“A Russian businessman has trademarked the emoticon — or combination of punctuation marks — used to convey a wink in text messages and e-mail.“

http://www.google.com/hostednews/ap/article/ALeqM5gE4Zc6sqIDqyfBQ1WlTKBf5GdInwD950NQ680

http://mnweekly.ru/business/20081211/55360588.html

This is fun. here are a few I made:

Takes about 10 minutes (on my 1.8GHz dual-core) to render each of these at a res of 1680×1050 at 1000 quality, but w/e, they make cool wallpapers

http://www.mugglenet.com/app/news/full_story/1973

Watch the high res one. It’s really AWEOSMEE!

http://www.theglobeandmail.com/servlet/story/RTGAM.20081101.wpalintranscript1011/BNStory/usElection2008/home?cid=al_gam_mostemail

It’s a 22-inch Dell E228WFP. Not actually from Dell though, got it at Future Shop for the same price. I didn’t bother trying to do a dual-screen thing with my laptop ’cause the laptop’s screen is too small and it takes too much effort looking from one to the other. The new monitor runs at 1680×1050 while the laptop is 1440×900 (the laptop screen is 14-ish inches btw. I think. Actually it might be 15. Whatever, who cares.)

In a previous post I showed how you can make an HDR image from 3 exposures - which is considered the “proper” way to do it. However, I also mentioned that this is possible using only one exposure. This is done using the RAW file created by the camera (you usually need to enable a setting to make it do that). The reason for this is that a RAW file stores much much more colour information (hence the larger size, usually 3x that of the JPG for the same photo). A JPG can only store a limited amount of colour information (24 bits is around 16 million colours). Now, if you don’t intend on doing anything with the image, then keeping only the JPG file is just fine - since computer monitors can’t display more colours than that anyways. However, with good cameras, the light sensor is actually a lot more sensitive and captures more detailed colour information than that which can be described by a JPG file (ie. 16 bits per channel rather than 8). Using the additional colour information available in a RAW file, it is possible to recreate conditions necessary to make an HDR image.

To my knowledge, Photoshop cannot do this, but Photomatix can. You just ‘drag-and-drop’ the RAW file (extensions vary, btw. My dad’s Nikon makes them .NEF, Canon I think does .CRW) into Photomatix, and it informs you that it has converted your RAW file into a pseudo-HDR image, and that using more exposures is recommended. Following that you take the same steps as you would have if you had been making a multiple-exposure HDR. Click “Tone Mapping”, and adjust stuff until you like it. Here is the result from one of mine:

Original Image (ISO: 250, Shutter speed: 1/250s, Aperture: f/8)

View of Eiger, Monch, and Jungfrau from Mannlichen / HDR with Photomatix

You can see that the original image is slightly overexposed. The sky and mountains are too light and appear sort of hazy and faded. In the HDR one however you can see that the colours are more saturated and the mountains and sky are more detailed and visible.

Here’s how it turned out with 2 exposures:

Original Exposure #1 (ISO: 250, Shutter speed: 1/250s, Aperture: f/8, 0.00 EV)

Original Exposure #2 (ISO: 250, Shutter speed: 1/500, Aperture: f/11, -2.00 EV)

HDR from 2 Exposures

Maybe it’s just my opinion, but I like the first result much more (using one exposure only). The multiple-exposure one seems kind of.. flat, and too yellow.

Anyway, since I’m posting this, I might as well talk about the .. place this was taken, in general, cause it was really awesome there and it was probably my favourite day in Switzerland. Here’s a map:

Jungfrau Region Map

The yellow circle is Mannlichen, where the above photos were taken from. The blue circle is Kleine Scheidegg, the town (if you can call it that) to which we walked from Mannlichen. This whole region has a privately-operated railway network (Jungfraubahnen), which one can use to get around between the places worth seeing. Of course, the major tourist attraction here is Jungfraujoch (the blue circle on the above map), which is the highest train station in Europe at 3471m. A round-trip ticket from Interlaken to Jungfraujoch and back costs about $160 per person (discounts if you have a Eurail Pass or a Swiss Pass). It’s expensive, but worth it.

Something else I noticed when I was looking through the photos is you can see Mannlichen (where these were taken from) on another photo taken from Jungfraujoch.

