On the matter of ANSI/ISO C and the purpose clc
petere
only incidentally relevant to ANSI/ISO C (which is what
comp.lang.c is dedicated to). Additionally, much platform
specific information is proferred as ANSI/ISO C information.
Further, redirecting the posters or pointing out errors seems to
be causing an unnaturally hostile response. I'd attribute the
vitriolic response to the (relative) anonymity of usenet.
What makes this particularly annoying is that it occurs
regularly.
If the posts to clc can be used as a guide, then few programmers
appear to appreciate what ANSI/ISO C consists of and the value
of knowing this for portability reasons.
Schildt, Lafore, Waite and Prata (the last three of "Waite
Group" fame) have contributed to this by presenting x86/DOS-
centric C as "C proper" (all of these authors books are beyond
there first editions so they must me popular, someone is buying
them). These books typically throw in a x86/DOS (or even worse
an MS or Borland) feature without flagging it as such.
Another contributing factor -- IMHO -- is the cost of the ISO C
standard and it's unavailablity through bookstores.
I would like to commend the key contributors to clc stance on
this matter. I think it important -- especially for those that
want to make a career of C programming -- to gain a firm
understanding of ANSI/ISO C. This is essential for learning to
write portable code and to the task of porting platform specific
code. comp.lang.c should be the forum in which this learning
can be assisted.
I cannot accept that anyone who says that they don't care about
the ISO C standard is serious about C programming. The standard
sarcasm is to refer to the Standard as "holy", "sacred" etc.
This is to imply an inappropriate adherence as per a religious
cult. The incontrovertible fact is that the standardisation of
C is what has made it available on almost all OSs and what has
simplified -- to some extent -- the support of more than one
platform by C shops.
A legalistic view of the ISO C standard -- and to all ISO
standards for that matter -- should be encouraged. The Standard
is what preserves the integrity of the C language and is _de
jure_ the last word on the C language. Where a vendor deviates
from the Standard they should be taken to task. A vendor's
deviance from the standard does not have any special status, the
deviation is an error, plain and simple (unless its deliberate
in which case it should be identified as a C variant).
There is an NG for almost all OSs, vendor implementations of C
compilers and SIGs. These NGs exist to discuss these matters
and that is where questions about sockets, far pointers,
cryptography, graphics, etc should be taken. Misdirected posts
diminish the value of comp.lang.c and the NG that the post
should have been directed to originally.
"Pragmatic men" who proudly exclaim that they have no concern
for standards and "only with getting the job done" should simply
take their knowledge to the appropriate NG.
This is my 2 cents worth.
-----------------------------------------------------------
Got questions? Get answers over the phone at Keen.com.
Up to 100 minutes free!
http://www.keen.com
petere
usenet.
Unfortunately remarq.com inserts those banners automatically.
keen.com appears to be one of remarq's sponsors so when I post
they add an advertising banner -- not unlike web mail --
_automatically_.
I do not even know what keen.com is !
Ben Pfaff
> > Got questions? Get answers over the phone at Keen.com.
> > Up to 100 minutes free!
> > http://www.keen.com
>
> Then you spoiled it by posting that COMMERCIAL SPAM in your signature.
Are you sure this isn't a banner inserted by his news host?
deja.com does something similar I think.
Jack Klein
<peliades...@sanderson.net.au.invalid> wrote in comp.lang.c:
[snip]
> This is my 2 cents worth.
>
>
>
>
> -----------------------------------------------------------
>
> Got questions? Get answers over the phone at Keen.com.
> Up to 100 minutes free!
> http://www.keen.com
A very nice post, which I totally agree with. :)
Then you spoiled it by posting that COMMERCIAL SPAM in your signature.
:(
Spam is spam, and is just as off-topic here and as abusive to this
group as anything you railed about.
[posted & mailed]
Jack Klein
--
Home: http://jackklein.home.att.net
Dave Vandervies
Jack Klein <jack...@spamcop.net> wrote:
>On Sun, 02 Jul 2000 03:41:36 -0700, petere
><peliades...@sanderson.net.au.invalid> wrote in comp.lang.c:
>
> [snip]
>
>> This is my 2 cents worth.
>>
>>
>>
>> -----------------------------------------------------------
>>
>> Got questions? Get answers over the phone at Keen.com.
>> Up to 100 minutes free!
>> http://www.keen.com
>
>A very nice post, which I totally agree with. :)
>
>Then you spoiled it by posting that COMMERCIAL SPAM in your signature.
>
>:(
>
>Spam is spam, and is just as off-topic here and as abusive to this
>group as anything you railed about.
Actually, I think that that sig is added by Remarq, not by the poster.
I haven't been keeping really close track, but it seems to me that
every post I've seen it on comes through Remarq (or at least has their
'Organization' line), so it may be that they're the people to complain
to about it and not the individual posters.
dave
--
Dave Vandervies dj3v...@student.math.uwaterloo.ca
"And it came to pass that in the hands of the ignorant, the words of the
Bible were used to beat plowshares into swords." --Alan Wilson Watts
Jack Klein
<peliades...@sanderson.net.au.invalid> wrote in comp.lang.c:
> I access usenet via remarq.com -- a web based interface to
> usenet.
>
> Unfortunately remarq.com inserts those banners automatically.
> keen.com appears to be one of remarq's sponsors so when I post
> they add an advertising banner -- not unlike web mail --
> _automatically_.
>
> I do not even know what keen.com is !
>
>
> -----------------------------------------------------------
>
> Got questions? Get answers over the phone at Keen.com.
> Up to 100 minutes free!
> http://www.keen.com
OK, then all is forgiven. :-o
Keen.com is a web site that advertises telephone answers to questions
for a fee, so it is certainly a commercial site. Perhaps you should
consider a different web-based server if you can't get regular nntp
access. Deja.com adds a banner at the bottom, but at least it does
not advertise a commercial site.
Mitch Vincent
It's made much more of a mess because people choose to make it a mess.
Without commercialization the Internet would be nothing so lets just accept
it (like we do junk postal mail and commercialism in every other part of our
puny little lives) and move on..
The only thing worse than SPAM is all the people that insist on stirring up
a huge stink about it, I really don't think it's a big deal if a post has an
extra 2 lines at the bottom advertising some cheesy web site that none of us
are going to pay any attention to at all... Nor do I think it's a big deal
that I receive a hundred unsolicited emails out of the fifteen hundred legit
emails I get a day -- it's a lot easier to hit a button than it is to get
all mad about it...
*steps off soap box*
Have a great holiday (for those of you that are American / Canadian)
Doot doot!
-Mitch
> > [snip]
> >
> >> This is my 2 cents worth.
> >>
> >>
> >>
> >> -----------------------------------------------------------
> >>
> >> Got questions? Get answers over the phone at Keen.com.
> >> Up to 100 minutes free!
> >> http://www.keen.com
> >
Decklin Foster
> I access usenet via remarq.com -- a web based interface to
> usenet.
>
> Unfortunately remarq.com inserts those banners automatically.
> keen.com appears to be one of remarq's sponsors so when I post
> they add an advertising banner -- not unlike web mail --
> _automatically_.
<OT excuse="fighting spam">
May I suggest news.cis.dfn.de? They require you to register and use
authentication to connect, but provide access for free. Point your web
browser at the same hostname to get signup instructions.
</OT>
--
There is no TRUTH. There is no REALITY. There is no CONSISTENCY. There
are no ABSOLUTE STATEMENTS. I'm very probably wrong. -- BSD fortune(6)
Selim Levy
>
> An even better idea would be -- just ignore it!
>
> It's made much more of a mess because people choose to make it a mess.
> Without commercialization the Internet would be nothing so lets just
> accept it (like we do junk postal mail and commercialism in every other
> part of our puny little lives) and move on..
Ahhh, but there is a big difference: junk mail is paid for by the people
advertising. If I receive junk mail in my (snail) mail box, I didn't
pay a penny for it. Hence, these advertising companies better think
before they send. (Email) spam, on the other hand, is not like so: it
doesn't cost the advertiser a penny, and the costs get placed upon me,
the user. I have to pay my ISP for longer hook-up times to download my
mail; more mail on the servers means more hardware (eg: bandwidth
related) at my ISP -- a cost that will eventually boil down to me; I may
even decide that my 300 baud modem is too slow :-). Even if you work
for an organization that gives you free internet/mail access, they will
incur more charges and hence either their profit or yours, as salary,
will be lower. So you see, there is a difference.
> The only thing worse than SPAM is all the people that insist on stirring
> up a huge stink about it, I really don't think it's a big deal if a post
> has an extra 2 lines at the bottom advertising some cheesy web site that
> none of us are going to pay any attention to at all... Nor do I think
> it's a big deal that I receive a hundred unsolicited emails out of the
> fifteen hundred legit emails I get a day -- it's a lot easier to hit a
> button than it is to get all mad about it...
Wow! I'm happy I'm not in your shoes! I would simply ignore *all*
incoming mail.
The one thing that I always find so frustrating and Americano-centric is
seeing the "this commercial email sent in accordance with Law xxx of
Congress..." (or whatever). As if I live in the US, ever have or the US
laws apply to me! Who do these spammers think they are??
For the people unaware, I suggest a great spam-fighting service: visit
www.spamcop.net for more information. (Is this email of mine considered
spam because of that? I hope not!! :-) )
> *steps off soap box*
>
> Have a great holiday (for those of you that are American / Canadian)
sob.... I'm in the wrong continent... :-(
Selim
--
"They are multiplying like flies. Let's be like happy baby maggots and
crawl all over them, licking up the filthy data they so lovingly
provide. Our bulging tummies and smiling faces will shine like a fat
turnip in the garden of life."
- Dann Corbit on comp.lang.c about search engines 2000-05-25
Dave Vandervies
Selim Levy <l...@beer.com> wrote:
>
>The one thing that I always find so frustrating and Americano-centric is
>seeing the "this commercial email sent in accordance with Law xxx of
>Congress..." (or whatever). As if I live in the US, ever have or the US
>laws apply to me! Who do these spammers think they are??
I'm not certain (not being American myself), but I think that the `law'
they cite ended up not even passing. A .usian (one who doesn't mind
contributing to off-topic discussions :) ) might be able to confirm
this.
>For the people unaware, I suggest a great spam-fighting service: visit
>www.spamcop.net for more information. (Is this email of mine considered
^^^
>spam because of that? I hope not!! :-) )
Looks to me like you're mentioning a network service, not advertising
something commercial; while that alone might not be enough to make it
not spam, I'm sure Somebody Who Cares would be able to use that to
weasel out of it somehow.
dave
--
Dave Vandervies dj3v...@student.math.uwaterloo.ca
``In other words, if we could get rid of the lusers, things could be done
right. This isn't exactly a newsworthy discovery...''
-Chris Adams in the scary devil monastery
Chris Torek
Dave Vandervies <dj3v...@student.math.uwaterloo.ca> writes:
>I'm not certain (not being American myself), but I think that the `law'
>they cite ended up not even passing. A .usian (one who doesn't mind
>contributing to off-topic discussions :) ) might be able to confirm
>this.
Yes, a lot of spam refers to "S.1618", which was "Senate Bill 1618".
A bunch use the phrase "passed by the 105th U.S. Congress", which
is not exactly truthful. "Congress" refers to both the House of
Representatives and the Senate -- these are roughly equivalent to
the Lower and Upper Houses in various other countries' Parliaments
-- and SB 1618 passed the Senate, but such a bill then has to be
reconciled with legislature from the House. Once a bill is approved
by both halves, it is sent on to the President, who may sign it,
veto it, or apply the so-called "pocket veto".
A great deal of US political grandstanding consists of getting some
bill passed in half of the Congress, knowing full well that it is
a bad idea and will be killed in conference or vetoed, just so that
one can claim to have voted for or against some particular issue.
In this particular case, however, the 105th Congress considered at
least three conflicting email "spam" bills, and passed none of them
on to the President. The spammers like to refer to SB 1618 because
it favors them so much.
A number of US states have signed into law legislation that treats
email spam pretty much the same as junk faxes. This makes all kinds
of good sense, which is no doubt why we have not gotten it done at
the Federal level yet. :-)
--
In-Real-Life: Chris Torek, Berkeley Software Design Inc
El Cerrito, CA, USA Domain: to...@bsdi.com +1 510 234 3167
http://claw.bsdi.com/torek/ (not always up) I report spam to abuse@.
