Blog

I’ve had quite a few messages on Facebook from previous users of Delete Checker that can no longer find it. So here’s a quick blog post that’ll hopefully get indexed by Google and therefore make Delete Checker findable for those who used to use it….or those that want a Delete Checker!

On the 16th of April, Facebook removed the Delete Checker application without notice. A little research indicated that Facebook had removed all applications that added the unintended functionality of checking if users had deleted you from their friends list. For those interested, Section II, 5a of the Facebook Developers Platform Policy outlines that “You must not circumvent our intended limitations on core Facebook features. For example: You must not inform a user that someone has removed the user as a friend.” What’s funny is that Delete Checker was approved for the Facebook Application Directory, which is not an automated process. Before removal, Delete Checker had approximately 2,500 users…now it has 30.

But don’t fret! I’ve recreated the application so that is accessible to all. However, I have not published the application to the directory. This basically means that unless you know where to look, you won’t find the application. I had two “Delete Checker” applications, one for production and one for development. The development application was not removed by Facebook showing that it’s unlikely they will remove applications that are not published to the directory. The only possibility of removal is if someone explicitly reports the application….but I’ll be sure to put it back up if it disappears, so check back here if it does!

So, here’s what you need to know:

What is Delete Checker?

Delete Checker keeps track of which friends deleted you on Facebook. It also keeps track of when friends deactivate their account or someone ignores your friend request. You can setup Delete Checker to automatically post insults about deleted friends to your wall.

Where is Delete Checker located?

The application page, where you can install, review, discuss features/bugs and become a fan is currently located here: http://www.facebook.com/apps/application.php?id=109342089106138

The direct link to the application is located here: http://apps.facebook.com/deletechecker_fo/

What do I need to do to make it work?

Both new users and previous users will need to re-authenticate the application for offline access (the application will prompt you, or follow the Instructions). If you were a previous user and you had automatic stream posts enabled, you will need to re-authenticate the application for stream posting by visiting the Settings tab as this is not enabled by default. Previous users will still have their log intact as it is stored on my own server and untouchable by Facebook.

I have a different question. Where’s more information?

You can visit the FAQ (highly recommended that you read this) for the application or post on the application’s discussions board.

The application is in violation of the developer policy. Is my account at risk?

While the application still violates the platform policy, users shouldn’t worry about the loss of their accounts. If anything, Facebook will just disable the application and possibly give me a warning about the application.

Today I noticed a significant privacy hole within Facebook. I always knew that I could enter an e-mail address into the Facebook search bar and it would search my friends or people with public profiles to bring up the user with those credentials. Today I received an e-mail with a bunch of e-mail addresses in the CC header. I had no idea who they were, so I went to Facebook to look them up. This is when I stumbled upon a rather large flaw.

Surprisingly, every single e-mail address I looked up returned a search result, even if the user was not my friend and their profile was private. I decided to use my dummy account, which has no friends and is not related to any network, to try the same thing. Again, Facebook returned the user given the particular e-mail address.

I tried a couple of experiments with my own account and the dummy account. I added an e-mail to my profile with the privacy setting of “Only Me” and then searched for it with my dummy account. Yup, it returned my profile relating my Facebook profile to the e-mail address that should never be seen publicly.

I then tried to guess the e-mail addresses of a couple of “web celebrities” (in particular, Kevin Rose and Alex Albrecht from Diggnation). Both users have strict privacy settings, however, using a little guess work  I was able to work out a couple of e-mail addresses that are related to their profile.

So, what does this mean? Well, a private e-mail on Facebook can be public. A little guess work can reveal any e-mail address specified on your Facebook profile. Not to mention you can use any of these e-mail addresses as the username to login to Facebook. Is this a flaw? I certainly think so. A little Googling didn’t bring up much, but Facebook did mess up on revealing e-mails recently. So, maybe you heard it here first.

The fix? If you don’t want your e-mail to be found, remove it from your list of e-mails from your profile. Your login e-mail will always be “guessable”, so change that if you wish too.

I like pretty pictures. Today, I saw my first 3D movie, Alice in Wonderland. After seeing it, I have decided that anyone that says 3D is an incredible experience either a) has vision problems and/or b) does not appreciate a high definition image. By no means is a “2D” movie high definition (speckles and grainyness), but it does look decent for its size. “3D” is a step backwards in terms of quality.

