Patrick Chanezon's weblog

Another VPS

...

What all this means is that I will be needing more bandwidth and storage space. I have chosen to use ServInt VPS instead of dragging a T1 line to my house and running my own Linux server farm which requires too much money way too early. ServInt has pretty good reputation so I can trust them not to disappoint me and 100GB/month transfer with 2 GB storage for $49/month is very attractive. As traffic and load grows, I can upgrade to ServInt dedicated servers.

If you decide to use ServInt as well, tell them Don sent ya. But then I just signed up so you might want to wait until I have used their service for a while. Life is full of surprises you know. ;-p

[Don Park's Daily Habit]

Don Park's choice of VPS is twice more expensive as Abe Fettig's, but the specs are much better. I think I'll use this one instead. Waiting one more week to see what fellow webloggers chose. It seems everyone's in the market for a VPS these days :-)

Virtual Dedicated Hosting

Hmm, this is interesting, $25/month for a whole virtual server - tummy.com. Looks very nearly as good as co-locating.

In use by Abe Fettig, must watch how he gets on.

[Raw Blog]

I was looking for a new hosting this summer but prices for dedicated, and even virutal private servers discouraged me. $25 is a good price.

Inner Child quiz

Fun quiz: I need to grow up in my head before my kids turns 16 themselves !-)

My inner child is sixteen years old!

Life's not fair! It's never been fair, but while
adults might just accept that, I know
something's gotta change. And it's gonna
change, just as soon as I become an adult and
get some power of my own.


How Old is Your Inner Child?
brought to you by Quizilla

A few classloader related links from my bookmarks, for Thesis weblog

The Thesis guys Looking into classloading

I've read many articles about classloading last year, when messing up with Jasper for our portal Server, but that was before I started weblogging, so I have no entry in my old weblog about that.

Here is an extract of my bookmarks related to classloaders. I hope this helps. I like reading your weblog: I especially like the idea of setting up a team weblog for a thesis.

One thing I would look at as well, because it seemed very promising to me at the time, is Bob McWhirter's Classworld project. Maybe it is not applicable to what you try to do, but the ideas developped there were of great interest to me, after a year trying to sort out the different classloading models for BEA, IBM and Sun ONE application servers.

Here are the rest of the links I pulled from my bookmarks.html. Hope this helps.

VM Spec Loading, Linking, and Initializing

Workarounds for Two Common BEA J2EE Connector Architecture Exceptions

WebLogic Server Application Classloading

http--servlet.java.sun.com-javaone-resources-content-sf2002-conf-sessions-pdfs-1311.pdf

http--www6.software.ibm.com-developerworks-education-j-classloader-j-classloader-a4.pdf

Understanding the Manifest

JDC Tech Tips October 31, 2000

Java Security Architecture

Java Security Architecture intro

http--staff.develop.com-halloway-code-gd.html

Dynamic Proxy Classes

WebSphere Application Server classloaders

Link: Toward a Weblogging Empire [Wired News]

Jason Calacanis was last seen as the editor of the Silicon Alley Reporter. Now he's resurfacing as the would-be czar of a weblogging clearinghouse. But is there any money in it? By Daniel Terdiman. [Wired News]

I don't really believe this is going to work, but let's follow this guy's venture.
Google's Adsense program seems to me like a better business model for webloggers desiring to get some revenue from their weblog.
The big problem I see in his attempt is that he does not really provides any value.
If their model is based on subscription, I think a majority of users are not ready to shell out any money for weblog reading. An example of this, not totally a weblog, but a site focused on a specific topic, is TheStreet.com: when they appeared I registered for a free trial, and found some interesting articles. At that time i was trading a little bit. But when it was time to pay for the content, I just stopped using the site because the value was not compelling for me. Maybe it is for other users, it should, since they are still in business today.
If their model is based on advertisement, then what is the incentive for one weblogger to join the site as opposed to using Google Adsense themselves and get rid of the middleman that a publisher represents ?
They would argue about the traditional role of a publisher to select and filter quality content for you. But this is specifically what is being automated by weblog-related technology. RSS readers, topical aggregators (Technorati, javablog), various social softwares, and other weblogger links achieve this selection today. In an imperfect way, but this is the area where I can see a lot of progress happen: collaborative filtering a-la-amazon, filtering based on geography, acquaintances, etc...
With Google's acquisition of blogger, and giving away the Pro service for free, I can see them scratch this specific itch in the near future.
Tough luck for old-style publishers who want to apply old business models to this new space !

Fabl: a new language based on RDF and DAML+OIL

Through Damien Stolarz, I just dicovered fabl, by Chris Goad. Fabl is "a programming language which is built within RDF", using DAML+OIL, and javascript as an external representation to input the code.

Damien blogged about RDFMapper, which is just a proof of concept application for fabl. I followed the link because RDFMapper implements an idea I had submitted to the LazyWeb last february in Apache maps its commiters based on GeoUrl: Lazyweb, give me a foaf based equivalent !. Not a very original idea, but a very useful functionality.

Chris sates his goals for Fabl as follows:

"The practicality of an RDF–based computational formalism is a central issue for this paper... ...

