|__ |__ |_/ RESEARCH
|___|__ /_| CENTER
Ray Dunn
>
>However, if you want input primarily from people who are competent in C and
>interested in its future development, it suffices to mention it quietly in
>the places such people frequent. Which is what was done.
>
Hmm. Yes and no.
If the standardizers of hammer design want to establish a better hammer
standard and want "user input" on the subject, should they advertise the
fact in the Journal of Tool and Die Making, or in the Communications of the
Association of Cabinet Makers?
If the keepers of 'C' are interested in its usefulness in various
applications areas, then one would expect them to *solicit* that input by
addressing those application people in their appropriate forums.
--
Ray Dunn. | UUCP: ..!philabs!micomvax!ray
Philips Electronics Ltd. | TEL : (514) 744-8200 Ext: 2347
600 Dr Frederik Philips Blvd | FAX : (514) 744-6455
St Laurent. Quebec. H4M 2S9 | TLX : 05-824090
Liber
>As a researcher, I find it
>necessary to glance at more than 200 journals. I certainly missed the
>announcement in CACM (one of my lower priority journals).
So you, as a *researcher*, wanted to know if anything spectacular was
happening in the C language development community. Did you start your
research into this subject by looking at a journal designed for this
group? From what you have posted so far, this does not appear to be
the case.
>I do not believe it appeared in _Science_, the journal of AAAS. Now most
>mathematicians and statisticians do not read any of the above named journals.
And most R&D computer people do not read Science. Do mathematicians
and statisticians make important announcements in the CACM? I think
not.
>How about asking
>the physicists and chemists and astronomers and geologists and biologists?
>(Apologies to the groups left out are in order.)
Apologies NOT accepted! If you can't come up with the definitive list
of places where this information should be published to reach everybody
who might even be remotely interested in this, why are you expecting
anyone else to be able to? Also, how many of these groups are really
interested in codifying existing C practice (as the X3J11 charter
clearly mandates), or are they just interested in getting their 'wish
list' kludged into the language (as you seem to be)?
--
_ __ NEVIN J. LIBER ..!att!ihlpb!nevin1 (312) 979-4751 IH 4F-410
' ) ) NEWS FLASH ... ... ... I GOT A PHONE!!
/ / _ , __o ____ (and there was much rejoicing ... yeaaa.)
/ (_</_\/ <__/ / <_ These are NOT AT&T's opinions; let them make their own.
Liber
|> My hat
|> off to the committee for not bowing to public pressure to try to make C all
|> things to all people (can you say PL/1... I knew you could.)
|That a badly designed language was rejected is irrelevant.
But the part that isn't irrelevant is *why* did PL/1 turn out, in your
words, to be a badly designed language? We don't want to go around
repeating the mistakes of the past if we don't have to.
Herman Rubin
> In article <8...@l.cc.purdue.edu> c...@l.cc.purdue.edu (Herman Rubin) writes:
> |In article <42...@adobe.COM>, bur...@steel.COM (Michael Burgett) writes:
> |That a badly designed language was rejected is irrelevant.
> But the part that isn't irrelevant is *why* did PL/1 turn out, in your
> words, to be a badly designed language? We don't want to go around
> repeating the mistakes of the past if we don't have to.
A language should be easy to read and as easy to write as possible. The
kludges made in PL/1 to allow the use of the properties of the machine were
to use the common assembler notation, which while it is precise, is difficult
to read and write.
HLLs such as C make heavy use of overloaded operators and infix notation for
operators. There are only a few assemblers which use infix notation, and I
know of none which use overloaded operators and weak typing. In addition,
HLLs allow multiple operations in a single statement, array handling, and
similar goodies.
The makers of PL/1, when they came to allowing the user to use the low-level
procedures, required the users to use the clumsy assembler notation or even
worse. I believe that a flexible HLL which comes close to accomplishing
what both C and FORTRAN accomplish, and a lot more, can be produced. It
might be necessary to require explicit operator precedence instead of implicit,
at least in some cases (it has been stated that this is one of the biggest
problems in a compiler; APL has completely dropped it), and possibly to
remove some of the implicities introduced in some of the languages.
