Locality Switch

I love a bargain; we all love a bargain. It’s even better when you get something really good for a dirt cheap price. One of the things I have been meaning to get for quite some time is a VPS, but I haven’t found a provider which meets my requirements in terms of price and reliability.

I’ve wanted a VPS to do so many things. As I work in radio, one of the primary things I’ve wanted to do was setup my own streaming media server. However, as I don’t have much cash to spend, it has been hard to find a reliable VPS provider within my price range.

The key word is here is ‘reliable’. There are countless dirt-cheap VPS providers out there, but when it comes to their network latency, uptime, and other factors, well, that is another thing all together. Not to mention the tendency for providers in this area to be here one day, and gone the next!

There are entire websites setup to track this whole ‘low end hosting’ market. One such website which I’ve been following for some time is low end box. They track the very low end of the market. In fact, their criteria for listing is under US$7/month. Now, that’s pretty cheap by anyone’s standard!

After umm-ing and err-ing for a while about singing up with one of these services, I finally decided last week to take the plunge. US$6/month and a few days waiting for PayPal to clear my money, I had signed up for Delimeter USA, which ironically, has their servers located in Germany.

For US$6/month, I got 256MB of RAM, 10GB of hard drive space, and a seemingly unlimited amount of data transfer (terms and conditions apply, of course). The data transfer was the deciding factor for me to choose this provider; I was happy when I found a provider which wouldn’t shut me down if I used a bit of extra data transfer.

As far as the stability of this service goes, I currently have no clue what to expect. They boast 99.9% uptime, but I’m not holding my breath. As long as they don’t go down for hours/days on end (as a previous Shoutcast host has done to me in the past), then I will be happy.

Right now, I have Icecast2 and Apache installed on my low end box, and it’s serving up 2CCR’s MP3 and AAC+ streams. I setup Apache as a reverse proxy, so I now have our station streaming on Port 80 (in addition to our legacy ports still working). I’ve also configured it to ‘pull’ the data from our encoder, rather than our encoder ‘pushing’ the stream out to the server constantly. The advantage of this is that if no one is listening, then we don’t use any bandwidth. So far, so good.

It’s great to finally have my own server running. Being able to do whatever Ubuntu supports is great. No longer am I restricted by what Plesk or cPanel wants me to do, or what my shared hosting provider dictates. I hope it stays up and running, and the provider doesn’t collapse. But if they do, there’s a whole world of other providers out there to choose from. To me, it’s a great adventure!

Crashing no less than six times in the space of fifteen minutes while trying to write yesterday’s blog post about Kwok. That’s right: Firefox crashed six times! Thank goodness for WordPress’ Auto Save.

Very embarrassing. Well, for Mozilla, at least. It’s just plain annoying for me.

This isn’t the first time I have had Firefox crash repeatedly on my Mac. Last time it happened, I was using Tiger. This time, I am using Snow Leopard.

Was I doing anything unusual when the crash occurred? No. I was basically just writing in WordPress, and occasionally uploading an image (with both the standard uploader and the Flash uploader). There have been reports of Firefox crashing when using WordPress, but that was resolved as an issue with Google Gears; I don’t use Gears, so this doesn’t help me.

I have used this experience as a chance to learn about how Mozilla deals with crash reports.

If you navigate your Firefox browser to about:crashes, you will get a list of GUIDs which reference to your crashes. Click on one of them, and you will be taken to the Mozilla crash reporting website, where you can see the gruesome details about the crash. Core dumps, running threads, modules, kernel details – it’s all there.

Bundle this with Bugzilla, and Mozilla really does have a great platform for dealing with crashes and bugs. It’s a really comprehensive system they have setup to deal with the sheer volume of issues they would have. Great job! Now, can you please fix my Firefox?

The Community Broadcasting Foundation announced in their July 2009 Email Newsletter that they would soon be offering grants for IP based Outside Broadcast equipment. Well, it’s finally arrived. Here’s the details.

