I went and bought the Samsung 2032WB I spotted some time ago.

Previously I had 15 inch Viewpoint CRT monitor only attached to this computer. Since it celebrated its 10 year birthday two months ago, you can imagine it being rather old already. Surely, lately it has had some weird pulsing issues where the image would suddenly contract and widen back to normal.

This new TFT was quite a kick to the wallet of a student, but this really seems like a quality display and actually is one of the cheaper ones that still offer DVI connection.

I still got the 15" one as my secondary display. I could for example put some chat windows or console log screens on it when I am playing full screen games. Of course there is still the free Dell monitor from University not connected to anything and my laptop sitting in the corner, serving as IRC and Xfire 24/7 watchdog.

Here's a photo of my current working corner:

(click the image to view in bigger resolution)

Just recently Xfire updated their website to have this new wide look. It fit just nicely to my old monitor at 1024x768 resolution. Now that I installed this new monitor with 1680x1050 resolution, the site looked so narrow when Opera was still maximized window; the blue page part in the middle with double the black space on both sides.

Well, from now on Opera shall not be maximized all the time, there is enough space to have more windows open around now

I have yet to test this new beauty in gaming and I wonder how well the games work if I put them to the native resolution of 1680x1050. Well, at least the 2ms latency time should not be a problem for gaming.

ADD: Oh yeah, forgot to mention originally. The display also came with a driver CD-ROM. I wonder what those drivers would do as the CD didn't work. It took 10 minutes for Windows to actually be able to list the files in the root of the CD and after another 10 minutes it still couldn't actually access any of the files, so I just removed the CD and thought of going without the drivers.

Yeah, I could possibly...

After fixing some compatibility problems with the latest Service Packs, Microsoft re-released the service packs for download. The SP3 for Windows XP also finally appeared to the Automatic Updates channel so everybody who has Windows XP with at least SP1a installed is now able to update to the latest state.

This blog entry is about my experiences of installing the latest service pack to two computers. This computer that I am using now had Internet Explorer 6 and Windows Media Player 11 installed prior upgrading. My laptop had Internet Explorer 7 but no Windows Media Player. Also a relevant info is that neither of the systems had Windows Messenger or Live Messenger installed.

Well, the installation on this computer went painlessly. I downloaded the network installation/for IT Professionals installer since the one from Automatic Updates didn't even download (that's why I also had only IE6, I did get IE7 through updates (actually 5 times so far!), but it never installed it).

Interestingly the download page for SP3 was labeled to be for IT Professionals and Developers. However, it seems like IT Pros nowadays are quite bad with computers. On the "Thank you for downloading" page it states instructions for saving the downloaded file, as if the people downloading didn't know it beforehand.

Also, what benefit does SP3 for Windows XP do if you do not have Windows XP? That's why the clever guys also included a "recommended download" link for Microsoft Windows XP, as seen in this screenshot: http://screencast.com/t/DUNnFcaR8

Well, since I am not a certified IT professional, the downloading was an easy task for me. I also uploaded the installer to my laptop after finishing the download on this computer, so I could update there also.

As I said earlier, I didn't have any problems installing SP3 on this computer. It just did backups, installed files and then asked reboot.

After restarting computer it asked me whether I want to enable Automatic Updates or not. Well, I have always had it on...

Today I held a small presentation in the course Academic English for Mathematics and Physics students. The subject was anti-gravity. As some people seemed more or less interested in this subject yesterday, I thought of writing about it on my blog.


Gravity is a force we all need to interact with every day. What gravity does is that it pulls two masses closer to each other. Measuring the spectre of distant galaxies the scientists have seen notable red shift telling us that the universe is expanding.

By the span of time it is apparent that the universe is expanding at accelerating rate. Now if there was only gravity trying to do the opposite, how could it even be possible that the expansion would get faster? There must be some kind of force pushing the galaxies apart from each other, doing exactly the opposite what gravity does. This force could be logically called anti-gravity.

