ISC Newsgroups - Unmatched )
John F. Morse
ftp://ftp.isc.org/usenet/CONFIG/newsgroups.bz2
When I ran the docheckgroups on it, I received four egrep errors:
news@optima5:/var/lib/news$ cat newsgroups | docheckgroups
egrep: Unmatched ) or \)
egrep: Unmatched ) or \)
egrep: Unmatched ) or \)
egrep: Unmatched ) or \)
This is from the bz2 archive. I don't know if the error exists in the
plain textfile or the gz archive.
--
John
No Microsoft, Apple, AT&T, Novell, Trend Micro, nor Ford products were used in the preparation or transmission of this message.
The EULA sounds like it was written by a team of lawyers who want to tell me what I can't do. The GPL sounds like it was written by a human being, who wants me to know what I can do.
Russ Allbery
> When I ran the docheckgroups on it, I received four egrep errors:
>
> news@optima5:/var/lib/news$ cat newsgroups | docheckgroups
> egrep: Unmatched ) or \)
> egrep: Unmatched ) or \)
> egrep: Unmatched ) or \)
> egrep: Unmatched ) or \)
>
> This is from the bz2 archive. I don't know if the error exists in the
> plain textfile or the gz archive.
That looks like a bug in docheckgroups where it's not escaping some
pattern properly.
--
Russ Allbery (r...@stanford.edu) <http://www.eyrie.org/~eagle/>
Please post questions rather than mailing me directly.
<http://www.eyrie.org/~eagle/faqs/questions.html> explains why.
John F. Morse
> "John F. Morse" <xanad...@example.invalid> writes:
>
>> When I ran the docheckgroups on it, I received four egrep errors:
>>
>> news@optima5:/var/lib/news$ cat newsgroups | docheckgroups
>> egrep: Unmatched ) or \)
>> egrep: Unmatched ) or \)
>> egrep: Unmatched ) or \)
>> egrep: Unmatched ) or \)
>>
>> This is from the bz2 archive. I don't know if the error exists in the
>> plain textfile or the gz archive.
>>
>
> That looks like a bug in docheckgroups where it's not escaping some
> pattern properly.
>
Although more than four, these may be what tripped the failures:
alt.bananen.home Een groep die NIET over B.A.N.A.N.E.N. gaat :)
alt.dur.fortran-is-evil For those of us who are forced to use it :)
alt.marianne.hier.zijn.je.mapjuhs De mapjes van Marianne. Zo raakt ze
ze tenminste niet meer kwijt. :)
alt.pl.comp.os.hacking.pomidor Pomidor :)
alt.pl.mszczonow OffTopic o Mszczonowie :)
alt.pl.tvn.bigbrother Alternatywa dla tvn.bigbrother. TWA mile
widziane. :)
alt.should.be.doing.labs Well, you should be, probably :)
chile.rec.musica.metal Tertulia sobre el Rock Pesado >:) .
hactar.test Post here in case you don't get enough email already :)
sk.sux.microsoft Diskusia o firme Microsoft jej programoch a preco su
zle :)
swnet.jobs help wanted :-)
ufra.flame ... der taegliche User-Kleinkrieg :-)
viwa.sobutilnik Have a drink? %).
Julien ÉLIE
> When I ran the docheckgroups on it, I received four egrep errors:
>
> news@optima5:/var/lib/news$ cat newsgroups | docheckgroups
> egrep: Unmatched ) or \)
> egrep: Unmatched ) or \)
> egrep: Unmatched ) or \)
> egrep: Unmatched ) or \)
I had already seen problems like that on a few systems (caused by
different grep or even sed implementations). I do not manage to
reproduce your problem with the docheckgroups shipped with INN 2.4.3.
Note that I have converted docheckgroups to use awk instead of egrep
since that time.
> This is from the bz2 archive. I don't know if the error exists in the
> plain textfile or the gz archive.
