I must not post about that blog. I must check my spelling. I must not post about that blog. I must check my spelling. I must not post about that blog. I must check my spelling. I must not post about that blog. I must check my spelling. I must not post about that blog. I must check my spelling.

It's time to talk about Yahoogroups. YG is the largest of a number of hosted mailing lists systems. Between them they've successfully taken all the pain out of starting and running an email mailing list. They handle subscriptions; bounces; they have a searchable web archive and so on. They've added some extra features like calendar, shared file store, a simple dbms. All the things that you can cobble together with Majordomo, Listserv, Mailman. But done moderately well in one place.

This has been so successful that there are a vast number of mailing lists hosted in these places. For many of us the first thing we do on starting a new project is to start another YG.

The price we paid for all this was a small amount of advertizing. And in the past you could pay a little ($50?) per year to have this removed. but in the last couple of months, Yahoo have bumped up the advertizing levels to the point where it's getting really intrusive. There are pop-under windows and adverts inserted every 5th page refresh when viewing messages. The advert appear at the top on send only lists. And as far as I can tell, you are no longer able to pay to have them removed from the emails.

Then there's the question of archives. Already there are archives for the old eGroups that have been lost as the list folded before the Yahoo merger. There is no mechanism now for backing up a YG list. And there is a vast amount of knowledge and thought wrapped up in these archives. If YG ever disappeared (not that there's any suggestion that it's going to, tomorrow) this would all be lost. Do we think that YG will last 20 years? How about 10 or 5? Will Google step in and rescue all the data, as they did for Usenet?

What to do? Is it time to push for a GPL set of code to replicate Yahoogroups and deliberately force them out of business by de-centralizing? Is it time to at least write a YG scraper that can archive a list to a dbms?

This has all been prompted by a recent post to another mailing list (not on YG!).
http://www.drop.org/node.php?id=763
http://kakkune.com/base/post.php?post_id=34
http://groups.yahoo.com/group/syndic8/message/1255
It's also prompted by a bunch of code I've written to mirror a mailing list to a module in Drupal.

SatireWire | Surprise Settlement Splits Microsoft : Redmond, Wash. (SatireWire.com) %u2014 In a surprise settlement today with nine U.S. states, Microsoft agreed to be split into two independent companies, one that will continue to make Microsoft operating systems, browsers, and server software, and another, potentially larger company that will make patches for Microsoft operating systems, browsers, and server software.

Re WSDL: Curiouser and curiouser, Dave writes: Sam Ruby believes[1] that WSDL is one of the cables of the bootstrap of Web Services. I don't. How will this get resolved? We'll try both ways. If he wants to support services that don't have WSDLs, he'll have to bend. If I want to connect my software to services that require WSDLs, I'll have to.

And Sam Ruby replies : And any service which is described with a WSDL can certainly be invoked without referencing the WSDL. No lock-in here! Which is certainly my understanding of WSDL.

There's a couple of issues here. The first is interop. As long as some SOAP interfaces and toolkits are missing WSDL, the others will have to cope with these situations. Which is roughly where we are today. The second as usual is MS. If MS client toolkits expect WSDL because MS Server toolkits always produce it, then they may drag the rest of the industry with them. And it's not just MS. IBM, BEA and The Mind Electric GLUE are also important players. But that's all currently speculation.

All this analysis avoids the real question. Is WSDL useful? Sam and Dave seem to disagree and I'm not sure. A formal documentation standard for SOAP interfaces seems like a good thing. But if it adds complication which actually slows up web service adoption, then maybe not.

[1]That's not a permalink I'm afraid as they seem to be broken. Check Friday, Jan 18.

City older than Mohenjodaro unearthed - The Times of India : Indian scientists have made an archaeological find dating back to 7500 BC suggesting the world's oldest cities came up about 4,000 years earlier than is currently believed Also Indian civilisation '9,000 years old' at the BBC and also a Radio article. This is all being discussed on Metafilter. The significance is that this goes back into the ice age and it appears that the city may have been swamped when the ice melted. Perhaps another source of the Great Flood myth? This city is 5*2 miles which is a substantial size that would have taken many years and perhaps centuries to develop. This has major implications for our views on the development of civilization and pre-history.

Here's UserLand's Winer talking at Infoworld's Conference on web services. He has some good points to make but then we get the old hobby horse. Pointing at IBM and Microsoft, Winer said WSDL (Web Services Description Language) was invented in such a way that it will only work in Java and .Net environments. "It can't work in a dynamic environment; it's a static interface," he said.

I have to say, I just don't get this. WSDL is a formalised documentation of a SOAP interface and the end points and locations at which that service is located. Because WSDL is XML it can be mostly generated directly from web service code in some environments. It can also be read and used to generate code and function stubs for a client to call it. I can't see anything to stop a web service written in a scripted environment like Perl, PHP, VB or UserLand's scripting language being documented in WSDL, even if it's done by hand. in fact Simon Fell has done exactly this for UserLand's Manila SOAP interface.