Any long term or temporary community station without any existing IP OB equipment may apply for the grant of up to $2,500. This opportunity is only open this year, so if you miss the 12th October 2009 deadline for submissions, then there is the chance you may never be able to apply for this type of grant.

Taking a look through the application form, you are asked how many OBs you did in 2009, and are asked to list the last five OBs you have performed. While it does say “if applicable” next to this bit, my guess is that stations with a previous history of OBs will be given priority.

The application form has a table with different categories of items, with a place to write in your quoted amount for each of the ones you are applying for. Categories include notebook computers, USB modems, Mixing consoles, Microphones, Audio interfaces and most curiously, Audio Streaming Software.

The Audio Streaming Software category is either badly worded, or implies that they don’t want to fund hardware solutions. It looks like AudioTX and Tieline are out of the question. I suppose this isn’t surprising, as the OB-over-IP Manual released by the CBF didn’t really cover any hardware solutions (or any open source software packages, either!).

What does this grant opportunity mean for the sector? It means that it may now be possible for many stations to move away from POTS codecs and leased lines which have been used for traditional outside broadcasts, and move towards IP equipment over 3G connections. It will allow for much more programming freedom, and will possibly allow for more OBs to occur.

Easier and cheaper OBs will hopefully allow for more community interaction and involvement, and that’s what we’re here for. Right?

Governments have a tendency to make things much harder than they have to be. Take, for example. social media. It’s a simple thing, isn’t it? Even a kid with elementary computer skills can understand how to post something to YouTube, a Blog, or Facebook. Why do our politicians get it so wrong?

Bureaucracy.

Let’s look at the short lived blog of Senator Conroy’s: The short lived Digital Economy blog. It looked and smelt like a government website, and got so many things wrong.

How about the Twitter account of my local Mayor, Councillor Larry Bolitho. The tweets are written in third person, each one starting with “Larry Bolitho”, and proceeding to tell us where he is. There aren’t any personal thoughts in there. Oh, and have you seen the instruction manual and disclaimer which comes with it?

Responses will be at the discretion of the Mayor… Who@HillsShireMayor chooses to follow on Twitter is at the discretion of the Mayor.

It would be great if our politicians could get with it on this whole social media thing. Some of their present attempts are just embarrassing.

However, there is someone who is up to date on this whole social media arena. Barack Obama. Have you seen the Whitehouse website? It’s beautiful, functional, accessable, and also integrates with social media platforms.

Anil Dash wrote a blog post about Government internet startups. He notes that they are now actually doing things right:

Now, .gov websites have historically been backwaters at best, a bunch of awkwardly-designed, poorly defined sites that only met the bare requirements of a web presence. But of course the current administration is comprised in great part of digital natives, and it’s remarkable how quickly they’ve remade the .gov world into not just a number of compelling websites, but into a broad set of platforms that are going to inspire as much technological innovation as Twitter, Facebook or the iPhone did when they unveiled their technology platforms.

It’s great to see what the Government of the United States of America is turning out, in terms of websites and useful web tools. They are being innovative, and are truly leading the way in government web presences. Countries around the world should use this as the benchmark, and strive to meet it.

Let’s hope the rest of the world can see the benefits of social media and the web, and start doing things properly. It can work to their advantage.

Twitter’s Trending Topics is where you can see the most popular topics on Twitter at the moment, and check out the stream of tweets related it it. It all works in real time, which opens it up to a whole heap of spam.

Twitter makes is really easy to get your message in front of a large audience through these trending topics, which spammers are utterly exploiting. To get your message in there, just include the trending key words in your tweet, and you magically appear in the stream for the topic.

It’s not only the spammy messages which are annoying – it’s also those foreign language messages which appear in the streams. I don’t need to see the tweets from those speaking another language – it doesn’t help me one bit.

How can this spam be stopped? Well, it would be fairly easy to remove the foreign language tweets from the stream – just have a set of common words for each language, and if the tweet matches another language’s key words, then Twitter can hide it from my stream.