And does the error still exist with this docheckgroups:
http://inn.eyrie.org/trac/export/8433/trunk/control/docheckgroups.in
Just download it, chmod 755 and put something like
. /usr/lib/news/innshellvars
at the second line of the file (the line is the same as what there is
in your current docheckgroups script).
Then, run with "newsgroups" and "docheckgroups.in" in the same directory:
cat newsgroups | ./docheckgroups.in
Is it better or not?
--
Julien ÉLIE
« Veni, uidi, et je n'en crois pas mes yeux ! » (César)
John F. Morse
Hi Julien.
Sorry this test and reply took so long. I've had thunderstorms all
night, and the sump pump float switch is acting up. Only got two hours
sleep before the alarm sounded at 0530 (1030Z), and I had been up 24
hours before that!
So, here's the output. It's nearly noon, and I'm going to bed. ;-)
news@optima5:~$ cat newsgroups | ./docheckgroups.in
awk: line 1: regular expression /^aaa[. ]| ... exceeds
implementation size limit
awk: line 1: regular expression /^aaa[. ]| ... exceeds
implementation size limit
awk: line 1: regular expression /^aaa[. ]| ... exceeds
implementation size limit
awk: line 1: regular expression /^aaa[. ]| ... exceeds
implementation size limit
awk: line 1: regular expression /^aaa[. ]| ... exceeds
implementation size limit
awk: line 1: regular expression /^aaa[. ]| ... exceeds
implementation size limit
news@optima5:~$
The above is INN version is 2.4.4 (on Optima5).
Out of curiosity, since you mentioned INN 2.4.3, I tried this on
Optima7, which is INN 2.4.3. I didn't see any errors like on Optima5.
news@optima7:/var/lib/news$ cat newsgroups | /usr/lib/news/bin/docheckgroups
<snip>
# Change these lines:
# xanadu.info Announcements about access (no posting)
# comp.parallel Massively parallel hardware/software. (Moderated)
# comp.protocols.dns.bind Berkeley Internet Name Domain (BIND).
(Moderated)
# comp.unix.bsd.freebsd.announce Announcements pertaining to
FreeBSD. (Moderated)
# comp.unix.bsd.netbsd.announce Announcements pertaining to NetBSD.
(Moderated)
# comp.unix.bsd.openbsd.announce OpenBSD announcements. (Moderated)
# info.bind *The Berkeley BIND server (bi...@arpa.berkeley.edu).
(Moderated)
# linux.debian.announce debian-...@lists.debian.org (Moderated)
# linux.debian.announce.devel
debian-dev...@lists.debian.org (Moderated)
# linux.debian.announce.security
debian-secur...@lists.debian.org (Moderated)
# linux.debian.beowulf debian-...@lists.debian.org (Moderated)
# linux.debian.doc debia...@lists.debian.org (Moderated)
# linux.debian.isp debia...@lists.debian.org (Moderated)
# linux.debian.laptop debian...@lists.debian.org (Moderated)
# linux.debian.maint.boot debia...@lists.debian.org (Moderated)
# linux.debian.maint.dpkg debia...@lists.debian.org (Moderated)
# linux.debian.maint.firewall debian-...@lists.debian.org
(Moderated)
# linux.debian.news debia...@lists.debian.org (Moderated)
# linux.debian.policy debian...@lists.debian.org (Moderated)
# linux.debian.ports.68k debia...@lists.debian.org (Moderated)
# linux.debian.ports.powerpc debian-...@lists.debian.org
(Moderated)
# linux.debian.project debian-...@lists.debian.org (Moderated)
# linux.debian.security debian-...@lists.debian.org (Moderated)
# linux.debian.user debia...@lists.debian.org (Moderated)
# linux.debian.www debia...@lists.debian.org (Moderated)
# linux.samba sa...@lists.samba.org (Moderated)
# linux.samba.announce samba-a...@lists.samba.org (Moderated)
# rec.radio.broadcasting Discussion of global domestic broadcast
radio. (Moderated)
Mon Apr 27 10:36:57 CDT 2009
news@optima7:/var/lib/news$
It took 01:15:28 to finish! Perhaps due to swap spooling, since this is
a production server and has only 128 MB of RAM.
