Date: Tue, 18 Jan 1994 14:08:32 -0800 From: mikew@cutthroat To: 590i@cutthroat Subject: Hotlist to HTML I have a shell script that allegedly converts a Mosaic hotlist into an HTML document. I haven't tried it yet -- I just grabbed it from Usenet earlier today -- but if anyone wants to give it a try, it's in: /homes/gws/mikew/www/hotlist2html I wish I'd had it a few days ago when I was manually converting my hotlist into an HTML document... -Mike
To: mikew@cutthroat Subject: Re: Hotlist to HTML Date: Tue, 18 Jan 94 14:11:15 -0800 From: Paul Barton-DavisIn a similar vein, I have a tool I wrote called, appallingly, mlist2html, that converts a mailing list (that is, a file with a userid-per line, no blank lines, etc) into an HTML file complete with pointers to people's home pages. I've been using it a lot. /homes/gws/pauld/www/mlist2html. At some point soon, I'll be setting up a more global area for WWW utilities. Promise. -- paul
Date: Tue, 18 Jan 1994 14:16:44 -0800 (PST) From: Aaron Kaleva PulkkaSubject: Re: Hotlist to HTML To: mikew@cutthroat On Tue, 18 Jan 1994 mikew@cutthroat wrote: > I wish I'd had it a few days ago when I was manually converting my > hotlist into an HTML document... Oddly enough, mailing your hotlist to yourself (from within Mosaic, call up the hotlist and hit the Mail button) will do this. You will need to tack on the , , , though it does add theautomatically and turns all the items into links (see /homes/pulkka/hotlist.html for an example of the result). I'd bet the shell script does something prettier, but... -- - - - - - - - - - - - - - - - - - - -- Aaron Kaleva Pulkka
CSE graduate student MUD aliases: teklon@kobra, kaleva@others HITLab research assistant sci.virtual-worlds & rec.arts.cinema co-moderator ACM chapter treasurer
Date: Wed, 19 Jan 1994 16:07:36 -0800 From: kepart@mako To: bershad@cs.washington.edu Subject: Re: Yesterday's 590I Brian Bershad writes: > Did anybody take any notes from yesterday's 590I? I asked levy what > happened and he said "so much I can't even remember." Some of it's online in the ``brainstorming page'' which is accessible from the 590I master page (http://www/research/projects/ai/590i). If other people could add/elaborate/improve upon what's in there (I know it's really incomplete & the organization isn't the greatest), that would be very welcome.
Date: Wed, 19 Jan 1994 16:52:57 -0800
From: mef@willow (Marc Fiuczynski)
To: 590i@cs
Subject: something to consider for WWW.
I don't have the data to back this up, but my guess is that there are
a reasonable number of WWW servers that all of us look at. If the
data on these servers changes slowly a simple caching scheme might buy
us alot in throughput and reduce latency.
Here is a blurb regarding WWFS that I found on comp.sources.misc.
------------------------------------------------------------------------
We are pleased to announce the availability of WWFS (WorldWide File
System, or Waku-Waku File System), a massively distributed file system
built upon existing anonymous ftp services.
The key component of WWFS is a daemon called "csd": a multi-protocol
hub for distributed filesystem, an NFS-FTP gateway software,
the Internet file cache, or the substrate for next-generation digital
media.
This research project has been funded and supported in part by NAIST
and WIDE Project.
Major features of WWFS:
- Save your time and network bandwidth -- shared, persistent file cache.
- Patch to add direct WWFS support in NCSA Mosaic for X (2.0 and 1.2).
- Server selection by geographical / topological region.
- World gets closer to you -- NFS-FTP gateway.
- High portability -- DEC Ultrix, OSF/1, SunOS 4, NeXT, BSD/386,
OMRON Luna2, Sony NEWS-OS 4, and MIPS RISC/os.
- No kernel modification.
- Migration tools from HTTP and Internet Gopher Protocol.
- Location-independent, uniform name space.
- Server can be restarted anytime.
- Server switch upon network failure.
- Freely re-distributable.
WWFS has been deployed campus-wide, 24-hour, and non-stop, at Nara
Institute of Science and Technology; OMRON Corporation has been
serving multi-level caching FTP server with WWFS.