Firstly, the glasses they give you have the same effect as wearing sunglasses. I’m sorry, but when did anyone wear sunglasses to a cinema? This significantly dulls the brightness and vibrancy of colour in the image. Throughout the whole movie I had the urge to remove the glasses to see the image with the brightness dial turned up. These 3D films, like Avatar, have “groundbreaking” CG. You would think that you’d like to view this image in it’s originally rendered, bright and beautiful colours, no?

The next complaint is about the focused portions of the image. Many will notice, when compared to the “2D” version, that a vast majority of the image is blurred to focus your eyes on the 3D portion. Although, this didn’t give me a headache (like some do suffer), I did find it very frustrating that I was unable to explore “groundbreaking” CG scenery. Instead, you are forced to focus on a portion of the image that, sometimes, isn’t even the character/object that is meant to be the centre of attention. To top it off, there are often objects that are presented in 3D but out of focus. Naturally, you are drawn to focus on that object but you become confused that the object looks both horrible and out of focus.

The movie was viewed at Hoyts Norwood (Adelaide, South Australia). The screen is most certainly not the biggest but it was the “Xtreme Screen” of the cinema. Now, I’m not quite sure if this is a 3D technology technicality or if it’s just how this movie was shot, but, the movie did not take up the whole screen (width wise). It was apparent to me that the curtain was still drawn over the edges of the screen. Firstly, this is supposed to be a spectacular image, why the hell isn’t it big? Secondly, this causes much disappointment when “3D” objects are flung off the side of the screen. Initially, the effect is cool, it looks like it’s coming towards you…but you are significantly disappointed when the border of the screen makes the image disappear and not the corner of your eye. Perhaps this would be more spectacular on an IMAX screen?

3D does, however, work in some cases. I found that when there is only one character on the screen, it worked well (an example in Alice in Wonderland is Cheshire Cat ). Also, ambient effects, such as smoke and fireflies, would add depth to the image. Fast moving objects, such as falling or flinged objects can also work, but I found that the fast moving object would often be blurred making it look very gimmicky. In Alice in Wonderland, I found that “real life” scenes were only partially in 3D and that the effect on “real life” scenes were underwhelming. CG generated scenes took more advantage of the 3D “technology”, significantly.

For non-glasses wearers, such as myself, these things are uncomfortable. I don’t choose to wear glasses for a reason and I’m pretty sure that most people that wear glasses only wear them because they have to.

For all this, you pay an extra AU$6 (Student). What a jip. Now, I understand that the technology to shoot the movie costs more, but I’m fairly sure the cinema does not need to purchase new projectors to show the film (I could be wrong). Why am I paying extra? New movies continuously use new and more expensive technology and we generally don’t pay extra. I can go see this movie in 2D and pay the normal price. Gimmick? Yup.

Where’s my damn hologram.

Putting a PC under your TV to play your pictures, movies, and all that jazz, sounds like an excellent idea until you actually try and implement it. I’ve been using a full blown PC since 2004 to record TV and stream media from my server….but it’s been hell.

Problem #1: Choosing the right hardware
Back in 2004, I thought that putting a PC under the TV to play media would be a fantastic idea. I started to plan a bunch of parts and on paper, it looked great. Being slightly more clueless than I am today, I didn’t do much research….which led to me buying a P4 LGA775 (the hottest of all Intel processors) processor and a set of loud components.

The TV tuner card (a Hauppauge PVR-MCE-350) was also very average. It’s an analogue card, but the quality of video it outputs is significantly worse than plugging it directly into the TV (via the same connector). Fast forward to today and I still use the same card. It works OK, but in the world of digital, the card needs updating.

The leads us to the next problem. Working with a FOXTEL box. In 2004, FOXTEL digital was introduced. Ironically, the set top box provided by FOXTEL does not output any digital video (only audio). FOXTEL boxes have only recently outputted in digital with the introduction of FOXTEL HD channels. That brings up a new problem. It’s impossible to create a “purely digital” media center PC…where there’s digital input and digital output. Why? Well firstly, HDMI and Component are protected by HDCP (and any capture cards to bypass HDCP are “illegal” or need heavy licensing fees). Secondly, it takes a beefy PC to capture and view HD video at the same time. Currently, the only way to capture HD video from FOXTEL would be via analogue component inputs. Capture cards that do this are reasonably expensive…$300+ The alternative is CableCard. A technology that has never taken off despite it being amazing and around for several years.