Jack Klein
<s810...@ms2.cc.ntu.edu.tw> wrote in comp.lang.c:
> The one thing that I always find so frustrating and Americano-centric is
> seeing the "this commercial email sent in accordance with Law xxx of
> Congress..." (or whatever). As if I live in the US, ever have or the US
> laws apply to me! Who do these spammers think they are??
>
> For the people unaware, I suggest a great spam-fighting service: visit
> www.spamcop.net for more information. (Is this email of mine considered
> spam because of that? I hope not!! :-) )
>
> > *steps off soap box*
> >
> > Have a great holiday (for those of you that are American / Canadian)
>
> sob.... I'm in the wrong continent... :-(
>
> Selim
The bill referred to was very friendly to spammers and justifiably
killed before it could become actual law.
Everyone of your points is absolutely correct, Selim. There are
estimates that as much as 30% of all email is SPAM, and the cost
affects everybody who uses the Internet, or does business with any
company or any person who uses the Internet, which amounts to just
about everybody.
The tone would quickly change if all that postal junk mail came
postage due, and you had to pay the postage whether you wanted the
junk or not!
If you are interested in legal efforts to combat SPAM, check out
http://www.cauce.org/. In addition to their efforts in the US, they
have links to similar groups on other continents and in other nations.
As for SpamCop, note my return address. I am a paid subscriber, and I
use their site for reporting every SPAM email I receive and every SPAM
posting I come across in newsgroups.
As for the US-centric attitude, it is a fact that most, but not all,
SPAM originates from US ISPs, points to web sites on US ISPs,
advertises US products (original or pirated). That does not make it
solely a US problem, but sadly most of it originates from here.
Dik T. Winter
In article <afZgOQ+ArMDwNC...@4ax.com> Jack Klein <jack...@spamcop.net> writes:
> The bill referred to was very friendly to spammers and justifiably
> killed before it could become actual law.
The most amusing reference to that bill I found in a mailing I received today
(two times, as I receive most) (I have split the lines, it came as one line):
****************************
This is a one time mailing: You will automatically be removed unless a
request is made to be added. EX. ADVERTISING Writer/Sender: EX. Advertising
Address: PO Box 67-0954, Pompano Beach, Fl 33067-0954 Advertising, Removal.
BULK E-MAIL SENATE BILL (We comply with this Bill) Title III: - Requires a
person who transmits an unsolicited commercial electronic mail message to
include at the beginning of the body of the message: (1) the name, physical
address, electronic mail address, and telephone number of the person who
initiates transmission of the message or who created the content of it;
and (2) a statement that further transmissions of such mail to the
recipient by the person may be stopped at no cost to the recipient.
It was at the end of the mailing, and of course they did not comply with
either (1), nor with (2). Or perhaps they thought the asterisks started
the body of the message.
And I find it mildly amusing that apparently my e-mail address is important
enough to be included in mailings to sell property in Florida, or CD's with
millions of e-mail addresses, including programs to send mail to all those
addresses. And of course those that think I can read Big5.
The most amusing however I found the one (already received many times)
advertising for warehouses where the request was made to print out the
mail and present the result to your CEO. Perhaps I will do that sometime;
I wonder what he will say.
--
dik t. winter, cwi, kruislaan 413, 1098 sj amsterdam, nederland, +31205924131
home: bovenover 215, 1025 jn amsterdam, nederland; http://www.cwi.nl/~dik/
Jack Klein
wrote in comp.lang.c:
Actually I disagree that it is off-topic. The original post, despite
the poster's original subject line stating that it was off-topic, was
about topicality, and discussions of topicality are always topical.
And discussions of what constitutes usenet SPAM in a particular group
can hardly be off-topic in that group, anymore than topicality is.
Even if it were not...
Just as there is a "spirit of C", not explicit in the standard but
understood all the same...
There is a "spirit of comp.lang.c".
Off-topic questions are certainly not welcome here. Nevertheless,
threads that start off with off-topic or topical issues sometimes spin
off into remote territory.
I notice that you haven't posted to the Euro 2000 thread to point out
that it is off-topic! :)
Chris Kuan
>On Tue, 4 Jul 2000 01:11:28 GMT, "Dik T. Winter" <Dik.W...@cwi.nl>
>wrote in comp.lang.c:
>
>> This is all of course blatantly off-topic.
>I notice that you haven't posted to the Euro 2000 thread to point out
>that it is off-topic! :)
Too depressed, probably...
--
Chris Kuan, CSC Technology Services, formerly BHP Information Technology
Concatenate for email: mr gazpacho @ hotmail . com
"Law is a repository for the aimlessly clever" - Tim Freedman
Richard Heathfield
>
> On Tue, 4 Jul 2000 01:11:28 GMT, "Dik T. Winter" <Dik.W...@cwi.nl>
> wrote in comp.lang.c:
>
> > This is all of course blatantly off-topic.
> >
<massive snip>
>
> I notice that you haven't posted to the Euro 2000 thread to point out
> that it is off-topic! :)
Let the children play. They'll soon tire of it. If the thread's still
extant in a day or so, though, it might be time to start banging heads
together.
--
Richard Heathfield
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
C FAQ: http://www.eskimo.com/~scs/C-faq/top.html
52 K&R Answers: http://users.powernet.co.uk/eton/kandr2/index.html (45
to go)
Jack Klein
<bin...@eton.powernet.co.uk> wrote in comp.lang.c:
> Jack Klein wrote:
> >
> > On Tue, 4 Jul 2000 01:11:28 GMT, "Dik T. Winter" <Dik.W...@cwi.nl>
> > wrote in comp.lang.c:
> >
> > > This is all of course blatantly off-topic.
> > >
>
> <massive snip>
> >
> > I notice that you haven't posted to the Euro 2000 thread to point out
> > that it is off-topic! :)
>
> Let the children play. They'll soon tire of it. If the thread's still
> extant in a day or so, though, it might be time to start banging heads
> together.
<absolutely off-topic>
Of course, if I had money on the game, it might look more relevant to
me.
</absolutely off-topic>
Chris Doran
T, but here goes:-
In article <08b6cd2c...@usw-ex0105-040.remarq.com>,
petere <peliades...@sanderson.net.au.invalid> wrote:
> There appear to be a disproportianate amount of posts that are
> only incidentally relevant to ANSI/ISO C (which is what
> comp.lang.c is dedicated to). Additionally, much platform
> specific information is proferred as ANSI/ISO C information.
The real problem is that this is called c.l._c_, and many people
outside academia just do not realise that there is any C other than
that which comes with their machine/compiler, let alone consider
whether or not write() is in the ANSI standard. Anyone just looking for
help with their C progam naturally lights on c.l.c.
The NG is being perceived to serve two incompatible functions: (a) to
help those who "just want to get the job done" incidentally in C, and
(b) those who want to discuss or promote pure ANSI C.
Is the solution to start a new c.l.ansic for the latter? The amount of
traffic on c.l.c seems to warrant a split, and such a name will be less
visible to the former.
> Further, redirecting the posters or pointing out errors seems to
> be causing an unnaturally hostile response. I'd attribute the
> vitriolic response to the (relative) anonymity of usenet.
The irritation is caused by the fact that all these flames redirecting
people and dissecting lines of code irrelevant to the actual problem
are taking more space than an OT message + reply would do. This costs
time reading them (and money for those of us who don't get free local
calls). Further, a question with several replies may be ignored on the
assumption that they're answers, which means someone may get no answer
at all.
A while back I started a campaign to get rid of "Test, please ignore"
postings by referring people to uk.test etc., but this also caused
irritation no matter how gently I tried to word it, so I stopped.
IMHO the best response to an OT posting is to answer it if you can,
adding a polite "BTW you might do better at comp.lang.yourcompiler". If
you can't answer it, don't reply at all or if no one has done so after
a few days, then and only then refer the OP to a better NG.
>...
As to other posts in this thread:-
keen.com took over remarq (or vice versa) a few weeks ago, and ever
since then the advert "signature" has appeared.
My most amusing SPAM was to be invited to have an entry in the
directory of prominent African Americans. What made them think I
qualified? I replied that I am neither American nor African :-)
Chris
Sent via Deja.com http://www.deja.com/
Before you buy.
Dik T. Winter
> >I notice that you haven't posted to the Euro 2000 thread to point out
> >that it is off-topic! :)
>
Not at all. But after having seen 22 soccer matches on television and
snatches of 3 others, I have temporarily enough. On for the Tour.
petere
original thread "On the matter of ANSI/ISO C and the purpose
clc" (sic) -- which should have been "On the matter of ANSI/ISO
C and the purpose of clc" -- I thought I'd re-enter the
discussion.
Decklin Foster wrote
>May I suggest news.cis.dfn.de? They require you to register and
>use authentication to connect, but provide access for free.
>Point your web browser at the same hostname to get signup
>instructions.
Thanks for the URL. I am waiting for my registration to complete
so you will have to put up with the spam in the signature line
until then.
Chris Doran wrote
>The real problem is that this is called c.l._c_, and many
>people outside academia just do not realise that there is any C
>other than that which comes with their machine/compiler, let
>alone consider whether or not write() is in the ANSI standard.
>Anyone just looking for help with their C progam naturally
>lights on c.l.c.
This is a good point. I'm sure the NGs name comp.lang.c by its
generality implies a "miscellaneous C issues forum" to the
casual observer.
Chris Doran wrote
>Is the solution to start a new c.l.ansic for the latter? The
>amount of traffic on c.l.c seems to warrant a split, and such a
>name will be less visible to the former.
"comp.lang.isoc" or "comp.lang.stdc" sound good, more specific
than "com.lang.c". I'd be interested to know what the main
contributors to the forum think of this.
Regis
> "comp.lang.isoc" or "comp.lang.stdc" sound good, more specific
> than "com.lang.c".
I would never guess that "isoc" is a related to C.
I would misread it as one word.
With "iso-c" I would understand.
(Or better "c-iso" because when we search for a newsgoup for C
we tend to search first at letter 'c'. Of course if the aim is to
decrease traffic then the first one is better)
"comp.lang.stdc" looks too much like "comp.std.c"
By the way, with the capital letter of 'C' has been dropped?
Régis
mike burrell
> The real problem is that this is called c.l._c_, and many people
> outside academia just do not realise that there is any C other than
> that which comes with their machine/compiler, let alone consider
> whether or not write() is in the ANSI standard. Anyone just looking for
> help with their C progam naturally lights on c.l.c.
they're wrong. big deal.
> The NG is being perceived to serve two incompatible functions: (a) to
> help those who "just want to get the job done" incidentally in C, and
> (b) those who want to discuss or promote pure ANSI C.
> Is the solution to start a new c.l.ansic for the latter? The amount of
> traffic on c.l.c seems to warrant a split, and such a name will be less
> visible to the former.
and what would propose happens on clc then? it would be about as useful as
something like comp.help. no one would be able to help anyone because no
one would be on the same wavelength. "hello, why can't i move my window to
the right?" "i don't know, maybe the robot arm isn't attached properly?" in
order for useful discussions to take place, there needs to be common ground,
and in the case of the C language, that is the ISO C standard.
--
/"\ m i k e b u r r e l l
\ / ASCII RIBBON CAMPAIGN mik...@home.com
X AGAINST HTML MAIL,
/ \ AND NEWS TOO, dammit
Richard Heathfield
>
>
> Is the solution to start a new c.l.ansic for the latter? The amount of
> traffic on c.l.c seems to warrant a split, and such a name will be less
> visible to the former.
>
Your chances of getting sufficient support for this to elicit a change,
on a scale of zero to one, are significantly less than zero.
comp.lang.c goes back a long way; this (along with comp.std.c) is where
dmr posts when he wants to post something C-related.
How would you feel about being the one to tell him to push off to
comp.lang.ansic? Telling dmr he's off-topic is not a task for the
faint-hearted, I assure you.
Thomas M. Sommers
>
> Actually I disagree that it is off-topic. The original post, despite
> the poster's original subject line stating that it was off-topic, was
> about topicality, and discussions of topicality are always topical.
Obviously, there must be a place to discuss the topicality of posts in
any particular newsgroup. Newgroups can be divided into two classes:
those that allow topicality discussions, and those that don't. Now
consider the newgroup that allows the discussion of the topicality of
posts in all newsgroups that don't allow topicality discussions. Where
does one discuss the topicality of posts in that newsgroup? For extra
credit, who shaves the barber?
Thomas M. Sommers
>
> petere wrote:
>
> > "comp.lang.isoc" or "comp.lang.stdc" sound good, more specific
> > than "com.lang.c".
