Older blog entries for mindcrime (starting at number 23)

New OpenQabal Release Coming Soon

The last month has been an interesting time for OpenQabal. Thanks to my previous employer graciously laying me off from my job about 5 weeks ago, I've had the opportunity to spend most of the last month essentially working on OpenQabal full-time. A lot of progress has been made, and I anticipate making a formal 0.0.3 release soon, probably within the next week.

Some of the changes that will appear (this does not, btw, map exactly to the previous road-map. As they say, "planning is essential, but a plan is useless.")

• New, more modular build system using Ivy for dependency management and incorporating useful tools such as FindBugs, JDepend, PMD, TestNG, Cobertura, etc.
• New "User Dashboard" component that provides the main point of entry and ties the various components together visually
• Concordantly with the introduction of the new User Dashboard, all of the old Sitemesh stuff has been ripped out
• New configure and install scripts to automate most of the tedious parts of building and installing OpenQabal
• Addition of many unit tests (we still don't have 100% test coverage, but progress has been made)
• Various minor bug fixes and tweaks
• Probably some other stuff I'm forgetting

Unfortunately a couple of things that I'd hoped to get in a 0.0.3 release are going to have to be deferred, with the most notable being OpenID login support. However, a subsequent 0.0.4 release should be around in relatively short order, and will include OpenID and - if things go well - a number of other big improvements.

Unfortunately I have to go back to work for "the man" now so that I can pay my rent, so I won't be able to spend quite as much time on OpenQabal. But things will continue to move forward. There may actually be some even bigger news in the near future, so stay tuned.

Syndicated 2008-03-29 07:23:28 (Updated 2008-03-29 07:24:20) from openqabal

Prolog? I'm going to learn Prolog?

Sometime ago I blogged about the availability of some great resources for learning Prolog. At the time, the available materials I'd found were:

• Artificial Intelligence through Prolog
• Introduction to Prolog for Mathematicians.
• Building Expert Systems in Prolog

Now, thanks to programming.reddit.com, I've found a couple of additional references on Prolog, which are also freely available online.

Simply Logical: Intelligent Reasoning by Example
and
Logic, Programming and Prolog

Luckily there are also a number of high quality implementations of Prolog available, including GNU Prolog, SWI Prolog and Ciao Prolog.

Now to find some time to dig in... and figure out what part(s) of OpenQabal to write in Prolog.

Edit (03-30-200): Apropos, this link just appeared at the top of programming.reddit.com. Good stuff.

Syndicated 2008-03-29 06:56:20 (Updated 2008-03-30 09:06:28) from openqabal

More computer science than you can shake a stick at

One thing about the Internet that never ceases to amaze me, is the sheer volume of incredible information that's available. I just regret that I don't have time to read and digest it all! But in case you have more spare cycles than I do, and are interested in drowning yourself in Computer Science, check out these links:

• Arxiv.org - Pre-Print Archive - New Listings in CS
• Directory of Open Access Journals - CS section
• Citeseer
• Pre-Prints of recent Donald Knuth papers
• Papers by Andrew S. Tanenbaum
• Marvin Minsky
• MIT OpenCourseWare - EE and CS section
• MIT CSAIL (Computer Science & Artificial Intelligence Laboratory) Publications
• Stony Brook Algorithm Repository
• Dictionary of Algorithms and Data Structures
• USENIX Proceedings
• Google TechTalks Channel

See what I mean. So much great knowledge, so little time... Then again, maybe if we spent less time blogging, we'd have more time to read. <smile />

Syndicated 2008-03-23 12:37:53 (Updated 2008-03-25 09:06:50) from openqabal

New from Google: Visualization API

Interesting news on the Google Code Blog recently. Seems that Google have released a new API, this time one for doing online data visualization. Google describe it as follows:

The Google Visualization API lets you access multiple sources of structured data that you can display, choosing from a large selection of visualizations. The Google Visualization API also provides a platform that can be used to create, share and reuse visualizations written by the developer community at large.

