-
This SupportPac allows you to use a WebSphere MQ queue manager as a JNDI repository. It can dynamically generate JMS administrable objects, based on actual queues on the queue manager - potentially saving you having to define any administrable objects for
-
The IBM Java Cryptography Extension (JCE) package shipped with GSKit by WebSphere MQ V5.3 may expire on May 18, 2006 at 21:59:19 GMT causing iKeyman administration functions to fail. GSKit is not used by WebSphere MQ V5.3 on platforms other than UNIX.
Stephan, our head support tech guy at work and favorite German butcher, got a blog. With the mixture of tech stuff in english, car trivia in swedish and some mixed in german it will be interesting to watch :-)
Lately I’ve been testing a bunch of the continous integration servers. I’ve gone through installations of CruiseControl, Parabuild, Continuum, Anthill, Luntbuild and now, Pulse.
Pulse is, for such a new product (current version is 1.0.4) an amazing piece of software. It got a very nice, intuative interface, support for the main SCMs (CVS, SVN, Perforce), SCM or timer triggered builds, support for the main build tools (Ant, Maven, Maven2, make) as well as custom builds. It also got a clever feature where the CI build defintion is stored with the source code in the SCM.
With the requirements we have at work, there are three major features missing, remote builds, project dependencies and LDAP integration. LDAP, which is a must have for us, is scheduled for 1.1. Remote builds is currently more of the nice-to-have type of requirement but based on the DamageControl matrix it seems like it’s planned for the near future. So, that leaves project dependencies (if project X has been updated, also build project Y and Z which depends on X). Feature has been requested (requires login) so hopefully it will be coming.
Even with these shortcoming, I think that Pulse got a bright future in this competitive space. It seems that Aussie developers just can’t fail these days.
I spent some time this weekend cleaning up my del.icio.us bookmarks. I have a decent, but not huge collection (616 links) and I find that every few months it’s a good thing to clean up duplicate tags (relax_ng vs relaxng), improve bundles and fix some misspelled tags. This time I also rearranged some of my tagging conventions. As most users, I add as many tags as makes sense for a certain link. I find the suggestions made by del.icio.us to be quite good and frequently choose among them.
Articles or blog posts I now tag with the author with the “by:” tag prefix, e.g. by:elliotte_rusty_harold. I find this a good way of keeping track of writing by my favorite authors.
I use the natively supported “for:” prefix to send links to friends who also use del.icio.us. I still got to gather some more experience on this but so far it looks good. I like to minimal approach del.icio.us takes to this type of functionality.
Links which are not HTML pages, especially PDFs, gets tagged with a “format:” prefix, e.g. format:pdf. This serves mostly as a hint/warning for myself while browsing my links. As linking to special formats like SVG might increase, this might be a good way of locating all of these at the same time.
Links which require some attention from me, gets a “to:” prefix. Currently, there are two of these, to:read and to:wish, but I expect further types in the future, e.g. to:blog for stuff I want to comment in my blog.
Since I spend a lot of time with different open source tools, lots of my links are for different OSS projects. As licenses are of great importance in this world, I’ve created a “license:” prefix, e.g. license:gpl. I haven’t really started using this yet (only one link) but I hope it will be useful.
As seen above, I use different prefixes for special types of tags. So far I find it a very good practice and I expect to define further prefixes as I go along. A de facto convention for these types of tags would be great. For example, if everyone would be using the by: prefix, finding all articles by an author, spread over different sites, would be very simple. Does anyone know of such a conventions that people use (it seems like some of the prefixes I use also get’s used by others, e.g. license:gpl)? I would certainly be happy to switch to whatever is the majority choice. If not, how would one establish such a convention in the best way?
-
A discussion on why RCP services over REST-like APIs don’t make sense.
-
Great presentation on Unicode, with some focus on the use in XML
-
A wizard to help developers figure out which Open Source licenses to use according to their preferences.
-
Great, detailed presentation on XML schema (XSD, WXS)
-
Very detailed presentation on Relax NG
-
An open source font that aims to cover all latin glyphs in Unicode.