How about comp.lang.c.iso?
Thomas M. Sommers
>
> Chris Doran <chris...@my-deja.com> wrote:
> > Is the solution to start a new c.l.ansic for the latter? The amount of
> > traffic on c.l.c seems to warrant a split, and such a name will be less
> > visible to the former.
>
> something like comp.help. no one would be able to help anyone because no
> one would be on the same wavelength. "hello, why can't i move my window to
> the right?" "i don't know, maybe the robot arm isn't attached properly?" in
> order for useful discussions to take place, there needs to be common ground,
> and in the case of the C language, that is the ISO C standard.
It couldn't be worse than it is now, with half of the posts complaining
that other posts are off-topic. At worst, the old (current) newsgroup
would wither and die. The advantage of a new newsgroup or a change of
name of the current group is that the purpose of the group would be more
obvious to the casual poster.
Thomas M. Sommers
>
> comp.lang.c goes back a long way; this (along with comp.std.c) is where
> dmr posts when he wants to post something C-related.
If I am not mistaken, comp.lang.c predates the standard. Omnia
mutantur, nos et mutamur in illis.
> How would you feel about being the one to tell him to push off to
> comp.lang.ansic? Telling dmr he's off-topic is not a task for the
> faint-hearted, I assure you.
I'm sure Mr. Ritchie could cope with a change. Anyway, C would still be
topical in comp.lang.c.
Kaz Kylheku
I think that there isn't any good reason to add another element to
a dotted name, unless other constrasting suffixes are introduced:
other comp.lang.c.* newsgroups.
What will stand in contrast to the .iso?
comp.lang.c.non-iso?
comp.lang.c.tangentially-related-to-c-by-six-degrees-of-separation-or-less?
comp.lang.c.gnu-msvisual-borland-objective-lcc-dec-sun-hp-ibm-zortech-lattice?
Humor me.
--
#exclude <windows.h>
Bryan Williams
> Jack Klein wrote:
> >
> > Actually I disagree that it is off-topic. The original post, despite
> > the poster's original subject line stating that it was off-topic, was
> > about topicality, and discussions of topicality are always topical.
>
> Obviously, there must be a place to discuss the topicality of posts in
> any particular newsgroup. Newgroups can be divided into two classes:
> those that allow topicality discussions,
This one.
> and those that don't.
There aren't any (at least not in comp.lang.c)
> Now consider the newgroup that allows the discussion of the topicality
> of
> posts in all newsgroups that don't allow topicality discussions. Where
> does one discuss the topicality of posts in that newsgroup? For extra
> credit, who shaves the barber?
The barber will shave himself, who else would he trust to do it right ?
--
Bryan Williams
"Of all the things I have lost I miss my mind the most."
- Ozzy O.
Jack Klein
<tm...@mail.ptd.net> wrote in comp.lang.c:
Surely only if it did not include any K&R1, C90, or C99 functions, and
certainly not unless it started with void main()!
Geoff Field
k...@ashi.footprints.net wrote:
> On Wed, 05 Jul 2000 00:32:20 GMT, Thomas M. Sommers
<tm...@mail.ptd.net> wrote:
> >Regis wrote:
> >>
> >> petere wrote:
> >>
> >> > "comp.lang.isoc" or "comp.lang.stdc" sound good, more specific
> >> > than "com.lang.c".
> >
> >How about comp.lang.c.iso?
How about comp.std.c? Oh, there's already one? Why not go there?
> I think that there isn't any good reason to add another element to
> a dotted name, unless other constrasting suffixes are introduced:
> other comp.lang.c.* newsgroups.
>
> What will stand in contrast to the .iso?
>
> comp.lang.c.non-iso?
>
> comp.lang.c.tangentially-related-to-c-by-six-degrees-of-separation-or-
less?
>
> comp.lang.c.gnu-msvisual-borland-objective-lcc-dec-sun-hp-ibm-zortech-
lattice?
comp.lang.c.pc-based
comp.lang.c.others?
> Humor me.
>
> --
> #exclude <windows.h>
comp.lang.c.windoze?
--
Geoff Field of XXXXia
Professional geek, amateur stage-levelling gauge
Regis
> > comp.lang.c.gnu-msvisual-borland-objective-lcc-dec-sun-hp-ibm-zortech-
> lattice?
>
> comp.lang.c.pc-based
> comp.lang.c.others?
> comp.lang.c.windoze?
clc.com-port
clc.move-cursor
clc.change-color
clc.is-key-pressed
clc.read-one-key
clc.clear-screen
clc.void-main
clc.tisdale
clc.barthez
clc.van-der-sar
clc.portugal-italy-in-final
clc.portugal-netherlands-in-final
clc.italy-netherlands-in-final
clc.france-sucks
Régis
Rick Dearman
If the casual poster had bothered to read the FAQ for usenet, or the
FAQ for this group, or Billy Chambless weekly posting, or the
multitude of information telling them how to find the correct location
to post their question. Or if they had even bothered in most cases to
define the problem properly they would have asked it in the correct
place.
The problem is really that given all the informatin above the "casual
poster" still tries to take the easy way out and just post to the
newsgroup that is "seems" to be the right one. They didn't bother to
find out so they are redirected, sometimes nicely, sometimes not.
Perhaps the problem is that given the multitude of newsgroups to
select from the "casual poster" is overwhelmed with choices and so
picks the one they think they want. So giving them two more choices
isn't really going to help them much.
Also this thread seems to rage along here every two months. This
thread will be followed closely by the "Why can't you be nice when
redirecting" thread. The gist of this thread will be that all posts to
newbies should be five pages long and explain in detail why they are
not posting in the right place, and we will all be required to use
smilely faces with each response.
Then there will be the "Why don't c.l.c have a charter", where
everyone will expound on the wonders of chartered newsgroups.
Of course the thread which hooked me on reading this newsgroup was
started by a poster who kept going on about how if Dennis Richie had
known how popular C would become he would have put in such&such a
feature, blah, blah, blah ... And then Mr. Richie replied explaining
why he'd left out such&such function. BAM! I must have laughed for
hours! ( Humm ... I appear to be digressing )
Anyway my point is that this subject comes up every other month, and
basicly most people feel it would serve no purpose. Now in the recent
past some people felt so strongly that newbies should have a place to
ask C & C++ related questions while learning they started their own
newsgroup comp.learn.c.c++ (or some such gubbins) anyway if you guys
all feel their is a need for comp.lang.c.ask-us-anything-you-like go
for it. There are procedures in place for that. Or you could start
comp.lang.c.iso but unless you get a vast majority of the regular
posters here to switch then you will not have many answers so people
will continue to post here, thereby not solving anything.
--
+ Rick Dearman, rd...@egomaniacs.net.nospam
+---------------------------------------------
When in Rome, do Roman numerals in math.
--anon
Chris Mears
scribbled the following:
>In article <slrn8m51e...@ashi.FootPrints.net>,
> k...@ashi.footprints.net wrote:
>> On Wed, 05 Jul 2000 00:32:20 GMT, Thomas M. Sommers
><tm...@mail.ptd.net> wrote:
>> >Regis wrote:
>> >>
>> >> petere wrote:
>> >>
>> >> > "comp.lang.isoc" or "comp.lang.stdc" sound good, more specific
>> >> > than "com.lang.c".
>> >
>> >How about comp.lang.c.iso?
>
>How about comp.std.c? Oh, there's already one? Why not go there?
Because that's about a different topic. You might as well go to
alt.fan.adjective-army.
--
Chris Mears
ICQ: 36697123
C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
Dann Corbit
news:m23dlpu...@cartmansmum.devteam.egomaniacs.net...
I agree. Pandering to the stupid does no one any good, including the stupid
themselves. If you split news:comp.lang.c into a thousand tiny fragments,
you will still get idiotic posts from people who didn't read the FAQ to
fragment <x> etc. What we have in news:comp.lang.c is a remarkable resource
for people who would like to learn about the C language. To try to "fix" it
is not going to improve that situation (which is already excellent) and it
won't solve the problems either.
There is nothing wrong with the current USENET organization. Many people
don't know how to use it yet. Can people really believe that the sensible
solution is to break it up to pander to idiots and (even more remarkably)
that afterwards they will somehow cease to be idiots?
For those who are not idiots (and there are plenty who simply stumbled by
the wrong newsgroups innocently for whatever reasons) they will assimilate
information and learn to use the resource properly under the current system.
Therefore, the frequent discussions of fragmenting c.l.c into splinters
really should be put to rest. Of course it won't, but it ought to be.
--
C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
"The C-FAQ Book" ISBN 0-201-84519-9
C.A.P. Newsgroup http://www.dejanews.com/~c_a_p
C.A.P. FAQ: ftp://38.168.214.175/pub/Chess%20Analysis%20Project%20FAQ.htm
Mark McIntyre
<tm...@mail.ptd.net> wrote:
>Richard Heathfield wrote:
>>
>> comp.lang.c goes back a long way; this (along with comp.std.c) is where
>> dmr posts when he wants to post something C-related.
>
>If I am not mistaken, comp.lang.c predates the standard. Omnia
>mutantur, nos et mutamur in illis.
I think Dennis & co were actually writing C before the standard too...
your point is slightly irrelevant
>> How would you feel about being the one to tell him to push off to
>> comp.lang.ansic? Telling dmr he's off-topic is not a task for the
>> faint-hearted, I assure you.
>
>I'm sure Mr. Ritchie could cope with a change. Anyway, C would still be
>topical in comp.lang.c.
but more to the point, why should he? clc is where we talk about
ansi/iso C. If you want a group to talk about something else, you can
easily create "comp.lang.c.extensions_to_standard" and that would be
nicely clear to everyone
--
Mark McIntyre
C- FAQ: http://www.eskimo.com/~scs/C-faq/top.html
Chris Torek
Geoff Field <geoff...@my-deja.com> writes:
>How about comp.std.c? Oh, there's already one? Why not go there?
I was pretty sure I had answered this kind of proposal before, so
I took a look in my "saved answers" file. Here it is -- and gosh,
it was from *you*! :-)
From: to...@elf.bsdi.com (Chris Torek)
Newsgroups: comp.lang.c
Subject: Re: HELP: Wrong answer when int / int!!!!
Message-ID: <89qdkt$793$1...@elf.bsdi.com>
Date: 3 Mar 2000 23:23:09 -0800
In article <89q2pk$jsa$1...@nnrp1.deja.com>
Geoff Field <geoff...@my-deja.com> writes, in part (mostly as
an aside):
>I would also like to point out that this is comp.*lang*.c rather than
>comp.*std*.c so those people who direct every poster with even slightly
>non-standard questions to other news groups without helping them are
>themselves guilty of decreasing the signal-to-noise ratio here.
(Note - topicality discussions are always on-topic :-) )
The group comp.std.c is not about programming in Standard C. That
(programming in Standard C) is what comp.lang.c is for -- comp.std.c
is for discussion about standardization of C. Are you confused yet? :-)
The distinction is perhaps best done by example. The question:
On page 47, line 13, of the current draft, is the "{"
character a typo?
is appropriate for comp.std.c, but not for comp.lang.c. Gripes
about missing index entries are appropriate for comp.std.c, but
not comp.lang.c. Questions about when and where the next
standardization committee meeting is to take place ... yep, you
guessed it.
There is some overlap, of course. One can ask whether some bit of
C code is "correct" in either group. The kind of answer one should,
at least in theory, get, is different, though: in comp.std.c, the
answer would be either "yes" or "no because of section 5.4.3, p. 41,
ll. 12--14, combined with 3.1.5, p. 14", while in comp.lang.c, the
answer might be "yes" or "no, but if you write this instead, it is
guaranteed to work and should do what you want".
Since not everyone reads both groups, sometimes it is wise to give
a comp.std.c answer in comp.lang.c, or vice versa. These should
probably come with a suggestion that further discussion occur in
the other group.
--
In-Real-Life: Chris Torek, Berkeley Software Design Inc
El Cerrito, CA, USA Domain: to...@bsdi.com +1 510 234 3167
http://claw.bsdi.com/torek/ (not always up) I report spam to abuse@.
Ben Pfaff
> How would you feel about being the one to tell him to push off to
> comp.lang.ansic? Telling dmr he's off-topic is not a task for the
> faint-hearted, I assure you.
Anyone who's about to flame dmr on clc should think first about
what happened to the last guy who tried that.
(He got a new quote for his .sig.)