View from Jungfraujoch, Switzerland (click and then mouse-over the image on the Flickr page to see where Mannlichen is labeled)

Anyway, that’s everything for now.

[Edit]: I also changed the background on this site to that image cause I was getting bored of the old one.

I blame this all on workers’ unions. Let’s first clarify what one is. “A trade union or labour union is an organization of workers who have banded together to achieve common goals in key areas such as wages, hours, and working conditions, forming a cartel of labor. The trade union, through its leadership, bargains with the employer on behalf of union members and negotiates labor contracts with employers.”

Basically, a labour union is an evil organisation which makes an agreement with the employers to represent the “rights” of the empoyees. There are several UN legislations which recognize the rights of labour unions and their members. One is Article 23 of the Universal Declaration of Human Rights which “identifies the ability to organise trade unions as a fundamental human right” and another is the Declaration on Fundamental Principles and Rights at Work, which defines the “freedom of association and the effective recognition of the right to collective bargaining as an essential right of workers.”

The most important reason for my disapproval of unions is the increase in prices of everyday commodities. In this case, bus tickets. Let’s look at the reasons for this strike. The Toronto Star quoted the leader of Amalgamated Transit Union Local 113 (the labour union which represents Viva bus drivers),  Bob Kinnear, as saying that “drivers are looking for paid sick leave, more consistent scheduling and a pay increase of 3 per cent in each year of a three-year agreement.” The employer evidently doesn’t want to go along with this price increase, so the union members decided to go on strike. And there is nothing anybody can do about it, as human rights legislations and labour laws seem to do nothing to prohibit strikes.

Think about how this situation would play out if it weren’t for unions. A number of people who are unhappy with the employer’s conditions refuse to work and go on strike - the employer fires them. The others learn from example and go back to work. However, the agreements made between the union and the employer don’t let the employer do anything about it, and neither does law. What will usually end up happening is that the employer will accept the union’s terms and raise the wage of the bus drivers. And how does this affect everybody else? The employer needs to make additional money to be able to pay the drivers more, so bus ticket prices go up. Not too long ago an adult ticket cost only $2.50, or even less than that. Now it’s $2.75 or $3.00, I don’t even remember.

The Toronto Star article also mentions another interesting piece of information - “The union proposal would put that rate up to $21.37, slightly more than the drivers at York Region’s three other contracted bus operators.” Yes, York Region Transit has more than one contractor working for them. Different operators run YRT and Viva services. So, if Viva drivers are making more all of a sudden, what will the YRT drivers think? Will they not react in much the same way? And will the employer not give in to the demands the same way? And then think about the TTC. Did they not also have a strike recently? Well, if Viva drivers will be making a few cents more than them, who’s to say they won’t want to make more $ than them? Accepting the union’s proposal would be another step in the already existing loop that pushes our transit prices higher and higher.

However, I do have some hope in the screwed up system in which we live. There was an incident where TTC drivers went on strike until the provincial government ordered them back to work. I just hope the same thing will happen here, and that the employers will not give in to the union’s demands.

Basically, it involves combining a number of exposures into one final image, especially useful in cases where you have light and dark parts in a scene which are impossible to capture in one single exposure. The best way to explain this is by example.

Here are three photos of the same thing taken in quick succession with different exposures (using the bracketing feature on my dad’s Nikon D80):

Airplane at YYZ - exposure 1

Airplane at YYZ - exposure 2

Airplane at YYZ - exposure 3

The first image is taken with the same aperture and shutter speed as would result from using the “Auto” setting on the camera (0.00 EV, 1/125s shutter speed, f/5.6 aperture). The second image is taken at -2.00 EV, 1/500s shutter speed, and f/5.6 aperture. The final image is +2.00EV, 1/30s, f/5.6. This is done using “exposure bracketing”, which is basically taking a number of exposure with varying exposure values (EVs). There are two ways of changing the exposure: changing the shutter speed (amount of time during which the shutter is open and light is being absorbed by the image sensor) and changing the aperture (an opening of variable size through which light passes and hits the sensor. You can see that the shutter speed is changing with each exposure in this example, while the aperture is staying the same. This is because changing the aperture can produce undesirable effects. One such effect is the changing of the depth of field (the part of the image which will appear in focus), and with different parts of the image being in focus it will be impossible to blend the exposures into one final image when producing the HDR image.