There’s also the problem of IR blasting. This isn’t FTA Digital/Analog, where you tune into a frequency depending on a channel, the capture card sees FOXTEL as a singular channel. Every time you want to change the channel you must blast infrared out to the FOXTEL box. I’ve found that this can often be inaccurate and laggy.

Problem #2: Choosing the software
Now the software. Generally, you’d use Windows Media Center. It comes with Windows. For a simplistic Media Center, it works pretty well. However, it’s hard to extend and skin. Windows 7 media center is particularly average when it comes to UI.

The alternatives? MythTV. XBMC. MediaPortal. MythTV is out, fuck Linux. XBMC is out, no capture card support. This leaves MediaPortal. An open source media center that’s highly extensible and skinable. MediaPortal, by default, looks like balls but it has a ton of cool features. Adding the StreamedMP skin and a couple of plugins makes MediaPortal look and feel like XBMC. It has excellent UIs for TV Shows and Movies (with a bit of work).

The downside? It’s buggy as fuck. Firstly, the StreamedMP skin doesn’t support 4:3. I guess this isn’t a big deal for most, alas, the parents refuse to venture into a world of new technology. This means you have to do a crapton of skin modifying to fit things on the screen. Secondly, I had to go ahead and check out the source from SVN to determine what was hanging MediaPortal for 2 minutes. It turns out there was no such timeout for devices like card readers. If there’s nothing in the card reader, there’d an extended I/O timeout which would hang MediaPortal. It’s an easy source fix, but there’s bugs like this that just make MediaPortal very unfriendly (without a helluva lot of work).

There’s also a problem with timeshifting. A feature that allows you to pause and rewind TV. With FTA, this is great. With FOXTEL, this is absolute pain. The way timeshifting works is that the capture card is saved to a temporary file before it is viewed on the screen. This means there’s usually a 3-4 second delay. This delay makes using onscreen components for the FOXTEL box (such as TV guide) impossible. Unfortunately, both MediaPortal and WMC do not allow you to disable timeshifting.

Finally, EPG (programme guide) data is nearly impossible to get for FOXTEL. There are sources such as OzTiVo, but it is often incomplete. In the end, I ended up writing a parser for YourTV (which has relatively complete FOXTEL data). It’s a very highly customized solution mainly because there is now “raw” data feed (such as XML). I have to parse HTML pages to get meta data. Each day downloads approximately 50-75MB of data (when parsed it’s only ~4MB, 600KB zipped)…which means I have to run this on my overseas VPS. Basically, the stars have to align to get any sort of consistent EPG data.

So, in conclusion?
a) If you want a purely digital media center: you’re going to have to use FTA Digital. You can use analogue component for FOXTEL but it’ll cost you.
b) EPG data for FTA is great! FOXTEL not so much.
c) FTA works great for timeshifting. You’re probably going to want to connect FOXTEL directly to your TV and capture card, so that when you need to record, you can.
d) Buy and research silent, cool and compatible hardware components.
e) 4:3 is out. 16:9 is here. Don’t even bother if you have an old TV.
f) Be prepared to pull your hair out.

You can probably tell this is just a spill of my brain. I haven’t gone in to much detail…but I’ve finally finished the rebuild project of my media center.

Exams and honours being over, I decided to revisit my multi-boot (7, SL, Ubuntu) and give it an upgrade. Ubuntu 9.10 “Karmic Koala” had been released but at the time I was too busy to upgrade (Windows being my main operating system).

I don’t use Linux too much. I’m proficient at it to get around, compile, install, even mess with kernel compilations, but I only use it when I really have to test something out. I had 9.04 installed; it was fresh, untouched, the only thing that had been installed is the proprietary NVIDIA drivers. Instead of doing a full reinstall, I thought I’d take the easy way out and use the GUI update manager. Huge mistake. After downloading about 700MB of packages, the installation began and failed approximately half way through. A completely optional component, OpenOffice, was able to bring the installation script to its knees with an error that said something along the lines off – Failed to install <insert random OpenOffice component here>. Your system may be in an usable state – Yes, such a minor package was able to cause such a gigantic problem. I attempted to re-open Package Manager and even tried apt-get…neither worked. Needless to say, rebooting resulted in an unmountable mounts. I am baffled as to how a clean 9.04 install failed to upgrade.