If there is a movement to produce a flexible HLL, I would be willing to
participate.
Herman Rubin
>
> >As a researcher, I find it
> >necessary to glance at more than 200 journals. I certainly missed the
> >announcement in CACM (one of my lower priority journals).
>
> So you, as a *researcher*, wanted to know if anything spectacular was
> happening in the C language development community. Did you start your
> research into this subject by looking at a journal designed for this
> group? From what you have posted so far, this does not appear to be
> the case.
>
> >I do not believe it appeared in _Science_, the journal of AAAS. Now most
> >mathematicians and statisticians do not read any of the above named journals.
>
> And most R&D computer people do not read Science. Do mathematicians
> and statisticians make important announcements in the CACM? I think
> not.
If mathematicians or statisticians attempt to produce a product to be used
by people in other fields, they surely make an effort to inform the users.
Research in computer science generally should not be advertised in other
journals. Research in programming languages are of this type. Mathematicians
and statisticians are not that arrogant that they would attempt to freeze
terminology for use by biologists, sociologists, physicists, computer
scientists, etc. The committee to establish C standards is doing this.
Therefore, they have the obligation to find out how that will impact the
users.
> >How about asking
> >the physicists and chemists and astronomers and geologists and biologists?
> >(Apologies to the groups left out are in order.)
>
> Apologies NOT accepted! If you can't come up with the definitive list
> of places where this information should be published to reach everybody
> who might even be remotely interested in this, why are you expecting
> anyone else to be able to? Also, how many of these groups are really
> interested in codifying existing C practice (as the X3J11 charter
> clearly mandates), or are they just interested in getting their 'wish
> list' kludged into the language (as you seem to be)?
The apologies are mine. However, when I post a reply to this group stating
that various users should be consulted, I do not feel obligated to come up
with a complete list. The C committee is composed of a larger number of
people and is more deliberative; it can and should be expected to come up
with a list which will leave out few users or potential users.
What does it mean to codify existing C practice? If it means what you are
implying, it is like the French Academy attempting to keep out English and
making a mess.
BTW, there are many statistical packages. I recommend that they not be used.
If the user does not understand the problem, they will give wrong answers.
Programming languages are somewhat similar. Fortunately, most of the time
they give correct answers, albeit slowly. Since a UTM can handle all problems,
any problem can be done clumsily on any machine with any remotely reasonable
HLL. You can walk from New York to Los Angeles, but you will probably use
an airplane instead.
Summarizing, languages (and operating systems) exist for the purpose of
enabling the user to efficiently use the capabilities of the machines to
solve problems. Currently, they fail to do this for people who can
understand the machine's capabilities.
Richard A. O'Keefe
>In article <86...@ihlpb.ATT.COM>, nev...@ihlpb.ATT.COM (Liber) writes:
>> But the part that isn't irrelevant is *why* did PL/1 turn out, in your
>> words, to be a badly designed language? We don't want to go around
>> repeating the mistakes of the past if we don't have to.
>
>A language should be easy to read and as easy to write as possible. The
>kludges made in PL/1 to allow the use of the properties of the machine were
>to use the common assembler notation, which while it is precise, is difficult
>to read and write.
>
quite know what Herman Rubin is getting at here. PL/I syntax, for those
who are fortunate enough not to know it, is full of things like
PUT FILE (OUTFILE) EDIT (THIS,THAT,THE_OTHER) (A(10)) PAGE;
Roughly,
<main keyword> {<sub keyword> [(<argument list>)]}... ;
For another example,
CALL PROCEDURE(ARG1, ..., ARGN);
For an example of arithmetic operations, to add two MxN arrays of
floating point numbers:
DECLARE (A,B) BINARY FLOAT DIMENSION (1:M, 1:N);
A = A+B;
I have no desire to praise PL/I, but I honestly don't see any resemblance
to any of the assembly languages I've ever used.
As for infix notation, I wish someone would come up with a standard
notation for sequence concatenation: I've seen "+" (which mathematical
convention reserves for commutative operations), "&" (which looks like
"and"), "*" (which makes the most sense, but is rare), and the theory
papers tend to use a sign which is a bit like ^ and a bit like the
intersection sign, and needless to say isn't in the ISO 8859/1 character set.