The issue of real spam would be a bit trickier to fix, but I believe it is a very doable. The first thing to do would be for Twitter to scan each tweet and see if it has a whole heap of trending topic keywords in it. These are the most obvious type of trending topic spam.

Second thing to do would be placing a little link on each Tweet in the stream of Trending Topics which says “spam”. If people click that, it flags it for moderation. Oh, and to avoid abuse of the spam button, it should be only be available to users who have been on the site for a few months.

Finally, I believe that it should take a certain amount of time of someone being a user before they can actually appear in the trending topics stream. It won’t hurt if the new members don’t get in the stream, will it?

These are a few simple ideas which I humbly present to Twitter to help remove the low quality contributions by people who feel it is necessary to spam up my tweet-box.

I’ve been playing around with MailChimp, which I am thinking may be used for a monthly member’s email at my local station. MailChimp, I have found, does a very nice job of allowing me to add images into a predefined template and write some content for it. However, what if I want to code my own emails?

Writing HTML to be displayed in email clients could be one of the riskiest things I ever do. Even more riskier than deploying a site written in Apache Tomcat (I never got along with that beast)!

Luckily, MailChimp has a nice little guide to help me out. It’s entitled How To Code HTML Emails, and contains advice such as this:

An HTML email is nothing but a web page. That’s it. I’m sorry if you thought there was more to it than that. So if you can code your own web page, you can code your own HTML email templates. There is a little catch, though. You have to code like it’s 1996 (I’ll explain later).

One little catch?

Do they even remember what the web was like in 1996? Incase you have forgotten what the web in ‘96 looks like, there is a handy guide avaliable for you:

the technology was different in 1996. Although Internet Explorer 3.0 could run Java applets and inline media, Netscape Navigator could not, and in any case nobody felt comfortable doing anything more complicated than making a few animated GIFs. Additionally, very few web designers had even the most rudimentary of aesthetic sensibilities, and nearly half of them were clinically retarded. The internet in 1996 looks like it had been created in its entirety by a panel of 13-year-olds with Geocities accounts who had about half an hour to spare each night before bedtime.

From: https://www.msu.edu/~karjalae/internet96.htm

If you follow the link to that page, you will see some beautiful screen shots of the web in ‘96. Wow!

Why would MailChimp tell us to code like that? It’s pathetic! Sadly, the safest way to code emails is to code like it’s 1996. Sadly.

Mail clients haven’t caught up with the rest of the world. Tables are the way to go, because email clients love to strip out half of your code, including CSS! I forgot the last time I wrote a page using a table for layout purposes – why are emails so backwards?!

There is the Email Standards Project which works “with email client developers and the design community to improve web standards support and accessibility in email”. Hey, they even publish a report on the status of each major client. Not surprisingly, Lotus Notes and Outlook 2007 are the worst, along with Gmail!

Their testing is based on their own Acid Test.

Let’s hope email clients can catch up with the rest of the world, and stop behaving like Internet Explorer 3.0!

In radio, highly accurate clocks are a must have. You can’t run without one. Even being out by a second or two can make timing out to the news a nightmare. Luckily, it’s really easy to synchronise your computer’s clock with highly accurate internet clocks.

It all works with NTP, or Network Time Protocool. It’s basically a protocool which allows you to synchronise one computer’s clock with another (hopefully accurate!) one. There’s heaps of clocks to synchronise with, and the NTP Pool Project is where you can find the network addresses for these clocks.

Signing up to the NTP Pool to get your clock synced with is pretty easy, and just requires some basic registry editing. Windowsnetworking.com explains the registry settings for NTP clients, and once you get one machine setup it’s just a matter of adding the registry changes to your logon script (or doing it manually, if you so desire!).

However, if you want more control over your synchronisation, you may setup an internal time server of your own which all of your network machines sync to. In the Windows world (read: Active Directory), it’s just a matter of telling your Windows Server to sync to the NTP pool at a fairly regular interval (once an hour?), and then tell your clients to sync to your own Windows Server at a regular interval – it can all be set with some logon scripts.