Floyd Davidson
>Richard Heathfield <bin...@eton.powernet.co.uk> wrote:
>>How would you feel about being the one to tell him to push off to
>>comp.lang.ansic? Telling dmr he's off-topic is not a task for the
>>faint-hearted, I assure you.
>
>announcement from DMR about the source code of some old UNIX compilers
>being made available for historic interest.
Hey, what kind of an asshole would give an old gentleman like DMR
a hard time about a thing like that, eh?
Tell us Richard... :-)
--
Floyd L. Davidson fl...@barrow.com
Ukpeagvik (Barrow, Alaska)
Thomas M. Sommers
>
> On Wed, 05 Jul 2000 00:47:19 GMT, "Thomas M. Sommers"
> <tm...@mail.ptd.net> wrote:
>
> >Richard Heathfield wrote:
> >>
> >> comp.lang.c goes back a long way; this (along with comp.std.c) is where
> >> dmr posts when he wants to post something C-related.
> >
> >If I am not mistaken, comp.lang.c predates the standard. Omnia
> >mutantur, nos et mutamur in illis.
>
> I think Dennis & co were actually writing C before the standard too...
> your point is slightly irrelevant
I merely point out that comp.lang.c did not always limit itself to
standard C.
> >> How would you feel about being the one to tell him to push off to
> >> comp.lang.ansic? Telling dmr he's off-topic is not a task for the
> >> faint-hearted, I assure you.
> >
> >I'm sure Mr. Ritchie could cope with a change. Anyway, C would still be
> >topical in comp.lang.c.
>
> but more to the point, why should he? clc is where we talk about
> ansi/iso C. If you want a group to talk about something else, you can
> easily create "comp.lang.c.extensions_to_standard" and that would be
> nicely clear to everyone
I am only suggesting that if the name of the newsgroup somehow indicated
more clearly that it was only concerned with standard C, the number of
off-topic posts (and the number of posts complaining about off-topic
posts) might be reduced. I would think that that would be a Good Thing.
Thomas M. Sommers
Never mind. It was a feeble attempt at humor, based on Russell's
Paradox.
Kaz Kylheku
<bin...@eton.powernet.co.uk> wrote:
>How would you feel about being the one to tell him to push off to
>comp.lang.ansic? Telling dmr he's off-topic is not a task for the
>faint-hearted, I assure you.
No problem. It has already happened, has it not? I vaguely remember an
announcement from DMR about the source code of some old UNIX compilers
being made available for historic interest.
--
#exclude <windows.h>
Richard Heathfield
>
> k...@ashi.footprints.net (Kaz Kylheku) wrote:
> >Richard Heathfield <bin...@eton.powernet.co.uk> wrote:
> >>How would you feel about being the one to tell him to push off to
> >>comp.lang.ansic? Telling dmr he's off-topic is not a task for the
> >>faint-hearted, I assure you.
> >
> >No problem. It has already happened, has it not? I vaguely remember an
> >announcement from DMR about the source code of some old UNIX compilers
> >being made available for historic interest.
Awwww, he was out of order. DEC tapes, indeed! :-)
>
> Hey, what kind of an asshole would give an old gentleman like DMR
> a hard time about a thing like that, eh?
>
> Tell us Richard... :-)
What are you looking at me for? :-)
For those of you who don't know, the incident in question took place on
the 29th July 1999, and started when dmr posted a thread with the
subject "A primeval C compiler".
At the time, comp.lang.c was getting rather strident in its search for
pure topicality, to a level which I thought was faintly ridiculous. Here
was the perfect opportunity to demonstrate this to them in a most
graphic way.
Many clueful people understood my purpose immediately. Some clueful
people didn't. Here was a flamewar in the making, but for once I had the
good sense to keep my mouth shut, so it died down quickly.
Look it up in Deja if you care.
The dmr quote in my sigblock is from an email in which he expressed
surprise at the reaction. (I just thought I'd slip that in casually...)
--
Richard Heathfield
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
"You will get a wide range of opinion on this ranging from "yes" to
"no"." - Lawrence Kirby, 4 July 2000.
Richard Heathfield
Bryan misread it (something he doesn't often do), and consequently
didn't "get" it. He does understand now; I know this because I thumped
him over the head with a large book (the book was about TCP/IP, as it
happens, so it's off-topic here) and shouted at him about all dogs not
being a dog, until he nodded.
Steve Gravrock
>Richard Heathfield wrote:
>>
>> comp.lang.c goes back a long way; this (along with comp.std.c) is where
>> dmr posts when he wants to post something C-related.
>
>If I am not mistaken, comp.lang.c predates the standard.
*nods* That would be why the newsgroup line, at least on my news server,
is:
comp.lang.c Discussion about C.
You are correct. Obviously a time existed when discussion of topics other
than Standard C was acceptable in comp.lang.c. I might point out, however,
that since then the number of platforms that run C has increased greatly.
I would speculate that the convention of only posting articles related to
Standard C arose because it became increasingly difficult to handle
discussion of such diverse platforms in a single newsgroup.
(Disclaimer: the key word here is "speculate". I wasn't around back then.
If someone who was could provide a bit of history I'd be grateful.)
Regardless of what was on-topic in the past, Richard's point stands.
comp.lang.c is an old and very well established newsgroup. In order for
the group to be renamed or a charter applied, there would need to be an
absolutely compelling reason. No such reason exists. Despite the high
volume and imperfect signal-to-noise ratio, comp.lang.c is a basically
functional newsgroup which provides an invaluable resource to a lot of
people.
If anyone who is advocating a renaming seriously thinks that it would be
pushed through, I suggest lurking in news.groups for two months before
proposing it again.
>Omnia
>mutantur, nos et mutamur in illis.
What does that mean? I'm afraid I haven't had much exposure to classical
languages.
>> How would you feel about being the one to tell him to push off to
>> comp.lang.ansic? Telling dmr he's off-topic is not a task for the
>> faint-hearted, I assure you.
>
>I'm sure Mr. Ritchie could cope with a change.
I'm sure he could too, but unless there is a consensus in the group that the
topic needs to be changed, he won't have to. No such consensus exists.
>Anyway, C would still be
>topical in comp.lang.c.
Hmm... where have I heard that before?
I have come to the conclusion that there aren't any problems in comp.lang.c
which would be best addressed by creating or deleting newsgroups. A name
change for comp.lang.c is out of the question, and appropriate newsgroups
already exist for all of the off-topic posts. What we have is not a poorly
named newsgroup or a shortage of newsgroups, but a user education problem.
Consider the typical scenario:
1. Beginning C programmer wants answers to a platform-specific question.
2. Because of unfamiliarity with C, our programmer doesn't realize that
the problem is platform-specific.
3. Our programmer sees "comp.lang.c" in a list of newsgroups.
4. Because of unfamiliarity with Usenet, our programmer posts to
comp.lang.c without first reading the FAQ or lurking.
There are two things that our newbie needs to learn: the limits of Standard
C, and the conventions of netiquite. The former is mainly accomplished by
redirecting posters and pointing them to the FAQ. This is important, even
if it's tangential to what we'd rather be discussing. The latter is a
problem that every newsgroup faces and none ever really solve. There will
always be clueless newbies who don't lurk or read FAQs. If you can't deal
with them, comp.lang.c.moderated is next door.
To say the same thing in far fewer words, I think the best course of action
is for comp.lang.c to carry on business as usual. Readers who have trouble
dealing with the number of short, off-topic threads should use their
newsreader's facility for killfiling by thread, or use a different
newsreader if theirs lacks such a facility.
/steve, who found this soapbox just sitting here
Richard Bos
I'm afraid you seriously underestimate the stupidity and perversity of
the common Usenet luser. It won't work.
Richard
Joona I Palaste
: In article <3962845B...@mail.ptd.net>, Thomas M. Sommers wrote:
:>Omnia
:>mutantur, nos et mutamur in illis.
: What does that mean? I'm afraid I haven't had much exposure to classical
: languages.
Everything changes, we too change in them.
"Omnia" means "everything", "mutantur" is the third person plural form
of "mutari", "to change", "nos" means "we", "et" in this case means
"too", "mutamur" is the first person plural form of "mutari", "in"
means "in", "illis" is the ablative form of "illis" (I think), meaning
"them".
--
/-- Joona Palaste (pal...@cc.helsinki.fi) ---------------------------\
| Kingpriest of "The Flying Lemon Tree" G++ FR FW+ M- #80 D+ ADA N+++ |
| http://www.helsinki.fi/~palaste W++ B OP+ |
\----------------------------------------- Finland rules! ------------/
"All that flower power is no match for my glower power!"
- Montgomery Burns
Steamer
> >If I am not mistaken, comp.lang.c predates the standard.
>
> *nods* That would be why the newsgroup line, at least on my news server,
> is:
>
> comp.lang.c Discussion about C.
This does appear to be the only official description of the purpose
of this newsgroup. But I think it postdates the standard. The file
at ftp://ftp.uu.net/usenet/control/comp/comp.lang.c.Z shows the date
of the first message as
1 Aug 91 09:01:44 GMT
> >Omnia
> >mutantur, nos et mutamur in illis.
>
> What does that mean? I'm afraid I haven't had much exposure to classical
> languages.
"All things change, and we change with them."
>appropriate newsgroups already exist for all of the off-topic posts
Really? Where is the appropriate place to discuss POSIX extensions?
The only newsgroup with the word "posix" in it on my ISP's news-server
is vmsnet.vms-posix, which doesn't look like a good place for cross-
platform discussions, and probably isn't even about C.
S.
Floyd Davidson
Lets be realistic though. It won't work, but not because some
mythical common Usenet user is a luser who is stupid and
perverse. It almost worked, but not quite, when Usenet was 95%
geeks with over-developed left-brained talents (and a lack of
right-brained perspective). It can't work when "normal" people
are invovled, and whether we like it or not, the typical Usenet
reader today is much closer to normal than when we geeks had
"our" thing going.
Anybody who thinks these various suggestions should work (name
changes, demanding newbies read the FAQ, lurk for a month,
etc. etc.) *is* a candidate for the title of Stupid, Perverse
Luser Pedant.
The only effective method I've found so far, is to filter out
what is of no value to me personally (which might be very
different than the filter criteria you or someone else might
use). Killfiles, scoring systems, and good newsreader software
which sorts and displays articles by subject, author, line
counts and so on is exceedingly helpful. But the bottom line is
always going to be a discerning reader!
Just don't read what you don't need to, and ignore the rest.
Floyd
Floyd Davidson
>Thomas M. Sommers wrote:
>>
>>If I am not mistaken, comp.lang.c predates the standard.
>
>*nods* That would be why the newsgroup line, at least on my
>news server, is:
>
>comp.lang.c Discussion about C.
>
>topics other than Standard C was acceptable in comp.lang.c. I
>might point out, however,
The newsgroup comp.lang.c has always been to discuss the
language C, and that has not changed even slightly. What has
changed is the definition of the C language (and indeed, even
what defines it has changed).
Originally C was defined in many ways, all derived from "The C
Programming Language" by K&R, but none of the definitions were
limited strictly to K&R. Even AT&T's UNIX compiler at the time
was significantly different that K&R. Since 1989 the ISO/ANSI
Standard has defined what C is, _and therefore has defined what
is on topic in comp.lang.c_.
But when the language was standardized, that single definition
_gradually_ became the one and the only on topic definition of C
on comp.lang.c. (The bitching about what was or not on topic
began quickly enough though!) There was a rather long
transition period to allow the implementation of the Standard.
That transition period is no different than what is taking place
today, as the 1989 Standard is still discussed and almost all
compilers actually used by readers of the group are based on the
1989 Standard, even though a new Standard exists and is in some
ways significantly different.
>that since then the number of platforms that run C has
>increased greatly. I would speculate that the convention of
>only posting articles related to Standard C arose because it
>became increasingly difficult to handle discussion of such
>diverse platforms in a single newsgroup.
>
>(Disclaimer: the key word here is "speculate". I wasn't around
>back then. If someone who was could provide a bit of history
>I'd be grateful.)
Adopting a Standard, or a new revision of such, has the side
effect of changing the specifics of what is on topic in a forum
such as comp.lang.c, and in some ways the transition is
difficult. (Here we are 10 years later and some people still want
to be in the transition from K&R to ANSI!) No doubt that same
debate will go on for the next decade in regard to the 1989
Standard as opposed to the 1999 revision.
