fixing a macbook pro,

well after my screen died a couple of weeks ago and I was felling hopeless without a second laptop, well I ordered a new screen and followed these instructions. And I don't want todo it again, its and absolutly night mare and shame on apple for gluing the display in. it was a major piece of work to get the display off without bending the frame.

well 220$ and 2 hours later its working again and I can say I'm able to replace every part of a mac book pro now...

mai taih de nase

• 1 shot 10 Cane rum
• 1 shot barcadi select - brown rum
  
• 1/2 shot amaretto
• 1 shot triple sec
• 2 shots creme de cassis
  
• some ice
• fill up with orange juice

ridiculous sweet, but Lindsay likes it

yay homemade pizza is the best pizza and it's so simle...

I finally managed to get the cuda program together to compile under windows and osx, generate the libraries and provide a working linking for windows and osx.

Well, I also got accepted at source forge and so another project is registered and running, which should be developed over the next couple of month.

https://sourceforge.net/projects/jacuda/

Well we also got a new 42" tv for the living room and a new bedroom. Ok the bed and we still need to get some bookshelf's, but at least the first room in the apartment starts to show that I start to settle here.

another day with C, C++,cuda and Makefiles.

slowly it starts to come back to me, the whole ideas about pointers, memory allocation and the whole stuff.

And I have to admit it's fun to program in a procedural style and not object oriented. But boy it's a weird way to think...

also I'm only able to use the nvcc compiler, windows keeps complaining if I use gcc/g++/cl/cc for the weirdest reasons.

well time to write some C+ functions, move them to cuda and write java and python wrappers.

yay

live is certainly interesting, it was the year 2001 and I told my self I never use c++ again, since java is so much nicer and c++/c is so error prone.

Now it happened that I found a paper about cuda, which basically allows you me to run calculations on the nvidia 8800 GPU's. Drawback, you have to write in a mixed style of C/C++.

Well, atleast I learn a lot, but why should I use the GPU in the first place? Now I use a mac pro with 2x2 cpus or a ibm X60 with 1x2 cpu for my daily work and both have quite some memory (8GB Ram and 2GB Ram) so I should be set with this for my daily live. Now a GT8800 has one GPU which has 128 Suprocessing units, basically meaning that you got 128 CPU's todo work with you. Now you could combine several of these cards together, in my case i could use up to 4 of these cards in my mac, which would give me 512 CPU's.

Now I'm sitting here und starting to learn good old C again and already registered a new project on sourceforge, you need to get references somehow...

the name, jacuda.

What do I want todo? Provide a java/python/groovy/C API to run scientific calculations on cuda enable graphic cards. And hope to speed up some calculations with this at work. Like comparing 2k vs 2k bins, my java programm only needs 5h todo this and maybe this makes it much much faster. The final goal is to write a cuda implementation of some CDK algorythm, since I got the feeling that many people actually need this feature. Some algorythms of the CDK are just to slow compared to other programs.

i got the distinct feeling that my oracle server does not like me

ok we got a box with 4CPU's and a load of 332, it should had not been higher than 4 - 8 maximum...

