lastest C standard version

[Keith Thompson]

I found on the web (don't remember where...) a pdf document whose
header (on each of the 500+ pages) says |left: (c)ISO/IEC right:
ISO/IEC 9899:1999(E)|
what is this document and is it an illegal copy of the final version
of the last standard? Or the (E) means a prior-to-release version?

That's the same page header I see on my legal copy of the C99
standard. If you found it on the web, it's probably an illegal copy.

In that case, what is the most up-to-date document : the WG14 C99
Rationale or the pdf I found?

The Rationale is a separate document, not covered by the license
restrictions covering the standard.

 

[Antoine Leca]

En (e-mail address removed), Thomas L. va escriure:

I found on the web (don't remember where...) a pdf document whose
header (on each of the 500+ pages) says |left: (c)ISO/IEC right:
ISO/IEC 9899:1999(E)|

Or the (E) means a prior-to-release version?

The "(E)" there means it is the English version of a document. A French
version would have a (F) in that place.
It has nothing to do with the status of a document (i.e., both an official
standard and a committee draft would have a similar marking.)


Hope it helps,

Antoine

 

[Irrwahn Grausewitz]

En (e-mail address removed), Thomas L. va escriure:

I found on the web (don't remember where...) a pdf document whose
header (on each of the 500+ pages) says |left: (c)ISO/IEC right:
ISO/IEC 9899:1999(E)|

Or the (E) means a prior-to-release version?

The "(E)" there means it is the English version of a document. A French
version would have a (F) in that place. [...]

Funny, I could've sworn the "(E)" stands for "electronic", since it
only appears in the electronic (.pdf) version of the standard.

Regards

 

[Douglas A. Gwyn]

Funny, I could've sworn the "(E)" stands for "electronic",

No, it does mean English.

 

[Thomas L.]

En (e-mail address removed), Thomas L. va escriure:

I found on the web (don't remember where...) a pdf document whose
header (on each of the 500+ pages) says |left: (c)ISO/IEC right:
ISO/IEC 9899:1999(E)|

Or the (E) means a prior-to-release version?

The "(E)" there means it is the English version of a document. A French
version would have a (F) in that place. [...]

Funny, I could've sworn the "(E)" stands for "electronic", since it
only appears in the electronic (.pdf) version of the standard.

Regards

In the end, I have no way of saying if it is a final (although in
English AND Electronic) version of the ISO/IEC 9899:1999 standard. It
has 538 pages.
Thanks anyway.

 

[lawrence.jones]

Funny, I could've sworn the "(E)" stands for "electronic", since it
only appears in the electronic (.pdf) version of the standard.

You're wrong on both counts: it stands for English and it appears in the
paper version of the standard, too.

-Larry Jones

They say winning isn't everything, and I've decided
to take their word for it. -- Calvin

 

[Keith Thompson]

En (e-mail address removed), Thomas L. va escriure:
I found on the web (don't remember where...) a pdf document whose
header (on each of the 500+ pages) says |left: (c)ISO/IEC right:
ISO/IEC 9899:1999(E)|

Or the (E) means a prior-to-release version?

The "(E)" there means it is the English version of a document. A French
version would have a (F) in that place. [...]

Funny, I could've sworn the "(E)" stands for "electronic", since it
only appears in the electronic (.pdf) version of the standard.

Regards

In the end, I have no way of saying if it is a final (although in
English AND Electronic) version of the ISO/IEC 9899:1999 standard. It
has 538 pages.
Thanks anyway.

My (legal) PDF copy of the C99 standard is 1411606 bytes; the md5sum
is ff8dc9d8f109111a91a70cc29fb16168.

 

[Paul Eggert]

My (legal) PDF copy of the C99 standard is 1411606 bytes; the md5sum
is ff8dc9d8f109111a91a70cc29fb16168.

That's odd. My authorized copy is 1844334 bytes long. Here is the
md5sum and file name.