top - 08:42:06 up 46 days, 11:58, 5 users, load average: 1.06, 1.08, 1.02
Tasks: 85 total, 3 running, 82 sleeping, 0 stopped, 0 zombie
Cpu(s): 41.0% user, 54.4% system, 0.0% nice, 4.6% idle
Mem: 127444k total, 124232k used, 3212k free, 23048k buffers
Swap: 200772k total, 16288k used, 184484k free, 37336k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+
COMMAND
30955 news 12 0 6652 6648 1032 S 3.6 5.2 2:03.83
docheckgroups
28845 news 19 0 548 544 432 R 3.3 0.4 0:00.10
grep
29736 john 15 0 1072 1072 844 R 2.3 0.8 1:12.64 top
Any thoughts about those lines I provided in the earlier message, which
had only closing parentheses (most were "smileys")? Perhaps they are a
genuine problem?
Julien ÉLIE
>>> news@optima5:/var/lib/news$ cat newsgroups | docheckgroups
>>> egrep: Unmatched ) or \)
>
> news@optima5:~$ cat newsgroups | ./docheckgroups.in
> awk: line 1: regular expression /^aaa[. ]| ... exceeds
> implementation size limit
I believe it is the same problem. The pattern in egrep was
too long.
Trying to search for this error in Google, I found out:
#define MIN_SPRINTF 400
EXTERN union {
STRING *_split_buff[MAX_SPLIT] ;
char _string_buff[MIN_SPRINTF] ;
} tempbuff ;
#define SPRINTF_SZ sizeof(tempbuff)
if (p == string_buff + SPRINTF_SZ - 2)
compile_error(
"regular expression /%.10s ..."
" exceeds implementation size limit",
string_buff) ;
It seems to be what happens in your case. And of course the search
pattern for all the hierarchies present in your newsgroups file
is more than 400 bytes!
> The above is INN version is 2.4.4 (on Optima5).
>
> Out of curiosity, since you mentioned INN 2.4.3, I tried this on
> Optima7, which is INN 2.4.3. I didn't see any errors like on Optima5.
Note that docheckgroups is the same in INN 2.4.3 and INN 2.4.4.
In fact, docheckgroups has not been changed since 2001; and I only
started to modify it in 2007.
It seems to be an issue in your configuration.
> It took 01:15:28 to finish!
It takes about 12 seconds on my server.
> Any thoughts about those lines I provided in the earlier message, which
> had only closing parentheses (most were "smileys")?
How could you be sure that the problem comes from smileys?
Especially when egrep reports an error when processing the *active* file!
2 egrep are against the active file (out of 4 [or 6 for INN 2.5.0] which
generated an error).
Maybe gawk works better than awk :)
--
Julien ÉLIE
« Nam et ipsa scientia potestas est. » (Francis Bacon)
John F. Morse
> Hi John,
>
> Note that docheckgroups is the same in INN 2.4.3 and INN 2.4.4.
> In fact, docheckgroups has not been changed since 2001; and I only
> started to modify it in 2007.
> It seems to be an issue in your configuration.
Would you have any clue as to what and where I could look?
>> It took 01:15:28 to finish!
>
> It takes about 12 seconds on my server.
Must be better than a 266 MHz AMD-K6. ;-)
>> had only closing parentheses (most were "smileys")?
> Any thoughts about those lines I provided in the earlier message, which
>
> How could you be sure that the problem comes from smileys?
> Especially when egrep reports an error when processing the *active* file!
> 2 egrep are against the active file (out of 4 [or 6 for INN 2.5.0] which
> generated an error).
Oh, I'm not sure at all -- just a guess, since there were many
colon+right parentheses found in the newsgroups file (which I referred
to as "smileys"). There were a couple with a different character and a
right parentheses as well.
