Curiouser and Curiouser!

Now that I want to write some Ruby code I'm looking for a good IDE. Using IntelliJ IDEA has rather spoiled me and I expect a lot of intelligence from an IDE these days. I'm not sure the Ruby market is mature enough to meet my expectations, nevertheless I went hunting and here's what I found:

• FreeRIDE
• VisualWX
• ArachnoRUBY
• Mondrian Ruby

FreeRIDE and Mondrian are very similar having been built with the same toolkits. Of the two Mondrian seems the more polished (the author was intending to make it a commercial product but seems to be veering towards open source now) however I couldn't find a way to debug code and the editor didn't support folding. FreeRIDE, although a little rougher round the edges, has both of these features, a plug-in architecture, and some experimental refactoring support too.

VisualWX initially appeared quite promising with the its integrated GUI builder (which uses the cross-platform wxRuby toolkit) however it continually complained about not finding something called Mingw32, I couldn't find the debugger, and I couldn't get the GUI builder to work anyway.

Finally there is ArachnoRuby. This is available as a 60 day preview of an expected commercial release. Pricing is $59 for the personal edition and $109 for the professional edition although there is no comparison chart available to tell you what's different about the two.

Although ArachnoRuby was superficially the most polished of the IDEs I tried, it did have a number of shortcomings: The editor did not supporting code folding, the editor tab was constantly getting split in a most irritating way, and indenting behaviour was not to my taste.

More seriously I got a strange warning:

(eval):219: warning: statement not reached

every time I ran any code which didn't happen with the other IDE's. More seriously still the application crashed while picking from a popup context menu. On the plus side the debugger looks very impressive and they claim to be able to do code insight (although I couldn't see any evidence of it in my short test). This is one I'd like to see again in a few builds time.

For now I'll be using FreeRIDE which is, for me, the pick of the bunch. But i'd love to find a Ruby IDE that I could hold up to my beloved IntelliJ.

Update: Austin Gilbert pointed out to me that Eclipse has a Ruby mode. Although I have always loathed Eclipse for Java development work I figured it wasn't up against any stiff competition here so I installed Eclipse 3.01 + Ruby mode.

Well Eclipse may have gotten prettier since I last used it but it's just as complicated and unfriendly. The Ruby mode doesn't offer anything beyond FreeRIDE (less even) and for the life of me I cannot even figure out how to get it to run my test case. I keep getting presented with dialogs for defining configurations and something about perspectives but as to actually running the code... well, no.

So I still hate Eclipse, but thanks anyway for the suggestion Austin.

Update updated: I'm not sure why but I decided to give Eclipse another try this morning before deleting it.  After a considerable amount of gnashing of teeth I have managed to get it to run & debug my test case.  Why do they make it so hard?  When I click Run I expect something sensible to happen not to get into arcane, self-referential, dialog boxes.  And preferences under the Window menu?  Of course it's the first place you look!  Anyway the debugger seemed to work apart from not being able to Terminate.  I'm not  sure how annoying that might prove to be.

At this point I'll grudgingly accept that Eclipse may be a viable platform for Ruby development.

Blog Research Repository. Wondering what would happen if I made Sigmund public, I have come to the conclusion that it is not as straightforward as it seems. The spidering problem, mentioned in the same post, is only a technical glitch. Matt Mower in... [Anjo Anjewierden]

Anjo posits the idea of a grand unified blog post repository where the text content (sans formatting) of every blog post could held. The idea is that authors of blog searching, and indexing tools could use the repository instead of spidering sites themselves. I'm working on a tool now which could make use of such a service and, if it existed, I probably would.

The likes of BlogLines, Feedster, and Technorati already have most, if not all, of the data but I think there are some significant challenges when you start thinking about how terms of service might have to work. Possibly an academic institution might be able to come up with something if they have the spare capacity to run such a thing on a non-commerical basis.

