pat@kla (02/11/90)
--- NOTE: This article has been cross-posted. If you post a follow-up,
--- PLEASE restrict it to the appropriate group. -Thanks.
I have just posted to alt.sources an improved arbitron program written
in perl. I am considering this to be a beta version until I hear
positive results from other sites.
I will gladly e-mail copies to anyone who does not get alt.sources.
Some of the advantages of the perl arbitron.
1) Greater portability.
Perl runs on a lot more systems than you can find compatible
versions of Bourne shell, sed, awk, etc.. (i.e. AmigaDOS)
2) More accurate reporting if the local expiration rate is very high.
The shell script compared .newsrc contents against the active
file. This version keeps an optional arb.last file to compare
against.
3) No temporary files.
4) Can be run on systems which do not support multi-tasking.
5) If your mailer allows the specification of the subject as
a command line parameter, this program will generate a
subject line of the form `arbitron data for MonYYYY'.
6) Direct connection to NNTP socket in distributed environment.
(Perl 3.0 is required for this option.)
7) Much easier to read.
(Unless you are an awk/sed/sh wizard :-)
Disadvantages:
1) This version is slower than the Bourne shell version.
(about 50% on a Sun3/60 running SunOS 3.5)
Other differences:
1) Obtaining count of valid users:
The script defaults to accepting all users within the specified
range of userids; but provides optional mechanisms for counting
actual userids stored in the wtmp(5) file.
This program provides a generalized filter capability which can
easily check the value of any of the fields from the password
file. It was felt that this is a more useful mechanism in
a networked environment.
The current version does not have support for nn history files or
notesfiles because I have no idea what the filenames or formats are,
and I'm not about to attempt to figure it out from the existing
arbitron scripts. If someone will mail me that information, I will
gladly include support in future versions. (My goal is a single,
readable, arbitron script suitable for any site which can support both
news and perl.)
-Pat
Copyright (c) 1989 PM Lashley under the terms of the GNU General Public License
PMLashley ...{sun | megatest | sts | zygot}!cohesive!kla!pat
<<< I haven't lost my mind. It's backed up on tape somewhere... >>>sean@ms.uky.edu (Sean Casey) (02/13/90)
Does arbitron support nn's datafile format yet? There is a huge base of nn users out there, including me. As more and more people convert, the statistic are going to have more and more error introduced. It's something you might want to think about. Sean -- *** Sean Casey sean@ms.uky.edu, sean@ukma.bitnet, ukma!sean *** "May I take this opportunity of emphasizing that there is no cannibalism *** in the British Navy. Absolutely none, and when I say none, I mean there *** is a certain amount, more than we are prepared to admit." -MP
nagel@ics.uci.edu (Mark Nagel) (02/13/90)
sean@ms.uky.edu (Sean Casey) writes: >Does arbitron support nn's datafile format yet? >There is a huge base of nn users out there, including me. As more and more >people convert, the statistic are going to have more and more error >introduced. I wouldn't worry about it. Kim Storm is preparing to release version 6.4 of nn and it will be using the .newsrc format only. This should happen within a month or so. In addition, if you configure the default init file properly (set newsrc, I think), you can have a .newsrc which mostly reflects the reading habits of nn users. -- Mark Nagel UC Irvine Depertment of ICS +----------------------------------------+ ARPA: nagel@ics.uci.edu | If you improve something long enough | UUCP: ucbvax!ucivax!nagel | eventually you will throw it away. |
tchrist@convex.COM (Tom Christiansen) (02/13/90)
In article <14116@s.ms.uky.edu> sean@ms.uky.edu (Sean Casey) writes: >Does arbitron support nn's datafile format yet? > >There is a huge base of nn users out there, including me. As more and more >people convert, the statistic are going to have more and more error >introduced. > >It's something you might want to think about. Johan Vromans's version of the arbitron knew about nn users. You meantion more and more people switching over the NN. I know this isn't the right forum for this question, but I need to ask anyway. What features made you convert? The speed of the local header cache? The subject selection menu? I once tried to convert from rn, and was not happy and went back. When I ask nn'ers why they like it, they say "it's easier than rn". When I ask them how you do some foo function in nn that I often perform in rn, they wonder why I would ever want to do that (like looking for the parent article according to the references line). I figure that since Larry Wall is sure to see this, he might see some responses to put away in his rn to-be-done list. Redirecting followups to a more appropriate newsgroup (if you can find one) would be welcome. --tom -- Tom Christiansen {uunet,uiucdcs,sun}!convex!tchrist Convex Computer Corporation tchrist@convex.COM "EMACS belongs in <sys/errno.h>: Editor too big!"
sean@ms.uky.edu (Sean Casey) (02/13/90)
tchrist@convex.COM (Tom Christiansen) writes: |You meantion more and more people switching over the NN. I know |this isn't the right forum for this question, but I need to ask |anyway. What features made you convert? I moved to nn after 4+ years of in-depth rn use. I like nn because one can very rapidly scan the list of subjects and select threads that are interesting. The thing I dislike most about rn, it's sequential nature, slows me down a great deal. Suppose there are 60 subject threads, and I will be reading three. We'll assume this is after any kill files have taken effect. My nn session might go like this: I see the first 20 or so threads on the first screen for this newsgroup. "Hmmm that one looks good." and I type its letter. -Space- I see page two and there's the other one. I type its letter. Page 3,4 and 5. None there. After a few spacebars I'm reading articles. Ok so I've quickly scanned 5 screens in addition to reading the articles. With rn, in addition to reading the articles, I gotta see 57 extra headers and type "k" to each one them. That wastes a lot of my time. What nn does best is help you to not read things you don't want. The slogan "No news is good news." is wll thought. Remember when you could read and keep up with all the newsgroups? Now there is so much stuff that nn saves me huge amounts of time by not reading it :-). I also think nn feels a lot faster than rn. It starts faster, and everything it does (at least on our system) seems to happen instantly. Remember how *wonderfully* fast rn was compared to readnews? That's how I think nn feels next to rn. Of course, everyone has to make their own choice. Readers might want to give nn a good test drive just to get exposure to some alternate software. Use whatever works best for you, and enjoy reading Usenet. Sean -- *** Sean Casey sean@ms.uky.edu, sean@ukma.bitnet, ukma!sean *** "May I take this opportunity of emphasizing that there is no cannibalism *** in the British Navy. Absolutely none, and when I say none, I mean there *** is a certain amount, more than we are prepared to admit." -MP
frotz@drivax.UUCP (Frotz) (02/16/90)
tchrist@convex.COM (Tom Christiansen) writes:
] What features made you convert? The speed of the local
] header cache? The subject selection menu? I once tried to convert
] from rn, and was not happy and went back. When I ask nn'ers why
] they like it, they say "it's easier than rn".
As other nn readers will no doubt attest, the subject selection mode
is the one big advantage of nn over rn. My time is limited and
therefore I can't read as much or as often as I would like to. Wayne
Davison maintains our local copy of nn and has sent patches to Kim
regarding better subject/thread selection and organization. If rn
were to incorporate this subject selection with a more visual
interface I might convert, but as it is, I have little spare time to
wade through groups like comp.sys.ibm.pc looking for the few (average
3-5) articles that I really want to read. Besides, it helps people
use better Subject/Summary lines. If I am not interested at the
subject level, then I don't have time to be interested in the article.
--
Frotz
#include <disclaimer/you_asked.h>