Another interesting side effect is that virtually all reference
documentation and tutorials become obsolete overnight. Today
there must be a thousand books in print on the C language that
are obsolete as of 1999. In 1989 there were relatively few, but
it happened that in the 3-4 years prior to '89 an accumulation
of some of the finest question and answer style explanations of
C programming had occurred right here on comp.lang.c. The
quality of the answers provided far exceeded any available
published material, and at least some fairly serious talk was
going on about gathering up the archives and publishing the
accumulated c.l.c articles of Chris Torek!
Unfortunately, with adoption of the Standard in 1989, it all
became obsolete...
(Maybe Lawrence Kirby and Chris Torek should get together
in 2-3 years and publish a collection of their posts between
now and then on the new Standard. I cannot imagine that it
would be less than the most on topic questions answered in the
most illuminating way. And three years from now will be the
right timing for the new Standard to hit full stride.)
>Regardless of what was on-topic in the past, Richard's point stands.
>comp.lang.c is an old and very well established newsgroup. In order for
>the group to be renamed or a charter applied, there would need to be an
>absolutely compelling reason. No such reason exists. Despite the high
>volume and imperfect signal-to-noise ratio, comp.lang.c is a basically
>functional newsgroup which provides an invaluable resource to a lot of
>people.
Plus, the well understood "charter", even if it consists only of
the three words "Discussion about C." is *perfect*.
>If anyone who is advocating a renaming seriously thinks that it would be
>pushed through, I suggest lurking in news.groups for two months before
>proposing it again.
(Translation: forget it!)
[large snip]
>4. Because of unfamiliarity with Usenet, our programmer posts to
> comp.lang.c without first reading the FAQ or lurking.
>
>... things that our newbie needs to learn: ... conventions of netiquite.
>... The latter is a problem that every newsgroup faces and none
>ever really solve. There will always be clueless newbies who
>don't lurk or read FAQs. If you can't deal with them,
>comp.lang.c.moderated is next door.
I would suggest it is indeed a fact that groups "never ever
really solve" that problem, which leans towards the concept that
it is not the newbies who are clueless, but those who think that
newbies necessarily should lurk, must read the FAQ, and must
become clueful prior to posting to the group. That is either a
pie in the sky or a head in the sand concept! *IT IS NOT GOING
TO HAPPEN!*
Demanding that it should happen provides therapeutic noise for
pedants, and nothing else.
>To say the same thing in far fewer words, I think the best course of action
>is for comp.lang.c to carry on business as usual. Readers who have trouble
>dealing with the number of short, off-topic threads should use their
>newsreader's facility for killfiling by thread, or use a different
>newsreader if theirs lacks such a facility.
>
>/steve, who found this soapbox just sitting here
Thank you for sharing it. I see that we agreed totally, and had
overlapping perspectives.
Dik T. Winter
> This does appear to be the only official description of the purpose
> of this newsgroup. But I think it postdates the standard. The file
> at ftp://ftp.uu.net/usenet/control/comp/comp.lang.c.Z shows the date
> of the first message as
>
> 1 Aug 91 09:01:44 GMT
Well, this is wrong. From the README file with the 1987 obfuscated c
contest results:
Rules and results were posted to comp.lang.c, comp.unix.wizards with
an announcement in news.announce.important.
In 1986 it was still net.lang.c, so the group was renamed around that
time. Why this control message of 1991 does exist I have no idea.
--
dik t. winter, cwi, kruislaan 413, 1098 sj amsterdam, nederland, +31205924131
home: bovenover 215, 1025 jn amsterdam, nederland; http://www.cwi.nl/~dik/
Floyd Davidson
>Steve Gravrock wrote:
>
>> >If I am not mistaken, comp.lang.c predates the standard.
>>
>> *nods* That would be why the newsgroup line, at least on my news server,
>> is:
>>
>> comp.lang.c Discussion about C.
>
>of this newsgroup. But I think it postdates the standard. The file
>at ftp://ftp.uu.net/usenet/control/comp/comp.lang.c.Z shows the date
>of the first message as
>
> 1 Aug 91 09:01:44 GMT
Comp.lang.c came into existance when "the great name change"
took place. Before that it was net.lang.c, which existed at
least as early as 1985 (and did not exist in 1982).
Richard Bos
> r...@fdhoekstra.nl (Richard Bos) wrote:
> >"Thomas M. Sommers" <tm...@mail.ptd.net> wrote:
> >
> >> I am only suggesting that if the name of the newsgroup somehow indicated
> >> more clearly that it was only concerned with standard C, the number of
> >> off-topic posts (and the number of posts complaining about off-topic
> >> posts) might be reduced. I would think that that would be a Good Thing.
> >
> >I'm afraid you seriously underestimate the stupidity and perversity of
> >the common Usenet luser. It won't work.
>
> Lets be realistic though. It won't work, but not because some
> mythical common Usenet user is a luser who is stupid and
> perverse.
No, but that's not what I said. The common Usenet user is probably not a
luser. But those of them that are are generally perfect examples of a
little learning being a dangerous thing; people with just enough
learning to post to Usenet, but not to do so according to netiquette,
just enough to spell "void main" correctly, but not to accept that one
shouldn't do so, _and_ with the luser attitude, are much more annoyingly
stupid than your average teachable newbie.
> It almost worked, but not quite, when Usenet was 95%
> geeks with over-developed left-brained talents (and a lack of
> right-brained perspective). It can't work when "normal" people
> are invovled, and whether we like it or not, the typical Usenet
> reader today is much closer to normal than when we geeks had
> "our" thing going.
You seem to forget that this is a programming newsgroup, not
soc.chatter.average-people. I think we're entitled to expect a modicum
of clue from people who post here.
> Anybody who thinks these various suggestions should work (name
> changes, demanding newbies read the FAQ, lurk for a month,
> etc. etc.) *is* a candidate for the title of Stupid, Perverse
> Luser Pedant.
Thank you. I think they should work. I think they _won't_, but I think
they should.
> Just don't read what you don't need to, and ignore the rest.
Unfortunately, if enough people do so and not enough people participate
in trying to keep the group useful, it won't be very long before there
will not be anything left that is useful; it will have been flooded away
by the off-topic drivel.
The only way comp.lang.c can be kept useful is by assuming that there
will always be a certain amount of off-topic posts, and taking the
trouble to direct these to other, better suited newsgroups. Massive
kill-filing is no solution; creating new newsgroups is no solution,
either.
Richard
Bryan Williams
> Bryan Williams wrote:
> > <silly bit snipped>
> Never mind. It was a feeble attempt at humor, based on Russell's
> Paradox.
My sincerest apologies, I was feeling 'tired and emotional' at the time and
completely misread what you had written. (I understand my press office has
already released a statement to this effect).
Kelsey Bjarnason
"Richard Bos" <r...@fdhoekstra.nl> wrote in message
news:3964a3e8....@news.worldonline.nl...
> You seem to forget that this is a programming newsgroup, not
> soc.chatter.average-people. I think we're entitled to expect a modicum
> of clue from people who post here.
Why should you be entitled to anything? Was there an admission fee to the
newsgroup I'm not aware of, one which includes some sort of contractual
obligation on the part of the newsgroup management to filter out those you
consider clueless? If so, you'd better take it up with the nresgroup
management; if you can't get them to do their job filtering things, at least
perhaps they will refund your ticket price.
Steamer
> >of this newsgroup. But I think it postdates the standard. The file
> >at ftp://ftp.uu.net/usenet/control/comp/comp.lang.c.Z shows the date
> >of the first message as
> >
> > 1 Aug 91 09:01:44 GMT
>
> Comp.lang.c came into existance when "the great name change"
> took place.
Yes, this is undoubtedly the case. Can anyone explain the message
archived on ftp.uu.net?
> Before that it was net.lang.c, which existed at
> least as early as 1985 (and did not exist in 1982).
The following URL mentions a net.lang.c post by Dennis Ritchie entitled
"Joy of Reproduction" and dated 4 Nov 1982:
Thomas M. Sommers
>
> "Thomas M. Sommers" wrote:
>
> > Bryan Williams wrote:
> > > <silly bit snipped>
> > Never mind. It was a feeble attempt at humor, based on Russell's
> > Paradox.
>
> My sincerest apologies, I was feeling 'tired and emotional' at the time and
> completely misread what you had written. (I understand my press office has
> already released a statement to this effect).
If anyone should apologize it should be me, for inflicting the "silly
bit" on the newsgroup.
Brian Evan Blank
>
> Floyd Davidson wrote:
[munched]
> > Before that it was net.lang.c, which existed at
> > least as early as 1985 (and did not exist in 1982).
I think that it is true that net.lang.c did not exist at the
start of 1982.
> The following URL mentions a net.lang.c post by Dennis Ritchie entitled
> "Joy of Reproduction" and dated 4 Nov 1982:
>
> http://citeseer.nj.nec.com/cidcontext/456072
Prior to the creation of net.lang.c, C questions were considered
topical in net.unix-wizards (with a hyphen, not two dots). A few
examples of C code were also posted in net.sources. The last
C related article from 1982 that I have a copy of was posted to
net.unix-wizards on 10 May 1982. Based on the traffic pattern
that preceded that post, I think that it is likely that
net.lang.c began soon afterwards.
This ng actually works very well. (Try sci.math for a technical
newsgroup that is broken.) There is nothing wrong with c.l.c
that would be solved by a name change. A person who will post
here for a VB decompiler will post anything anywhere.
(To reach me, replace my last name with my first.)
--
Brian Blank * barf [ba:rf] 2."He suggested using
Department of Mathematics FORTRAN, and everybody barfed."
Washington University in St. Louis -The Shogakukan DICTIONARY OF
St. Louis, MO 63130 NEW ENGLISH (Second edition)
Dann Corbit
news:8k37um$o99$1...@nnrp1.deja.com...
[snip]
> 4) c.std.c is very much on the bottom of my list of ng's to read -
> after c.l.c. I haven't read the charter of the ng for quite a while.
> Given the charter, I even wonder why the ng's still around - after all,
> the various versions of the C standards have been out for a while now.
news:comp.std.c is *the* place to go to get a definitive answer to a
difficult question about _interpretation_ of the standard.
I post there about once per month (always with questions and never with
answers!)
;-)
Floyd Davidson
>Steamer wrote:
>>
>> Floyd Davidson wrote:
>> > Before that it was net.lang.c, which existed at
>> > least as early as 1985 (and did not exist in 1982).
>
>I think that it is true that net.lang.c did not exist at the
>start of 1982.
That is correct. See <http://communication.ucsd.edu/A-News/index.html>
which is an archive of Usenet messages from May 1981 through May 1982.
>> The following URL mentions a net.lang.c post by Dennis Ritchie entitled
>> "Joy of Reproduction" and dated 4 Nov 1982:
>>
>> http://citeseer.nj.nec.com/cidcontext/456072
>
>Prior to the creation of net.lang.c, C questions were considered
>topical in net.unix-wizards (with a hyphen, not two dots). A few
>examples of C code were also posted in net.sources. The last
>C related article from 1982 that I have a copy of was posted to
>net.unix-wizards on 10 May 1982. Based on the traffic pattern
>that preceded that post, I think that it is likely that
>net.lang.c began soon afterwards.
May 1982 does seem to be about when net.lang.c was created,
though I couldn't find anything more definitive than is
described here.
>This ng actually works very well. (Try sci.math for a technical
>newsgroup that is broken.) There is nothing wrong with c.l.c
>that would be solved by a name change. A person who will post
>here for a VB decompiler will post anything anywhere.
Exactly.
Geoff Field
I bow to your greater on-line presence.
In my defence, I will mention a few factors:
1) I tend to have a Write-Only Memory
2) I browse the newsgroups as a background task at work.
This means that I tend to view those newsgroups specific to my
current employment and project first - specifically PowerPC, Fire
industry stuff, realtime and embedded architectures. After all of
these have been viewed I *might* get around to c.l.c.
3) There's a *lot* of traffic in c.l.c about relevance and a tendency
for posters who are slightly platform-specific to be abused. The
atmosphere here is sometimes one of - to be frank - arrogant snobbery.
If it wasn't for the fact that some of the arrogance is justified
(there are some very knowledgeable posters here), I wouldn't even
bother with c.l.c.
4) c.std.c is very much on the bottom of my list of ng's to read -
after c.l.c. I haven't read the charter of the ng for quite a while.
Given the charter, I even wonder why the ng's still around - after all,
the various versions of the C standards have been out for a while now.
Nonetheless, thanks for your clarification.
