Main
Latest
- Office Space
- UDP Port 18919 (DDoS?)
- South African Ubuntu Mirror
- Semantic Web Slideshow
- Chisimba Feediscovery Module
- Chisimba Skype Module
- Sun Java 6 on Ubuntu Jaunty
- Zoopy PHP CLI Upload
- Modern Spam
- Afrihost Capping #fail
Archives
- June 2004
- July 2004
- August 2004
- September 2004
- October 2004
- November 2004
- December 2004
- January 2005
- February 2005
- March 2005
- April 2005
- May 2005
- June 2005
- July 2005
- August 2005
- September 2005
- October 2005
- November 2005
- December 2005
- January 2006
- February 2006
- March 2006
- April 2006
- May 2006
- June 2006
- July 2006
- August 2006
- September 2006
- November 2006
- December 2006
- January 2007
- February 2007
- March 2007
- April 2007
- May 2007
- June 2007
- July 2007
- August 2007
- September 2007
- October 2007
- November 2007
- December 2007
- January 2008
- February 2008
- March 2008
- April 2008
- May 2008
- June 2008
- July 2008
- August 2008
- September 2008
- October 2008
- November 2008
- December 2008
- January 2009
- February 2009
- March 2009
- April 2009
- May 2009
- June 2009
- July 2009
- August 2009
- September 2009
- October 2009
- November 2009
- December 2009
- January 2010
- February 2010
Hotmail on open source?
There are outdated reports, circa 1998, stating that Hotmail was last known to use Solaris and BSD—notably non-Microsoft products—for its mission-critical components [7]. More recent data show that in 2001 the web services was rewritten to run on the NT based IIS platform, creating a small decrease in the throughput utilizing the same server hardware that previously hosted Apache on top of BSD. By 2002, the incoming SMTP services were transitioned off of the Solaris platform onto NT, requiring significantly more hardware to maintain the same mail receiving service. As of the beginning of 2005, Hotmail has ceased all development on non-Microsoft components and is now executing on plans to finalize the transition off of the last few remaining Solaris based services.
- Wikipedia: Hotmail: Miscellaneous
Microsoft always keeps reminding us that they don't develop for other platforms because they believe in Windows, but acts are more important than words in my opinion.
Maybe this is one of the inspirations for the SSCLI?
I have nothing against Microsoft using open source software to power their solutions; as a matter of fact I am very happy they are! What does upset me a bit though is that they don't embrace open source openly; they want to use it but keep telling the world how bad it is. Didn't you know open source is pure communism?
Take the SSCLI for example. I can't even find a good homepage on Microsoft's own websites about it; the wikipedia article is the most informative I have seen so far.
Sometimes I really wish that the various open source communities and Microsoft can stop fighting and work together.
SSCLI
The SSCLI is probably the most interesting development so far to come out of Microsoft's Shared Source Initiative in my opinion.
It is essentially an open source (if you can call this open source), cross-platform implementation of the .NET Framework. Yes, this means .NET running on Unix too.
Of course there is also the Mono project and DotGNU, if you prefer real open source stuff. :)
Microsoft & Open Source
Very interesting! Microsoft is giving away source code with the Microsoft Shared Source Initiative.
This is still far from "free software" (as defined by the FSF) but this is an interesting move nonetheless. Is even Microsoft slowly realising that open source is the future?
Good wikipedia article about this.
Tuesday Updates
- FC 5 has been released.
- Pure Pwnage Episode 10 is out.
- In case you're looking for work in Cape Town, there is a web development job on offer.
websitewriters.co.za
websitewriters.co.za is the first web standards community website I know of in South Africa. Web standards awareness is, as in most other countries, greatly lacking here.
Congrats and many thanks to Raoul for taking the initiative! I hope to see the project grow from strength to strength. :)
Alive in Joburg
This is actually a little old, but I thought I should post about it anyway.
Alive in Joburg is a short movie made by Neill Blomkamp.
Aliens land in the city of Johannesburg, South Africa. They come to stay among us, but the locals don't like them. They are foreign, and are treated as such. And what's more, the Apartheid government is out to get them!
Really funny; good CGI too! Worth a watch; I think especially South African expats will like this one. :)
Thanks to Aryan Blaauw for the link!
Zooomr
Flickr just got competition... Zooomr is in the house!
It's really much like Flickr, just simpler and yet has a couple of very interesting features. The first thing you'll notice though is that the interface is much less cluttered.
The login system is fantastic. You don't need to use your Yahoo account as with Flickr; you have a variety of options because it integrates with different login systems including Gmail. Personally I especially like the MyOpenID integration, although I just used my Gmail account.