top - 11:45:56 up 61 days, 1:14, 1 user, load average: 332.86, 332.35, 330.72
Tasks: 453 total, 334 running, 119 sleeping, 0 stopped, 0 zombie
Cpu0 : 8.7% us, 11.5% sy, 0.0% ni, 78.2% id, 1.4% wa, 0.0% hi, 0.1% si
Cpu1 : 8.0% us, 11.5% sy, 0.0% ni, 80.3% id, 0.1% wa, 0.0% hi, 0.0% si
Cpu2 : 7.9% us, 11.5% sy, 0.0% ni, 80.4% id, 0.1% wa, 0.0% hi, 0.0% si
Cpu3 : 8.0% us, 11.4% sy, 0.0% ni, 80.4% id, 0.1% wa, 0.0% hi, 0.0% si
Mem: 2355340k total, 2344524k used, 10816k free, 201168k buffers
Swap: 1048568k total, 19612k used, 1028956k free, 607180k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
9666 oracle 25 0 23044 7704 4096 R 8.8 0.3 183:34.83 perl
9143 oracle 25 0 23044 7704 4096 R 8.8 0.3 165:01.08 perl
14558 oracle 25 0 23044 7704 4096 R 8.8 0.3 162:30.51 perl
24927 oracle 25 0 23044 7708 4096 R 8.8 0.3 152:36.97 perl
13659 oracle 25 0 592m 58m 57m R 8.5 2.6 302:04.46 oracle
13997 oracle 25 0 601m 46g 46g R 8.5 2064.9 340:51.62 oracle
22099 oracle 25 0 92160 6600 5736 R 8.5 0.3 300:50.44 oracle
25858 oracle 25 0 23044 7704 4096 R 8.5 0.3 287:52.72 perl
28081 oracle 25 0 23040 7704 4096 R 8.5 0.3 278:28.88 perl
9844 oracle 25 0 23044 7704 4096 R 8.5 0.3 245:16.83 perl
10397 oracle 25 0 23040 7704 4096 R 8.5 0.3 244:08.57 perl
10962 oracle 25 0 23040 7704 4096 R 8.5 0.3 243:00.47 perl
12069 oracle 25 0 23044 7708 4096 R 8.5 0.3 240:51.97 perl
4691 oracle 25 0 23040 7704 4096 R 8.5 0.3 186:51.45 perl

wounder full gas is getting more and more expensive (4.35 right now) and I seriously started to consider selling it.

well after checking around I got a quote for around 1600 - 2000$ for my truck and so it's just not worth to sell it anymore.

seems like I'm driving it till it falls apart.

putting grails and maven together so we can use our dependencies and all are happy?

it took some time to figure it out, but thanks to this plugin, it actually works.

Here is a complete working pom

a couple of days later, I discovered some serious drawbacks with the oh so great groovy language

1. it's slow as hell, and I mean slow
   
2. no strong typing, i get so many class cast exceptions because I have no idea what kind of object is returned, so if you junit test what you write before you run it you are good, could be a good thing
   
3. the eclipse plugin for groovy plain sucks and is to slow, time to join the project and optimize it
4. hard to get the principle of closures, but can only blame my self for this, give it time

about the junits tests, normally you know you have todo them, but it just takes so much time to write that code. Now if you use groovy to write your tests and your java tests, you are having a lot of fun, since it's a matter of minutes to write a complete test to validate the parsing of an xml file.

and we all want to have well tested software...

groovy on rails = grails = cool

finally a webframework which just works. Not like appfuse, which turned out to be a nightmare or spring + spring mvc which was very nice, but the development was just to much xml configuration and so slowed me down a lot.

grail on the other hand, not one single line of configuration yet. It just worked...

Now that work calmed a little bit down and I'm now under wedding stress (actually it's pretty relaxed for me, i just need to decide which people we are going to invite and where are we putting all the people)

Anyway back to what I wanted to write about, a new day and a new idea. Last night I was sitting in the office till 4 am and trying to fix a problem. Of the spend hours the most time I was just sitting there and waiting for stuff to compile, deploy, start application server, install clusters, short I spend most of the time waiting.

Now who want's to wait, because he just changed some variables and has to redeploy everything to see if it works?

Not me for this reason I'm learning now yet another programming language. This time it's groovy.

Why?

because you can include into java or java into groovy, since they both generate the same byte code which will be executed.

And I don't really like to mess around with python ==> jython ==> java. It's well to much of a mess.

The next thing is, I have to write a website. Which I will no do with the hype product rails. Now rails is written in ruby, which I have zero intention to learn. Ruby is just to slow and I'm missing the point of ruby.

