Tuesday, April 26, 2011

virtual keyword

If you want to declare a different behavior for a method in your derived class - compared to a method declared in the base class, you should use the virtual keyword. This way, you are allowed to adapt your derived class to specific conditions that requires changes in the base methods. Yeah, use virtual, ok? (So you gave the impression that you know everything about the class declarations in C++. I don't buy it alright!  -C'mon, I know that stuff ok - virtual is important :s)

Monday, March 14, 2011

mod+#

Once more, "mod + #" (i.e #=1:9, don't be such a geek, just say # represents any number from 1 to 9) does only change your desktop; it does not take you back to the previous file or web page or other program or the iPhone that you hold one minute ago. It just changes the view. I know it looks completely ridiculous for you if you are not using xmonad (nobody in your audience uses xmonad you moron; do you even have followers, anyway? - I do have ok. He is, no she is a friend of mine! - I don't believe. You even don't have an imaginery girlfriend  - Yes you are right, it is because I have imaginary girlfriends) (Hey, I'm just kidding, ok! - No you are not.) but for those of you using xmonad might have experienced that (you are even not sure, you just say "might"; this ''might" happen only to you my friend).

So, it is like you were using ctrl+z to take back your special friendship (no, you did not have such thing) or or, like, err like you were using it to reverse your mistakes or - oh this is just lame. This even does not deserve any discussion at all. All you read is just a debate of mine with my inner voice.

Next time, a discussion on Hg perhaps. Kinda useful, right? (right...)

Tuesday, March 1, 2011

xmonad

My colleague Cyril suggested (and I guess earlier Yves suggested to Cyril) me to use Xmonad instead of Gnome. At first,  they both were like - dude you should install it: So I installed it. They suggested me some layouts and Cyril shared his setup and tweaks and etc. We started xmonad together and I used it straight  - two consecutive days (this is because you couldn't figure out how to log out - no, no I knew it - it was Mod+Something+SomeOtherThing). And yes, when they saw me xmonad they should have automatically assumed that I liked it. Well, not really. Actually, yeah; it was cool to use an interface like in sci-fi movies but where are the desktop icons, and where is "my computer"?

Now, after long days - thanks a lot guys. Xmonad it the best. It rocks, literally (no rock music of course. or maybe you couldn't figure that out, as well - huh pal?). I understood that the graphics driver was the issue on my computer  - xmonad was responding slowly. As soon as this issue is solved (you realized that after two months of course. but hey, it was such a lightening moment) xmonad starts to perform very well. Even, I changed the console into a transparent mode and used a nice background (which is listed here) via feh.

Here is my current (terms are subject to change) display:



What are those things written on the screen. Trilinos. More on Trilinos to come.

Monday, February 21, 2011

typeinfo for Epetra_VbrMatrix error

Stupid Trilinos error, but I was more stupid. The reason for that (as I found out  -no I found it out) is the order of the libraries. This case was a simple linear solve based on Epetra and AztecOO. I knew that there was an order to call the libraries but since my all previous cases were functional - I forgot that issue completely. The solution is to put "-lepetra" after "-laztecoo" (not after -lamesos you moron, don't confuse the packages that start with the same letter. - yeah, sorry about that) in your makefile and you are done. I believe my all other trilinos based compile errors was because of this ordering. Now I'm gonna try them as well. God, please let them compile smoothly, as well.

...
...

Yes, yes. It does. I'm crying right now (no, you're not).

Thursday, February 17, 2011

parallel vs. sequential

One thing clear,
I'm working on parallel processing, and I also have some stuff waiting to be sent to some journal yet I have a sequential life! What a dilemma :|  I wish a week for me would be something like,
Monday
Monday
Tuesday
Tuesday
...
Saturday
Saturday
Sunday
Sunday


So, I could have been doing my current research parallel to my yet to be completed studies. Unfortunately,  MPI_COMM_WORLD does not print the weekdays,  only  if(myPID==0) ...

Where to start?

"Hello world!" (I hate this phrase)

Why to start into such a blog? (because your all other previous blogs failed to draw attention)
Here are the facts about me and my current situation:

Catching up with the guys in computer science is not easy if the staff is good (they are the best) and you are not a computer scientist but an engineer, mechanical. I should admit that I worked on Computational Fluid Dynamics (CFD) in my PhD which brings me close to computational science. So, I know a lot on iterative solution techniques, nonlinear equations, preconditioning, programming and also used - tried various commercial & open source software. On top of that, since linux is a must in computational science (this is true, who does everything in windows is not a computational guy, sorry. Yeah, Mac counts up to a point, but fails is advanced levels ok. I'm aware of the fact that Apple products look cool - like iMac. So, I got myself an iPhone even if I was keen to have an HTC just for Android - I use google services very often) I worked with Fedora, Suse and Ubuntu (was my favorite but not anymore even if I'm currently using it) hence I have some experience on being an intermediate OS user. If one should have to select a mechanical engineer into computer science - that would be me (This is not being arrogant but being confident about yourself and giving the value you deserve). At least, that was my intention. However, I don't fully agree with my own opinion before; now I see I know crap! 

Well of course, the people in my new department comes from a different disciple - they even don't call them engineer. So, it is quite natural that they are way better on various aspects of computing: databases, networking, protocols, useful tools - commands. A number of programming languages. Bla bla and bla. So, how can I compete with that? Hey, I know entropy! - So what?, could they say. Entropy does not help me here, probably would not be anyway. Most of the experience that I got was from my struggles in computing. I became an expert of trial and error and asked myself "Why?" to question the things that was presented me. That is, ladies and gentlemen, is vital to improve oneself. Still, in some cases I look like a dummy. Apparently, all of precious my efforts was not enough to go step higher - was just enough to get here - into my current position: post-doc.

Yet I'm a quick learner. Yet an important part of learning is to teach, at least sharing what you have learned and telling the new info in such a way that you are talking against a class full of students (any girls in the class - not sure, sorry. damn.) Yet I like blogger. 

Summing up, here I post not only the stuff that I know then and believe that is imporant, also the tricks that I learned from my journey, here in computer science, will be a part of this new blog of mine. New approaches,  methodologies, jokes :b (you are funny, already - yes I am, right? - yes you are - don't be that insecure). And also a bit of Zurich and Switzerland (girls? - oh please shut up now.)

Before finishing up this first post, one thing you should now is that I'm not crazy but there might be some frequent encounters with my inner voice (that's me. - yes you). You are free to ignore him (no no please, not) but time to time he can talk about some important stuff or look things from the other side (yes I do that).  Still, don't freak out, ok?

Where was I, ahh. Where do I start?