Geoff
In article <8k0ilk$oh2$1...@elf.bsdi.com>,
--
Geoff Field of XXXXia
Professional geek, amateur stage-levelling gauge
Sent via Deja.com http://www.deja.com/
Before you buy.
Thomas M. Sommers
>
> "Thomas M. Sommers" <tm...@mail.ptd.net> writes:
>
> > mike burrell wrote:
> > >
> > > Chris Doran <chris...@my-deja.com> wrote:
> > > > Is the solution to start a new c.l.ansic for the latter? The amount of
> > > > traffic on c.l.c seems to warrant a split, and such a name will be less
> > > > visible to the former.
> > >
> > > and what would propose happens on clc then? it would be about as useful as
> > > something like comp.help. no one would be able to help anyone because no
> > > one would be on the same wavelength. "hello, why can't i move my window to
> > > the right?" "i don't know, maybe the robot arm isn't attached properly?" in
> > > order for useful discussions to take place, there needs to be common ground,
> > > and in the case of the C language, that is the ISO C standard.
> >
> > It couldn't be worse than it is now, with half of the posts complaining
> > that other posts are off-topic. At worst, the old (current) newsgroup
> > would wither and die. The advantage of a new newsgroup or a change of
> > name of the current group is that the purpose of the group would be more
> > obvious to the casual poster.
>
> If the casual poster had bothered to read the FAQ for usenet, or the
> FAQ for this group, or Billy Chambless weekly posting, or the
> multitude of information telling them how to find the correct location
> to post their question. Or if they had even bothered in most cases to
> define the problem properly they would have asked it in the correct
> place.
>
> The problem is really that given all the informatin above the "casual
> poster" still tries to take the easy way out and just post to the
> newsgroup that is "seems" to be the right one. They didn't bother to
> find out so they are redirected, sometimes nicely, sometimes not.
By definition, the casual, or novice, poster doesn't do these things,
and likely doesn't know that he should do these things. That is part of
the problem, and it won't go away.
It is not really unreasonable for a person who is trying to program in
C, and who has a question, to expect comp.lang.c to be the place to ask
the question. A person who is just starting out can't be expected to
know the contents of the standard as well as people who are much more
experienced.
> Perhaps the problem is that given the multitude of newsgroups to
> select from the "casual poster" is overwhelmed with choices and so
> picks the one they think they want. So giving them two more choices
> isn't really going to help them much.
It might help, if the names more accurately described the purposes of
the groups. Or it might not. But no one will know for sure unless it
is tried.
Thomas M. Sommers
>
> >Omnia
> >mutantur, nos et mutamur in illis.
>
> What does that mean? I'm afraid I haven't had much exposure to classical
> languages.
My apologies. I got the Oxford Dictionary of Quotations for Christmas,
and have been abusing it. It is attributed to Emperor Lothar I
(795-855), and means "All things change, and we change with them."
Often given as "Times change, and we change with them.", in which form
it dates to the late 16th century.
> I have come to the conclusion that there aren't any problems in comp.lang.c
> which would be best addressed by creating or deleting newsgroups. A name
> change for comp.lang.c is out of the question, and appropriate newsgroups
> already exist for all of the off-topic posts. What we have is not a poorly
> named newsgroup or a shortage of newsgroups, but a user education problem.
But the imprecise name of the newsgroup contributes to the education
problem.
> Consider the typical scenario:
>
> 1. Beginning C programmer wants answers to a platform-specific question.
> 2. Because of unfamiliarity with C, our programmer doesn't realize that
> the problem is platform-specific.
> 3. Our programmer sees "comp.lang.c" in a list of newsgroups.
> 4. Because of unfamiliarity with Usenet, our programmer posts to
> comp.lang.c without first reading the FAQ or lurking.
>
> There are two things that our newbie needs to learn: the limits of Standard
> C, and the conventions of netiquite. The former is mainly accomplished by
> redirecting posters and pointing them to the FAQ. This is important, even
> if it's tangential to what we'd rather be discussing. The latter is a
> problem that every newsgroup faces and none ever really solve. There will
> always be clueless newbies who don't lurk or read FAQs. If you can't deal
> with them, comp.lang.c.moderated is next door.
All true, but because of #2, even a novice who does lurk and read the
FAQ might still make a mistake about what is in the standard and what
isn't. And that person, who has made a good-faith effort, is likely to
be met with a "You can't do that in C" response.
<rant>
"You can't do that in C" is just wrong. You can do almost anything in
C. Of course, those who say that really mean "You can't do that using
only the variables and functions defined in standard C", which is true,
but pointless. You can do almost nothing useful without using something
not defined in the standard. You can't even say `int i;', because the
variable `i' is not defined in the standard.
</rant>
Thomas M. Sommers
>
> "Thomas M. Sommers" <tm...@mail.ptd.net> wrote:
>
> > I am only suggesting that if the name of the newsgroup somehow indicated
> > more clearly that it was only concerned with standard C, the number of
> > off-topic posts (and the number of posts complaining about off-topic
> > posts) might be reduced. I would think that that would be a Good Thing.
>
> I'm afraid you seriously underestimate the stupidity and perversity of
> the common Usenet luser. It won't work.
Nothing will reduce the noise level to zero, but it might be reduced,
even reduced significantly. No one will know how much unless it is
tried, though.
mike burrell
> variable `i' is not defined in the standard.
yes you can; the standard allows you to define variables.
there are some things which cannot be done entirely in C, though, such as
clearing the screen. in order to clear the screen in a C program, something
must be done (such as calling a function that is implemented in machine
language) which is not part of C. defining a variable does not fall in this
realm, and is part of C.
--
/"\ m i k e b u r r e l l
\ / ASCII RIBBON CAMPAIGN mik...@home.com
X AGAINST HTML MAIL,
/ \ AND NEWS TOO, dammit
Richard Heathfield
>
<much snippage - just one point I wanted to answer>
> 4) c.std.c is very much on the bottom of my list of ng's to read -
> after c.l.c. I haven't read the charter of the ng for quite a while.
> Given the charter, I even wonder why the ng's still around - after all,
> the various versions of the C standards have been out for a while now.
Geoff, this is a thinko, isn't it?
The C9899:1999 ISO Standard was only published this January, and
ratified by ANSI even later, so I'd have thought this would be a very
bad time for comp.std.c to close its doors. Many of us are still coming
to grips with the implications of the new Standard, a process which is
slower than it perhaps need be due to the lack of C99 compilers. The
comp.std.c newsgroup is going to loom large in our lives, I expect, over
the next year or three.
--
Richard Heathfield
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
C FAQ: http://www.eskimo.com/~scs/C-faq/top.html
Richard Heathfield
>
> It might help, if the names more accurately described the purposes of
> the groups. Or it might not. But no one will know for sure unless it
> is tried.
I don't have a problem with the existing name of comp.lang.c, and I
suspect you'll find most people who read it don't have a problem with it
either. But try it if you wish.
If you succeed (by some miracle) in creating a comp.lang.ansic or a
comp.lang.isoc, who will post there, hmmm? Nobody, that's who. We'll all
still post here, and the new group will fail. That's people for you, I'm
afraid. We're stubborn like that.
Nick Keighley
Richard Heathfield <bin...@eton.powernet.co.uk> wrote:
> Chris Doran wrote:
> >
> > Is the solution to start a new c.l.ansic for the latter? The amount
> > of traffic on c.l.c seems to warrant a split, and such a name will
> > be less visible to the former.
so the people who use the group correctly are supposed to move?
> Your chances of getting sufficient support for this to elicit a
> change, on a scale of zero to one, are significantly less than zero.
I was wondering if it had an imaginary component...
> comp.lang.c goes back a long way; this (along with comp.std.c) is
> where dmr posts when he wants to post something C-related.
>
> How would you feel about being the one to tell him to push off to
> comp.lang.ansic? Telling dmr he's off-topic is not a task for the
> faint-hearted, I assure you.
Richard is in a position to express an opinion on this... :-)
See his sig.
> "Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
--
I construct in my mind some unheard-of machine to move and carry
weights, making it possible to create great and wonderful things.
-- Leon Battista Albereti (1441)
Thomas M. Sommers
>
> "Thomas M. Sommers" wrote:
> >
> > It might help, if the names more accurately described the purposes of
> > the groups. Or it might not. But no one will know for sure unless it
> > is tried.
>
> I don't have a problem with the existing name of comp.lang.c, and I
> suspect you'll find most people who read it don't have a problem with it
> either. But try it if you wish.
>
> If you succeed (by some miracle) in creating a comp.lang.ansic or a
> comp.lang.isoc, who will post there, hmmm? Nobody, that's who. We'll all
> still post here, and the new group will fail. That's people for you, I'm
> afraid. We're stubborn like that.
Obviously it wouldn't work if people don't want it to work. I
personally don't care enough one way or another to do anything about
it. When I get sufficiently annoyed at some of the antics here I just
go away for a while.
Thomas M. Sommers
>
> Thomas M. Sommers <tm...@mail.ptd.net> wrote:
> > You can't even say `int i;', because the
> > variable `i' is not defined in the standard.
>
> yes you can; the standard allows you to define variables.
And functions, and to include other source files, yet some people here
see `#include <unistd.h>', or `write(...)', and say it isn't C.
> there are some things which cannot be done entirely in C, though, such as
> clearing the screen. in order to clear the screen in a C program, something
> must be done (such as calling a function that is implemented in machine
> language) which is not part of C.
The same applies to large chunks of the standard library, yet no one
says that fopen() is not C.
mike burrell
> mike burrell wrote:
>> Thomas M. Sommers <tm...@mail.ptd.net> wrote:
>> > You can't even say `int i;', because the
>> > variable `i' is not defined in the standard.
>>
>> yes you can; the standard allows you to define variables.
> And functions, and to include other source files, yet some people here
> see `#include <unistd.h>', or `write(...)', and say it isn't C.
indeed. if you can find a way to implement write() in C, i'd like to see
it.
for importantly, though, it's not that "write() isn't C" so much as "nobody
knows wtf write() actually does". the behaviour of the Unix write() is not
the only write() function ever written; the unistd.h found in Unix is not
the only unistd.h ever written. we're not mind readers, and we're not
omniscient to all C APIs ever written.
if someone were to post this:
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#define BL 50
void
bar(int fd, void *c)
{
if (write(fd, c, BL) == -1) {
fprintf(stderr, "errno: %d\n", errno);
exit(EXIT_FAILURE);
}
}
help!! why does write() fail and set errno to EFAULT?!@
there is no way in hell we can be expected to answer that. there is just no
way to expect us to know all common C APIs (such as the Posix API presumably
used in this case). if, however, they were to append:
my documentation says this:
ssize_t write(int fd, const void *buf, size_t count);
EFAULT buf is outside your accessible address space.
*then* we can be of some assistance. if there is stuff being done that is
not described by the standard (such as the meaning of EFAULT), then we
*must* be explicitly told about it if any help is to be expected.
>> there are some things which cannot be done entirely in C, though, such as
>> clearing the screen. in order to clear the screen in a C program, something
>> must be done (such as calling a function that is implemented in machine
>> language) which is not part of C.
> The same applies to large chunks of the standard library, yet no one
> says that fopen() is not C.
the difference is that fopen()'s behaviour is defined in the standard, so we
are all expected to know it. if people come in here asking questions about
outp() and vline() and we say they would be better off asking somewhere
else, it's not because we're mean: it's because we honestly have no idea
what the behaviour of those functions is unless we're told. i can guess:
vline() probably draws a vertical line, but what kinds of arguments does it
take? does it have side effects? what significance does the return value
have? no idea. the poster has to mention those things explicitly if he
wants help.
FWIW, if i were in charge of C, there would be no standard library, but
that's another story. i'm stuck with it :D
Thomas M. Sommers
>
> Thomas M. Sommers <tm...@mail.ptd.net> wrote:
> > mike burrell wrote:
> >> Thomas M. Sommers <tm...@mail.ptd.net> wrote:
> >> > You can't even say `int i;', because the
> >> > variable `i' is not defined in the standard.
> >>
> >> yes you can; the standard allows you to define variables.
>
> > And functions, and to include other source files, yet some people here
> > see `#include <unistd.h>', or `write(...)', and say it isn't C.
>
> indeed. if you can find a way to implement write() in C, i'd like to see
> it.
But the implementation shouldn't matter when you are just talking about
using the function. Nothing (as far as I know) stops functions in the
standard library from being implemented in assembler. Many will have to
be, or (more likely) will have to call functions such as write() that
are.
