emv@msen.com (Edward Vielmetti, moderator) (04/26/91)
Archive-name: archives/admin/administrivia/0-- Original-posting-by: emv@msen.com (Edward Vielmetti, moderator) Original-subject: comp.archives disappears for a while Reposted-by: emv@msen.com (Edward Vielmetti, MSEN) comp.archives will be on hiatus until some time in May; if you're not getting your usual feed of articles, it's because there aren't any being posted. I'll still be screening incoming articles for likely candidates, so nothing will be completely missed; the challenge will be for me to weed though an enormous number of postings to find the good stuff. Here's how you, as someone who wants to have their postings noticed and distributed through comp.archives, can help. - Informative subject headers. They should include the name of the package and the word "available", e.g. Subject: (package) -- (bright shiny wrapped package) available - A magic phrase in the body of the posting. This is what I would prefer, all on one line: (package) is available via anonymous ftp from - A description of the location of the package, in the same unambiguous notation that comp.archives uses, all on one line: ftp.domain.org:/pub/package-1.00.tar.Z - A nice long healthy hunk of text describing what the package does, who worked on it, what other packages it is similar too, maybe a little bit about how it works. About 5K of text is more or less right. Pretend that there's a full text indexer that is going to be storing the text, and lard it with all of the appropriate buzzwords that people in your field would be looking for if they were going to search for it. If you follow these guidelines, then it's a good chance your announcement will be picked out from 1000-odd other random things that will no doubt pile up over time. The other reasonable way to get my attention is to drop a short note to me (emv@msen.com) and let me know when and where you posted something; that should let me find it quickly too. comp.archives is not (as I might have mislead some people into believing!) completely automated! there's a fair amount of work that goes into the article location that's automated, but the tools that disambiguate between "package is available via anonymous ftp from" and "where can i ftp a copy from" and "we need a net connection so we can telnet, ftp, etc" are in their formative stages. there's a database of package to category types that makes the Archive-name production reasonably straightforward; my date parser goofs up every so often, and it's sometimes hard to categorize things. don't know how useful that all is, I try mostly to keep related packages together to the extent that that's even possible. A sticky part is getting the Archive: or Archive-directory: piece right so that people can use comp.archives to drive an automatic ftp fetcher. archie helps if something's been around for a while, but for new stuff it's relatively time-consuming to have to track just exactly where on a site and how it's spelled if the poster gets it wrong. comp.archives in the future... MSEN (as you've noticed from the headers) is slowly becoming the primary place from which I will be producing comp.archives. This moves it out of the realm of "done on company time and they don't mind that much" to "done as a service of a for-profit company with the hope of getting a return on investment". MSEN, Inc. is a small startup company which has as its primary goal not to lose too much money, to put itself on the Internet, and to provide a range of network services. As such I'll be trying to put comp.archives on a self-supporting basis, i.e. charging money to someone somewhere, or getting money from someone somewhere. What that means is that I need a credible threat :-) and to say "if you like comp.archives, support it, because if it's too much of a drain on MSEN's time and my time it'll go away in a year". What exactly that support is is left intentionally vague for the moment because I don't have a good answer. It might mean spin-off products, like putting some extracts of the collection on paper and selling them on late-night TV "THOUSANDS of programs absolutely FREE with your paid Internet subscription!". There's a reasonable chance of providing pay-per-view or subscription access to a nice full-text full-screen browser for the collection; that would make a lot of comp.sources.wanted obsolete, sort of an "archie" for comp.archives. I suppose that I could make comp.archives into "msen.archives", with a Clarinet-like distribution policy, though the techno-legal nonsense of restricting distribution of other people's words puts a sour taste in my mouth. Indeed there are any number of things that would be awful nice to have but which would take a lot of work. *It's rapidly approaching the stage where it's going to take a serious infusion of technology to keep comp.archives running.* I really want to end up with something much more generalizable, something that you could say to it "look for all of the interesting articles about frame relay and whatever related subjects on whatever lists, and just show me the interesting ones", and it would find them for you; the newsreaders of the 1990's have huge piles of news to weed through and only the puniest of newsgroups to split them down with. Thus, I see no choice but to try and force the issue and say "if the technology doesn't get better within a year, I give up, I can't keep up with it all." Technology costs money, at least the technology that I'm aware of. Topic (by Verity) could be a very useful tool for scanning through news. There's some very good full-text searching and X11 based presentation software made by the folks who did the New OED which would be fun to throw at the task. No doubt there are expert systems and lexical analyzers which could chew on articles and spit out just exactly the ones that I would pick, all neatly tagged like I would tag them, and send them to comp.archives. (i wish.) whew. comments should go direct to me (emv@msen.com) or to comp.sources.d, and you should all vote YES on comp.archives.admin (send votes to kent@uunet.uu.net). -- Msen Edward Vielmetti /|--- moderator, comp.archives emv@msen.com "(6) The Plan shall identify how agencies and departments can collaborate to ... expand efforts to improve, document, and evaluate unclassified public-domain software developed by federally-funded researchers and other software, including federally-funded educational and training software; " High-Performance Computing Act of 1991, S. 218