For instance, close to the end of the newsgroups file was "%)" at the
end of the description for the viwa.sobutilnik group. Further up, the
chile.rec.musica.metal group has ">:)" in the description.
After finding many right parentheses, I narrowed my search for ":)"
only, instead of just all right parentheses. There likely are other
patterns with single right-parentheses, which is what the original
"Unmatched" failure indicated.
news@optima5:/var/lib/news$ cat newsgroups | docheckgroups
egrep: Unmatched ) or \)
egrep: Unmatched ) or \)
egrep: Unmatched ) or \)
egrep: Unmatched ) or \)
According to
http://www.eyrie.org/~eagle/software/inn/docs/docheckgroups.html
docheckgroups does look at the newsgroups file:
" ... Then, docheckgroups checks the active and newsgroups files and
displays on its standard output a list of changes, if any. It does not
change anything by default; it only points out what should be changed: ...."
After searching the newsgroups file, I noticed these single parentheses
were used in a few of the newsgroup descriptions, and suspected they
might be the cause.
My active file does not have parentheses in it, nor does it have any of
the newsgroups that were listed as failures. Therefore the failure
couldn't have come from the active file.
Thoughts?
Julien ÉLIE
> Would you have any clue as to what and where I could look?
If, at the beginning of your docheckgroups script (that is to say after
the third line -- after the load of innshellvars), you add:
LC_ALL=C; export LC_ALL
does it change something?
It reminds me of William's issue with non-ASCII characters:
https://lists.isc.org/pipermail/inn-workers/2009-January/016410.html
https://lists.isc.org/pipermail/inn-workers/2009-January/016435.html
> My active file does not have parentheses in it, nor does it have any of
> the newsgroups that were listed as failures. Therefore the failure
> couldn't have come from the active file.
"egrep patterns file" is how egrep works. The problem comes from patterns
in your case, not the file.
--
Julien ÉLIE
« À l'école, en algèbre, j'étais du genre Einstein.
Mais plutôt Franck qu'Albert. » (Philippe Geluck)
John F. Morse
> Hi John,
>
>> Would you have any clue as to what and where I could look?
>
> If, at the beginning of your docheckgroups script (that is to say after
> the third line -- after the load of innshellvars), you add:
>
> LC_ALL=C; export LC_ALL
>
> does it change something?
>
> It reminds me of William's issue with non-ASCII characters:
> https://lists.isc.org/pipermail/inn-workers/2009-January/016410.html
> https://lists.isc.org/pipermail/inn-workers/2009-January/016435.html
>
>
>> My active file does not have parentheses in it, nor does it have any of
>> the newsgroups that were listed as failures. Therefore the failure
>> couldn't have come from the active file.
>
> "egrep patterns file" is how egrep works. The problem comes from
> patterns
> in your case, not the file.
Hi, Julien.
I think we have some kind of language problem here. Either you are
looking in the wrong direction, or I am not expressing the problem I see
well enough for you to understand.
I base this on your comments about the active file and the egrep
patterns. I am not interested on the egrep patterns, nor am I interested
in the active file (/var/lib/news/active).
What my question is (or was) is based on what I see as an error message
"egrep: Unmatched ) or \)" which appears to me as pointing to the right
(closing) parentheses without a left (opening) parentheses. "Unmatched"
is the error.
I check the newsgroups file (/var/lib/news/newsgroups) which I
downloaded from ftp://ftp.isc.org/usenet/CONFIG/newsgroups.bz2 and
installed, and it does, indeed, have occurrences of "unmatched" parentheses.
If egrep is catching these, why do you believe something is wrong with
egrep -- or innshellvars?
It appears to me innshellvars, docheckgroups and egrep are doing exactly
what they are expected to do: catch these errors in the newsgroups file
(not the active file in this case).
Should someone check the three newsgroup files on ISC and remove the
unmatched parentheses? Why would newsgroup descriptions need a "smiley"
anyway? ;-)
Julien ÉLIE
> What my question is (or was) is based on what I see as an error message
> "egrep: Unmatched ) or \)" which appears to me as pointing to the right
> (closing) parentheses without a left (opening) parentheses. "Unmatched"
> is the error.
