C to Java Byte Code

[Programmer Dude]

I'll be interested to see what Agent makes of this post
when it reads it (that is, will it color code the quotes).

No colors. Looks just like an "old fashioned" post in tin or
something. Oooooh,... 80s flashback!!

 

[Floyd L. Davidson]

Dik T. Winter coughed up:

I saw this as well yesterday----Not sure what it means.

Some newsreaders, and some news servers, will refuse to post a
message that contains more quoted text than non-quoted text.
The mechanism used is extremely simple, and merely counts lines
that begin with the most commonly used quote prefix characters.

The method for overriding the count is to use a quote prefix
that begins with a space or any other character that is rarely
used (and thus does not trigger the mechanism). The
newsreader/server will then allow the message to be posted.

Of course it should be realized that when newsreaders and
servers began implementing that mechanism it caused a
significant number of people to permanently override it by
making their default quote prefix something that would not be
counted. That resulted in the now common practice of using a
variety of characters other than '>'. (Other methods to
override it are even worse...)

 

[=?iso-8859-1?q?M=E5ns_Rullg=E5rd?=]

No colors. Looks just like an "old fashioned" post in tin or
something. Oooooh,... 80s flashback!!

FWIW, gnus colored it properly.

 

[Michael Wojcik]

[Dropped two of the four off-topic groups.]

Is there a document somewhere which suggests how to form a standard usenet
quoted reply line? I cannot find it via google.

There are a great many such documents. Probably a majority of them
are Usenet posts. In some cases "suggests" might be too mild a term.

Thus far in this discussion (such as it is), I haven't seen anyone
point out that there is a current set of Internet Drafts looking to
update the Netnews standard (RFC 1036).[1] I shall leave reading and
interpreting them (the "Usenet Best Practices" draft in particular),
and questions of their status and relevance, as a task for the
reader. (I *have* read and interpreted them, for the benefit of
similar discussions in other groups, and found it not the most
rewarding task to which I might dedicate my time.)

I might suggest that anyone (say, Mr Newmane) making bold claims about
applicable "standards" without doing the basic homework to discover
those documents is perhaps not worth listening to on such matters.
(I might also levy criticism against people like Mr MacHonahey, who
insists that one convention is "universal" in Usenet, then in the
same breath complains that it is not universally followed.) But
perhaps my standards are too high.

Is there someplace that defines (unofficially of course) the guideline for
it?

Of course. Unofficial definitions abound. Far more interesting, in
my opinion, are attempts at officially defining it - none of which
have yet succeeded, AFAIK.

Somewhat less interesting, but still more useful than the random
stabs of hoi polloi, are the various netiquette and net.usage guides
that touch on the topic. Brendan Kehoe's "Zen and the Art of the
Internet", for example, notes that "many newsreaders" use the "> "
quotation marker, but does not specifically endorse it.

At this point, though, we do not even have a clear consensus, much
less a recognized standard. If "Usenet Best Practices" becomes an
RFC, that will clarify things somewhat, but only somewhat.


1. http://www.ietf.org/ids.by.wg/usefor.html

 

[Michael Wojcik]

Well, most don't allow you to put white space before your quote token.

You've done a comprehensive survey of newsreaders to determine this?

You are right it is a /tradition/, it has always bene tradition to start
a line with a quote character, lest it be counted as local text.

You've done a comprehensive survey of Usenet usage to determine this?

furthur
more, most readers that allow for od and unusual quoting styles are
usually features created by people who probably haven't spent a lot of
time in UseNet.

You've done a comprehensive survey of newsreaders to determine this?

Or are you merely full of crap?

 

[Floyd L. Davidson]

>

AZN> Well, most don't allow you to put white space before your
AZN> quote token.

You continue to advocate the use of *broken* newsreaders as a
default standard for Usenet... (The /reasonable/ solution is to
use a better newsreader.)

Just for grins (and groans), I reconfigured Gnus just to post
this one article. You'll note that it allows the quote prefix
to begin with whitespace, /and/ that is also allows two
different quote prefixes. In this example the text *quoted*
*from* the previous article is prefixed by " AZN>", and text
*quoted* *in* the previous article is prefixed by " > ".

I haven't used this facility before (and won't likely again),
and I assume that it is perhaps handy for dealing with some of
the odd things people do with /supersite/. Note that I also let
Gnus reformat the paragraphs to maintain proper line lengths,
and it had no problem with the odd quote prefix or the leading
whitespace. (Good newsreaders really are *nice*! Or, they are
if you define "good" as a measure of functionality rather than
glitter. You'll be pleased, though, to hear that the spell
checker was indeed confused, and wanted to correct your spelling
errors as well as mine. But that's a bug, not a Usenet
Standard!)

AZN> You are right it is a /tradition/, it has always bene
AZN> tradition to start a line with a quote character, lest it
AZN> be counted as local text. furthur more, most readers that
AZN> allow for od and unusual quoting styles are usually
AZN> features created by people who probably haven't spent a
AZN> lot of time in UseNet. (Not always the case, but ther are
AZN> some poor readers out there.)