0b5cc55c3239c3d4e0c71cba784b2ba5 ISO_IEC_9899;1999(E)-Character_PDF_document.pdf

I got my copy from the ISO. Acrobat Reader gives the following
properties for this file (which I had to transcribe by hand, sigh):

Title: ISO/IEC 9899:1999
Subject: Programming Languages - C
Author: JTC 1/SC 22/WG 14
Binding: Left Edge
Producer: Acrobat Distiller Command 3.0 for HP-UX A.09.01 and later (HPPA)
Created: Thu Apr 22 17:51:25 1999
Modified: Tue Nov 30 16:32:20 1999
File Size: (Bytes): 1844334
Security: None
PDF Version: 1.2 (Acrobat 3.x)
Page SIze: 8.3 in x 11.7 in
Number of Pages: 554
Tagged PDF: No
Optimized: Yes

I hope the only real difference is that my copy is for A4 paper, and
probably your copy is for US letter. Still, that difference in size
is worrisome. (Are you sure your copy is complete?

 

[Keith Thompson]

That's odd. My authorized copy is 1844334 bytes long. Here is the
md5sum and file name.

0b5cc55c3239c3d4e0c71cba784b2ba5 ISO_IEC_9899;1999(E)-Character_PDF_document.pdf

I got my copy from the ISO. Acrobat Reader gives the following
properties for this file (which I had to transcribe by hand, sigh):

Title: ISO/IEC 9899:1999
Subject: Programming Languages - C
Author: JTC 1/SC 22/WG 14
Binding: Left Edge
Producer: Acrobat Distiller Command 3.0 for HP-UX A.09.01 and later (HPPA)
Created: Thu Apr 22 17:51:25 1999
Modified: Tue Nov 30 16:32:20 1999
File Size: (Bytes): 1844334
Security: None
PDF Version: 1.2 (Acrobat 3.x)
Page SIze: 8.3 in x 11.7 in
Number of Pages: 554
Tagged PDF: No
Optimized: Yes

I hope the only real difference is that my copy is for A4 paper, and
probably your copy is for US letter. Still, that difference in size
is worrisome. (Are you sure your copy is complete?

My copy is also A4 (I've never printed it, so that's not a problem,
even though A4 paper is almost nonexistent around here). I got it
from ANSI; there's some ANSI boilerplate text on the first page, which
I presume your copy doesn't have.

The properties on mine are almost identical, except for:

Modified: 2000-07-17 14:07:57
File Size: 1.35 MB (1,411,606 Bytes)
Security: 40-bit RC4 (Acrobat 3.x, 4.x)
PDF Version: 1.3 (Acrobat 4.x)
Page Size: 8.26 in x 11.69 in

The number of pages is the same. I'm using Acrobat Reader 5.0 under
Windows XP.

I'm not sure what the 40-bit RC4 is about. I never asks me for a
password when I open the file. (I may have had to do something when I
first downloaded it; it's been a long time, and I've forgotten the
details.)

Does your copy have bookmarks?

 

[Jonathan Leffler]

My copy is also A4 (I've never printed it, so that's not a problem,
even though A4 paper is almost nonexistent around here). I got it
from ANSI; there's some ANSI boilerplate text on the first page, which
I presume your copy doesn't have.

The properties on mine are almost identical, except for:

Modified: 2000-07-17 14:07:57
File Size: 1.35 MB (1,411,606 Bytes)
Security: 40-bit RC4 (Acrobat 3.x, 4.x)
PDF Version: 1.3 (Acrobat 4.x)
Page Size: 8.26 in x 11.69 in

The number of pages is the same. I'm using Acrobat Reader 5.0 under
Windows XP.

I'm not sure what the 40-bit RC4 is about. I never asks me for a
password when I open the file. (I may have had to do something when I
first downloaded it; it's been a long time, and I've forgotten the
details.)

Does your copy have bookmarks?

Aye, well, they've been busy. My (legitimate) copy has a different
modification time (same date as Keith, but off a few hours), a
different size, and does have bookmarks.

Size: 1412026
MD5 Checksum: bbba3981291c7000b1f6ffc8030d66a1
Modified: 7/17/2000 11:07:57 AM

(Precisely three hours difference? I'm on US/Pacific time; is Keith
on US/Eastern time?) But still a different size...weird!

Clearly, the PDF files change a lot more than anyone would have
expected. In fact, I'm left wondering if each separate copy
downloaded isn't unique somehow... No concrete evidence - just that
everyone seems to have a different files.

 

[Paul Eggert]

Does your copy have bookmarks?

Yes.

Possibly the difference in size is because ANSI stripped out the
embedded fonts -- they may not have been as worried or careful about
the exact appearance of the standard on your screen.