But I still think that RSS Archives are a good idea. I think they solve two problems in a decentralized way:

1. To be able to access, in a standardized way, the entire contents of a weblog. Even with blogs which produce a page-per-post you have the problem that the home page will only link to recent posts. You have no reliable way to determine where the rest of the posts are and indexing the whole blog, Google style, is wasteful and likely to include lots of stuff which isn't really part of the blog content.
2. To be able to access, in a standardized way, the content of each post. Which bit of, say, a Wordpress generated blog page contains the actual content of the post? The idea is to encode each post into a separate RSS file which is, theoretically, easy to read using standard RSS tools.

These two goals are achieved through the opml map file connecting the permalink of each post to the URL of the RSS archive for that post with the effect that, from my RSS archive, you have a predictable means of accessing only & all of the content of my weblog posts.

Further enhancements could include adding the publication date to the map which would make it trivial to retrieve intelligent subsets of the data (e.g., all posts from 2003). At this point I would probably start heading towards an XTM or XFML map instead to leverage what those standards can do (which is the direction I was heading in with liveTopics many moons ago).

One year ago. One year ago I posted a list of Italian blogger girls posing nude. In the next 2 days I got the all time traffic peak of this site. :-)

Thanks to Steve Kirks for his excellent "On This Day" radio macro.
[Paolo Valdemarin: Paolo's Weblog]

I missed this macro, it's neat.  Thanks Steve.

Who is Reviewing My PGP Purchase?. I'm a long-time user of PGP encryption software. Though I don't use it for everything, I use it to sign and/or encrypt e-mails with sensitive or critical information. I also use it to encrypt confidential files on my computers. And sometimes I use the PGPWipe function to make sure things are really deleted. I have for years, so many years that I was still on version 4.something.

But with the new ThinkPad running WindozeXP Pro my old version of PGPMail wouldn't load. I also had a copy of the newer PGP 7.0 freeware, but it wouldn't load either. Both gave me error messages about being incompatible with the OS. Most likely not, since my old version ran fine on W2k, but still...

So I went looking for the latest freeware version of PGP. I downloaded the ZIP file but it contained a corrupt archive of the .exe file. I tried it again, figuring maybe there was some network error, but got the same corrupt result. I figured "What the heck? I haven't paid for this stuff in a long time, it's ok if I buy a new copy." So I went to the PGPStore and ordered a new copy of PGPDesktop 8.1. I filled in the forms, paid my money, and got my confirmation. This is what it said...

I wasn't aware we had reinstituted export restrictions on encryption tools. Six hours later I still don't have my software. And I want to know just who the hell is reviewing my purchase, and what are they looking at? Are they checking my name against the "Do Not Sell" list? Or maybe they're checking it against the same bogus list the airlines are using to trap terrorists on airplanes. Maybe they're looking at my credit card records to see if I made any bulk fertilizer purchases recently. Who knows?

I know one thing -- if PGP Corporation is the only one reviewing my name they don't have authority to look at anything and the whole idea of claiming a review is ludicrous. If someone other than PGP Corp. is reviewing this purchase data I want to know who, how, what, and why. And I want to know now.
[b.cognosco]

Terry's asking some pretty good questions.

I'm currently reading Information Retrieval: Data Structures & Algorithms by Frakes & Baeza-Yates which I found in the LondonMet library (when I should have been looking for psychology books).  It's the first book I've found which has practical information on implementing algorithms for ranking & clustering data.  As books go I've found it reasonably easy to read although my math is very rusty which trips me up a lot.

What i'd really like to do is find someone working in this field already who I can learn from.  Anyone know anyone (academics maybe?) in London or surrounds who works in the text analysis and/or data-clustering?

I'm also looking into techniques like the Singular Value Decomposition (used in Latent Semantic Analysis) but that requires some matrices-related heavy lifting which I'm slowly working my way back up to (with lots of interesting diversions along the way).  I also came across something called Support Vector Machines which seems to have applications in this area.

I think I need an external brain pack, anyone got any liver..?