The full value of formalizing computation within RDF will be realized only by an open standard. We regard Fabl as a proof–of–concept for such a formalization."

The paper is a fascinating read. It's been years I've been reading about RDF, and I begin to get the state of mind you need to be into to think in terms of triples, subject, predicate, object, resources, and litterals; but this paper goes one step further, applying RDF to computation, which is not a small scope of application, and using the whole salsa: RDF Schemas and DAML+OIL !

RDF syntax and semantics can be viewed as having three layers: (1) a layer which assigns concrete syntax (usually XML) to RDF assertions, (2) the data model layer, in which RDF content is represented as a set of triples over URIs and literals, and (3) a semantic model, consisting of the objects and properties to which RDF assertions refer. DAML+OIL specifies semantics[7] constraining the relationship between the data model and the semantic model.

The proper level of description for computation over RDF is the data model; the state of an RDF computation is a set of triples . This triple set in turn can be construed as a directed labeled graph whose nodes are URIs and literals, and whose arcs are labeled by the URIs of properties.

Fabl is executed by a virtual machine. An invocation of the Fabl VM creates an initial RDF graph which is in effect Fabl's own self description: the graph contains nodes for the basic functions and constants making up the Fabl language. Subsequent activity modifies the RDF graph maintained by the VM, called the "active graph". The Fabl interpreter can accept input from a command shell, or can be configured as a server in a manner appropriate to the application.

One problem that I see for immediate use is verbosity.

Consider the simplest of data structures, a point on the plane with two coordinates, which can be expressed in Java by:

public class Point {
    double xc;
    double yc;
}


Here is an extract from a Fabl RDF file at http://purl.oclc.org/net/fabl/examples/geom defining the same structure:

<rdf:RDF
  xmlns:rdf ="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
  xmlns:xsd ="http://www.w3.org/2000/10/XMLSchema#"
  xmlns:daml="http://www.daml.org/2001/03/daml+oil#"
  xmlns:fabl="http://purl.oclc.org/net/nurl/fabl/"
  xmlns:nurl="http://purl.oclc.org/net/nurl/"
>

<daml:DatatypeProperty rdf:ID="xc"/>

<daml:DatatypeProperty rdf:ID="yc"/>

<daml:Class rdf:ID="Point">
  <rdfs:subClassOf>    
    <daml:Restriction>
      <daml:onProperty rdf:resource="#xc"/>
      <daml:toClass rdf:resource=
          "http://www.w3.org/2000/10/XMLSchema#double"/>
      <daml:cardinality>1</daml:cardinality>

    </daml:Restriction>
  </rdfs:subClassOf>    
  <rdfs:subClassOf>    
    <daml:Restriction>
      <daml:onProperty rdf:resource="#yc"/>
      <daml:toClass rdf:resource=
          "http://www.w3.org/2000/10/XMLSchema#double"/>
      <daml:cardinality>1</daml:cardinality>

    </daml:Restriction>
  </rdfs:subClassOf>    
</daml:Class>  

At a time when Ruby or Python programmers, and scripters in general find java too verbose, and that RDF doesn't take off in part because the syntax is too complicated to use, this RDF syntax for programming construct doesn't appear to me as very enticing ! But I guess tools need to appear to make that less painful. Maybe Danny Ayer's IdeaGraph could be a good base for that. He should turn it into an Eclipse plugin (it seems like Danny already invoked the LazyWeb for this one).

I wish Chris good luck in his endeavour: it's a pleasure to see this kind of really weird project blossoming, far from the endless controversies about RDF.

Link: The Big Sync [Joi Ito's Web]

The Big Sync [Joi Ito's Web]

Interesting post about Joi's blogging habits.
Since I quit Radio, I did not form new efficient blogging habits yet.
Just playing with many tools to determine the ideal set for me.
Right now MT and NetNewsWire seem to fit the bill pretty well, but I still need to find the right format and style for writing posts, while being able to do that quickly, and customize my MT templates.
I need to try Kung-Log
But when this is done, I'll definitely write something about it.

Link: The Big Sync [Joi Ito's Web]

The Big Sync [Joi Ito's Web]

Interesting post about Joi's blogging habits.
Since I quit Radio, I did not form new efficient blogging habits yet.
Just playing with many tools to determine the ideal set for me.
Right now MT and NetNewsWire seem to fit the bill pretty well, but I still need to find the right format and style for writing posts, while being able to do that quickly, and customize my MT templates.
I need to try Kung-Log
But when this is done, I'll definitely write something about it.

Link: Neal Stephenson launches a Wiki to explain his new novel [Joi Ito's Web]

Neal Stephenson launches a Wiki to explain his new novel Quicksilver. Very cool.

via Boing Boing via Jeremy [Joi Ito's Web]

subethaedit, a collaborative editor on the Mac

http://www.codingmonkeys.de/subethaedit/

Palm synchronization in java

http://www.jsyncmanager.org/

Sun's new chip design

http://www.raibledesigns.com/page/rd?anchor=sun_to_present_chip_design

Bart Simpson and Google

http://www.raibledesigns.com/page/rd?anchor=good_stuff_simpsons

OPML to SVG

http://www.protocol7.com/archives/2002/03/17/opml_to_svg/
neat demo !