Seems that you can embed these visualizations into your own website, embed them as a Google Gadget and/or use them with Google Docs in some fashion. This seems like something that could be pretty interesting to play with.

Not sure that it has any particular relevance to OpenQabal, other than serving as one more demonstration of the value of Google Gadgets. In that regard, if something like this catches on, it argues even more strongly for us to move forward with implementing a Google Gadgets / Open Social container as part of OpenQabal dashboard. Of course, that is already being strongly considered anyway.

Syndicated 2008-03-20 03:56:00 from openqabal

Widgets, Portlets, W3C, Google, oh my...

Interesting how different threads of thought come together, often about the same time you find some interesting new information. Earlier this morning I stumbled across the fact that the W3C are working on a Widgets specification. With that in mind, and the earlier discussion about JSR-286 in mind, I decided to start researching what other people were saying about the intersection of portlets and widgets. That led directly to this TSS discussion. Some more researching led to the discovery of Netvibes and their UWA API for Widgets as well.

So, Widgets, Porlets, both or neither? ?

Certainly there is some overlap in the intention of some of these different approaches. That the W3C spec will be an open standard makes it very appealing, and those of us at OpenQabal will certainly be rooting for it to catch on. Following up on what I said before about the need for Google to submit Google Gadgets to a standards body, I find myself wondering how much common ground there is between Google Gadgets and W3C Widgets 1.0, and wondering if Google might just get involved with the W3C effort, instead of maintaining a proprietary standard. One can hope...

And all of this still leaves one wondering what is going to happen with WSRP, which seems to fit into this space somewhere as well.

Syndicated 2008-03-17 09:21:20 (Updated 2008-03-17 10:07:44) from openqabal

Yahoo goes Semantic

As announced Thursday, Yahoo are moving to embrace Semantic Web technologies and microformats in their search platform. Good news for advocates of semantic web technologies, as this should help spur increased adoption of RDF, RDFa, Dublin Core, OWL and the various microformats.

What it means for the long-term development of the Semantic Web is TBD, of course. The overall vision of the power of the Semantic Web goes well beyond search and includes the use of personal agents and reasoning engines that can answer questions in an "intelligent" manner, rather than simply returning a list of web pages. How far Yahoo plans to go with this will remain to be seen.

More discussion and analysis of this story can be found on the many semantic web blogs aggregated at Planet RDF.

Syndicated 2008-03-15 10:45:39 (Updated 2008-03-15 10:50:06) from openqabal

Yahoo to adopt Google OpenSocial?

According to this ZDNet blog, Yahoo may be poised to adopt Google's OpenSocial API soon. If this happens, it would mean that essentially every important player in the social-networking / online application space had adopted OS, with the exception of Facebook. This would, of course, put a little additional pressure on Facebook to adopt OpenSocial as well, or risk losing developer mind-share to what would be a defacto industry standard.

And so what is the relevance of this to OpenQabal? In short, the possibility of OpenSocial becoming the industry standard for plugging social applications into a container means it's that much more likely that we will also adopt OpenSocial as our strategy (or at least a strategy) for dealing with this same issue.

With a (more or less) open standard out there, widely adopted by oodles of developers and platforms, it just makes sense to support it. Now this would probably matter more to somebody who was interested in putting an OQ based site on the public internet and running a general purpose social-networking site, than to somebody looking at OQ for enterprise use. And the enterprise really is more what OQ is aimed at, at least in the short-term. But even there, as more actual useful applications emerge (as opposed to the "vampire slayer" and "rate my hot friends" type apps) even enterprise users will want the ability to plug this stuff into their platform.

There is one thing that we should hope for however... Google really needs to take the Google Gadgets specs / api and the OpenSocial specs / api and submit this stuff to a recognized international standards body for adoption. No matter how much you trust Google now, it is really always better to work with standards that are controlled by a neutral (well, more or less neutral) standards body than by one vendor.