Looking from the client end, I can't see anything inherent in a scripting language that would stop you writing something to parse a WSDL file, extract the key information and build function calls. More likely in both scripted and compiled languages is that you'd define a system that understood the function calls after reading the WSDL. You'd then only use the WSDL at run time to determine the location of the end points. This mainly due to the difficulties in understanding the semantics of the function calls. This is mostly because (for example) knowing that PostID is an Int doesn't tell you where the data comes from or why you should use it.

Our Dave keeps making these statements about WSDL being unsuitable for dynamic environments but he doesn't seem to be able to say why. I've just spent the last few minutes searching for an example of this in a mailing list late last year, but failed. If I find it again, I'll post it. Ah! Google to the rescue. Start here and follow the thread.

Wired had the RH image on their Rants and Raves page last month next to a couple of letters about the open source movement. I was hoping they'd put it on the web so I could point at it, but they didn't, so here's a scan.

Tasty poster from The Modern Humorist. Shazaam! Kazaa Shuts Down What Wired really means to say is that Kazaa has turned off downloading of their client, but Grokster is still downloadable and the network is still running and serving more MP3s than Napster at it's height. I've said it before, and I'll say it again. Kazaa should just declare bankruptcy (or something) and voluntarily shut themselves down. They cannot win the legal battle and it's not worth the large amounts of money to try. Just release the protocol as open source and run for the hills. We're in the middle of a 30 year war to re-define what we mean by copyright and there's no point in being a casualty.

Hello from Radio - Test2

An entertaining satire on the British tendency to grumble. [thanks, Blog.org] Oh, yes. LOL.   "George Farthing, an expatriate British man living in America, was recently diagnosed as clinically depressed, tanked up on anti-depressants and scheduled for controversial Shock Therapy when doctors realised he wasn't depressed at all - only British." While you're at The Brains Trust, check out their guide and product warning on How to eat pretzels.

OpenP2P.com: Interoperability, Not Standards [Mar. 15, 2001] Another great article from Clay Shirky.

What's this all about then? eBusiness Club Website : A fusion of leading technology providers and business support organisations deliver ultimate eBusiness information solutions to the SME community. That's a good line in MarComms-speak... Clearly some mission critical, ultimately scalable, industrial strength, business-technology fusions happening there. I particularly liked this bit under "Find your nearest club", "A full list of the Clubs will be available shortly."

If the tech world runs at 3 times normal life speed, and the government runs at 1/3 of normal life speed, they're a difference of about an order of magnitude. It's hardly surprising that this sort of thing is happening about 3 years too late and after the internet bubble has well and truly burst.

What's The Worst Idea For A Website You've Heard? - seethru And here's another one. websites no one has thought of yet
SeeThru continues to improve... I particularly liked "Amiwearingahatornot". but then I'm still waiting for the "Which dinosaur are you - Quiz".

The Times discovers blogs.

Shrub Junior: Wasssup!
Shrub Senior: Wasssup!
Shrub Junior: Kicking back, watching the game.
Shrub Senior: Drinking a Bud.
Shrub Junior: Eating pretzels.
Shrub Senior: Wasssup! Wasssup?
Shrub Junior: Don't know. I choked on a pretzel, fainted and hit my head.   Damn foreign foods, it's a Jewish conspiracy I tell you. I never choked on Momma's good old Texas Chilli.
Shrub Senior: Or Enron.
Shrub Junior: Wasssup!
Shrub Senior: Wasssup!

Mike Krus has done an interesting thing. He's made all the scraped RSS feeds on Newsisfree into RSS 0.92 and implemented the cloud feature. So taking a random feed, there's this entry in the channel section.

<cloud domain="www.newsisfree.com" port="80" path="/RPC" registerProcedure="hpe.rssPleaseNotify" protocol="xml-rpc" />

If you're aggregator is always on the net, then you can make an XML-RPC call to this interface, giving it your callback interface. Now when newsisfree notices that the feed has changed, it will call you back. You can then go and collect the RSS with the new data. This is much more efficient than collecting the feed once an hour to see if it's changed. The structure of the calls and callbacks are detailed on the RSS 092 documentation page.

Slashdot | Philips Says Compact Discs Can't be Copyprotected Fascinating. Philips hold the rights to the CD logo. They have come out and said that copy-protected CDs do not abide by the Red book standard and so cannot carry the logo they are in fact "silver disks that look similar to audio CDs but are not audio CDs". Further if they produce a CD/RW device for copying CDs and the customer puts "silver disks that look similar to audio CDs but are not audio CDs" in and the copy works, it's not their problem because they never said the device would work with these, only with CDs.

Wouldn't it be great if one of the co-authors of the standard (and a major company with big pockets) sued the big record companies for using the logo illegally because they had abused the standard? Even if they just got a temporary injunction freezing sales, it would kill copy protected CDs immediately.

TUX: Term Unit X : Issue Seven - Reboot The Adventures of Tux, the comic book.