Instead of messing around in recovery mode I decided to chuck in my 9.10 DVD and go for the install….since I had nothing important installed anyway. Installation seems to take a lengthy amount of time, and to this day the Ubuntu installer STILL makes no logical sense. Seriously, why is an installer removing packages when installing the operating system for the first time? I don’t care about the technicalities, this is NOT good user interface design.

At the time of installation, my disk configuration looked something like this:
Boot Priority 0: DVD Drive
Boot Priority 1: 160GB Ubuntu Drive
Boot Priority 2: 300GB Windows 7 RAID0
Boot Priority 3: 250GB Mac OS X Snow Leopard Drive

During installation I chose to install Ubuntu to the Ubuntu drive (funnily enough). This drive at the time was set to the highest boot priority. Alas, after installation, Grub seemed to have made itself on to the Windows 7 drive (for no particular reason) AND the Ubuntu drive. I mean what the hell? This drive had nothing to do with the installation and the Grub installation on the Windows 7 drive kept on failing with “File Not Found” errors. It’s almost as if the Ubuntu installer goes something along the lines of:

foreach (Drive d in drives) {
if (windows_installed(d))
// Make Windows partition unusable to piss off Microsoft fanboys.
// Who needs Windows anyway?!
install_broken_grub(d);
}

I decided to go ahead and reinstate the 7 MBR, disconnect all drives (except the Ubuntu drive) and do yet another install. Hoorah! We’re in. Straight off the bat, X-Fi Digital Out no longer works (it did in 9.10). The ALSA mixer seems to have disappeared and whatever the hell the new sound configuration manager is, it does not allow 5.1 Digital Out. Attempting to install the proprietary Creative drivers fails due to missing includes. OK, no problem I’ll solve that later.

Instead I decided to go ahead and update Ubuntu and install the proprietary NVIDIA drivers. All done via the Ubuntu UI. What seems to have been success turned into utter failure upon reboot. I was greeted by a constantly flickering console screen. Booting to recovery mode revealed that apparently “the NVIDIA drivers do not support the GPU at PCI0:0″. For the record, both cards are relatively new (GTX 260 and 8500GT) and worked perfectly in 9.10. OK, so, why the hell hasn’t the generic driver been reinstated so the GUI is accessible? It actually turns out that Ubuntu had installed version 173 of the NVIDIA driver rather than 185. Why? (you might ask) Well, it turns out that upon being prompted to install the driver, the default selection is version 173, even though the one tagged as “[Recommended]” is 185. Why the hell isn’t the recommended selection the default?! Why the hell doesn’t Ubuntu gracefully recover on such a catastrophic failure. Surely this is simple to detect. Either way, I am now at the point at which I’ve installed version 185 and rebooted to find Ubuntu is freezing on boot (every time).

After several forceful reboots, Ubuntu decided to boot….and hoorah, wobly windows. Now to fix the X-Fi.

After an hour or so of tinkering…I did three things…who knows which of these actually made the thing work, but here goes (X-Fi Digital Out on Ubuntu 9.10):

1. Open sound preferences, make sure the hardware configuration for the X-Fi device is set to analogue 5.1 out + analogue stereo in. (This makes a lot of sense, right?)
2. Make sure the input and output device are set to the X-Fi device (and not some integrated sound card)
3. Open up a Terminal window, type alsamixer. Scroll left (with the left arrow) till you find “Digit-IO”. If it’s Off, toggle it to On by pressing ‘m’.
4. You might have sound now…you might not. The only other thing I did was install the Creative X-Fi drivers. Note: These don’t compile by default on 9.10!

So here’s how to compile them:
1. Download them here: http://support.creative.com/downloads/download.aspx?nDownloadId=10792
2. Extract them to your desktop
3. You’re going to modify two files, let’s start with ctatc.h
4. Look for the line “#include <sound/driver.h>” and comment it out like so: “//#include <sound /driver.h>”
5. Now the next file, xfi.c
6. Again comment out “#include <sound/driver.h>” like so: “//#include <sound/driver.h>”
7. Now look for the line: “card = snd_card_new(index[dev], id[dev], THIS_MODULE, 0);” and replace it with:
“snd_card_create(index[dev], id[dev], THIS_MODULE, 0, &card);”
8. You can now follow the README included with the drivers….sudo make, sudo make install, sudo reboot

Frankly, this user experience is pathetic. Something needs to be done. The Ubuntu team need to stop wasting their time thinking up obscure codenames and ripping off the Windows 7 pulsing logo. Quality control is much to be desired.