Syndicated 2008-03-13 08:29:04 from openqabal

Just noticed that JSR-286 (Portlet 2.0 ) specification was approved

Somehow I had missed the news until just now, but it appears that the JSR-286 spec has been approved. Congrats to the JSR team on getting this out the door. A lot of people have been longing for the new features of Portlet 2.0 and looking to move beyond JSR-168 portlets. Hopefully this will help drive even more adoption of portlets, which are really pretty cool technology.

And how does this relate to OpenQabal? Well, let's just say that we're looking at the possibility of integrating a lightweight portlet container as a way to allow 3rd party apps to plugin to our "user dashboard" application. What's going to be fun for the next while is working out exactly what technology or technologies are going to be useful for this. The Google Gadget / Shindig / OpenSocial stuff may be applicable, portlets certainly deserve a look, and there is also WSRP to look into.

The fun never stops around here...

Syndicated 2008-03-11 11:23:59 (Updated 2008-03-11 11:53:41) from openqabal

Big Project Reorg Underway

So tonight I started a fairly radical restructuring of OpenQabal, relative to how things are packaged. Up until now we've been packing everything into one big honkin' EAR file... all the webapps, all utility jars needed by all the webapps, the MDB for single-sign-out, etc. Truth be told this was always just a convenience thing, since it's quick and easy to deploy one EAR file while you're in development mode... but there has always been a thought that the eventual model would bundle things in a more granular fashion, to allow for separating components across servers, to make it possible to cluster individual modules, etc.

Well, as it turns out a weird glitch in the J2EE packaging spec is forcing this issue to be dealt with sooner than later. The issue is, when you bundle dependencies into the EAR as utility jars and reference them through the Class-Path: entry of the WAR file's MANIFEST.MF, you would intuitively expect (well, I would anyway) that a given WAR will only be able to access resources from a JAR which is explicitly has a reference too. Unfortunately, it turns out that this is a gray area in the spec, and some app servers (such as JBoss AS 4.0.5, which is my default dev/test environment) allow a WAR to "touch" things from other JARs. This becomes an issue when, for example, you have two webapps, one of which uses WebWork1 and the other of which uses Struts 2. What winds up happening (intermittently of course) is that the Struts 2 app picks up the Freemarker templates which are bundled into the WebWork1 distro. Or vice versa. And needless to say, the files aren't compatible. And if this problem didn't expose itself as a conflict over Freemarker templates, it would crop up sooner or alter as a class version conflict or something similar.

So, starting tonight I'm reworking all of the OQ projects so that each webapp is it's own deployable bundle (an EAR if needed, a WAR otherwise).

At the same time, I've also cut everything over to use the "new" build system by default, which temporarily renders all the the build and deploy documentation obsolete. The new build system is a big improvement over the old one though, and once all of the documentation is in place, it should make development on OQ ever so slightly easier.

Also, don't bother looking at the CruiseControl reports for the next day or two as well, since I haven't even started to tackle re-jiggering that stuff. Probably by the end of the week at latest everything should be stable again.

And then some real work can start on getting new features and functionality implemented.

And yes, I will eventually be posting more documentation, specifically use-cases and capability-cases, etc. that help elaborate on what OQ is, what scenarios it is meant to be useful in, etc. No, really, I promise. <smile>

Syndicated 2008-03-10 09:01:01 (Updated 2008-03-10 09:02:12) from openqabal

DataPortability.org Logo Contest!

In response to the Cease & Desist letter from Red Hat, the DataPortability Working Group have announced a contest to create a new logo. They are inviting all DP fans and interested parties to contribute logo artwork for possible adoption as the new DataPortability logo.

Sounds like a good way to resolve the conflict to me. Here's hoping the new logo is even more rad than the existing one.

Syndicated 2008-02-22 22:40:38 from openqabal