The only problem is that, at the moment, they seem to be suffering from bandwidth issues. I didn't even manage to get anything uploaded. But as soon as the growth pains are over, I'm sure they'll be a force to reckon with.
Thanks to calisza for the link.
The wikipedia page is very informative.
Pure Pwnage
This is too funny not to blog about. I actually knew about this site for a long time, but never got around to blogging about it.
Pure Pwnage (pronounced: Pure Ownage) is a show all about Jeremy (also known as "teh_pwnerer") and his life as a "pro gamer" pwning noobs. A total waste of time, but quite funny. :)
Highly recommended for anybody with too much time on their hands, and even for those (like me) that don't have the luxury of being bored! :P
Also see the Wikipedia page for more info.
PHP 6
Currently, I have to agree that PHP sucks. The language has originally been invented to do "cool things" (read: quick dirty hacks) instead of to allow professional developers to build proper, structured web applications. Yes, it was (and still should be) aimed purely at the hobbyist.
It is now sometimes being used to build massive systems, and its focus has changed a bit to be more structured, but the more I use PHP the more I realise that the original "dirty nature" still runs strong in its blood.
I am very excited after reading a post from Jeroen van der Meer about PHP 6 though. The improved Unicode support should help it catch up with JSP and ASP.NET and the addition of namespaces is going to be really great for OOP.
PHP 5 has already seen vast improvements, and slowly things seem to be going in the right direction. Let's hope they get there.
Thanks to Jean-Paul for the link!
Google Mars
First we got Google Maps, then Google Moon, and now Google Mars. Spotted any men yet? I have to be honest; I'm not so particularly interested in the guys. I'm actually waiting for Google Venus so that I can spy on the girls! ;)
Read more on the official Google blog.
Hustle
Yesterday evening I watched an episode of Hustle on SABC 3.
The show is all about high-level con artists. When I mean "high level", I mean that 100 000 British Pounds it just a joke to these guys. As Mickey (one of the main characters) explained, it's not even really about the money.
These okes are just brilliant. I mean, they will even con the con artist! The way they use people's greed again them...
So far, I think the series is going to be seriously interesting; definitely gives me insight into a life I don't know shit of. AFAIC, recommended watching.
The Big Turnoff
One thing I believe should come standard in any operating system is basic power management.
In the "old days", the computer's power switch was directly connected to the power and when it's turned off the computer's power drops entirely. Even the power outlet on the power supply (which was normally connected to the screen) was turned off.
However, those days are over now; today's computers have "intelligent" power systems. When you press the power button and the computer is off the computer turns on and the BIOS takes over. Normally, when the switch is pressed again the computer will turn off. However, the operating system is allowed to "override" the power switch so that when the button is pressed, it sends a signal to the operating system instead of just turning off the computer. This first allows the operating system to go through the proper shutdown routine and then lets the operating system issue a command so that the entire computer turns off.
Also, the cards inside the computer can now stay "alive" while the computer is being fed with power, even though the computer might be "turned off". This means NICs can stay on the LAN and wait for Wake on LAN signals.
Windows handles this nicely (for a change). You use the normal operating system "Shut Down" feature, the shutdown routine follows, and then issues the right command to the BIOS to let the computer turn itself off. Otherwise, press the power switch and a fast shutdown routine will follow, after which the computer is also turned off.
Many Linux distributions also handle this nicely. For example, Ubuntu does it much like Windows.
FreeBSD (surprisingly) also handles it quite well. It will not turn the computer off at the end of a normal shutdown (halt), but it will first do a shutdown routine and turn the computer off afterwards when the power button is pressed.
Many others, like OpenBSD, Debian, SuSE (although I'm not sure about the latest versions), etc don't seem to handle it at all that well. I don't like having to wait around until the operating system's shutdown procedure finishes so that I can manually turn off the computer at the end of it. I also hate it when the computer just immediately turns off without doing a proper shutdown procedure first when I press the power button.
So a cry to all the operating system developers: Please wake up and don't turn us off your platform. :)
Object-relational mapping
Many thanks to Jean-Paul for telling me about this.
ORM seems to be one of those upcoming technologies that look very promising.
The problem with OOP is that the objects are all kept in main memory and are thus wiped out of existence when the program ends or the computer turns off.
Therefore, any data that needs to be saved for future use will have to be stored into and later retrieved from a database. This usually requires the programmer to manually write some SQL for lots of different classes.
