tjaard.nl

Has been a while since I last posted something, hasn’t it…? I finally found a tad of time in which I thought it would be a good idea to fix the permalinks of my blog. My hosting provider upgraded their servers and initially they forgot to transfer hidden files, among which, shockingly, my .htaccess files. They were really helpful in recovering it, but I was a bit too busy with all kinds of things to do the final bit, namely putting it in place correctly and with the right permissions :) .

But enough chatter about my laziness… Since two months and a half, I’ve been working at my new job! Since 1 June I’m working as a technical consultant for ORTEC in Groningen. I am active in two different teams here: one maintains specialised gas field simulation software for the NAM and the other offers planning solutions to a variety of customers. I must say that I am quite happy with this, as the things that I am working on are very diverse. I’m tracking down bugs in pieces of Fortran (that may be over 30 years old) on one day and may be working on ASP reports the other. At the moment, most of the things that I do are aimed at getting me to know the software better.

While I may spend quite a lot of my time programming, I will also be involved in architecture design and consultancy. Developing myself in these things is exactly the wish that I expressed when I applied for the job and these are the things that I am going to get. And so far I got immense freedom to come up with ideas and to learn about customer relations, contracts, cutting good deals… In many respects I feel that I want to be able to develop vision and to think a few steps ahead to get customers exactly what they need and I feel that this is appreciated by my co-workers.

Ah well… a couple of days ago I realised that I am really glad that I did not choose a career in academia and got this job instead. I couldn’t see myself working solely on another thesis for four years and rather wanted to develop myself so that I can be of direct use to others and as a bonus I got a really nice atmosphere and great colleagues

I may share things that I learn every once in a while here – that is, as far that doesn’t cross company interests ;) . But this is all work-related stuff for now, folks… cheers!

Finally, I’m done with all the studying kerfuffle!

Had my colloquium a couple of days ago, got a nice grade… (9/10) and today I had my thesis printed and I applied for getting my diploma. Which I will receive in June… Apart from perhaps helping another master student with some stuff I made I’m done. Here is the abstact of my thesis:

When one designs an architecture, this involves a lot of knowledge about this architecture. Such archictectural knowledge (AK) often remains tacit, but it has been shown that making it explicit and explicitly using it can greatly help a qualitative architecting process. The nature of AK in quantitative architectural analysis is still largely unknown. This master thesis investigates this matter in an industrial case study at Astron. Astron is an institute that develops radio telescopes. Modern radio telescopes are very large and complex embedded systems. Their development involves a lot of quantitative analysis, of which the results are used to make design decisions. The various parts and aspects of the developed systems are analyzed by domain experts. Astron currently experiences problems in their analysis and decision making process. A lot of AK remains tacit, which makes it hard to share analysis between domain experts and other stakeholders. It also makes it hard to trace back the reasons for design decisions. This master thesis presents an analysis of the analysis and decision making in Astron’s development process. The types of AK involved in the various activities within the process, as well as their relations, are condensed in a domain model, which provides a common ground for capturing and sharing all relevant AK. To facilitate the capturing and sharing of this AK, The Knowledge Architect tool suite is proposed, which implements the domain model. Part of this tool suite is implemented in The Knowledge Architect Excel plug-in, which provides a means to make AK explicit in Excel-based quantitative analysis models. One asset this plug-in has is a visualization of the contents of an analysis model which should provide analysts with extra and quicker insight into one another’s analysis models. To which extent the plug-in manages to do this is validated in a controlled experiment.

If any of you people would perchance care to read the whole: my master thesis! Yes, I know that there is a grammar error in the abstract in the pdf… but hey, I had to make the thing final sometime and it was too late when I noticed…

As you already might have read, I wrote a bit about all this a while ago, so you’ll be able to read more when you follow the link. And to be complete I perhaps should mention the group’s home page again as well as a direct link to the tool I made.

Have fun! I’ll have… finding a job ;) ! PhD, no PhD… choices…

I have not been blogging a lot as of late. I’ve been rather busy with things on one hand and did not feel like writing much on my blog on the other. My master thesis work for example is keeping me rather busy. But it is starting to pay off: my thesis is nearing completion, the tool I’ve made has now been put online, the technical report about an experiment I did with this tool has almost been cleared so that it can be made public and I’ve helped my supervisor write a paper about my work that has been submitted to QoSA. Now let’s hope it gets accepted :) .