Incidentally, given these "OE folks are using OE_QuoteFix" are
the makings for de facto standards and tradition claims, have
you ever taken a look at /supercite/ and the way it cites
previous articles?

Since /supercite/ existed long before OE began to "standardize"
Usenet with non-functionality, it would seem that the authors of
OE and OE_QuoteFix *should* have looked at examples of similar
programs to avoid the same mistakes and at least to include the
same features. (Of course, since they had little Usenet
experience at the time, they may just have not understood...)

 

[joe]

FWIW, gnus colored it properly.

Gnus also wraps it correctly when reformatting.

Joe

 

[Michael Wojcik]

The only official guidelin on the general subject is RFC 1036, which
is ancient and provides no real guidance in this specific area.
Somewhat newer (but still roughly 10 years old) is Henry Spencer's
"Son of RFC 1036", which he published to Usenet as an "RFC Draft to
be" -- i.e. he didn't even consider it sufficiently finished to
qualify as a draft. OTOH, there isn't anything better either...

Most of the recommendations of Son of 1036 - in fact, much of the
text, more or less verbatim - is included in the current set of
Internet Drafts for revising the Netnews standard. See my other post
in this thread for the reference. (Or look it up yourself; anyone
prepared to debate questions of Internet standards ought to know how
to do that. I imagine Jerry does, since he is familiar with Son of
1036. How many of Dik's detractors here do?)

 

[Floyd L. Davidson]

Gnus also wraps it correctly when reformatting.

I did find a "bug" in Gnus though... with the
message-yank-prefix and message-yank-cited-prefix variables set
to start with whitespace, invoking ispell on the message buffer
checks the spelling of the quoted text as well as the message
text (which it normally does not). I guess ispell.el isn't as
smart as message.el.

Maybe Alfred Z. Newmane can tell us how OE's spell check
functionality works in those circumstances? ;-)

 

[Dik T. Winter]

> That because, unlike Dik, other people have actually got a clue of
> what and WHY some things are done on UseNet in the way they have been
> for decades.

At least I have a clue *why* long ago people suggested to put a space
in front of the quotation symbol.

 

[Dik T. Winter]

> Is there a document somewhere which suggests how to form a standard usenet
> quoted reply line? I cannot find it via google.

No, there is no standard about that.

> My understanding was that the line to be quoted has a quote character
> (usually >, sometimes | ) inserted at the /beginning/ of the line. Not just
> before the first non whitespace character.

Originally the advice was not to use the square brackets, the curly
brackets, the vertical bar, the backslash, the underscore or the
backtick. The reason was that that could make quotations difficult
to read for people that used another incantation of ISO-646 than
ASCII.

 

[Dik T. Winter]

> Well, most don't allow you to put white space before your quote token.
> You are right it is a /tradition/, it has always bene tradition to start
> a line with a quote character, lest it be counted as local text.

Yup, that is what rn (the second newsreader ever) did do indeed, by default.
It came public in september 1984. (I do not know when I started to use it,
it must have been somewhere in 1985 or 1986, before that time I used the
first newsreader ever: readnews.) But with rn you could change the quoting
mechanism easily, put in your .rninit a line like
-F" > "
and you would get my quoting style. There was a pretty good reason to
change it and it was largely advised...

> most readers that allow for od and unusual quoting styles are
> usually features created by people who probably haven't spent a lot of
> time in UseNet.

Eh? Larry Wall (of perl fame) not having spent a lot of time in Usenet
back in 1984? Where were *you* at that time? Actually you would be
right. At that time you could (and I did) read *all* articles posted on
the net without problems. Darn, all the news came here, nightly, through
a 1200 baud dial-up line from Amsterdam to the US (seismo), the first
newslink outside the US and Canada.

> (Not always the case, but there are some poor readers out
> there.)

Yes, there are (OE?) I still think rn is a good newsreader, and still do
use it. (Although it has since that time been updated quite a bit.)

Funny, in 12 days it will be 20 years that I posted my first article on
Usenet. <[email protected]>.

 

[Dik T. Winter]

> Taking the Son of RFC 1036 as our guidance, we find only that lines
> should be broken "as appropriate", and "Although styles vary widely,
> for plain text it is usual to use no left margin..." That's
> _extremely_ weak guidance at best -- even reading it as strongly as
> possible, I can't see how it could be taken as much more than a
> suggestion that Dik's style is mildly unusual.

I do not even think that. Plain text is something different from
quotation. The intention was to stop people who indented *all* text
they entered by preceding their text with either a space or a tab.
And, yes, that did occur.

 

[Dik T. Winter]

> See my other post
> in this thread for the reference. (Or look it up yourself; anyone
> prepared to debate questions of Internet standards ought to know how
> to do that. I imagine Jerry does, since he is familiar with Son of
> 1036. How many of Dik's detractors here do?)

I did know Henry Spencer's stuff already a long time. Now Charles
Lindsey is apparently trying to draft a new version (is that the
same Charles Lindsey I met in connection with Algol 68?). I would
think the 'gnus' people would have something to say about it.

