Thread View: comp.archives.admin
4 messages
4 total messages
Started by Nathan Torkington
Tue, 04 Jun 1991 11:09
Description programs
Author: Nathan Torkington
Date: Tue, 04 Jun 1991 11:09
Date: Tue, 04 Jun 1991 11:09
117 lines
5537 bytes
5537 bytes
What programs are people using to index their archives? I am currently using "Describe" by Tim Cook. I will include the README after my .sig, and I believe it is ftp-able from admin.viccol.edu.au - it is for Unix. It is a good program, and gives excellent results - even if it required a little hacking to get to work on our platform :-) It allows you to store a text string in a hidden dbm file and then use 'dls' to give the descriptions in ls form. Heck, read the README and find out what it does ... What are other people using? [Please note, you happy FTP-ers, that the archive site I maintain does *NOT* want overseas FTPing. We have nothing which cannot be gotten faster and easier from America or other sites - besides which, we pay for overseas traffic. It is for internal New Zealand use only ...] Nat. -- Nathan Torkington Contracting to CSC, Internet: gnat@kauri.vuw.ac.nz Victoria University Of Wellington, Vocal Abuse: +64 4 853 881 (post-chronos) P.O. Box 600, New Zealand. This is not an official communication of any part of Victoria University. --README FOLLOWS-- Description ----------- Have you ever wondered what on earth "xfrg2.2.tar.Z" was? Are you wondering whether "foo" is really a scratch file? Does it ashame you especially, seeing that these files were created by you? I have the solution for you... File Descriptions. I'm sure some operating system or another once had the ability to set a descriptive comment on a file, but Unix wasn't one of them. It is possible, however, to retro-fit onto Unix an implementation that works quite well. In this package, I have provided a utility to set/delete/list file descriptions (describe), and a directory listing utility that shows any file descriptions that are set next to the respective files (dls). Implementation -------------- The storage/retrieval of file-descriptions is implemented in the following way: All file descriptions are stored in a DBM file called .desc, in the same directory as the files described. This in fact means that you have two files, .desc.pag and .desc.dir. Each description is indexed by the name of the file (excluding any path information). To provide a better chance of finding the description after the file has been renamed (within the same directory), the name is also indexed by the inode-number. This means that the file's description will still be found after it has been renamed (within the same directory) or re-written by an editor. To prevent the two indexes from becoming mixed up, whenever a name/description pair are stored with a name that is the same size as an inode-number, the name is stored with a trailing null byte. Purpose ------- My reason for coming up with this idea is simple. I got tired of logging on to some distant site's anonymous FTP area, and coming across something like: -rw-r--r-- 1 root 214807 May 22 1989 3c503.tar.Z -rw-r--r-- 1 bin 27751 Aug 22 16:35 bonnie -rw-rw-r-- 1 uucp 65383 Aug 23 02:54 dynafeed.tar.Z -rw-r--r-- 1 bin 10444 Nov 3 21:58 fpipe -rw-r--r-- 1 bin 47138 Aug 22 16:41 opaque (good name!) -rw-r--r-- 1 don 683759 May 29 23:21 psiber.tar.Z You know? You sit there and think ... "some of those sound interesting, but it wouldn't be nice for me to download them just to find out what they are...". If there is no other demand for describe/dls, there should be as a replacement for ls in anonymous FTP areas. And that's how simple it is, once you have dls installed, simply copy it to ~ftp/bin/ls, and the FTP daemon will use it instead of standard ls for anonymous logins. The output of dls was actually designed with anonymous FTP use in mind (who cares who owns the file or which group it belongs to?). Of course, the site administrator has to provide meaningful descriptions for what has been put in the anonymous FTP area. I don't think it would be too hard to generate descriptions files from an index of something like comp.sources.unix, though. A dozen lines of perl should do it. An Example ---------- As an example of what an anonymous FTP site that uses dls looks like, try my back yard, admin.viccol.edu.au. Copyright --------- The whole system is Copyright (c) Tim Cook 1991, apart from a courtesy copy of strpbrk.c which is Copyright (c) 1985 Regents of the University of California. All together, the package can be distributed provided no profit is made from any distribution, and all copyright notices remain intact. The Author ---------- Tim Cook, Systems Administrator, Victoria College Computer Services. I am the administrator of a Sequent S27 (Symmetry), admin.viccol.edu.au, which runs administrative systems under Oracle. Could I ask all those who install dls in their Anonymous FTP areas to drop me a line? Any criticisms, suggestions gladly accepted. I am quite interested in what people think of this idea.