------------------------------------------------------------------------
Marc
Date: Sun, 23 Jan 1994 15:51:13 -0800 From: voelker@edgar To: 590i@cs Subject: server side include files I sent out a message (that I should have sent to 590i) asking how to get executables invoked when html links are traversed, and Terri sent a good response that I've included below. I'm still having trouble getting the server include mechanism to work with the current web server, though, and I'm guessing that this server option has been turned off on the department's server. (Another reason why I'm guessing this is the question at the end of the /cse/lab/projects/www/issues file, which asks "How do we let people get access to the server's ability to run their own gateway scripts?") Does anyone know if the option has been turned off? -geoff From: Terri WatsonSender: Terri Watson To: voelker@cs.washington.edu cc: bershad@cs.washington.edu, savage@cs.washington.edu, zahorjan@cs.washington.edu, levy@cs.washington.edu, lazowska@cs.washington.edu, gaetano@cs.washington.edu, mef@cs.washington.edu, forman@cs.washington.edu Subject: Re: invoking scripts through html links Date: Sun, 23 Jan 1994 15:28:39 PST In message <199401232315.PAA22276@edgar.cs.washington.edu> you write: > >Do any of you know how to have a script/executable invoked when an >html link is traversed (when a user clicks on a link)? Sorry, meant to post this sooner: Command in html file is: An example from my personal page is: Location
------- Another example, where the URL is a script:
Send a message to Terri
calls the script zsend which currently looks like: #!/bin/sh echo Content-type: text/html echo cat << EOMSend a zephyr message to Terri Send a zephyr message to Terri
Date: Sun, 23 Jan 1994 16:35:25 -0800 From: pauld@stowe (Paul Barton-Davis) To: voelker@stowe Subject: server side includes You are right to suppose that this option has been turned off. There are, as I mentioned last week in 590i, a number of rather difficult issues associated with the use of URL's that refer to the output of some executable file. The HTTP protocol makes it possible to supply not only arguments but standard input to these executables, and thus opens up the potential for all kinds of security issues that I would love to find solutions for. I don't want to make any great claims about my programming ability, but it seems reasonable to assume that its indicative of many people within the department. My second attempt at writing a server-side script contained an unbelievably stupid and extremely large security hole. There isn't much a precedent for this kind of system, in which random, unidentified remote clients cause a potentially large set of local executables to be run, with arguments and input. Thats why for now, nobody except me has the ability to install this kind of script, and I'm being rather slow (read: cautious) about using it. I sincerely would like to resolve these issues, both because I don't want to stifle people's creativity, and because I don't want to create a motivation for multiple HTTP servers within the cs.washington.edu domain. So, if anyone has any good ideas, please pass them around. -- paul
Date: Wed, 2 Feb 1994 15:54:18 -0800 From: pauld@stowe (Paul Barton-Davis) To: 590i@stowe Subject: SEW: the dumb approach to handling the web Don't ask me why its called SEW, because I don't know. You can, however, read about how to use it: http://www.cs.washington.edu/homes/pauld/sew.html The software is all in the 590i area, so you are welcome to use it. The current implementation is not particularly speedy, not totally suprising given its existence as a suitably complex awk script, but if anyone wants to write a C version ... be my guest. Remember, SEW relies on your social skills as much as my finess with awk: if you don't know anyone who sends you URL references, then SEW isn't going to help you at all. With that in mind, I invite you all to send me URL references - just make sure the body of the messages looks like this: URL: the-url-of-the-document Title: the-title-of-the-document descriptive text possibly lasting several lines It would nice if the subject said "URL reference" too. In the next few days, /usr/local/bin will contain a mosaic that will make this latter step easy. PLEASE, PLEASE send me bug reports. Also, if you figure out cool optimizations for the awk script, let me hear about them as well. -- paul
Date: Thu, 27 Jan 1994 12:01:41 -0800 From: pauld@stowe (Paul Barton-Davis) To: 590i@stowe, www-folk@stowe Subject: server-side executables The HTTP server on www.cs.washington.edu is now configured to allow people to try their hand at providing server-side executables (i.e. when someone using a HTTP client follows a link, it causes a program to be executed, and the output of that program to be sent back to the client). Your programs must reside in /projects/ai/590i/post-bin. See the 590i page for a pointer to a document describing how to do this. Please pay attention to the security issues and guidelines described there. Access to these programs is restricted to CS department machines only. -- paul
Date: Wed, 2 Feb 1994 16:14:56 -0800 From: pauld@stowe (Paul Barton-Davis) To: 590i@stowe Subject: while on the subject of SEWing ... ... I should mention /projects/ai/590i/pauld/murl. This is for use from your favorite rn-like newsreader, wherein you type (from the article selection level): | murl pauld excellent article on throbbing gristle and it will promptly send a SEW-compatible message to "pauld", that looks like: To: pauld Subject: URL reference URL: news:message-id-of-news-article Title: news article from newsgroup-article-was-posted-to excellent article on throbbing gristle -- paul "no apologies for unix names" barton-davis
Date: Tue, 8 Feb 1994 14:13:56 -0800 (PST) From: Mike Perkowitz
Date: Tue, 8 Feb 1994 14:14:30 -0800 (PST) From: Mike Perkowitz
Date: Tue, 8 Feb 1994 14:17:52 -0800 (PST) From: Mike Perkowitz
Date: Mon, 14 Feb 1994 11:50:56 -0800 (PST) From: Mike Perkowitz
Date: Mon, 14 Feb 1994 14:39:20 -0800 From: pauld@stowe (Paul Barton-Davis) To: 590i@stowe Subject: Mike's sew index, part II So, I really liked this idea, but not the details. Now, I just do this: Incoming Suggestions to create a link to a script that will create the index for me on the fly. Why keep it around in a file if it takes 2-3 seconds to create it ? I kept Mike's file layout (~/www/incoming/*[-index].html), and tried to show my novicehood when it comes to perl. Details in 590i/post-bin/sewindex. Obviously, for your own personal use, change the query string from ?pauld to ?yourname, whatever yourname is. -- paul
Date: Mon, 14 Feb 1994 15:02:57 -0800 (PST) From: Mike Perkowitz