 

[Dik T. Winter]

> Dik T. Winter coughed up: ....
>
> I saw this as well yesterday----Not sure what it means.

First I was wrong. Next, let me explain. When in 1984 rn came out to
replace readnews as a newsreader (they were, I think, second and first
respectively), the system would reject articles where the number of lines
of quoted text was larger than the number of lines of new text. This
according to a guideline that had come into force around that time
(readnews did not impose such limits). In some cases this was too
severe (but it helped quite a bit, as it suggested severely trimming the
quoted material). So around that time ways were thought to allow such
postings if you did put in enough effort. One way was to follow the
advice above: when you were put into 'vi' (as text editor) to create
your follow-up, first delete any stuff that was irrelevant, next change
the quotation character. (The command ":%s/^>/</" is a 'vi' command
to change any ">" that occurs at the start of a line to "<".) Doing
that would interfere with the check of new text against quoted text.
By the time that advice got to the guidelines (I think December 1987)
many people had already done various things with the quoting string,
as it was easy with rn to change the quoting string to anything you
fancied (about three years of frustration have gone into it). That
is why you see the proliferation of quoting styles, none of them have
ever been disallowed (but it was suggested that you did not use
specific symbols).

BTW, one of the big advantages of 'rn' over 'readnews' was the
introduction of 'kill-files', another one was that now you no longer
needed to see *all* header lines of an article.

 

[J. J. Farrell]

J. J. Farrell coughed up:

Is there a document somewhere which suggests how to form a standard usenet
quoted reply line? I cannot find it via google.

None that I know of. This is largely a matter of tradition.

My understanding was that the line to be quoted has a quote character
(usually >, sometimes | ) inserted at the /beginning/ of the line. Not just
before the first non whitespace character.

That is the most common tradition. Dik's style is certainly
unusual, but has not previously been a problem. In any case,
Dik's long-standing unusual quoting style is effectively
"protected" by another long-standing (though less widely
known) tradition - that lines beginning with a space character
should not be reformatted. This makes it practicable to quote
posted C code, for example.

Is there someplace that defines (unofficially of course) the guideline for
it?

Not that I know of.

 

[Thomas G. Marshall]

Dik T. Winter coughed up:

I saw this as well yesterday----Not sure what it means.

First I was wrong. Next, let me explain. When in 1984 rn came out to
replace readnews as a newsreader (they were, I think, second and first
respectively), the system would reject articles where the number of
lines
of quoted text was larger than the number of lines of new text. This
according to a guideline that had come into force around that time
(readnews did not impose such limits). In some cases this was too
severe (but it helped quite a bit, as it suggested severely trimming
the
quoted material). So around that time ways were thought to allow such
postings if you did put in enough effort. One way was to follow the
advice above: when you were put into 'vi' (as text editor) to create
your follow-up, first delete any stuff that was irrelevant, next
change
the quotation character. (The command ":%s/^>/</" is a 'vi' command
to change any ">" that occurs at the start of a line to "<".) Doing
that would interfere with the check of new text against quoted text.
By the time that advice got to the guidelines (I think December 1987)
many people had already done various things with the quoting string,
as it was easy with rn to change the quoting string to anything you
fancied (about three years of frustration have gone into it). That
is why you see the proliferation of quoting styles, none of them have
ever been disallowed (but it was suggested that you did not use
specific symbols).

BTW, one of the big advantages of 'rn' over 'readnews' was the
introduction of 'kill-files', another one was that now you no longer
needed to see *all* header lines of an article.[/QUOTE]

I recognized the vi / ex / sed-like command, and knew what it did per se. I
just didn't know what they were trying to accomplish. To defeat the more
quote than substance rule makes sense. (I hope we're way past ever needing
that silly rule again).

 

[Thomas G. Marshall]

J. J. Farrell coughed up:

"Thomas G. Marshall"


None that I know of. This is largely a matter of tradition.


That is the most common tradition. Dik's style is certainly
unusual, but has not previously been a problem. In any case,
Dik's long-standing unusual quoting style is effectively
"protected" by another long-standing (though less widely
known) tradition - that lines beginning with a space character
should not be reformatted. This makes it practicable to quote
posted C code, for example.

Understood, but that doesn't "protect" his style policy for the simple
reason that he will be adding a space in front of lines that would only get
cutoff grotesquely if there weren't allowed to be formatted. Unless such
lines are never truncated either. {shrug}

 

[Programmer Dude]

To defeat the more quote than substance rule makes sense.
(I hope we're way past ever needing that silly rule again).

In this age of nuubs who can't, or won't, edit, I think the
rule makes more sense than ever!

 

[Antony Scriven]

> Dik T. Winter coughed up:
>

> > [About the rejection of news articles with more quoted
> > text than original text.]
> >
> > [Full, lengthy quotation snipped, including incorrect
> > line wrapping by OE.]

>
> [...] To defeat the more quote than substance rule makes
> sense. (I hope we're way past ever needing that silly
> rule again).

For some, perhaps. Might others need encouragement?

Antony