I'm not surprised that each PDF copy would differ: this is standard
practice these days, as publishers put in watermarks. It does raise
the issue about how one knows that one's copy is valid, but I guess
the standards organizations don't care all that much about that issue.

But hundreds of kilobytes' worth of differences? Wow.

I'm not sure what the 40-bit RC4 is about. I never asks me for a
password when I open the file.

The password may be protecting just some functions: for example,
it could be preventing you from printing or selecting text.

My copy from ISO has no encryption and no restrictions imposed by
Acrobat Reader. For example, I can select the entire PDF document and
copy the result into an Emacs temporary text buffer for viewing; this
consumes 985607 bytes in the Emacs buffer.

there's some ANSI boilerplate text on the first page, which
I presume your copy doesn't have.

Correct. My page 1 looks like this:


INTERNATIONAL ISO/IEC
STANDARD 9899

Second edition
1999-12-01

=============================================

Programming languages --- C

Langages de programmation --- C


=============================================

ISO IEC Reference number
logo logo ISO/IEC 9899:1999(E)
(C) ISO/IEC 1999


Page 2 has a more-detailed copyright notice at the bottom of the page,
and has the ISO's Geneva address and says "Printed in Switzerland".
Above that it has a PDF disclaimer that says it may contain embedded
typefaces that are licensed from Adobe and 3 lines of legal mumbo
jumbo that say I have to obey Adobe's font licenses and it's not ISO's
fault if I don't. (Which is fine with me.)

 

[Richard Kettlewell]

Clearly, the PDF files change a lot more than anyone would have
expected. In fact, I'm left wondering if each separate copy
downloaded isn't unique somehow... No concrete evidence - just that
everyone seems to have a different files.

My copy has the same size and MD5 hash as Paul's.

 

[Brian Inglis]

Yes.

Possibly the difference in size is because ANSI stripped out the
embedded fonts -- they may not have been as worried or careful about
the exact appearance of the standard on your screen.

I'm not surprised that each PDF copy would differ: this is standard
practice these days, as publishers put in watermarks. It does raise
the issue about how one knows that one's copy is valid, but I guess
the standards organizations don't care all that much about that issue.

But hundreds of kilobytes' worth of differences? Wow.



The password may be protecting just some functions: for example,
it could be preventing you from printing or selecting text.

My copy from ISO has no encryption and no restrictions imposed by
Acrobat Reader. For example, I can select the entire PDF document and
copy the result into an Emacs temporary text buffer for viewing; this
consumes 985607 bytes in the Emacs buffer.



Correct. My page 1 looks like this:


INTERNATIONAL ISO/IEC
STANDARD 9899

Second edition
1999-12-01

=============================================

Programming languages --- C

Langages de programmation --- C


=============================================

ISO IEC Reference number
logo logo ISO/IEC 9899:1999(E)
(C) ISO/IEC 1999


Page 2 has a more-detailed copyright notice at the bottom of the page,
and has the ISO's Geneva address and says "Printed in Switzerland".
Above that it has a PDF disclaimer that says it may contain embedded
typefaces that are licensed from Adobe and 3 lines of legal mumbo
jumbo that say I have to obey Adobe's font licenses and it's not ISO's
fault if I don't. (Which is fine with me.)

My latest ANSI copy contains all the above and the embedded fonts,
security only prevents modification with a master password, and the
following properties between Binding and Page Size:

Creator: Not Available
Producer: PDF PT 2.33 (pdf-tools.com)
Modified: 2004-03-15 21:11:03
File Size: 1.38 MB (1,443,802 Bytes)
Security: 128-bit RC4 (Acrobat 5.0)
PDF Version: 1.4 (Acrobat 5.x)

plus a watermark in the bottom margin of each page:

"Licensed to Systematic Software/Brian W. Inglis
ANSI Store order #X124089 Downloaded: 3/15/2004 9:11:01 PM ET
Single user license only. Copying and networking prohibited."

Perhaps the independent commercial PDF Tools products provide better
encryption and compression than Acrobat Distiller?

 

[Paul Eggert]

Perhaps the independent commercial PDF Tools products provide better
encryption and compression than Acrobat Distiller?

Compression sounds plausible, but I'd expect to see a difference more
with images, not a text document, because images can be tweaked more
for better compression/worse quality. So I'm still surprised to see
the size differences.

