Alastair.com.au

A new Wikipedia function I noticed today: Wikipedia Book Creator. It allows you to piece a collection of articles together in one easy lot, which you can then browse through, download as a PDF or get suggestions for additional pages.

Link:

• Wikipedia Book Creator

As has been all over the Internet, Apple are holding a Press Conference in approximately six and a half hours about the ‘iPhone4’. Exactly what it’s about no one knows. Over at WhirlPool there is a LOT of conjecture (see below links).

I’m up doing some work for overtime tonight, but who knows. I may still be up at 3am…

Whirlpool threads:

• http://forums.whirlpool.net.au/forum-replies.cfm?t=1468237&p=1
• http://forums.whirlpool.net.au/forum-replies.cfm?t=1428850&p=1
• http://forums.whirlpool.net.au/forum-replies.cfm?t=1474097&p=1
• http://forums.whirlpool.net.au/forum-replies.cfm?t=1488015&p=1

After chatting with my mate Mike about my previous investigations into HTML5, he responded with:

"This, IS what is going to suck about HTML 5.0. You know this is going to cause compatibility hacks to make things work for your video tags, or you are going to keep using flash or silverlight (probably flash)."

For video it’s pretty simple. You can to cater for any browsers you want to support… you have to encode the video multiple times but the browser goes through until it finds one it can handle.

e.g..

<video width="320" height="240">
<source src=".....mp4"...>
<source src=".....ogv"...>
</video>

As far as the other compatibilty issues go, there’s a javascript library called Modernizr that can take care of all that for you.. e.g.

if (Modernizr.video && Modernizr.video.ogg){
// preload ogg video assets
}
else if (Modernizr.video && Modernizr.video.h264){
// preload h264 assets
}

or for Geolocation for example:

if (Modernizr.geolocation){
navigator.geolocation.getCurrentPosition(function(position) {
// pass the lat and long values to an application
// e.g. a setUserLatandLong() function may find the closest bodega
setUserLatandLong(position.coords.latitude,position.coords.longitude);
});
}


It’s all pretty cool. There’s even a JS library out there to help with the IE compatibility (http://code.google.com/p/html5shiv/)..

Better yet, have a read of http://www.diveintohtml5.org/. It should fill in any gaps… ones that can be filled at this stage anyway

As mentioned previously, browser support is a bit iffy, but there are ways to hack them into shape. I’ve heard it suggested that one of the ways of going about it is to code in HTML5 and hack everything else to fit, so that when the browsers catch up you won’t have to work as hard to make it compliant.

Over the weekend I thought I’d see what all the fuss was about HTML5.. I’ve read a dozen or so articles and I’m still not that clear. At the moment it seems not that many browsers support that many aspects of it, with the nightly builds of ones like Chrome being the most advanced (supporting <section&gt, <header&gt and other semantic ones.. It’s got some functional new tags (<video>, <audio>,<canvas>, etc).. and some semantic ones (<header>,<nav>,<footer>, etc) which at the moment are more like glorified <div> tags..

I digress.. of the articles I’ve read the best one so far is http://diveintohtml5.org/, written by Mark Pilgrim which is more of an online book than an article. The way it reads is similar to Masters of Doom about the rise and fall of iD software. It has a large section on the history of HTML and how it came to be, as well as how it relates to xHTML and standards groups in general.

I’m only about 3 chapters in and we finally get to see some HTML5 in action. At this stage Pilgrim is hillarious. He starts of the chapter:

“This chapter will take an HTML page that has absolutely nothing wrong with it, and improve it. Parts of it will become shorter. Parts will become longer. All of it will become more semantic. It’ll be awesome.”

A little bit later he’s talking about DOCTYPEs and how they can tell the browser how far to break the standards as far as rendering a web page goes. In particular he mentions how, when Microsoft were developing IE5 for Mac, they built it so it was to the HTML standard. However that didn’t render web pages correctly so they had to introduce ‘quirks’ mode. And it grew from there. Various browser have different quirks modes, etc.

