The beaver is a proud and noble animal
Notes from a bemused canuck
I'm going to be adding a few more frequent flyer miles to my body. I'm going to 3 conferences this year:
– “HABATOS” consortium meeting in Lodz, Poland (June 25th). BTW, HABATOS means “Healthy Ageing by Adaptation to Oxidative Stress”. They want me to give a talk on PRIDE and how it can be used to store their data.
– PSI meeting in Washington DC (September 25-27)
– HUPO meeting in Long Beach, CA (October 28-November 1)
Going to be a busy little beaver for a while.
Phil and I have been working on cleaning up the PRIDE codebase. It's long overdue for a good overhauling because as we've been doing this exercise, we've found a cubic assload of code whose only apparent purpose is to be tested (meaning that it's only used in the unit tests, but nowhere else in actual production code and as such, is completely useless, really).
So for the last few weeks, I've been combing through 191 java classes (35562 lines of code), pruning out bits of junk, improving code structure and usability and generally going nuts.
It's tedious work because as soon as you fix something in one place, you need to cascade the changes to everywhere else that uses the bit of code you just re-wrote. It's the worst type of IT work: one that is boring as hell but requires a lot of concentration to figure out what stuff you wrote 6 months ago (or worse, was written by somebody else) is supposed to be doing and how it should use the improvements you're making as you go along.
At least I have the gay dance to keep me sane.
The work I've been doing for the last week, trying to massage somebody else's data to make a square peg fit into a round hole, turned out to be a complete and utter waste of time. The data is shit, corrupted and inconsistent and I have just informed the person who gave it to me that it is so. I could say that I hope he doesn't hate me for it, but honestly, that would be a lie.
Day 2 of the PRIDE/OLS Hackathon is done and over. We've been inducing small brain seizures in 10 people for the last 2 days with PowerPoint overdoses. The good news is that we have a good spread of cookies and tea and coffee. The bad news is that we still have more powerpoint.
On the plus side, this exercise has forced us to write tons of good documentation – which any decent programmer is loathe to do. This will also be a good preparation for the Thursday morning talk I have been volunteered to do in a month for the whole sequence database group. But I digress.
The hackathon has also been a good training ground to see what kind of mistakes/problems people will experience in the wild. We know how to run the programs – we should, we built them from the ground up. What takes us a little less than an hour to do from scratch is proving to take more than a day for the average user. Mostly, it's because everything works on our laptops. In the wild, we need to deal with older/newer versions of stuff we need, different languages, different operating systems, different ways of thinking of how stuff should work. It's proving enlightning. And frustrating. But that's a whole other story.
On a completely different train of thought, I think that we've actually beginning to be regulars at the pub. We had dinner with the course people there tonight (paid by the EBI). I wanted to bring back a chocolate tart for Katy (as she was feeling a bit down) so I asked for one to take away. The waitress went to place the order and came back a minute later saying that the chef wanted to decorate the plate (as they normally do when you eat it at the restaurant), if they did garnish it prettily, would I mind bringing the plate back tomorrow?
In a word… sa-weet!
The Ontology Lookup Service, a lightweight cross-platform tool for controlled vocabulary queries
Richard G Cote, Philip Jones, Rolf Apweiler, Henning Hermjakob
BMC Bioinformatics 2006, 7:97 (28 February 2006)
Total accesses to this article: 246
Accesses within last 30 days: 246
Your article is also available through PubMed Central, and overall statistics indicate that your article will have been accessed on PubMed Central a roughly equivalent number of times, doubling its total visibility.
This makes me happy.
you're off email for 2 days and you have 115 emails in your inbox.
I've been answering emails for the last 3 hours and I'm still not done!
Dear brain.
I'm sorry I put you through the last 2.5 days. Although I'm glad I managed to get a last-minute spot in the java design pattern training course, I didn't expect to see this much material this quickly and I apologize for dumping all this stuff on you at such short notice. I completely understand your sudden desire to shut down to prevent circuit overload. If you can wait just another hour, it's going to be the weekend and we can relax with an assload of Buffy the Vampire Slayer.
Thank you,
me.
One of the things that I like about my job is that I usually don't need to waste time fiddling with system administration.
Usually.
Except for now.
I have a bitchin' piece of code that works perfectly on my laptop setup, where I am the master of my domain. I can install dependencies on the fly and make sure that the latest libraries are there.
Of course, this is not so on our production machines where I am a lowly user and at the whims of the BOFHs.
For the past 2 days, I have been trying to compile and install a piece of software. For the past 2 days, I have been trying to compile and install the libraries that I need to compile and install the first piece of software. This would be so much easier if I had root access and didn't have to install to non-standard paths and having to ensure that the friggin compiler uses the latest versions of the libraries, not the obsolete ones.
I need a hug.
I'm applying to a part-time job as a technical copy writer for a biotech firm here in Cambridge. It's 5-10 hours a week, and it might give us some more leeway in trying to get the capital we need for a down payment on a mortgage. In that vein, I have updated my resume. You can see it here if you're curious: http://www.flubu.com/resume/rc-cv-2006-en.pdf
It ain't looking too shabby, if I do say so myself :)
My paper has been accepted pending minor revisions.
Woohah!