But well there is groovy on rails, which seems to be a very cool package.

this week one of my coworkers approached me the question:

can you please take a nice picture for my poster with wine bottles and so?

So I sat down and starting shooting away. Firstly I wanted to do some low key pictures since they did not tell me how the color scheme of the poster is. But after I saw the poster it was clear, we need a high key picture, so that it matches the background.

Anyway here are the two best shots of this experiment and I just love the high key pictures.

Collections can be found here

• low key
• high key
  

just playing around with two flashes, which overheated after 20 minutes or so since we got like 90F in our apartment.
And lets say so the AC is not keeping up...

as promised here is the M&M picture.


sadly they were already kinda melted and the font is pretty ugly.

But hey it worked!

-correction they worked for 3 month than she ended it -

all pictures provided by the blog belong to Gert Wohlgemuth. For reproduction and publication please ask me before you use them. If the pictures do not belong to me they will have a caption or a link to the owner.

thank you very much

BinBase,

the project i'm working on since years is finally at a stage were people start to ask me questions about it and I had today the opportunity's to give a small talk about the basic idea of the system.

• white shirt + dress shirt at 40 degree Celsius bad idea
• to breath during a talk is a good idea

I also discovered the eclipse zest api and promptly integrated it into the BinBase system. The first idea I got for it was actually to visualize dependencies of compounds to species and other way around.

bbq season started and so a small post to make the perfect medium rare tri-tip steak.

The thing about tri tips is that they are so big that you can have all kinds of rareness in the steak at the same time. So there is a part which is perfect for everybody. Normally the ends are well done and the more you get to the middle the more rare the steak becomes.

Anyway here is a recipe how to cook it on a gas bbq.

1. get a 1.5 - 1.8 lbs tri-tip steak
2. marinate it for 12 - 24h in a marinade of your choice
3. preheat the bbq as high as possible
4. drop the steak on the bbq on very high temperature and let the lid open
5. wait 5 minutes
6. turn the steak and let the lid open for another 5 minutes
7. reduce the heat to medium low and turn the steak
8. close the lid for the next 10 minutes
9. flip the steak
10. close the lid and wait another 10 minutes
11. now its done.
    

afterwards place the steak in a wooden board and let it rest for 5 - 10 minutes before you cut it. Since we don't want that it drys out.

enjoy.

the lack of code quality in open source software, particular maven.

• why is it that you can't download a program, follow the installation procedure and it just works?
• why is it that a program version is marked at stable and released to the public and does not do what it's supposed todo?
• i don't expect it to be flawless, but it should provide some working results
• why is it that maven 2.0.6 is not compatible to the plugins of 2.0.5?
• there are standards to version software
• a 1.0.2 version has to be able to work on the data of a 1.0.1 version without errors
  
• a 2.0.0 version does not have to be compatible to a 1.9.9 version.
• a 1.2.* version should be compatible to a 1.1.9 but it doesn't have to be
  
• why does maven fail to install particular packages in a multi module project
• i use the software and build software it, it makes me look bad if I have to work around bugs in very unprofessional ways
• a user should not be expected to get a documentation for compiling a software like this
  

short why is there not a certain amount of code quality in maven, every time I use a new plugin I spend at least 2 hours to work around it bugs and get it started.

For example eclipse does not have these problems, they have a stable system to release projects and to make sure that they are stable. Why can't be maven like this?


What is the result of these issues? I consider to download the latest maven sourcecode and start to develop my own maven2 fork. just to get to a working standard which works like it's expected.


...yes I hate to waste my time on things I shouldn't have to, specially if it's my weekend.

Now that my mac book is dead an I'm frustrated for other reason I decided there is only one thing todo,

time to buy toys!

In this case I finally bought a second flash for my camera and start learning to use it.

My first idea to shoot throw a glass plate did not really work that well. Basically there was barley any difference to using only one flash.

Now using it as a rim light, is another story, I actually really like the effect of a rim light in this case.

time todo dishes.