> for importantly, though, it's not that "write() isn't C" so much as "nobody
> knows wtf write() actually does". the behaviour of the Unix write() is not
> the only write() function ever written; the unistd.h found in Unix is not
> the only unistd.h ever written. we're not mind readers, and we're not
> omniscient to all C APIs ever written.
>
> if someone were to post this:
<code snipped>
> there is no way in hell we can be expected to answer that. there is just no
> way to expect us to know all common C APIs (such as the Posix API presumably
> used in this case). if, however, they were to append:
>
> my documentation says this:
>
> ssize_t write(int fd, const void *buf, size_t count);
>
> EFAULT buf is outside your accessible address space.
>
> *then* we can be of some assistance. if there is stuff being done that is
> not described by the standard (such as the meaning of EFAULT), then we
> *must* be explicitly told about it if any help is to be expected.
All true, but in all likelihood if someone were to give this extra
information, he would still be flamed here. Even though, while not
everyone might know the system being used, many probably do and could
answer the question.
> > The same applies to large chunks of the standard library, yet no one
> > says that fopen() is not C.
>
> the difference is that fopen()'s behaviour is defined in the standard, so we
> are all expected to know it.
Of course, there is no guarantee that the fopen() being called is
actually the standard fopen().
> if people come in here asking questions about
> outp() and vline() and we say they would be better off asking somewhere
> else, it's not because we're mean: it's because we honestly have no idea
> what the behaviour of those functions is unless we're told.
A person with an ungenerous soul might well get the impression that some
people here derive pleasure from being rude to newcomers.
> FWIW, if i were in charge of C, there would be no standard library, but
> that's another story. i'm stuck with it :D
Here I must disagree strongly. I remember trying to port a program from
Version 7 to VMS using Whitesmith's C. Even some of the most common
stdio functions were different. Since C is such a minimalist language
(a Good Thing), it needs a standard library.
Michael Powe
Hash: SHA1
>>>>> "petere" == petere <peliades...@sanderson.net.au.invalid> writes:
petere> There appear to be a disproportianate amount of posts that
petere> are only incidentally relevant to ANSI/ISO C (which is
petere> what comp.lang.c is dedicated to). Additionally, much
petere> platform specific information is proferred as ANSI/ISO C
petere> information.
petere> Further, redirecting the posters or pointing out errors
petere> seems to be causing an unnaturally hostile response. I'd
petere> attribute the vitriolic response to the (relative)
petere> anonymity of usenet.
petere> What makes this particularly annoying is that it occurs
petere> regularly.
petere> If the posts to clc can be used as a guide, then few
petere> programmers appear to appreciate what ANSI/ISO C consists
petere> of and the value of knowing this for portability reasons.
petere> Schildt, Lafore, Waite and Prata (the last three of "Waite
petere> Group" fame) have contributed to this by presenting
petere> x86/DOS- centric C as "C proper" (all of these authors
petere> books are beyond there first editions so they must me
petere> popular, someone is buying them). These books typically
petere> throw in a x86/DOS (or even worse an MS or Borland)
petere> feature without flagging it as such.
This is a completely inaccurate description of Prata's <C Primer Plus>
from the Waite Group. Have you actually even read it?
Let's look at an excerpt from the Preface to the 3rd edition.
[ ... ] what's new:
* coverage of the ANSI C library has been expanded, and Appendix F,
"The Standard ANSI C library," describes the entire library.
* Now that the ANSI C standard has been around for several years, this
edition devotes less space to the discussion of pre-ANSI C.
[ ... ]
* The book discusses many of the additions to the ANSI C standard
proposed by the C9X committee, some of which are already available
on many compilers.
As a person who actually has read this book (and therefore is
qualified to discuss what is included in it), I can tell you that
there are many discussions of platform-specific items, problems of
portability and warnings about using functions that are not ANSI C.
The first chapter also includes a long section describing differences
in the programming process on different platforms.
I find it somewhat ludicrous to see a writer accused of being
"DOS-centric" whose book <UNIX System V Bible> is on the recommended
list at ACCU. I know that in the larger picture of things, it is to
be expected that people will spout off about books about which they
know nothing. But I find it annoying when readers who need a useful,
well-written book are being turned away from that book by blind
prejudice.
mp,who programs almost entirely in linux and unix and has never typed
#include <conio.h> in a source file
PS There's something wrong with your email address. You should fix
that.
- --
"Well, obviously the best programming is LISP." -- Richard Stallman
Michael Powe Portland, Oregon USA
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Mailcrypt 3.5.5/GnuPG v1.0.1 http://www.gnupg.org
iD8DBQE5aMUz755rgEMD+T8RAq7ZAKCm/Vjw0WBYmAWkdmGaDefrBaq6IwCbBjlP
cwh0cZrk071qbTlsk1geT3s=
=GNjA
-----END PGP SIGNATURE-----
Geoff Field
Richard Heathfield <bin...@eton.powernet.co.uk> wrote:
> Geoff Field wrote:
> >
>
> <much snippage - just one point I wanted to answer>
>
> > 4) c.std.c is very much on the bottom of my list of ng's to read -
> > after c.l.c. I haven't read the charter of the ng for quite a
while.
> > Given the charter, I even wonder why the ng's still around - after
all,
> > the various versions of the C standards have been out for a while
now.
>
> Geoff, this is a thinko, isn't it?
>
> The C9899:1999 ISO Standard was only published this January, and
> ratified by ANSI even later, so I'd have thought this would be a very
> bad time for comp.std.c to close its doors. Many of us are still
coming
> to grips with the implications of the new Standard, a process which is
> slower than it perhaps need be due to the lack of C99 compilers. The
> comp.std.c newsgroup is going to loom large in our lives, I expect,
over
> the next year or three.
Fair enough, I suppose. OTOH, it seems a bit odd to give a document a
date of 1999 and then release it in 2000. Nonetheless, since the
release date is only about six months ago, there is still plenty of
scope for discussion.
Frankly, though, it's *not* likely to be a big part of _my_ life. I
work in embedded software, where functionality comes WAY before
portability. Also, in this industry (fire protection), documentation
comes WAY before coding. The language is almost irrelevant. Since my
current project will be coded in UML, C++ and PowerPC assembly, C
standards, I'm afraid to say, are the last thing on my mind. My C will
be basically the old ANSI standard where it's needed.
Geoff
--
Geoff Field of XXXXia
Professional geek, amateur stage-levelling gauge
Richard Heathfield
>
> In article <396575CB...@eton.powernet.co.uk>,
> Richard Heathfield <bin...@eton.powernet.co.uk> wrote:
> >
> > The C9899:1999 ISO Standard was only published this January, and
> > ratified by ANSI even later, so I'd have thought this would be a very
> > bad time for comp.std.c to close its doors. Many of us are still
> coming
> > to grips with the implications of the new Standard, a process which is
> > slower than it perhaps need be due to the lack of C99 compilers. The
> > comp.std.c newsgroup is going to loom large in our lives, I expect,
> over
> > the next year or three.
>
> Fair enough, I suppose. OTOH, it seems a bit odd to give a document a
> date of 1999 and then release it in 2000.
Yes, it does seem odd. It's easy enough to understand, however, when one
realises that the year reference (1999) relates to the date of
ratification by ISO (late 1999), *not* publication by ISO (Jan 2000) or
ratification by ANSI (spring 2000).
> Nonetheless, since the
> release date is only about six months ago, there is still plenty of
> scope for discussion.
A masterful understatement, sirrah. :-)
> Frankly, though, it's *not* likely to be a big part of _my_ life. I
> work in embedded software, where functionality comes WAY before
> portability. Also, in this industry (fire protection), documentation
> comes WAY before coding. The language is almost irrelevant. Since my
> current project will be coded in UML, C++ and PowerPC assembly, C
> standards, I'm afraid to say, are the last thing on my mind. My C will
> be basically the old ANSI standard where it's needed.
A perfectly reasonable point of view, but of course not one that is
particularly well-matched to comp.lang.c (which, no doubt, explains why
it's one of the last newsgroups you read). Don't worry - we aren't
offended! But we're not going to change what we talk about for your
benefit, and I'm sure you wouldn't expect us to.
--
Richard Heathfield
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
C FAQ: http://www.eskimo.com/~scs/C-faq/top.html
Geoff Field
Richard Heathfield <bin...@eton.powernet.co.uk> wrote:
> Geoff Field wrote:
> >
> > In article <396575CB...@eton.powernet.co.uk>,
> > Richard Heathfield <bin...@eton.powernet.co.uk> wrote:
> > Fair enough, I suppose. OTOH, it seems a bit odd to give a
document a
> > date of 1999 and then release it in 2000.
> Yes, it does seem odd. It's easy enough to understand, however, when
one
> realises that the year reference (1999) relates to the date of
> ratification by ISO (late 1999), *not* publication by ISO (Jan 2000)
or
> ratification by ANSI (spring 2000).
"spring?" Not in this hemisphere (unless it's going to happen in about
three months' time).
> > Nonetheless, since the
> > release date is only about six months ago, there is still plenty of
> > scope for discussion.
> A masterful understatement, sirrah. :-)
We're full of that down here in Oz - or full of something, anyway ;-)
[My dissertation on functionality vs portability, UML, C++, etc snipped]
> A perfectly reasonable point of view, but of course not one that is
> particularly well-matched to comp.lang.c (which, no doubt, explains
why
> it's one of the last newsgroups you read).
Absolutely. Nonetheless, I have found a detailed knowledge of portable
C very handy in the past, and still do. One project in particular
required communications code that would compile and run on any *nix
platform, as well as the embedded platform, and even on a Mac. This
sort of thing requires fairly strict adherence to the standard. A lot
of embedded projects are also tested on a host as well, because
embedded hardware often isn't available until late in the cycle.
> Don't worry - we aren't offended!
I'm glad of that. It's simply my POV at the moment.
> But we're not going to change what we talk about for your
> benefit, and I'm sure you wouldn't expect us to.
Quite right. If any one person (or small group of people) tried to
dictate what was going on in a global newsgroup, there would be all
sorts of ruckus!
Dann Corbit
[snip]
> Quite right. If any one person (or small group of people) tried to
> dictate what was going on in a global newsgroup, there would be all
> sorts of ruckus!
Good thing that will never happen. After all, this is USENET where it's
never September and where logic and reason reign supreme.
petere
>C Primer Plus
>from the Waite Group. Have you actually even read it?
Yes. I have read an earlier edition "Waite Group's C Primer Plus"
-- it may have been the first -- cover to cover. I have since
thrown it out. I recall it being a book with a flashy cover and
lots of void main(void) stuff and no references to other OS.
The other books I mentoned I still have and have read.
>Let's look at an excerpt from the Preface to the 3rd edition.
>* coverage of the ANSI C library has been expanded, and
Appendix F,
> "The Standard ANSI C library," describes the entire library.
>* Now that the ANSI C standard has been around for several
years, this
> edition devotes less space to the discussion of pre-ANSI C.
(Pretty late coming to the party and now C has been "isofied".)
Yeah, it may have changed since the edition I read. So what.
The earlier editions were read by someone, were they not ?
Comment on them is still relevant.
>As a person who actually has read this book (and therefore is
>qualified to discuss what is included in it), I can tell you
>that
>there are many discussions of platform-specific items, problems
>of
>portability and warnings about using functions that are not
>ANSI C.
>The first chapter also includes a long section describing
differences
>in the programming process on different platforms.
I have read many Waite Group books and I recently purchased two
of Mr Lafore's and I can say that most of the Sams books suck.
(Incidentally, I own most of the C tutorials and references on
the market, I'm something of a collector :-) so I think I am
well placed to comparative critiques of C books.)
>I find it somewhat ludicrous to see a writer accused of being
>"DOS-centric" whose book <UNIX System V Bible> is on the
recommended
>list at ACCU.
I didn't say that the writer is DOS-centric, I was refering to
the bulk of the Sams and Schildt books. All of the older Sams
and Schildt books taught C in a DOS centric manner, the current
crop teach it with a Win32 orientation.
>I know that in the larger picture of things, it is to
>be expected that people will spout off about books about which
>they
>know nothing.
You know nothing about what I have read so the assumptions are
all yours.
>But I find it annoying when readers who need a useful,
>well-written book are being turned away from that book by blind
>prejudice.
There are many better books than Prata's on C and Unix. From
first hand experience I can tell you that most of the Sams
books: Unix, TCP, HTML, C and even application packages such as
Photoshop all suck big-time. They are all voluminous but empty
at the same time.