That's fair enough, but when you have complex data structures, things can become hectic. For example, if you have objects that link to other objects (even of a different type) that link to yet other objects, it can sometimes become a very difficult task to make sure all the necessary data gets saved.
Ok, so can't we just have a system that handles that kind of stuff for us? Of course we can; and on to the world of ORM. :)
A good implementation of an ORM system should allow the programmer to "map" objects to a relational database. In this sense, it will handle the storing and retrieving of data in the objects and the database for us so we don't even need to worry about writing any SQL.
Dependencies between objects (one object linking to the next) should also automatically be handled with relationships in the database. So to the programmer, it will feel just as if he/she is working with normal objects, while in fact he/she is working with the database.
Another important advantage would be database platform independence. Because you're writing no SQL, you can have the same system easily work with both MySQL and Postgres.
I will be doing more research into the various PHP implementations. Yes, for Fidelis. :)
Amount of Items in Feeds
My del.icio.us account is running over. :)
Many feed readers reload each feed every 30 minutes to 2 hours. And many other syndication services like Suprglu only reloads once a day it seems.
Therefore, when you expose a feed, please make sure it displays enough items. For example, my del.icio.us feed only exposes the last 31 items to be entered in my public bookmarks. However, I sometimes add twice as many items within just a couple of hours. So now things could start to get behind; some feed readers / syndication services that don't reload my feed often enough will loose some of the items.
(SuprGlu doesn't even seem to care about anything other than the first 15 items. Yeah, they suck, but then again I probably shouldn't have it syndicate my public bookmarks because there's so many.)
What I am trying to say though is that one be careful how many items one has in one's feeds. I have seen hundreds of items in a single feed, taking lots of bandwidth/time to download, and in the meantime many of those items are months old. That's bad. However, don't just keep the last couple - at least keep the last 2 days' items in the feed (unless this turns out to really be too much).
I think a fairly good rule of thumb for a busy feed would be this: Take the amount of links added on an average day, double that, and put that amount of items in your feed.
If you are coding your own system, or your system allows this, there is an even better approach.
Firstly, work out a threshold. This should be the amount of items you would like the user to see the first time they add the feed in their feed reader. Then the rules should be this: If the last two days' items are more than the threshold, display all of them; otherwise display the last amount of items equal to the threshold.
Just some random thoughts...
Google Pages
Update: I don't know how I missed this earlier, but there is indeed an upload function to allow you to upload your own content. I have updated this page accordingly.
Google Pages is now in public beta. The idea is to give you free webspace to easily publish pages using Google's own online WYSIWYG editor or to upload your own content.
Unlike most free web hosting services, there are no ads on the pages.
When you use the WYSIWYG editor, you can choose from a couple of themes and layouts, but that's about as much as you can customise. I couldn't find any way to edit the source of those pages. You also can't edit the pages you uploaded using the WYSIWYG editor, which I guess is fair.
The sad thing is that this is yet another flop for web standards. The WYSIWYG editor creates very "nice" markup. The doctype is XHTML 1.0 Strict, yet font elements are liberally used. Don't mention validation, please.
The CSS is inline, needless to say. Who said it isn't fun to waste bandwidth?
It should be noted though that this is still in beta, like most other things Google. However, this is no excuse for the poor markup being generated. Let's hope it improves though before final launch.
An "interesting" development, nevertheless.
Lonely Finger: Solved
And there the mystery finally was solved: Saturday I saw their latest advert on the TV. The lonely finger found some other fingers and is now very happy. And then the 5FM logo popped up.
Congrats to Justin Harrison for his excellent detective work!
Probably the funniest part of this was that many people (including myself) were convinced this was a hack into the SABC itself. :D
Accessible Alternate Content
For accessibility reasons, it’s important to specify alternate content for any multimedia including images, audio, video, etc. When a text alternative is provided for an image, it should serve the same purpose as the image itself and not just describe what it looks like; and similarly for other multimedia content. However, there are some cases where it’s important to know that the alternate content is replacing an image in order to give some context, even if the user can’t actually see the image (either by choice or phyisical limitation).
- excerpt from Lachy’s Log: Accessible Alternate Content
I always love to discuss everyday markup issues. As far as I'm concerned, Lachy's post should be highly recommended reading for any web developer.
Newsvine Launches!
Earlier I blogged about Newsvine; today I find out they just came out of beta yesterday. Of course I didn't notice, since I was off voting. You can read the official release blog article too if you like.
Copyright © 2004-2009 Charl van Niekerk. All articles are released under the Creative Commons Attribution 2.5 South Africa licence, unless where otherwise stated.

