Tag: me

EAI

i’m being interviewed by open enterprise trends about my recent piece on advogato.

1. What “lesson learned” about getting different open Source projects to work together did you learn from SlideML?
– micro steps are preferable to lofty goals that you never reach anyway
– the same technical issues pops up in different environments, you just need to find a common language to recognize that you are sitting in the same boat

2. What are the “hazards” a developer team should be aware of when they begin to introduce Open Source into a commercial enterprise? (For instance, with commercial software, the conclusion is almost ALWAYS that the software from Oracle, IBM, Microsoft and others WILL work together — even if they have to spend millions with a professional integration services firm to do it).

not all open source software has a sufficiently large community to support its further development. becoming a respected open source citizen takes some work, but is crucial for organizations if they ever want to roll back their modifications into the main line of development. getting support is also much easier if a community respects you. this usually means organizations have to adopt a humble approach towards open source communities. ibm was not welcomed to apache by virtue of its brand, but rather each ibm employee had to prove his worth by valuable contributions. given enough resources, open source software of course integrates even better than proprietary software because all pieces can be molded as needed. for practical purposes it makes much more sense to look out for standards support (see below)

3. You mention a key lesson learned from CMSML: [From your article “Lesson: Don’t be afraid of the proprietary world, and seek collaboration for open standards with it wherever possible.”] Can we drill down on that idea a bit? How would you suggest an Open Source and/or commercial developer team “seek” such collaboration? Would you suggest how to do that, for instance, when working with Apache/Tomcat. PHP or MySQL?

there are many areas (file formats, internet standards, naming conventions) that are the low-hanging fruit for both sides. if you observe the blog ecosystem you will notice that individuals from competing firms nevertheless collaborate on technical issues and share implementation experiences. (sam ruby (ibm) links to don box (microsoft) and so on) let the individuals touch base. engineers can relate to each other, never mind the politics of open vs closed. that’s how we did it.

4. What’s your take on the emergence of “open standards” for commercial products, such as C#, JBoss for instance? Does this “middle ground” present something that open Source developers should be in favor of, rather than against?

absolutely. open standards are crucial. applied correctly, they enable the emergence of vast software ecosystems. the emergence of the internet both wouldn’t have been possible without, and greatly promoted the value of open standards. open standards are not really a middle ground. i view them as somewhat orthogonal to the question of whether the source is open or not. there are open source programs with undocumented file formats that are in essence no better than any proprietary program because it is just not cost-effective to glean the format from the source code. the oft-heard argument that because you can reverse-engineer a format with the help of the source code, that software is “open” is fallacious.
for all practical purposes, organizations do not have the resources to undertake that effort, and they thus are not really able to hedge against closed formats by picking open source programs.

Transparent calendaring

I’m currently researching tools that allow me to publish and maintain my schedule online. Features I need:

  • integration with outlook
  • automated synchronization (preferably 2way)
  • standards support (iCal, vCal etc)
  • public / private events

Of course there are wish list items like group scheduling, rss feeds, FOAF support, location-awareness. I believe that opening up my schedule for others to take a peak at is most useful (with the exception of some events that may be too private). Interestingly, there is a Semantic Web Calendar Agent that

provides interoperability between RDF based calendar descriptions on the web, and Personal Information Manager (PIM) Systems such as Microsoft’s Outlook. Schedules and events can be described on the web in RDF, using existing ontologies such as the Hybrid iCal-like RDF Schema or the Dublin Core ontology, and linked to individual’s contact information described, for example, at their home page.

its conveniently written by some of the guys from #foaf 🙂 There is also a RDF Calendar task force

#foaf

i’m hanging out on #foaf these days, trying to learn as much about FOAF and semantic web killer apps as possible.
<gregor> do you think mobloggers could be tricked into expanding the dataset
<zool> that’s what my plan is with noderunner london – not tricking though, just playing games, without exposing the underlying mapping agenda 🙂
<gregor> “a picture tells a 1000 words”
<zool> maps are stories…
<gregor> tricking in the sense that they wouldn’t care if they knew 🙂
zool smiles yes
<gregor> word. heard that some mobiles can determine their location
<gregor> useful for location -> text mapping
<gregor> or piccies
zool nods, convergence of tech, decentralization of data, hopes

bayes mind

A Bayes net is a graph structure whose nodes are terms or concepts with attached truth values or probabilities, and whose links are conditional probabilities.

i’m experiencing the value of bayesian networks for information filtering. while the obvious spam filtering is nice, there is much more to it. pretty soon the lines will be rather blurred between serendipity and info glut.
as matt griffith puts it:

My problem is information overload. I’m much more interested in seeing the same thing for RSS. Instead of blocking stuff I don’t want I want it to highlight the stuff I might want. I’ve been out of the loop lately because I can’t keep up with all of the feeds I would like to monitor. I need help.

highly agreed. my blogroll is growing, and time is always short. i’m inclined to think that with a better data set, automated metadata will go much farther. encouraging to see that some EU money is spent on this.

Quack

<miko> gregor you’re such a geek 😉
miko wonders when gregor will start speaking in XML
<miko> would be no problems with interruptions tho cos everyone would have to wait until the </comment> tag
<drew> That won’t be long.
<gregor> i would have long ago, but can’t make up my mind whether i should emit atrtributes or elements
<miko> and getting last word would be easy with </argument>
<miko> hehe oh dear
<gregor> hm, i should reserve a namespace
<johnny> gjr ?
<drew> gregorspeaksxml.com
<johnny> that’s not specific enough
<gregor> xmlns:quack=”http://greg.abstrakt.ch/2003/speech&#8221;
<johnny> quack 🙂 hehe
<johnny> you need a personal namespace though
<johnny> unless you wanna share your namespace with the rest of us 🙂
<johnny> then we can all quack
<miko> see we laugh but i bet he goes and does it 🙂
<miko> that would make us quackers
<gregor> the incentive is there for sure