Then:
wget ftp://ftp.isc.org/usenet/CONFIG/newsgroups.bz2
bunzip2 newsgroups.bz2
grep '^chile\.' newsgroups | docheckgroups
What's the error you see?
I chose chile.* because you told me:
chile.rec.musica.metal Tertulia sobre el Rock Pesado >:) .
> If egrep is catching these, why do you believe something is wrong with
> egrep -- or innshellvars?
It would be weird that I could not egrep into a file containing only ")".
(I did not say something was wrong with innshellvars.)
> It appears to me innshellvars, docheckgroups and egrep are doing exactly
> what they are expected to do: catch these errors in the newsgroups file
> (not the active file in this case).
There's no error in the newsgroups file!
--
Julien ÉLIE
« Chaque chêne est envahi par quantité de druides qui
cueillent le gui en travaillant dur de la serpe... » (Astérix)
Russ Allbery
> What my question is (or was) is based on what I see as an error
> message "egrep: Unmatched ) or \)" which appears to me as pointing to
> the right (closing) parentheses without a left (opening)
> parentheses. "Unmatched" is the error.
>
> I check the newsgroups file (/var/lib/news/newsgroups) which I
> downloaded from ftp://ftp.isc.org/usenet/CONFIG/newsgroups.bz2 and
> installed, and it does, indeed, have occurrences of "unmatched"
> parentheses.
Yeah, I know, this looks plausible, but we don't think this is the cause
of the problem. Instead, we think the problem is that your egrep can't
handle unlimited-length patterns.
The way docheckgroups works is that it first scans the list of
newsgroups to determine what top-level hierarchies are defined in it.
It then turns that list of top-level hierarchies into a regex, and uses
it to extract the relevant groups from your local active file so that it
can compare which groups need to be changed.
If you run docheckgroups on the entire newsgroups file from ftp.isc.org,
you end up with a *lot* of top-level hierarchies, and therefore a very
long regular expression. It looks like your egrep only has a limited
length of pattern that it supports and it runs out of room in its
pattern buffer before it sees the end of the regular expression and
hence the closing parenthesis. That's where the error message is coming
from.
The smileys in the newsgroup descriptions are probably just a red
herring. docheckgroups doesn't care about the contents of the
description section.
John F. Morse
> Hi John,
>
>> What my question is (or was) is based on what I see as an error message
>> "egrep: Unmatched ) or \)" which appears to me as pointing to the right
>> (closing) parentheses without a left (opening) parentheses. "Unmatched"
>> is the error.
>
> Then:
>
> wget ftp://ftp.isc.org/usenet/CONFIG/newsgroups.bz2
> bunzip2 newsgroups.bz2
> grep '^chile\.' newsgroups | docheckgroups
>
> What's the error you see?
I received no error.
I can see why -- after reading Russ's description of the list of
top-level hierarchies turned into a long length regex which is too long
for my implementation of egrep.
> I chose chile.* because you told me:
>
> chile.rec.musica.metal Tertulia sobre el Rock Pesado >:) .
>
>> If egrep is catching these, why do you believe something is wrong with
>> egrep -- or innshellvars?
>
> It would be weird that I could not egrep into a file containing only ")".
> (I did not say something was wrong with innshellvars.)
I only mentioned innshellvars because it was one of the three that was
mentioned earlier.
>> It appears to me innshellvars, docheckgroups and egrep are doing exactly
>> what they are expected to do: catch these errors in the newsgroups file
>> (not the active file in this case).
>
> There's no error in the newsgroups file!
I see now (the above test proved it).
Thanks for taking the time to try and explain this to me, Julien. I was
missing the key part (long regex) and couldn't understand what was
happening.
Also, please forgive me for being slow getting back to you (and Russ).
