M. Edward (Ed) Borasky wrote:
“grep” standing for “Generalized Regular Expression Processor”, etc.
Being the pedant that I am I have to clarify this statement. The
‘grep’ name came from use in ‘ed’ and similar editors. Of course
ed-like editors were the standard editors at the time that grep was
written. Why were line editors used? Because a screen editor just
does not function well on the paper printing terminals commonly in use
at that time. 
A common operation by anyone who uses a line editor is to print lines
matching a pattern in the file. The way to do this in ‘ed’ is the
following:
$ ed somefile
g/RE/p
q
The ‘q’ quits the editor (in case you try this and need to know how to
get out of the program). The /RE/ is a regular expression pattern.
If the RE matches then invoke the command-list. The ‘g’ is a suffix
to globally operate on every line in the file. Without the ‘g’
(global) suffix the command would operate only on the current line.
The ‘p’ is the command-list item to print.
Putting this all together the command reads, globally operate on every
line of the file, if a line matches the regular expression then print
it. Global Regular Expression Print. grep.
Bob
On Wed, Sep 05, 2007 at 08:31:35AM +0900, Peter C. wrote:
On 9/4/07, Chad P. [email protected] wrote:
I don’t think I’ve ever seen this much wailing and moaning over a handful
of characters in a subject width in my life.
To be fair, you’ve said somewhat more in this thread than anyone else
(providing 15 of 50ish posts). Matz has suggested this is a decision of the
“people” and so would appear to be reliant on general concord. So far I
count about 7 or 8 definite nays and 3 or 4 ayes.
To be fair . . . you’re about right.
I just wish the “nay” answers actually provided answers that amounted to
more than hand-waving and insults. (I generalize, of course.)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Chad P. wrote:
(or even [R], now that I think about it – though
maybe that should be reserved for the R language, which makes me think
[RB] might be more appropriate to this list).
Actually, there are several R lists, and they all use subject line
tagging, for example, [R-help]. I don’t believe they’ve co-opted [R]
yet, but the main R developers’ list is [Rd].
that going on around here.
Well … it’s at least a unique thread, rather than yet another
discussion of IDEs and editors, yet another person who wishes floating
point arithmetic followed the same axioms as real numbers, or worse yet,
several pages of incomprehensible regular expressions that I’m supposed
to debug. 
Which reminds me … I keep meaning to ask … how did regular
expressions morph from an undergraduate computer science concept into a
“language” that every working programmer is expected to be able to
read and write in his or her sleep? Let’s debate that! 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFG3eHn8fKMegVjSM8RAjaPAKCSyBdIemxG5ei8I2/DMVm/hcu8fgCgsdL9
apfPpFGYZsXKUsLU6pqPt7A=
=h8kh
-----END PGP SIGNATURE-----
James Edward G. II wrote:
I watch over the newsgroup gateway and I can say that it would not be
affected. It uses the headers to do it’s magic, because I feel that’s
more reliable. The gateway existed back when the numbers were still
around and it could do so again, though it has been rewritten since then.
I cannot speak for the other forums, of course.
James Edward G. II
I notice that some messages are still not being propagated properly
through the gateway. Is this still a function of the message type? I
think I remember something about HTML causing problems. For example,
the first message of this thread was not propagated, just the reply from
Matz.
On 05/09/07, Peter C. [email protected] wrote:
So far I count about 7 or 8 definite nays and 3 or 4 ayes.
Add my definite nay, for the reasons stated elsewhere in the thread.
Roy.
I’ve got an idea. Forgive me if this issue has been resolved, or if
this
idea has been suggested before. This thread is really long and mostly
irrelevant to the initial question, so I haven’t read it all. But
here’s
what I understand:
- The main argument against prepending [RUBY] is that it pushes the
subject line a few characters to the right.
- The main argument for it is that a mail filter – a computer, not a
human – can identify these emails by the subject line.
So why not append [RUBY] to the subject? That way, folks with skinny
screens won’t lose actual subject content (the only thing that gets
pushed
off the screen is the label), and folks who want to filter by subject
can
still do so.
Unless there’s some technical reason why this won’t work, the only
problem I
see is aesthetic. =P
Altay
On Sep 4, 2007, at 5:40 PM, Chad P. wrote:
The only argument against it that I’ve seen that might convince me
is the
one that relates to forum and newsgroup gateways. Too bad all that
has
happened there is someone said it might be a problem, and nobody has
confirmed it.
I’m done with this no-longer-civil-conversation, but I need to add a
note about the above as it might be relevant to any decision made.
I watch over the newsgroup gateway and I can say that it would not be
affected. It uses the headers to do it’s magic, because I feel
that’s more reliable. The gateway existed back when the numbers were
still around and it could do so again, though it has been rewritten
since then.
I cannot speak for the other forums, of course.
James Edward G. II
M. Edward (Ed) Borasky wrote:
?
Yes … that part of it I know … UNIX as a way for researchers to get
lots of work done with text files, “grep” standing for “Generalized
Regular Expression Processor”, etc. The part I don’t know is how it went
from Kleene, an obscure computational structure equivalent to finite
state machines, an exercise in undergraduate automata theory textbooks,
etc., to a practical tool like “grep”, and I don’t know why it “caught on”.