To get to the physical characteristics of anti-gravity, let's first see those about more commonly known force, gravity, first. In classical mechanics gravity is generated by mass. Albert Einstein with the famous formula E=mc² shows that mass is just a form of energy. We can say that it is energy that creates gravity. This energy could be mass or vector fields such as electric fields.

Gravity can also be related to pressure. The pressure on Earth is so small that it does not cause any noticeable gravity force, but take for example a massive collapsing star where the pressure could be millions or billions of that on the Earth's atmosphere. There it is strong enough to actually form gravity.

Now let's look at anti-gravity then. It could be caused by something called scalar fields. No evidence of them existing has been found yet, but they have the characteristics to form negative pressure, which in turn would cause force opposite to that of gravity. Since scalar fields also have energy, they also create gravity. The point is that the pressure by scalar fields is big enough to form three times more anti-gravity than its energy forms...

Just in a few days a couple of web browser updates were pushed by both Mozilla and Opera:

Firefox 3 beta 5

• Over 750 changes from previous beta
  
• Security improvements
  
• Better integration to the underlying Operating System for more native look
  
• Smart folders with backup options
  
• JavaScript speed optimizations

Opera 9.27

• Security: Fixed an issue where newsfeed prompts could cause Opera to execute arbitrary code
  
• Security: Solved an issue where resized canvas patterns could cause Opera to execute arbitrary code
  
• Security: Improved keyboard handling of password inputs
  
• Stability fixes relating to BitTorrent transfers and Acid3 test
  
• Fixed the print dialog on OS X (Leopard).

Opera Mini 4.1 beta

• Improved speed (some pages load 50% faster)
  
• Find in page -feature
  
• Web address auto-completion
  
• Download files including web pages for offline reading
  
• Upload files in web forms such as attachments for webmail

On a more or less related note, the Epiphany web browser used in Gnome desktop environment will drop support for Mozilla's Gecko engine and will continue using WebKit exclusively. (source)

So PWN3D's website and my t3H 1337 was hacked on the 1st of April 2008. Everything suddenly changed to be like l337-sp33k.

Thanks to all those freaked out people in WH1N3RZ it appears that this year's April was - as they sing in Portal's ending song Still Alive - A HUGE SUCCESS.

Many people requested for n00bz! as their R0X0RZ were messed up, the names of all the 0WN4GE were changed.

Why stop it on website, though? Take it to the client as well! I took some screenshots of how my Xfire client changed during this hectic day of amusement, you can find the story with related W4R3Z at this topic which was locked in one point by [xf1r3] Mf1V3 and then reopened by the power of the haxor #0MI3Z.

As apparently my client was hacked, also my name was changed to 0WN3D! for the rest of the day with a custom status saying T#15 4C(0UN7 \/\/45 H4Xt. There were some people concerned about my safety, thank you for taking care of me. Here's one of the chat logs:

Quote

ConV3r5a71On 57Ar73d 837W33n 0WN3D! ANd =scum= Phreak[AoG] 13INF: Tue, Apr 01 19:24:24 2008
[19:24] =scum= Phreak[AoG] 13INF: Wow, not really hacked tbh, you guessed the password?
[19:24] =scum= Phreak[AoG] 13INF: So technically not hacking
[19:24] 0WN3D!: scum
[19:24] =scum= Phreak[AoG] 13INF: Yes?
[19:25] =scum= Phreak[AoG] 13INF: you must be really cool?
[19:26] 0WN3D!: too much for you
[19:26] =scum= Phreak[AoG] 13INF: you must have alot of friends, if you take time "guessing" peoples passwords
[19:26] =scum= Phreak[AoG] 13INF: Especially on xfire, thats quite sad, is it not?
[19:26] 0WN3D!: you are
[19:26] =scum= Phreak[AoG] 13INF: I am what?
[19:27] =scum= Phreak[AoG] 13INF: So was it hard to, guess the password?
[19:27] =scum= Phreak[AoG] 13INF: If your that "pro" a "H4x0r" then hack me please?
[19:27] 0WN3D!: you have no idea what you are talking about
[19:28] =scum= Phreak[AoG] 13INF: Well apparently you "hacked" that account, am i correct?
[19:28] 0WN3D!: that one you got right, almost
[19:28] =scum=...