I'm fighting a cron and/or syslogd problem, which is only on this one
server. If I determine it may be something in INN, ... I'll be back. ;-)
John F. Morse
OK. That makes sense. Thanks for taking the time to explain this, Russ.
> The smileys in the newsgroup descriptions are probably just a red
> herring. docheckgroups doesn't care about the contents of the
> description section.
Then perhaps this should be corrected?:
According to
http://www.eyrie.org/~eagle/software/inn/docs/docheckgroups.html
docheckgroups does look at the newsgroups file:
" ... Then, docheckgroups checks the active and newsgroups files and
displays on its standard output a list of changes, if any. It does not
change anything by default; it only points out what should be changed: ...."
So, are you saying that, if the above is indeed correct, then
docheckgroups looks in the /var/lib/news/active file, and then only
looks at the newsgroup names in the /var/lib/news/newsgroups file, and
not the description portion?
Russ Allbery
> Then perhaps this should be corrected?:
>
> According to
> http://www.eyrie.org/~eagle/software/inn/docs/docheckgroups.html
> docheckgroups does look at the newsgroups file:
>
> " ... Then, docheckgroups checks the active and newsgroups files and
> displays on its standard output a list of changes, if any. It does not
> change anything by default; it only points out what should be changed: ...."
>
> So, are you saying that, if the above is indeed correct, then
> docheckgroups looks in the /var/lib/news/active file, and then only
> looks at the newsgroup names in the /var/lib/news/newsgroups file, and
> not the description portion?
It looks at it in the sense that it will use it to update your
newsgroups file, but it doesn't care about the contents. The contents
are just copied verbatim into your newsgroups file for groups that are
added.
Julien ÉLIE
> Thanks for taking the time to try and explain this to me, Julien. I was
> missing the key part (long regex) and couldn't understand what was
> happening.
I thought it was clear when *you* wrote:
news@optima5:~$ cat newsgroups | ./docheckgroups.in
awk: line 1: regular expression /^aaa[. ]| ... exceeds
implementation size limit
And then I answered:
I believe it is the same problem. The pattern in egrep was
too long.
Trying to search for this error in Google, I found out:
#define MIN_SPRINTF 400
EXTERN union {
STRING *_split_buff[MAX_SPLIT] ;
char _string_buff[MIN_SPRINTF] ;
} tempbuff ;
#define SPRINTF_SZ sizeof(tempbuff)
if (p == string_buff + SPRINTF_SZ - 2)
compile_error(
"regular expression /%.10s ..."
" exceeds implementation size limit",
string_buff) ;
It seems to be what happens in your case. And of course the search
pattern for all the hierarchies present in your newsgroups file
is more than 400 bytes!
But well, I should have explained more. I do not have Russ' skills
to explain something clearly in English :)
--
Julien ÉLIE
« Un sourire coûte moins cher que l'électricité
mais donne autant de lumière. » (Abbé Pierre)
John F. Morse
You English is fine, Julien. Better than most I've seen in Usenet.
Actually, better than most Americans I've noticed!
The "trouble" was my looking for something unimportant, and in the wrong
place.
What you wrote above is now clear once the reason was explained.
John F. Morse
> "John F. Morse" <xanad...@example.invalid> writes:
>
>
>> Then perhaps this should be corrected?:
>>
>> According to
>> http://www.eyrie.org/~eagle/software/inn/docs/docheckgroups.html
>> docheckgroups does look at the newsgroups file:
>>
>> " ... Then, docheckgroups checks the active and newsgroups files and
>> displays on its standard output a list of changes, if any. It does not
>> change anything by default; it only points out what should be changed: ...."
>>
>> So, are you saying that, if the above is indeed correct, then
>> docheckgroups looks in the /var/lib/news/active file, and then only
>> looks at the newsgroup names in the /var/lib/news/newsgroups file, and
>> not the description portion?
>>
>
> It looks at it in the sense that it will use it to update your
> newsgroups file, but it doesn't care about the contents. The contents
> are just copied verbatim into your newsgroups file for groups that are
> added.
