Monday, July 26, 2010

Rebol reverence and followership

I noted to the Rebol3 group at the Rebol world that
>> test: 42
== 42
>> test:none
== test:none
>> testnone
** Script error: testnone has no value
comment  {test remains 42 but no error unlike, say, simple undefined testnone That missing space looks nasty when test:none should have been test: none Is this a bug in R3? test:none does not return none but instead returns test:none}
A tolerated bug in R2 becomes a revered feature in R3, no? If that is not a bug, then I may be the pope. Or is it "the caprice of the colon", otherwise known as 'the big C" ? You must excuse me, Jon Stewart just explained that science is faith and I am still in awe. Ohmmmmmmm.

Friday, July 23, 2010

disposable 10$ laptop to throw-away into recycling?

Years ago in a Byte article Alan Kaye conjectured on electronic paper - the prospect of a 10$ laptop is now in the BBC news. But it will run linux ... and which linux?

Wednesday, July 21, 2010

Java Smalltalk

When I'm between contracts I can end up in some odd "HR" conversations called "technical interviews."

Young Java team leads generally know nothing of Smalltalk (but they've heard tell of it), so they cannot believe that someone outside the fold would understand the difference of
  if  ( x == y )  {}
  if ( o.equals( obj ) ) {}
And yet this is the very difference that makes Smalltalk's IdentityDictionary suitable on one occasion and a mere Dictionary on another.

But how to talk about that to a Java interviewer? Better not to mention it. A Smalltalk Set of integer keys is not a Java Set.

And as for generics and Collections, best not to mention how long Curl (of ) had typed collections ( a Smalltalk abomination) and an {Iterator-of t:Type} class.  Or that Curl abstract classes with multiple inheritance an interface framework make.

There once was a day when COBOL developers could opt to work in Smalltalk on some banking and financial systems.  But would one of today's few Smalltalk teams be any too happy to be joined by someone with years in C++? Hmmm.  And some wonder why so little gets done in Logtalk or other specialized languages while talk goes on of "dialects" and "language-oriented programming".  What can multiple-paradigm languages such as Oz be taken for in the monoculture of C# and J2EE shops?  And all the while the scripting hares nibble quietly on their Python deployment scripts hidden from CIO- and CTO-view in over-grown weed patches ... called "integration" and "minifying" and "monitoring".

Perhaps it all stems from the obsession with mowing and trimming the chemically-perfected green lawn.  What humble vegetable gardener grows only the green bean, or only spuds? 

Monoculture.  Best practiced in the dark.  Now where was that Erlang posting? Hares, rabbits, meerkats ... it must be the O'R libra colophony.

Sunday, July 4, 2010

RDFa Drupal Open Graph

 I have added a page to collect RDFa links over at

There will be a separate parsing section for PEG's, Icon 9.5, ObjectIcon 2.3, Rebol 3, Logtalk and PROLOG (and SNOBOL in C ?)

The number of essential links is now near twenty - and this is in English only.

It is still not clear to me how searchmonkey will fare as Yahoo! moves to using Bing for search.

SearchMonkey for documents was already infuriating for me due to SearchMonkey being restricted to Flash documents ( given my work in Curl markup for literary documents as and .)

I still have to get a post together on RDF with Curl documents.  RDF is largely associated with the one XML syntax - but there is nothing about RDF triples that requires XML.  Curl documents which are embedded in HTML pose no problem in general since the RDF can be expressed in the embedding object element.  Curl documents which are themselves .curl applet files offer unrestricted tagging in the initial {applet } expression so the Curl runtime itself could be adapted to export RDF.  The matter is less obvious for the {include } of documents which are .scurl text files at the top-level within applet files.

The root Curl document, when unspecified, is a TextFlowBox, but with Curl 7.0 there are now ParagraphVisual and TextVisual classes. To quote from the documentation for the parent class,
This class provides a public API for accessing the contents of text-containing objects to be rendered by text containers such as TextFlowBox.
At the moment there is no option available for a popup action such as "Print" which would offer RDF triples rather than the content - and these for the Visual container "hit" as opposed to the entire "page". A Visual would need an option such as exports-rdf? and possibly a tie to a triples validator.

The whole matter ties back to whether Curl is suited for CMS - if a suitable server-side partner emerges as IT continues its slouch towards HTML5.  Or Flash 12.1 or ...

I had thought that Drupal could be a candidate, but after some exploratory efforts I lean towards a Gemstone + Seaside 3.0 option ( pul-ease do not mention Scala Lift ... )  I still keep an eye on Wicket (Hippo?), Zope/Plone and Ruby candidates such as Nesta.

Saturday, July 3, 2010

Oblog or not (Prolog Logtalk)

Sometime ago I added some bits to on Logtalk (since updated by the author of Logtalk himself.)  In the Prolog wikipedia article, I now see mention of Oblog. Or not.

Every link on the web that I can see says the same thing of Margaret McDougall, Oblog, object-oriented Prolog and EdCAAD.

Now I believe that EdCAAD is ESALA.

And Oblog?  A wiki obus and not an opus?