Looking at the first image, you can probably notice several problems. While the nose of the plane appears to have a good exposure - not too dark, not too light, the upper portion of the sky is over-exposed. Here, clipping occurs, which means that colour information is lost. The parts of the sky that are completely white are overexposed, and there is no software that will be able to restore what the colour details that existed in the original scene. You can try to open the image in Photoshop and move the white point around, but that will only result in a shade of grey - you will not get the same details that you see in the image with a lower exposure value.

The next exposure has quite the opposite problem - while the sky has been captured much better, the bottom half of the photo is under-exposed. Shadows clipping doesn’t occur as easily as highlights clipping, however if you try to restore the dark portions of the image you might be somewhat successful, but chances are it will turn out very grainy and will look bad.

The final exposure depicts the area under the plane (and the plane itself) very well, but the sky is completely gone. Without using HDR imaging, a photographer would have to choose something to sacrifice - the highlights or the shadows. Alternatively one could go with the first exposure and keep the midtones, but lose some of the shadows and some of the highlights.

The purpose of HDR is to avoid losing any details in the photo. It works by taking those three (could be done with more or fewer) exposures, and combining them into one image. There are numerous software applications which work with HDR. My favourite one is Photomatix, however Photoshop is also capable of this (I use CS3). Here is the result from Photoshop:

HDR Airplane at YYZ - HDR with Photoshop

Personally, I don’t like it. Maybe I’m just not great with Photoshop’s HDR features, but to me that image looks really flat and the colours are somewhat lacking. Here is my result from Photomatix:

HDR Airplane at YYZ - HDR with Photomatix

In the rendered HDR image you can see both the sky and the dark details at the bottom. I realize it does not look realistic, and very much edited/photoshopped, but it is very rare for an HDR image to look convincingly lifelike. The purpose is to capture a large amount of light information, which you can’t do by regular means - and that’s what is done. There are many aspects of the resulting image which Photomatix lets you control when combining exposures into HDR, and it takes quite a while to get it to look the way you want to. As a general rule, I always post-process all of my HDR photos in Photoshop after Photomatix, as I don’t often like the result produced by Photomatix - but only very minor adjustments.

There are many things you can do with HDR - and not all so drastically different from original exposures but much more subtle. It is also not always necessary to use multiple exposures for HDR - it is possible to create an HDR image from one single RAW file (however it will not be as effective). In fact, my most “popular” photo as suggested by Flickr of the cow in the Alps is a single-exposure HDR - but maybe more on that later.

[Edit]: How did I just post this without mentioning what kind of plane it is?! It’s an Air Transat Airbus A310-300. K, that’s better.

asdf-OS Screenshot

So far, here’s what it can do:

Hardware Support:

• PS/2 Keyboard
• PS/2 Mouse
• Serial Port (mostly for debugging purposes)
• PCI (only tested device listing so far)
• PIT (Programmable Interval Timer - used for task scheduling)
• ATA Hard Disk Drive (using LBA with PIO - can read and write sectors to/from the drive, parse partition table, and working on FAT32 drivers)
• Video/screen functions with VESA/VBE
• RTC (Real Time Clock - to get the system time from BIOS)
• PC Speaker beeping (I was bored, lol)

Kernel/Internal Stuff:

• Initialisation of GDT, IDT, ISRs, etc. (basic stuff needed for the OS to run)
• Exception Handling (”RSOD”)
• System calls (used by ring3 - or usermode - programs to communicate with the kernel)
• Multitasking! Yay. (This took a really long time to get working, but is important to any functioning modern OS)
• Message queues for inter-process communication

Memory Management

• Physical memory manager - keeps track of used/free pages
• Paging! Virtual addressing is used to access memory. Each process has its own page directory, thus cannot access the memory of other processes. Also ensures protection of kernel pages from usermode processes.