The first time I used them was in a mainframe editor that used a
superset of them. You could go to the beginning of a file, skip 20
occurrences of a pattern, change the next 50 occurrences of a different
pattern, etc. This was a major improvement when you had to edit a
10,000 line file on a mechanical tty. In today’s multi-core, multi-GHz
personal computers something like this would not be too noticeable time
wise, but on a 60 cps tty the ability to do something like this was a
huge time saver.
On Sep 4, 2007, at 18:31 , Peter C. wrote:
So far I count about 7 or 8 definite nays and 3 or 4 ayes.
You may add another definite “nay” on my behalf.
Michael G.
grzm seespotcode net
From: “M. Edward (Ed) Borasky” [email protected]
Which reminds me … I keep meaning to ask … how did regular
expressions morph from an undergraduate computer science concept into a
“language” that every working programmer is expected to be able to
read and write in his or her sleep? Let’s debate that!
grep, sed, awk, . . . perl, ruby, . . .
?
Regards,
Bill
Chad P. wrote:
I just wish the “nay” answers actually provided answers that amounted to
more than hand-waving and insults. (I generalize, of course.)
I just realized, looking over my mailing list subscripions, that I think
[SOME_LIST] tags are ugly and make it harder to scan over the subjects.
Is this a matter of taste? Sure, but that doesn’t make it any less
valid. For me, this is a stronger reason than losing four to seven
characters (even though I hate it when threads overflow beyond the space
thunderbird allots for the subject).
I hope you see this as a valid reason, though I obviously don’t expect
you to agree.
Dan
On Fri, Sep 07, 2007 at 06:02:48PM +0900, Todd B. wrote:
chooses to not sequester away ruby emails might want to quickly
glance at the screen and visually see which emails are ruby/non-ruby.
That’s my case. I don’t filter by the subject line tag – I use headers
for that (but I generally filter by views rather than by “folders” or
“bins”).
Has anyone who wanted a [RUBY] tag (or equivalent) actually said
anything
about wanting to filter by it? I don’t recall seeing such a thing, so I
imagine that if someone said as much in this thread it must have been
one
of the first few responses (giving me enough time to forget it).
Respectfully, no, unless it’s very short. I suggest sort by the To:
line. If that is trouble, then your mail client needs to change, any
mailing list like this generates enough mail, that you need either a
dedicated e-mail account for the list and/or a mail client capable of
simple sorting rules.
On 9/4/07, Chad P. [email protected] wrote:
I don’t think I’ve ever seen this much wailing and moaning over a handful
of characters in a subject width in my life.
To be fair, you’ve said somewhat more in this thread than anyone else
(providing 15 of 50ish posts). Matz has suggested this is a decision of
the
“people” and so would appear to be reliant on general concord. So far I
count about 7 or 8 definite nays and 3 or 4 ayes.
Regards,
Peter C.
On Wed, Sep 05, 2007 at 10:51:35AM +0900, Dan Z. wrote:
thunderbird allots for the subject).
I hope you see this as a valid reason, though I obviously don’t expect
you to agree.
Actually, I can respect that response, if only because it doesn’t
pretend
to be more than it is.
I vote against. Those with the firehose of ruby-talk gushing into their
singular inbox is probably an e-mail masochist and enjoys it That
said,
whatever the majority decides, although this all rather sounds like
change
for change’s sake.
Cheers,
Peter C.
I vote against, basically because I believe the issue can be solved
differently (and people are actually doing it). Note that the OP is
using a gmail account which has nice filtering capabilities (which I
use as well).
Plus, there is a certain chance that the header mangling conflicts
with various other forums that are interconnected with this mailing
list (notably comp.lang.ruby and the ruby forum). For these forums a
changed subject would be redundant because in a way they are
filtered already. So we would either have to live with those
redundancies or make gateways smarter to remove that tag again. I
don’t feel it’s worth the effort.
Kind regards
robert
M. Edward (Ed) Borasky wrote:
wise, but on a 60 cps tty the ability to do something like this was a
huge time saver.
I remember something like this in the text editors on Xerox CP-V and
earlier systems, but I don’t think they pre-date Unix.
I was using UCEDIT (University of Calgagy EDIT) on a CDC Cyber mainframe
in the late 1970’s. I don’t remember when the program was actually
created as I recently lost the documentation for the program.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Bill K. wrote:
Regards,
Bill
Yes … that part of it I know … UNIX as a way for researchers to get
lots of work done with text files, “grep” standing for “Generalized
Regular Expression Processor”, etc. The part I don’t know is how it went
from Kleene, an obscure computational structure equivalent to finite
state machines, an exercise in undergraduate automata theory textbooks,
etc., to a practical tool like “grep”, and I don’t know why it “caught
on”.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFG3iMf8fKMegVjSM8RArx/AKCqt2vruuQlnredgjsLf4Uhe2WywACgxhp3
QjkcSZ99tPpyHdPcpLLS1T4=
=euON
-----END PGP SIGNATURE-----