Incidentally, I own Stephen Prata's "C++ Primer Plus" 3rd ed --
and have read it -- and it is pretty ordinary.
>mp,who programs almost entirely in linux and unix and has never
typed
> #include <conio.h> in a source file
Most of my work is Unix systems programming and none of the
Waite Groups books are respected by Unix programmers. They are
generally regarded as being in the same league as the
Sams "Unleashed" series (eg. Unix Unleashed) or the 7-days
series (eg. Teach Yourself C in 7 days) and so on. I have yet
to see a reference to any Sams book in authoritative Unix texts
(I can give you a bibliography if you like).
Further, Stephen Prata is an unknown in the Unix world, his Unix
books are irrelevant. That you would quote him as a Unix
authority demonstates how little you know. SPs two Unix books
cover elementary Unix topics in a pretty poor way. I have yet
to read a paper or book by the C and/or Unix gurus W. Richard
Stevens (RIP), P.J Plauger, Harbisson/Steele, Holub that refered
me to a book by Prata.
>mp
isn't it SP, Stephen Prata that we are talking about.
>PS There's something wrong with your email address. You should
fix
>that.
There's something wrong with your bookshelf. You should fix
that. I don't think you have read a good book on C or Unix.
-----------------------------------------------------------
Got questions? Get answers over the phone at Keen.com.
Up to 100 minutes free!
http://www.keen.com
Brian Evan Blank
[in response to the > > of Michael Powe]
> >This is a completely inaccurate description of Prata's
> >C Primer Plus
> >from the Waite Group. Have you actually even read it?
>
> Yes. I have read an earlier edition "Waite Group's C Primer Plus"
> -- it may have been the first -- cover to cover. I have since
> thrown it out. I recall it being a book with a flashy cover and
> lots of void main(void) stuff and no references to other OS.
Flashy cover? Are you judging the book by its cover? Let us
turn to the interior. You seem to be basing your criticism of
"The Waite Group's C Primer Plus" on its first edition. I
say that not because of what you conjecture but through a
process of elimination. There was an earlier 1984 book that
did not have "The Waite Group" in its title. The second edition
added a "New" to its title. The third edition does not fit your
description at all, so the first edition is all that remains.
It came out in 1990 when the ISO standard was new and non-
compliant compilers were still commonplace. However, there is
no need to excuse any sins Prata may have committed in 1990.
That work is two editions and ten years in the past. It is
unfair to level a charge now based on your recollections of
ancient history. And when I see your statement that you read
the book cover to cover, I am reminded of the person who
complains, item by item, about the unpalatable offerings of
a restaurant and then concludes the critique with his final
condemnation: "And the portions are so small!"
> >Let's look at an excerpt from the Preface to the 3rd edition.
> >* coverage of the ANSI C library has been expanded, and
> Appendix F,
> > "The Standard ANSI C library," describes the entire library.
> >* Now that the ANSI C standard has been around for several
> years, this
> > edition devotes less space to the discussion of pre-ANSI C.
>
> (Pretty late coming to the party and now C has been "isofied".)
An author must consider his audience, if only because his
editor will insist on it. When the second edition of Prata's
book came out K&R C was still common. True, Prata does
generally refer to ANSI C and not ISO C but he also mentions
that C has been "isofied."
> Yeah, it may have changed since the edition I read. So what.
> The earlier editions were read by someone, were they not ?
> Comment on them is still relevant.
Relevant as long as you make clear that you are referring
to an outdated edition. The 1984 edition was also read by
someone. Is it relevant to comment on its lack of coverage
of the standard without also mentioning its vintage?
> >As a person who actually has read this book (and therefore is
> >qualified to discuss what is included in it), I can tell you
> >that
> >there are many discussions of platform-specific items, problems
> >of
> >portability and warnings about using functions that are not
> >ANSI C.
> >The first chapter also includes a long section describing
> differences
> >in the programming process on different platforms.
Yes. At appropriate places (command line arguments etc.) there
is platform-specific discussion that includes UNIX, LINUX, DOS,
Windows, and Macintosh. There is not that much else that is
platform-specific and what there is is not "DOS-centric." BTW,
the words to which Michael Powe refers have been dropped. Here
they are:
petere> Schildt, Lafore, Waite and Prata (the last three of "Waite
petere> Group" fame) have contributed to this by presenting
petere> x86/DOS- centric C as "C proper" (all of these authors
petere> books are beyond there first editions so they must me
petere> popular, someone is buying them). These books typically
petere> throw in a x86/DOS (or even worse an MS or Borland)
petere> feature without flagging it as such.
> I have read many Waite Group books and I recently purchased two
> of Mr Lafore's and I can say that most of the Sams books suck.
So not only do you judge a book by its cover, you also judge it
by the other books put out by the same publisher. If Sams were
to publish a book by noted experts on ISO C such as Lawrence
Kirby, Stephan Wilms, et al., would you denigrate it because
"most of the Sams books suck"?
> (Incidentally, I own most of the C tutorials and references on
> the market, I'm something of a collector :-) so I think I am
> well placed to comparative critiques of C books.)
But you are not well placed to criticize the current edition
of Prata's book. Nor do you qualify your statements to indicate
that they might not apply to the edition of Prata's C book
that is now available.
> >I find it somewhat ludicrous to see a writer accused of being
> >"DOS-centric" whose book <UNIX System V Bible> is on the
> recommended
> >list at ACCU.
> I didn't say that the writer is DOS-centric, I was refering to
> the bulk of the Sams and Schildt books. All of the older Sams
> and Schildt books taught C in a DOS centric manner, the current
> crop teach it with a Win32 orientation.
For your convenience, I have quoted above what you actually
said. You should have included it at the start of your reply
to Michael Powe so readers would have known what his statement
"This is a completely inaccurate description ... " refers to.
And you are wrong again. The current crop of Sams books do
not necessarily teach C with a Win32 orientation. Prata's book
is a counterexample. There may be another.
[Tiresome squabling munched]
> There are many better books than Prata's on C and Unix.
IYHO. I usually recommend the books by Deitel-Deitel,
Kelley-Pohl, and King. I have not (yet) evaluated Prata
as carefully as those and therefore I do not recommend it
(yet). OTOH, from what I have read so far, I think that
it _might_ be the best of the bunch for many newcomers
to C. If you know many better books, then feel free to
list them the next time the question arises here (as
it undoubtedly will in a few days). Remember that many
of the books that are better for some readers are not
better for all newcomers - especially those with
backgrounds in business or science who will have a
serious need for programming skills but who will not
be professional programmers.
[more munched]
> Most of my work is Unix systems programming and none of the
> Waite Groups books are respected by Unix programmers. They are
> generally regarded as being in the same league as the
> Sams "Unleashed" series (eg. Unix Unleashed) or the 7-days
> series (eg. Teach Yourself C in 7 days) and so on. I have yet
> to see a reference to any Sams book in authoritative Unix texts
> (I can give you a bibliography if you like).
More guilt by association. What will you say if Sams ever
publishes a book titled "C Unleashed" that is authored by
known experts on standard C?
(To reach me, replace my last name by my first.)
petere
No. As I said I read it cover-to-cover.
>Let us
>turn to the interior. You seem to be basing your criticism of
>"The Waite Group's C Primer Plus" on its first edition.
I think you are right about the edition. Unfortunately, I did
(literally) throw it away so I am unable to confirm this.
>And when I see your statement that you read
>the book cover to cover, I am reminded of the person who
>complains, item by item, about the unpalatable offerings of
>a restaurant and then concludes the critique with his final
>condemnation: "And the portions are so small!"
This is an unfair statement. If I said I had read half of it
you would have complained that I judged a book without having
read it. Don't mistake cute analogies with reasoned argument.
>Relevant as long as you make clear that you are referring
>to an outdated edition. The 1984 edition was also read by
>someone. Is it relevant to comment on its lack of coverage
>of the standard without also mentioning its vintage?
You concede that the original edition was read by someone. Is
it not reasonable for me to conjecture that some programmers
are "programming under the influence" of that book. If so, then
my point still stands.
>So not only do you judge a book by its cover, you also judge it
>by the other books put out by the same publisher.
No I don't judge a book by its cover. As I said, I read it in
in its entirety. If you cannot understand this then there is no
point in continuing this discussion.
I think it is fair and reasonable to judge publishers in broad
terms. All of the books put out by a publisher on a given topic
are subjected to the same stylistic template. O'Reilly's books -
- for example -- are consistently excellent. I have read
several of them and they are exemplary. The same goes for
Addison-Wesley.
>If Sams were
>to publish a book by noted experts on ISO C such as Lawrence
>Kirby, Stephan Wilms, et al., would you denigrate it because
>"most of the Sams books suck"?
If this did occur then the statement "most of the Sams books
suck" would still be true, hence I would assert it.
>But you are not well placed to criticize the current edition
>of Prata's book.
Perhaps not. But I was not offering a review of the current
edition.
>Nor do you qualify your statements to indicate
>that they might not apply to the edition of Prata's C book
>that is now available.
No, I didn't specify any editions because the subject of my
posting was topicality re clc, bad books were only peripheral to
the matter.
>And you are wrong again. The current crop of Sams books do
>not necessarily teach C with a Win32 orientation. Prata's book
>is a counterexample. There may be another.
I wasn't referring specifically to Sams' books. I am referring
broadly to Sam's, QUE, Wrox and the like. Most of these
publishers books do teach OS and even compiler specific C.
You seem to have trouble differentiating between "all", "some"
and "most". I did say "most" not "all".
>IYHO. I usually recommend the books by Deitel-Deitel,
>Kelley-Pohl, and King.
I like those too.
>If you know many better books, then feel free to
>list them the next time the question arises here (as
>it undoubtedly will in a few days).
I'm not interested in engaging in a discussion about C books. I
threw in the comment about those voluminous books put out by
Sams, QUE, Wrox etc as a likely contributor to the blurring of
the notion of generic ie. ANSI/ISO C as opposed to some vendors
implementation and proprietray extensions to the language.
I think you and Mr Powe have demonstrated your small mindedness
by not contributing anything to the central concern of the
thread but by jumping on the statement about Stephen Prata's
book. BTW I still think the first edition sucked.
>More guilt by association. What will you say if Sams ever
>publishes a book titled "C Unleashed" that is authored by
>known experts on standard C?
I will say what I originally said, namely that most of the Sams
titles on C and Unix suck.
>(To reach me, replace my last name by my first.)
I have no interest in reaching you.
Brian Evan Blank
[munched]
[The > > part that follows is from me.
Why should I have to clarify this? ]
> >And you are wrong again. The current crop of Sams books do
> >not necessarily teach C with a Win32 orientation. Prata's book
> >is a counterexample. There may be another.
>
> I wasn't referring specifically to Sams' books. I am referring
> broadly to Sam's, QUE, Wrox and the like. Most of these
> publishers books do teach OS and even compiler specific C.
>
> You seem to have trouble differentiating between "all", "some"
> and "most". I did say "most" not "all".
"I understand very well the English, as she is spoken."
I would like to suggest that you reconsider your modus
operandi. Michael Powe responded to a direct quote of
yours. When you responded to his reply, you snipped your
original comment and claimed to have said something else.
I objected to that practice and suggested that you refrain
from it. Now you have applied your debating style to me.
Here is your string that I was commenting on and which
you snipped in your reply to my response:
petere> "All of the older Sams and Schildt books taught C in
a DOS centric manner, the current crop teach it with
a Win32 orientation."
You say that I "seem to have trouble differentiating between
`all', `some', and `most'. I do _not_ see a "some" or "most"
in your assertion. You say that you "did say `most' not `all'."
I _do_ see an "all" at the start of your run-on sentence. It may
or may not be intended as the implicit subject of "teach", but
even if it is not, "current crop" is not otherwise modified. Only
one counterexample is necessary to invalidate your statement
as written. (I suspect that another counterexample will be
released any day now but as a counterexample it will be
superfluous.) Now go back and reread my usage of the
adverb "necessarily." It was put there to signify that I
was not responding to an assertion qualified by "some" or
"most."
My point here is this: had you not snipped your original quote,
everyone would have easily been able to see for themselves
what you did or did not originally write. I am content to let
you have the last word on the discussion of Prata, Sams, and
the Waite Group and concede that my restaurant critic wise-
cracking was unfair. But I do think that you should adjust your
manner of posting follow-ups.
(To reach me, replace my middle name with my first.)