Video

• VESA Driver
• Double-buffering (draw operations are done to a double-buffer in dynamically-allocated memory rather than writing directly to video hardware memory. This avoids flickering)
• 800×600x32 mode by default, but can be changed to a different resolution (running 1024×768 in a VM is difficult to work with because of the screen space it takes up on my 1440×900 laptop screen)
• Various graphics functions (ie. fillrect, drawline, fillgradient, and so on)
• Alpha-Transparency!
• Mouse cursor rendering
• Bitmap drawing functions
• Rendering of custom image format based on XPM (X PixMap), used for window title bars, close button, etc.
• Font rendering (font data stored in a bitmap-like thing)

GUI

• Windows and event handling based on the Windows design (using message callback procedures for each registered window-class)
• Resizeable and moveable windows.
• Window Controls: Button, Edit, Text.
• Message boxes
• Taskbar
• “Desktop” - you can click and drag and draw rectangles just like in Windows, lol
• Message queues for event handling in usermode

Usermode

• Loading and running PE executables in a separate process and address space.
• System calls communication with the OS/Kernel.
• Libraries with C runtime stuff, memory allocation, GUI window creation / various GUI-related features.
• Error handling! If a usermode program executes an illegal instruction (ie. divide by 0, access memory it doesn’t own, etc.), only the process will be killed, the entire OS will not crash! Same as on other protected mode operating systems - if your browser crashes on Windows you don’t get a BSOD.
• Sample test usermode calculator program.

And that’s about everything I can think of right now. The project is currently 12,282 lines of code in 84 files (doing a line count of all files with the extensions: c, h, asm - excludes Makefile and linker script).

asdf-OS runs fine with 16MB of RAM, but will crash with an out-of-memory error if you spawn a large number of windows and resize them to make them take up a lot of screen space. The main memory consumer in my OS is, of course, the GUI. Every window (this includes controls such as buttons) takes up height * width * 4 (* 2 if usermode). And that’s an unavoidable fact of making GUIs in a 32-bit colour mode. The (bootable) kernel image fits on a 1.44MB floppy. I develop on my own Windows XP machine and run in VMWare, but I have tested it on an ancient computer in the basement and it appears to work just fine.

The only piece of software which is not my own that asdf-OS depends on is GRUB, which I use as my bootloader. It’s great because it does all the necessary stuff to set up protected mode and start my kernel in protected mode. It also sets up the graphics mode that I ask for, which would otherwise be a pain in the ass to do - going back to real mode, calling the BIOS interrupts, and back to protected mode. Other than that, though, you can say this was coded “from scratch”, perhaps only taking a few snippets from here and there.

I have a few screenshots that I took during the development of it at various stages. Here’s what it looked like in January,  when I was working on the memory management stuff:

asdf-OS Screenshot - January 2007, text-mode

And here it is during GUI development (pre-transparency and after transparency):

asdf-OS Screenshot - GUI 1

asdf-OS Screenshot - GUI 2

Finally, here’s a recent video demonstrating the features of usermode error handling (division by 0 exception in usermode calculator kills the process and system continues running, but division by 0 exception in kernel-mode calculator crashes the OS):

Or view the higher resolution one here: asdf-OS Screencap on blip.tv

Edit: Thanks Frederick, I didn’t know you could embed these things.

Edit 2: It’s worth mentioning that lots of credit goes to Napalm for help and suggestions on various things.

I obviously didn’t upload all of them cause that’d take forever but here are some of the ones I thought were best that I picked out - http://www.flickr.com/photos/weasel707/collections/72157607066362777/

Here are some of my favourites:

Eiffel Tower

Eiffel Tower at Night

Louvre, inside

View from Jungfraujoch, Switzerland, highest train station in Europe at 3471m

Lauterbrunnen Waterfall

View of Lauterbrunnen Valley from Mannlichen / HDR

View of Wetterhorn near Mannlichen / HDR

View of Eiger, Monch, and Jungfrau from Mannlichen / HDR

View near Kleine Scheidegg

Sign Post near Kleine Scheidegg, Wetterhorn in background

Sunset in Bern, Switzerland / HDR

Zurich, Switzerland / HDR

And finally, my personal favourite:

Cow in Swiss Alps, near Mannlichen / HDR

Anyway there’s a lot more, those are just some of my faves. The best pics are probably in the Alps set, cause the Alps are really pretty. I’ve been asked for the high res (original) version of the pic with the cow in the mountains, here it is - Cow in Swiss Alps - Original Size (3900×2613).