overby@plains.UUCP (Glen Overby) (08/03/90)
In article <1265@sirius.ucs.adelaide.edu.au> cagney@cs.adelaide.edu.au (Andrew Cagney - aka Noid) writes: >The'll [FP package] be automatically ftp'd to plains.nodak.edu (USA) shortly. they have been (ah, it's so nice to have daemons to do my work for me). >Glen (Overby), do you already have the fp (floating point) stuff on >plains.nodak? Its not really an `Australian' product. This is an organization question. As it happens, I do not have a copy of that particular package on Plains. I have found it difficult to organize things posted to comp.os.minix, because it is not always obvious what runs on what machines (and this situation is going to get worse when we get Mac and Amiga people on!). So far, I've attempted to stuff most things in the "all.contrib" for apparently portable programs, "pc.contrib" for things related to Intel brain-dammage, and "st-contrib" for things related to the Space-Invaders company's machines. I'd like to propose a posting standard. Namely, that used by many other Usenet groups (most of which are moderated). These groups add a couple lines at the beginning of the posting containing, for example: X-Checksum-Snefru: e80a5aa3 4d17bf71 233d6222 ad37b7d8 Submitted-by: "Glen Overby" <rtfm@blizzard.ndsu.nodak.edu> Posting-number: Volume 22, Issue 112 Archive-name: {pc,st,mac,amiga,68K,8086}/some_name/part01 The "snefru" program is some checsum program that Rich $alz uses on Comp.sources.unix; maybe we shouldn't bother with it. The Posting-number line is inappropriate since this group isn't moderated and sequenced. The important two lines are the "Submitted-by:" and "Archive-name:" lines. If would put a {machine/cpu-name}/program-name/part-number on the "Archive-name:" line, people who use and save this stuff can use a program such as "rkive" to save the articles in an appropriate place. I know I'm probably just blowing a lot of hot air; 10% of the people posting to the group will take my suggestion and that few won't make much difference. -- Glen Overby <overby@plains.nodak.edu> uunet!plains!overby (UUCP) overby@plains (Bitnet)
cagney@chook.ua.oz (Andrew Cagney - aka Noid) (08/04/90)
From article <5389@plains.UUCP>, by overby@plains.UUCP (Glen Overby):
> I'd like to propose a posting standard....
[Good proposal followed]
Any way, I have a further proposal relating to the management of the archives.
At present when someone posts an article to the net it is dropped into an
archive as is. When patches come round they are dropped into the same
directory. The poor end user is then left to fetch them & apply them in
succession :-)
For the australian archive I've been trying to avoid doing this. If a major
local product is posted I contact the poster and organize a compressed
tar file of what was posted. If the originator posts an update I get them
to send to me (using ftp), for the archive, an updated posting of their work.
By doing this I am managing to keep the archive tidy with out too much work by
the originator and me much while there is greater convenience for the any one
that uses the archive.
Some notes are:
- This is not suitable for posters (sp?) who do not have ftp access
- It's probably ok for small archives (like oz) but not so usefull
for large archives.
- The poster gets a little extra work
- The archiver may have a little less work :-)
- Who is responsible for postings that get taken over?
Any comments on this?
Andrew Cagneycagney@chook.ua.oz (Andrew Cagney - aka Noid) (08/04/90)
From article <5389@plains.UUCP>, by overby@plains.UUCP (Glen Overby):
> I'd like to propose a posting standard....
If everyone adopts it you'll put me out of a job :-)
Andrew Cagneyhyc@math.lsa.umich.edu (Howard Chu) (08/04/90)
In article <1280@sirius.ucs.adelaide.edu.au> cagney@chook.ua.oz (Andrew Cagney - aka Noid) writes: >At present when someone posts an article to the net it is dropped into an >archive as is. When patches come round they are dropped into the same >directory. The poor end user is then left to fetch them & apply them in >succession :-) Yep, kind of a drag. I know that the version of arc archived on uunet doesn't even have all of my updates in the directory, it just has my original posting. >For the australian archive I've been trying to avoid doing this. If a major >local product is posted I contact the poster and organize a compressed >tar file of what was posted. If the originator posts an update I get them >to send to me (using ftp), for the archive, an updated posting of their work. >By doing this I am managing to keep the archive tidy with out too much work by >the originator and me much while there is greater convenience for the any one >that uses the archive. I try to do this with the atari archive here as well, sort of. We do all the work, and don't usually keep in contact with the poster. When a shar file is posted to one of the atari program groups, I repackage it into an arc file, usually. I decode all uuencoded postings, as well. (Don't need uuencode for an ftp service, and our mail server automatically uuencodes files on the fly.) Upgrades to source postings tend to be complete replacements, so the issue of applying patches to archive files doesn't usually arise. >Some notes are: > - This is not suitable for posters (sp?) who do not have ftp access It can work fine if the poster mails cdiffs or somesuch. Just puts more load on the archivers... > - It's probably ok for small archives (like oz) but not so usefull > for large archives. Hey, going thru this effort saves space, at least. And a lot of time for people using the archive. > - The poster gets a little extra work > - The archiver may have a little less work :-) uh... right. }-) > - Who is responsible for postings that get taken over? Taken over? Whoever did the takeover, presumably. -- -- Howard Chu @ University of Michigan one million data bits stored on a chip, one million bits per chip if one of those data bits happens to flip, one million data bits stored on the chip...