The project of this night for me was to code a C function that returns SHA1 hex digest of a string.

I know that there are already several free implementations even with source code with various licences available, but all of them seemed overally complex to me.

Trying to find out how the SHA1 algorithm goes from those source codes was just pain, so I digged up the definition of how to calculate the digest. Transforming the definition there to C code applied to my special data types where this is to be used it was actually quite easy.

There were some things I had to debug, in the end the problems were that I forgot to multiply the length with 8 to get bits from bytes and that in one of the initial constants two numbers were swapped by mistake.

There was also a complicated case of Big Endian vs. Little Endian integer presentation that I replaced with explicit bit shifting operations.

So now I have a working, straight-forward function that takes a string as a parameter and returns a 40 byte string containing the SHA1 hex digest of the given string.

And the source code has less lines than in any of the other C implementations I found with Google. Keep it simple, stupid!

Update: but Safari beat Opera due to error in Acid3 itself, see comments.

Opera developers have been working hard on improving dynamic web content handling in their browser since the Acid3 test came out.

While the test itself is new and not completely finalized, most web browsers including Firefox, Opera and Safari (Internet Explorer is still stuck somewhere below 20%), have been improved in internal development to try to reach the full 100 points of this test. Previously Apple's Safari held the record of being the first browser build to get over 90 %, but now Opera with their internal test build took the grand full score.

While this Opera build is only a test build and not publicly available, this marks some sort of milestone in web browser development. A technology preview of this new build will supposedly be available next week.

See the Opera Desktop team blog for more information.

I thought it was easy from the past experiences. I am not sure if these were bugs in the linker or what, but I had some weird problems trying to get a bit more complex code to compile into a DLL file.


Problem 1: Undefined (*UND*). Symbol

I got these very weird errors, stating some symbols I declared being dllexport were undefined. If I didn't have __declspec(dllexport) then those symbols were found fine by the other internal functions, but I wanted them to be usable from external applications loading the library.

The cause seemed to be that for some reason you cannot have a reference to the function before it appears in the source code, e.g.

Code:



That would say that myexport is undefined. OK, in that example it is as easy as to move the function to be before it is first called, but my situation was more complex; I had two externalized functions that called each other, so no matter in which order I had them in the source code, it would always complain on either of them being undefined.

As the code is supposed to be platform-independent (it compiled and linked just fine under Linux and Solaris already, just Windows was problematic) I made some dirty preprocessor hacks to work around this annoying error.

Since this error only occurred when the functions were __declspec(dllexport), I made one of the problematic exports a normal internal function with different name and made a wrapper with the original name.

This also included that I renamed the function call to be to the internal function everywhere inside the library to prevent extra function calls.

In the private header file (one used when compiling the library itself, not the one to be given to programs using the library):
Code:



Now in the main program the actual function is named myexport_internal and the wrapper is created conditionally:

Code:

So I got a new 320 GB Hard drive to replace my half-dead 160 GB one. Now that I have more space I could experiment with new space-hungry programs. One of them is andLinux, Ubuntu Linux inside Windows.

While it works quite well (once you get it to work...) do not expect to run video players or fast graphical games through it; the X server does not contain all the features the full X server in stand-alone distributions usually do and all the data exchange between Windows environment and the Linux system is done using virtual Ethernet link.


(Click the image for full screen shot)

If you try to play videos at 30 fps, that just won't work too well. It used 100 % of a single core on my CPU and the video was lagging behind the audio unless frame drop was used. And that was with a relatively small video (some 320x240 resolution I guess). With a bigger video, or full screen, you can see it update the video window in blocks, it is that slow to emulate X window system without direct access components.

But I mainly am going to use it to practice some Linux commands and most of all to compile my C programs in Linux environment to see they really are platform-independent and manage to compile with GCC on Linux.

While the link is still valid: http://www.xfire.com/xf/modules.php?name=Forums&file=viewtopic&t=168273