The trick to this method if selecting a Stratum One or Stratum Two NTP server on the web, from the public list of NTP Project servers. The reason for this is so you get a steady and reliable time source, rather than just selecting one at random from the NTP Pool.

I decided to sign up for FeedBurner. It couldn’t have been much easier. Here’s what happened:

1. I went to the Feedburner site – it already had me signed in to my Google Account
2. I entered the Feed URL of my RSS feed
3. I pressed next a couple of times

Simple, eh?

My new feed is up and running! If you’re interested, please subscribe.

Within the FeedBurner console, there’s a few tabs which allow me to see how many subscribers I have, etc. It’s neat. Even the copy they have to show the stats are being gathered is smooth:

The hard part was actually trying to redirect my Wordpress feed URL to the new FeedBurner URL. There was no simple way to do it, at least that I could find up front. The FeedBurner Help sent me through all sorts of different things, before I could find what I was looking for. Even a Google Search wasn’t helpful.

The FeedBurner Help provided a hint to redirect any URL ending with .xml, but as I use the custom Permalink URLs, that won’t work. No luck there.

After some further digging, I discovered the FeedSmith plugin which does what I want it to do: redirect all of my existing feeds to FeedBurner. Perfect. I uploaded that plugin, activated it, and then typed in my new Feed URL. It’s all done.

What I want to know is why this FeedSmith plugin isn’t on the Wordpress Plugins Directory? Would it have something to do with Google taking over the plugin’s development, and the fact that they own Blogger?

Wikipedia has had, in my opinion, one of the oddest user interfaces in the mainstream web of today. I say oddest because it isn’t necessarily bad, but it just didn’t seem to fit in with the style of the rest of the web.

While the west of the web has been opting for clean layouts with heaps of space, Wikipedia has had a very cluttered look and feel to it. And, it’s been primarily Grey. I mean, c’mon! Grey?!

The Wikipedia Usability Initiative has apparently been working on a new design for a while and been testing users to see how they interact with the design. The new interface is now ready for users to try out, as long as you have a Wikipedia account.

The design changes are minimal, which is good. I like what they have done. They Realigned, not Redesigned.

However, I believe Wikipedia have not gone far enough in the improvements of their site. Here’s two things they should have done:

• Some sort of AJAX thing for opening up the Edit tab, rather than loading another page
• WYSIWYG editor for the Edit tab. Even if it was only for some core functions, not the advanced ones, that would be good

This is a chance for Wikipedia to refine and update their interface to improve the user’s experiences. I hope they can make the most of it.

There’s a few websites I go on the internet for insight into design and usability trends: A List Apart, Alertbox, Sitepoint, Web Design from Scratch, and Signals vs Noise. I generally find these sites very useful, and I respect the opinions of the writers there. However, sometimes people are just plain wrong.

The offender today is Jakob Nielsen, who writes for Alertbox. He published an article back in July about password masking, and why we should do away with it:

Most websites (and many other applications) mask passwords as users type them, and thereby theoretically prevent miscreants from looking over users’ shoulders. Of course, a truly skilled criminal can simply look at the keyboard and note which keys are being pressed. So, password masking doesn’t even protect fully against snoopers.

In essence, he is saying we should do away with those dots or asterisks which appear when we type in our passwords. What a bad idea.

The only place I believe this may be valid for is private offices. Anywhere else, it is a bad idea. Schools, internet cafe’s, open plan offices and public kiosks are just a few example of where people type in passwords where people seeing the screen is common place.

Password masking doesn’t protect against malicious criminals who want to hack your bank account. It protects from casual hacking – where people you know look over your shoulder and grab your password to have a bit of fun with you. With password masking, this isn’t a problem. Without it, we are all opening ourselves up.

Perhaps there is a solution to help Jakob see his password on the screen: a browser extension which toggles the visibility of your password on websites. Simple. He can use it, and the rest of us can keep living our lives as if nothing ever happened.

(Just a note: I still respect Jakob Nielsen – I think most of his articles are great. It’s just this one thing I have a problem with. Even others agree with my point of view.)