In the absence of an agreed notation for such a fundamental operation,
the use of functional notation has a lot to commend it.
Andrew Koenig
> A language should be easy to read and as easy to write as possible. The
> kludges made in PL/1 to allow the use of the properties of the machine were
> to use the common assembler notation, which while it is precise, is difficult
> to read and write.
Would you mind explaining this a little more?
I don't understand what you're trying to say.
--
--Andrew Koenig
a...@europa.att.com
SH...@utdalvm1.bitnet
>Does C++ solve the oft-mentioned problems with C for numerical work? Are
>vectorizing C++ compilers available on "crunching" machines, e.g. Cray, Convex
>and Alliant? (In fact, are vectorizing _C_ compilers available for the latter
>two)?
>Brian Glendenning INTERNET - br...@radio.astro.toronto.edu
>Radio Astronomy, U. Toronto UUCP - {uunet,pyramid}!utai!radio!brian
Yes, there is a vectorizing C compiler for Convex C1; it's called vc. It is
compatible with 4.2BSD C, with an extension of a 64-bit integer data type
(long long int). About all of the regular cc options are supported.
--Shane Davis
Systems Programmer, Univ. of Texas at Dallas Academic Computer Ctr.
SHANE@UTDALVM1{.BITNET|.dal.utexas.edu},r...@engc1.dal.utexas.edu
Dan Schlitt
in the results of the standardization process for both fortran and C
didn't (and don't) know about it. I have learned by sad experience
that that is just the way it is. "I didn't know that the city was
planning to put in that dump right next door to me!" "Well, you
should have. There was the required legal notice in the paper and
there were stories on TV and in the newspaper." "Yes, but I didn't
think they meant next door to ME." You can lead a student to
information but you can't make him think.
I'm very cynical about the utility of public notice about events of
potential significance. If you think that an organization of which
you are a member let you down by not giving you information about the
C standardization then write to the organization and complain. Then
maybe it won't happen next time. And maybe you will pay more
attention next time -- I know that I will.
ACM SIGNUM has given attention to the floating point arithmetic and
fortran standardization. I think they missed out on C. That is too
bad. Those of us who are members probably goofed by not getting
something in the Newsletter. Of course, the Fortran Forum of ACM gave
good coverage of the fortran process. USENIX covered the C
standardization in ;login and also POSIX etc. A quick look through
the back issues of SIAMNEWS makes me think that SIAM let its members
down on the standarization issues.
So next time let's not make the same mistake; make a new one.
Dan Schlitt
>These discussions about the flaws of the C language in dealing with complex
>floating point ops, and the *failure* of X3J11 to solicit input and rectify
>these things are getting _old_....
>
>1) C is not now, has not been in the past, and (hopefully) will not be in
>the future, a lanugage designed for writing scientific applications
>
>2) C was designed and implemented to remove the onus of using assembly language
>to write operating systems, utilities, device drivers and the ilk. In this
>regard, it has no equal.
>
>In light of 1 & 2... where's the beef? C is doing what it is designed to do,
>and from what I've seen of the ANSI standard, will continue to do so. My hat
>off to the committee for not bowing to public pressure to try to make C all
>things to all people (can you say PL/1... I knew you could.)
>
>
If, as you claim, C is designed to "remove the onus of using assembly
language" then you should take note that that is frequently what must
be done to get good numerical code -- write in assembly language.
If floating point isn't necessary for writing your operating systems,
device drivers, etc. then why not take ALL of the floating point
stuff out of C. Then the numerical people will go away and not bother
you. On the other hand .... if you need the floating point then do it
right. That is all many of us are asking for. If you don't use
floating point then the changes shouldn't get in your way anyhow.
There is no need to make C into FORTRAN or PL/1. I personally don't
see the need for adding the fortran ** to C. The people who argued
for a function call made good _numerical analysis_ sense in their
arguments. But remember that the numerical people weren't the only
ones who wanted to add useless binary operators to the language.
MY hat is off to the committee for making a number of changes in C
which will make it do floating point better.