>
Understood. Thanks.
Who has responsibility or authority to clean up the newsgroups
descriptions? Is it a person listed in a charter, or can anyone attempt
standardization of newsgroup descriptions?
They do look like a hodgepodge of various attempts of humor, and
carelessness in punctuation, etc., and I'm not even considering the
alt.* hierarchy. =-O
Perhaps this has always been on the back burner because not many
newsreaders can even display these descriptions? They are rather useless
once you have satisfied your own curiosity about a group. They also take
a significant amount of time to download.
Perhaps posting them on the Website would be sufficient? That likely
would make them too hard to find for people who don't know they exist in
the first place.
Russ Allbery
> Understood. Thanks.
>
> Who has responsibility or authority to clean up the newsgroups
> descriptions? Is it a person listed in a charter, or can anyone
> attempt standardization of newsgroup descriptions?
The hierarchy administrator. New newsgroup descriptions will be updated
from the checkgroups or new newgroups for that hierarchy automatically.
John F. Morse
> "John F. Morse" <xanad...@example.invalid> writes:
>
>> Understood. Thanks.
>>
>> Who has responsibility or authority to clean up the newsgroups
>> descriptions? Is it a person listed in a charter, or can anyone
>> attempt standardization of newsgroup descriptions?
>>
>
> The hierarchy administrator. New newsgroup descriptions will be updated
> from the checkgroups or new newgroups for that hierarchy automatically.
>
I'd suspect that would kind of eliminate any chance of a consensus on
standardization. ;-)
To get standardization, probably someone from ISC would have to ramrod
the work. Otherwise there would be many arguments, like between
Americans and the British about where a period should be in a sentence.
I guess it's best to sweep it under the rug, and move on to another
topic which has some kind of chance to be solved. ;-)
D. Stussy
news:gtfqm1$otm$1...@optima7.xanadu-bbs.net...
> Russ Allbery wrote:
> > "John F. Morse" <xanad...@example.invalid> writes:
> >> Understood. Thanks.
> >>
> >> Who has responsibility or authority to clean up the newsgroups
> >> descriptions? Is it a person listed in a charter, or can anyone
> >> attempt standardization of newsgroup descriptions?
> >
> > The hierarchy administrator. New newsgroup descriptions will be
updated
> > from the checkgroups or new newgroups for that hierarchy automatically.
>
> I'd suspect that would kind of eliminate any chance of a consensus on
> standardization. ;-)
>
> To get standardization, probably someone from ISC would have to ramrod
> the work. Otherwise there would be many arguments, like between
> Americans and the British about where a period should be in a sentence.
>
> I guess it's best to sweep it under the rug, and move on to another
> topic which has some kind of chance to be solved. ;-)
Or put a woman in charge. They hate periods! ;-) Eliminate them all.
John F. Morse
;-)
Julien ÉLIE
> Your English is fine, Julien. Better than most I've seen in Usenet.
> Actually, better than most Americans I've noticed!
That's very kind. Thanks!
As for your question about descriptions, you will for instance
notice there:
http://usenet.trigofacile.com/hierarchies/index.py?see=NEWS&showactions=all
2007-08-15 15:00:11 changedesc news.admin.announce (by checkgroups)
It means that the description of that newsgroup changed after a checkgroups
(a list of all valid newsgroups along with their descriptions for a given
hierarchy).
Have a nice day,
--
Julien ÉLIE
« -- Nous parlerons quand l'interprète dormira.
[Bong !]
-- Il dort. On peut parler. » (Astérix)
John F. Morse
> Hi John,
>
>> Your English is fine, Julien. Better than most I've seen in Usenet.
>> Actually, better than most Americans I've noticed!
>
> That's very kind. Thanks!
You are welcome, Julien.
I had always heard English is a difficult language for a foreigner to learn.
Might be, as well as for a lot of Americans too! Especially the
teenagers. I don't have a clue about what they are saying. ;-)
Do teenagers in your part of the world also seen to encode their native
language?