Wattenberg in MITs Tech 100 (Clay Shirky)

Many-to-Many: Wattenberg in MITs Tech 100 (Clay Shirky)
Martin Wattenberg, co-reator of historyflow, has been named one of MIT’s Tech Review 100 Innovators under 35. Congratulations!...
Read Complete Article
http://www.corante.com/many/archives/2003/09/18/wattenberg_in_mits_tech_100.php
Guid: 159@http://www.corante.com/many/
-----------------------------------------------------------------------------------------------------------
You can manage your account by logging into http://rss.blogstreet.com

Now we just gotta get it to work!

Marc's Voice: Now we just gotta get it to work!

Clay Shirky has discovered what we've been up to.....

Marc Canter, who has been talking for some time about the need for a service independent method of handling digital profiles, has launched People Aggregator. In addition to support for handles/pointers to multiple services, Marc has done a lot of work around granularity of FOAF files (Friend of a Friend -- a markup language for relationships.) Marc says the system will have:

[...] control over FOAF files - I've read a lot of feedback myself and I've read other folks have the impression that as soon as their personal data is in FOAF, that it'll be ripped off, scraped, used by others, etc. This is (obviously) a main reason why (you) researchers want to keep each person's FOAF file to themselves, but I hope you realize is that that's just not practical in the real world - so we're going to implement:

- private FOAF files - where NO ONE can access them, associate themselves with or basically even know about (why do you want that? Please bear with me)
- consensually controlled FOAF files - where only be consent can one associate or do anything with your FOAF
- public FOAF files - status quo, how they're dealt with today

In alpha, but worth a look...[Many-to-Many]
Read Complete Article
http://www.corante.com/many/archives/2003/09/18/peopleaggregatorcom_is_out.php
Comments: http://comments.blogs.it/comments?u=100198&p=1727&link=http%3A%2F%2Fblogs.it%2F0100198%2F2003%2F09%2F18.html%23a1727
PubDate: Fri, 19 Sep 2003 03:34:58 GMT
Guid: http://blogs.it/0100198/2003/09/18.html#a1727
Source: Many-to-Many
-----------------------------------------------------------------------------------------------------------
You can manage your account by logging into http://rss.blogstreet.com

AOL to Be Dropped by Time Warner

New York Times: Technology: AOL to Be Dropped by Time Warner
AOL Time Warner plans to change its name to Time Warner and its stock symbol to "TWX" over the next several weeks.
Read Complete Article
http://www.nytimes.com/2003/09/19/business/media/19AOL.html?ex=1379304000&en=db5b60ca26f08af4&ei=5007&partner=USERLAND
Author: Andrew Ross Sorkin and David D. Kirkpatrick
PubDate: Fri, 19 Sep 2003 09:05:46 GMT
Guid: http://www.nytimes.com/2003/09/19/business/media/19AOL.html
-----------------------------------------------------------------------------------------------------------
You can manage your account by logging into http://rss.blogstreet.com

Weblogs, prior art, and virtual machines

Jon's Radio: Weblogs, prior art, and virtual machines
Ray Ozzie recently posted what may prove to be the single most influential weblog item ever written: Saving the Browser. As you probably already know, Ray makes a compelling argument that the 1993-era Lotus Notes should have been considered prior art for the Eolas patent filed in 1994 and issued in 1998. Ray's extraordinary essay might conceivably save Microsoft ten times what it invested in Groove, should the argument prove decisive in an appeal of the recent ruling in favor of Eolas. Of more interest to those who weep only crocodile tears for Microsoft in this case, it might prevent a bunch of other applecarts from being upset: Flash, Mozilla, Safari. ...
Read Complete Article
http://weblog.infoworld.com/udell/2003/09/17.html#a798
-----------------------------------------------------------------------------------------------------------
You can manage your account by logging into http://rss.blogstreet.com

New PowerBooks!

Raible Designs ~ We Build Web Apps: New PowerBooks!
Charles has more . Too bad these suckers aren't faster - but the price is nice - and its twice as fast as my current one. It sure is tempting. As I'm writing this, I can't even price a 17" bad boy b/c the Apple Store appears to be getting hammered.... [883 characters]
Read Complete Article
http://www.raibledesigns.com/page/rd/20030916#new_powerbooks2
Comments: http://www.raibledesigns.com?anchor=new_powerbooks2
Guid: http://www.raibledesigns.com/page/rd/20030916#new_powerbooks2
-----------------------------------------------------------------------------------------------------------
You can manage your account by logging into http://rss.blogstreet.com

[Many-to-Many: Codifying Relationships]

[Many-to-Many: Codifying Relationships]

[Paolo Valdemarin Weblog]

[Paolo Valdemarin Weblog]

[Blogging Roller]

[Blogging Roller]

[E-Commerce News: Google as New Enterprise Superstar]

[E-Commerce News: Google as New Enterprise Superstar]

The Google Calculator

[kuro5hin.org || Fun with the Google Calculator]
I'm impressed that it knows the answer to life the universe and everything !