Paul Holbrook's Radio Weblog : Worth $40 a year? You decide ..
Updated: 4/8/2003; 8:54:40 PM.

 

Subscribe to "Paul Holbrook's Radio Weblog" in Radio UserLand.

Click to see the XML version of this web page.

Click here to send an email to the editor of this weblog.

 
 

Thursday, April 11, 2002
DECtapes - ^C, and wait for the rewind ...

The item from Slashdot about old computers reminds me of one of the first systems I was ever around: a PDP-8/E running OS/8 off a DECtape.  The OS and all the applications were stored on the DECtape, which was this little (4") tape that had a file system stored on, much like a floppy disk. 

What I remember about it was hitting control-c to interupt whatever program was running, and then waiting for the tape to rewind back to the beginning to reload the command prompt.

We're all spoiled now.


11:24:57 PM      comment []
Old system simulators and old media

The Computer History Simulation Project [Slashdot: News for nerds, stuff that matters]  "The Computer History Simulation Project is a loose Internet-based collective of people interested in restoring historically significant computer hardware and software systems by simulation."

This is an interesting idea.  Of course, even if you've got a simulator, you need to get the programs themselves that ran on these machines, and that means dealing with all kinds of now-ancient media.

When I was going to college back at UC Irvine '78-'82, we used a system called the Terak that used USCD Pascal as the OS.  It's barely possible that I still have programs for the Terak down in the basement someplace - but I haven't seen an 8" floppy disk drive in years, so I don't know how I'd ever get them back again.

About the only thing I've been able to keep moving forward is email.  When I left Xerox in 1989, I took copies of some of my mail files.  I have no idea what form I took them in then, but ever since then, they've ended up as files in various /home directories on various UNIX and Linux systems, and I've been able to move them from one system to another over the Net.   By now I must have those same files on multiple ZIP disks and now CDs. It's like storing things on a boomarang - throw it out into the net for long enough (but not too long) and then repackage it when it comes back around again.


11:19:12 PM      comment []
s l am : site for sysadmins
About s l a m.... A site for sysadmins of centralized sites.  The author admits that his site was outed beore he was ready for it.
My original idea for developing this site was to keep well below radar coverage, that is until I was ready to launch. Dave made me modify my plans in a hurry yesterday morning. Well, it's nice to have company :-) For readers who may be wondering what I'm up to, here is a short explanation. No doubt it will change over time, so I've placed a link in the home section of s l a m 's sidebar. [s l a m]

7:38:03 PM      comment []
Library Cats!

Library cats in the United States.  [Scripting News]

Anyone who knows me would know that that I had to post this item.


7:19:18 PM      comment []
Limiting RSS summaries to the first paragraph

Following up on Jon Udell's discussion of reducing the size of RSS summaries, and later Dave Winer's implementation of a callback to make this possible, and then my note on CNN.com's practice of using the first 'graph as a summary, I believe I've made it all work.

As Dave suggested, I defined my own callback in weblogData.callbacks.rssFilterDescription:

on firstPara(description, adrpost) {
 return (html.getOneTagValue(description,"P"))}
 

The description is stored in HTML, so getOneTagValue conveniently does exactly what I want.

Of course, if the first tag in my description isn't a P, this won't do what I want it to do.  I'll work on that another time. 

Side note on figuring this out: I spent what seemed like hours trying to make this work with regex.split (which basically worked if you split at </P>) and regex.extract (which never seemed to work.)  Extract was particularly annoying: I had a pattern that looked something like "<P>.*</P>".  Although the documentation of regex.extract suggested that the * would grab the minimum number of characters to make a match work, the * in fact seemed to be greedy: it appeard to match the entire description, which of course starts with a <P> (for the first 'graph) and ends with a </P> (which is actually part of the last 'graph, not the first.)

5:52:16 PM      comment []
RSS summaries: lead sentence or paragraph? (and how CNN.com did it.)

Shortening RSS descriptions to lead sentences. I'm really enjoying my ability to scan a lot of sources in my Radio news aggregator. [Jon's Radio]

Jon Udell writes about how he hacked the code that writes his RSS file to include only the first sentence instead of the entire item.  He contrasts this to another approach he's seen that cuts off the description after an arbritrary number of characters.  (I've seen that one, and I don't like it.)

I'd rather have it another way.  CNN.com has a product they referred to as "Quick News."  It's a syndication product, suitable for sending out to WAP phones, airplane seatbacks, or any other place that might want short CNN headlines.  Up until early last year, the Quick News product was produced by a separate staff who re-wrote a certain number of CNN.com stories into this shorter format.  This was deemed as not the most efficient use of staff, so instead they sought to modify the CNN.com internally-built CMS tools to take the first paragraph and turn that into a quick news story.   (They also had to train staff to make sure the first 'graph was something that could be syndicated.) 

I think that approach would work very well here.  If you want a single sentence as your description, great; if you need a couple of sentences, that would work, too.


10:22:46 AM      comment []

© Copyright 2003 Paul Holbrook.



Click here to visit the Radio UserLand website.

 


April 2002
Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30        
Mar   May


 4/7/03
 1/29/03
 1/26/03
 1/23/03
 1/23/03
 1/23/03
 1/21/03
 1/21/03
 1/16/03
 1/16/03
 1/15/03
 1/14/03
 1/14/03
 1/14/03
 1/5/03
 1/5/03
 1/3/03
 1/2/03
 1/2/03
 1/1/03
 12/30/02
 12/28/02
 12/26/02
 12/24/02
 10/8/02
 10/8/02

Home Page