> As for your question about descriptions, you will for instance
> notice there:
>
> http://usenet.trigofacile.com/hierarchies/index.py?see=NEWS&showactions=all
>
>
> 2007-08-15 15:00:11 changedesc news.admin.announce (by checkgroups)
>
> It means that the description of that newsgroup changed after a
> checkgroups
> (a list of all valid newsgroups along with their descriptions for a given
> hierarchy).
OK. Thanks for that link. It puts a lot of good info in one place where
it may be easier to find.
Isn't Trigofacile your domain? I have that NNTP server configured in my
newsreader from months ago, but rarely go there.
The word "trigofacile" doesn't translate from French into English when
using Google Translate, or Yahoo (Alta Vista) Babel Fish. What does it mean?
("tri" usually indicates three [from Latin and Greek] or "sorting" in
French, while "facile" = easy. Of course "trig" is short for trigonometry.)
Julien ÉLIE
> I had always heard English is a difficult language for a foreigner to learn.
Pronouncing English well is also not easy for French people :)
In Europe, it is said that they who speak French best are Russian people
(it is true that when they study French, they study it properly and
pronounce it fine).
> Do teenagers in your part of the world also seen to encode their native
> language?
Oh yes, unfortunately. They keep sending SMS and are not keen at all on spelling!
> Isn't Trigofacile your domain?
Yes.
> The word "trigofacile" doesn't translate from French into English when
> using Google Translate, or Yahoo (Alta Vista) Babel Fish. What does it mean?
> ("tri" usually indicates three [from Latin and Greek] or "sorting" in
> French, while "facile" = easy. Of course "trig" is short for trigonometry.)
You guessed right!
I started my web site in 1998 when I was in secondary school. As I was really
interested in maths, it was the main subject of my web site.
"Easy trigo[nometry]". "Trigo" with a pronounced "o" is short for trigonometry
in French.
One of the most interesting page is the one on the CORDIC algorithm (how a pocket
calculator computes trigonometric functions):
http://www.trigofacile.com/maths/trigo/calcul/cordic/cordic.htm
(Written a few years ago -- not when I was in secondary school, this one!)
Since that time, I have widened the subjects of my pages.
--
Julien ÉLIE
« Mais pourquoi courent-ils si vite ? Pour gagner du temps ! Comme le temps,
c'est de l'argent... plus ils courent vite, plus ils en gagnent. » (Raymond Devos)
Joshua Cranmer
> Do teenagers in your part of the world also seen to encode their native
> language?
Although I am a native English speaker (well, American, if I must be
precise), I have taken a few years of French. I can, however, definitely
affirm this statement, as I saw when I stumbled across a French quote
database. It's practically gibberish.
Jokes also seem to lose their value after you have to spend several
seconds translating them...
--
Beware of bugs in the above code; I have only proved it correct, not
tried it. -- Donald E. Knuth
Julien ÉLIE
> news@optima5:~$ cat newsgroups | ./docheckgroups.in
> awk: line 1: regular expression /^aaa[. ]| ... exceeds
> implementation size limit
If you have gawk installed on your system, you can put its path
in inshellvars:
AWK='/usr/bin/gawk'
instead of
AWK='/usr/bin/awk'
and the new docheckgroups script I gave you in that thread
will normally work (because gawk has no size limit).
--
Julien ÉLIE
« César, c'est un Jules tout de même ! » (Astérix)
John F. Morse
> Hi John,
>
>> news@optima5:~$ cat newsgroups | ./docheckgroups.in
>> awk: line 1: regular expression /^aaa[. ]| ... exceeds
>> implementation size limit
>
> If you have gawk installed on your system, you can put its path
> in inshellvars:
>
> AWK='/usr/bin/gawk'
> instead of
> AWK='/usr/bin/awk'
>
> and the new docheckgroups script I gave you in that thread
> will normally work (because gawk has no size limit).
Thanks for the update on this Julien.