Since I haven’t really yet told a lot about exactly what I’ve been doing I’ll try to explain the scope of my work in short. My assignment takes place within the context of the Griffin project. The Griffin project is all about so-called architectural knowledge (AK in short). AK is knowledge that people use when they design systems and can incorporate all sorts of things, such as requirements, design rationale, decisions… Usually, such knowledge is used implicitly, but it can be very beneficial to make it explicit so that more knowledge and insight about the architecture and the architecting process can be shared.

My assignment specifically looks into the AK involved in quantitative architectural analysis and in this light I’ve done a case study at Astron; a research institute on the moor in Drenthe (nice environment :) ) where they develop really fancy radio telescopes. I’ve researched what kinds of AK are present in their analysis process and have developed a tool with the intention that it helps them with this, preferable by presenting the AK of the analysis model in such a way that more insight into this analysis model is gained.

The tool is a plug-in for Excel that enables the user to make AK explicit by adding semantic information to the cells. This information then can be extracted from the Excel file using a piece of software which puts the AK into a central repository called the knowledge base. Other tools then can connect to this knowledge base to perform all kinds of neat operations on the captured knowledge, such as relate pieces of analysis models to design decisions. The knowledge base namely not only stores information from Excel; all kinds of analysis and architecting tools are envisioned to be connected to it so that one central repository is created from all AK produced in a project. So far, for Word (for architecture documents) and Excel plug-ins have been created and a new master student just got started investigating the development of a plug-in for Python.

The tool is not just about making annotations. As I mentioned, a goal is also to present the AK of an analysis model in a attractive and insightful manner. That’s why the tool offers a fancy visualization which displays a graph that reflects the structure of the analysis model and allows one to investigate which relations exist and which cells actually depend on one another. It can be downloaded from the SEARCH Griffin page (or directly here if you’re lazy). A manual is included, so by all means toy around with it if you like. I’ll probably post up my master thesis within a couple of weeks, blessings until then :) .

Wereldgelijkvormigheid is een begrip dat de meesten van ons waarschijnlijk wel zullen kennen. Johannes schrijft dat we de wereld niet moeten liefhebben (1 Joh. 2:15). Jakobus spreekt zelfs van vijandschap tegen God als je voor de wereld kiest (Jak. 4:4). Jezus heeft zelf gezegd dat we een lichtend licht en een zoutend zout zijn. Om kort te gaan: wij, kinderen Gods, zijn niet net als de rest van de wereld. Wij zijn anders. Wij volgen Jezus en niet onze eigen begeerten en strijden ervoor dit elke dag weer te zo doen. We zijn radicaal en niet lauw of halfhartig. En ook wij genieten uiteraard van de goede dingen die er zijn en nemen regelmatig de tijd voor ontspanning.

Zoals bijvoorbeeld bij een leuk filmpje. Beetje actie, al dan niet op een komische manier, moet ook kunnen. Zolang het bloed niet in het rond spettert en het geweld niet buitensporig grof is kan het prima, onschuldig vermaak zijn. Maar er is wel een probleempje. In de gemiddelde Hollywoodproductie wordt ongans veel gevloekt. Het f-woord staat uiteraard met stip op één, maar het komt ook vaak genoeg voor dat de naam van onze Heer ontheiligd wordt. En dit gebeurt in het hele spectrum van genres dat ons aangeboden wordt, waarbij een vleugje losse omgang met seks en propagatie van ander onchristelijk gedrag ook nogal eens voorkomt.

Wat kan een arme christen doen? Ik heb laatst een film gezien waarin welgeteld één keer het derde gebod op grove manier werd overtreden toen iemand schrok. Die film was verder best leuk… Maar ik denk dat we hier ernstig moeten nadenken over wie we willen dienen. Wat weegt zwaarder? Onze relatie met God of ons gerief door een stukje vermaak? Wie het weet mag het zeggen…

My brother (biologically ;) ) Jan is a strange guy in some ways. One of which is the irresistible urge to toy around modifying all sorts of consumer electronics. There are days on which I have the impression that his screwdrivers and soldering iron are his best friends. Which of course is very convenient when you need to have something fixed yourself ;) .

Jan spent quite a lot of time making his Gamecube more visually appealing. I believe he has finished this about a year ago now… he’s more apt handling hardware than handling code (quite my opposite, actually), so getting the site up took a while longer. The code may look awful, but it’s the contents that count, isn’t it ;) ?