So far we have:

bytes modified pub. notes
1,844,334 1999-11-30 ISO (produced by Acrobat Distiller Command 3.0)
1,411,606 2000-07-17 ANSI (Keith Thompson's, produced by Acrobat Distiller?)
1,412,026 2000-07-17 ANSI (Jonathan Leffler's copy, unknown producer)
1,443,802 2004-03-15 ANSI (Brian Inglis's copy, produced by PDF PT)

Both ISO copies are the same (that's a relief! it makes it quite
implausible that my ISO copy is corrupted), but the ANSI copies all
differ, and they are quite a bit shorter than the ISO copy.

It still looks like the ANSI PDF version left something out, but what?
They have the same number of pages....

Perhaps ISO disabled some forms of compression? For example, they may
not have used LZW compression because of the Unisys patent. Is there
an easy way to determine what forms of compression were actually used
in a PDF file?

 

[Marcel van Kervinck]

My (legal) PDF copy of the C99 standard is 1411606 bytes; the md5sum
is ff8dc9d8f109111a91a70cc29fb16168.

Same here:

MD5 (ANSI+ISO+IEC+9899-1999.pdf) = ff8dc9d8f109111a91a70cc29fb16168

Which makes me confident there is no
watermark in the document.

Marcel
-- _ _
_| |_|_|
|_ |_ (e-mail address removed)
|_| Marcel van Kervinck

 

[David Hopwood]

No, my ANSI copy is also produced by Acrobat Distiller Command 3.0.

Compression sounds plausible, but I'd expect to see a difference more
with images, not a text document, because images can be tweaked more
for better compression/worse quality. So I'm still surprised to see
the size differences.

So far we have:

bytes modified downloaded pub. notes
1,844,334 1999-11-30 ? ISO produced by Acrobat Distiller Command 3.0
1,411,606 2000-07-17 ? ANSI Keith Thompson's, Acrobat Distiller?
1,412,026 2000-07-17 ? ANSI Jonathan Leffler's, unknown producer
1,443,802 2004-03-15 ? ANSI Brian Inglis's, produced by PDF PT
1,410,200 2002-06-20 2002-12-28 ANSI David Hopwood's, Distiller Cmd 3.0 554pp

Both ISO copies are the same (that's a relief! it makes it quite
implausible that my ISO copy is corrupted), but the ANSI copies all
differ, and they are quite a bit shorter than the ISO copy.

It still looks like the ANSI PDF version left something out, but what?

Maybe embedded fonts?

They have the same number of pages....

Perhaps ISO disabled some forms of compression? For example, they may
not have used LZW compression because of the Unisys patent. Is there
an easy way to determine what forms of compression were actually used
in a PDF file?

Yes, open the file in a text editor that doesn't mind binary data (not
notepad) and scroll down to the first non-text section. Before it will
be something like:

<< /S 11004 /O 14737 /Filter /FlateDecode /Length 5117 0 R >>

/FlateDecode means DEFLATE compression. Also check the PDF version (first
line of the file). Mine is %PDF-1.4; later versions may produce shorter
files.

David Hopwood <[email protected]>

 

[Paul Eggert]

Yes, open the file in a text editor that doesn't mind binary data (not
notepad) and scroll down to the first non-text section. Before it will
be something like:

<< /S 11004 /O 14737 /Filter /FlateDecode /Length 5117 0 R >>

My (ISO) copy says this:

<< /S 10997 /O 14719 /Filter /FlateDecode /Length 5107 0 R >>

which apparently is the same compression technology as yours.

Also check the PDF version (first line of the file). Mine is
%PDF-1.4; later versions may produce shorter files.

My copy says %PDF-1.2. So that might explain things. But is it
reasonable that 1.4 produces files that are so much shorter, even if
the same compression is used?

 

[David Holland]

> Two more points. First, be sure your suggestion is
> for something that is significantly useful. There
> was a long list of features considered but not
> adopted for the 1989 C standard, and it would be
> wise to review those before bringing up the same
> old proposals all over again.

Is this material posted anywhere?

 

[Douglas A. Gwyn]

Is this material posted anywhere?

I think it was printed near the front of the 1989 C standard.

 

[lawrence.jones]

I think it was printed near the front of the 1989 C standard.

I thought it was in the Rationale, but it doesn't seem to be in either
of those places.

-Larry Jones

When you're as far ahead of the class as I am, it doesn't take much time.
-- Calvin