“The last time I tried to count, there were 5 doctypes that triggered “almost standards mode,” and 73 that triggered “quirks mode.” But I probably missed some, and I’m not even going to talk about the crazy shit that Internet Explorer 8 does to switch between its four — four! — different rendering modes. Here’s a flowchart. Kill it. Kill it with fire.)”

I used to use http://www.goodpassword.com for generating random passwords for random accounts I created. However when I went there this morning I noticed that the site has gone to a Pay-For-Service model (rather than advertising). You only find this out after clicking the “Generate Password” button.

While it’s a good application, it’s not something that couldn’t be done without a few minutes by one’s self.

So I give you my attempt as a Password Generator

Edit: Thanks to Mike for pointing out a typo in my password generator link.

I’ve just noticed that Google are cashing in on the success of 3D blockbusters such as Avatar.

In what looks like a recent addition to Google Maps they give you the opportunity to toggle between normal (2D) and red-blue 3D street view. 

Along the left-hand edge of the screen there is now a small person wearing the red-blue 3D glasses (unfortunately the polarised kind for Avatar won’t work). Clicking on this toggles the different modes.

I haven’t tried it yet but it could be quite interesting, especially at those locations where people have come across Street View phenomena.

Flinders St in 2D

Flinders St in 3D

I ran into a problem recently where someone needed me to compress a large number of IIS web log files (~40GB) for them to download.

For some reason they specifically wanted me to use Bzip2 compression. Unfortunately bzip2 doesn’t have a recursive option like gzip, 7-zip, etc.

To solve the problem I joined it (using Cygwin and the find command) like so:

find –name *.log –exec bzip2 –v {} :

There are obviously other options that allow you to keep the original file or choose the quality of compression you want, but you can check out the man page for that.

Every now and again I set up a Linux server for some reason (DNS, nagios, web, etc). My choice for distribution is CentOS, mainly due to the fact I just like RedHat based distros.

There are usually a number of actions I perform on every server and I thought I’d collate them here so I don’t have to go looking around the web each time I need to remember the sequence of commands I need to run. There won’t be much here to start with but over time I hope this will build into a decent resource.

Disable IPv6

From: http://wiki.centos.org/FAQ/CentOS5#head-47912ebdae3b5ac10ff76053ef057c366b421dc4

Nothing I’ve done so far requires IPv6 so I disable it as having to sort through lists of IPv6 addresses as well as IPv4 ones can be confusing.

Just a quick one: When installing Exchange 2010 on Server 2008 R2 you need to make sure the IIS 6 Compatibility features of the Web Server role are installed. Otherwise during Exchange 2010 setup you will get errors indicating it can’t access the IIS Metabase and/or Default website even though it’s functioning perfectly.

There are plenty of blogs and articles out there about setting up nagios and not quite as many covering distributed monitoring. However there isn’t much out there surrounding trouble shooting of nagios-nsca. I had been having a problem where it would only accept one request before it started failing. I went through all the possibilities including multiple configuration problems, file permissions (probably screwing a few things up there in the process) and account permission issues.

In the end it turned out to be a simple problem ostensibly caused by the SuSE guys (the Operating System I was working with).

In the default init script (under /etc/init.d/nsca) it had the following:

The line starting with the ‘#’ is the original. The -s option (the only difference) meant it start in “standalone single-process daemon” mode. From what I can tell means “only accept one response”, which is pretty useless. Changing it to daemon mode fixed all my problems, and ended about 3-4 days of hair-pulling stress.

Some helpful Nagios links:

• http://nagios.sourceforge.net/docs/2_0/distributed.html
• http://www.packtpub.com/article/passive-checks-nsca-nagios-service-check-acceptor
• http://www.mail-archive.com/nagios-users@lists.sourceforge.net/msg06776.html (this is the one that solved my problems).