Re: Description programs
Author: David Bremner
Date: Tue, 04 Jun 1991 17:31
Date: Tue, 04 Jun 1991 17:31
19 lines
1436 bytes
1436 bytes
In a related vein, I have started keeping bits I grab from comp.archives in a full text database ( Liam Quin's lq-text ). With the compressed file support I hacked in, this takes about the same amount of space as storing an uncompressed copy of the text. I think that something like this could provide a very useful service; like archie's new whatis facility, but more powerful. In case your wondering, I'm not ready to make my database available; primarily I don't have the resources, also, my complete database is only a couple of megabytes; I'm pretty selective. However, with this caveat, I might be interested in helping get something like this set up somewhere. What do people think, is it worth the trouble to be able to type "news reader" and pull up the readmes for a bunch of news readers? David -- bremner@cs.sfu.ca ubc-cs!fornax!bremner
Re: Description programs
Author: Albert Langer
Date: Wed, 05 Jun 1991 12:37
Date: Wed, 05 Jun 1991 12:37
13 lines
1115 bytes
1115 bytes
In article <1991Jun4.173107.4964@cs.sfu.ca> bremner@cs.sfu.ca (David Bremner) writes: >What do people think, is it worth the trouble to be able to type "news >reader" and pull up the readmes for a bunch of news readers? YES YES YES PLEASE! (Could you make available your mods to lq-text and/or scripts for inputting data from comp.archives etc?) -- Opinions disclaimed (Authoritative answer from opinion server) Header reply address wrong. Use cmf851@csc2.anu.edu.au
Re: Description programs -- lq-text
Author: Liam R. E. Quin
Date: Sun, 09 Jun 1991 00:05
Date: Sun, 09 Jun 1991 00:05
51 lines
2889 bytes
2889 bytes
bremner@cs.sfu.ca (David Bremner) writes:
> What do people think, is it worth the trouble to be able to type "news
> reader" and pull up the readmes for a bunch of news readers?
cmf851@anu.oz.au (Albert Langer) enthuses:
> YES YES YES PLEASE!
>(Could you make available your mods to lq-text and/or scripts for
>inputting data from comp.archives etc?)
It seems only fair to David Bremner to say that I have his changes to my
lq-text, and you can get them from me if you prefer. I certainy prefer (!),
because I find it useful to know who's using lq-text and what they're doing
with it, as this helps me to decide what to do next...
The next release (the current version is 1.10) *will* include support for
externally compressed files and for archives. I am also hoping to include
the shared (distributed) wordlist server and the X11/xview front end, as well
as a number of contributed programs.
Having said all that, I've run out of words. No, that's not what I wanted
to say (he blathered wildy), sorry!
I'm also hoping to get some better result-sorting (`ranking' in the literature)
into the next release. As with compression, the hooks have been there for
ages, and I have experimented with both these things already.
For comp.archives I can see that one might want to add a file at a time --
that's fine -- and to rename files from time to time as they're archived.
So I'll have to have the long-awaited lqadmin stuff as well...
I've also done a bit more work on making installation easier, as although it
is already easy for programmers and people who have ported usenet software,
a number of others have had difficulty forgetting to change things. So the
new defaults either configure themselves or are ludicrously wrong so you
*have* to cange them.
Note: You can get lq-text by anonymous ftp from ftp.cs.toronto.edu, in the
pub/lq-text directory. Check the version number in the filename -- the
last version I put there was: lq-text1.10.tar.Z
The README file is included in the tar.
Note to Ed: there ought to be some way of marking a news article as
explicitly containing a pointer to an archive (as this one does), and also
a way of explicitly marking an article as not intended for comp.archives even
though it contains trigger-phrases...
Liam
--
Liam Quin, lee@sq.com, SoftQuad, Toronto, +1 416 963 8337
the barefoot programmer
Thread Navigation
This is a paginated view of messages in the thread with full content displayed inline.
Messages are displayed in chronological order, with the original post highlighted in green.
Use pagination controls to navigate through all messages in large threads.
Back to All Threads