Internet Things

Bloglines closing down

by
Bloglines is down!

Bloglines, a web site which allows you to subscribe to and read RSS/Atom feeds online, is closing on October 1st.

It’s a bit of a shame – I started using Bloglines almost exactly 6 years ago, and it was great to be able to read all of my feeds on one central web site. That being said, a couple of year later I moved to Google Reader and have stuck with it ever since, due to its better user experience. Bloglines never seemed to catch up and I only know a few people still using it now.

Even so, the number of people using feed readers is apparently falling, and according to paidContent Google Reader is also experiencing a fall in traffic. Whether this is because of feed features in desktop browsers like Firefox and later versions of Internet Explorer have become people’s preferred way of using feeds, or that people prefer Facebook or Twitter, I’m not sure, but it would seem that their popularity is waning.

Up and running again!

by

At last, the move to the new host is complete!

I’m now hosting with Bytemark, a well-known hosting company based in York with data centres in Manchester and London. The domain is managed by 123-Reg – I’ve decided to keep the domain and hosting separate to make any future host changes easier.

The hosting package is quite different to the package I had previously. Whereas before I had a simple virtual host setup with FTP, email and databases, this package offers SSH access to a complete virtual machine running Debian. In other words, I can configure everything on the server essentially as I wish – even changing to a different Linux distro if I wanted to.

I’ll be sticking with Debian for now. It brings with it Apache 2.2 and MySQL 5, and I have the ability to add things like FastCGI should I wish to at a later date. I’ve also been able to install every optional Perl module that Movable Type can use, including ones like Image::Magick and XML::Parser which require compiled code.
I’m still using Movable Type 4.3x for now – it’s still supported, with version 4.34 having come out a few days ago, and having tried MT5 for a bit I think it’s a bit overkill for what I need it for.

The move to the new host wasn’t quite as smooth as I’d hoped – some issues with having the domain released from the old host, for example, and the new machine that the site runs on had a disk failure last night, literally about 30 minutes after I updated the DNS to point to the new server. It’s also taken some time to get used to configuring the server myself, rather than having a fancy-pants admin interface to do it for me. Suffice to say I’ve learnt a lot about the Unix command line over the past couple of weeks.

Hopefully now everything will work, but give me a heads-up if it doesn’t!

New host

by

I have just ordered a new hosting package with a different hosting company. The package offers faster and more up-to-date servers for around £5 extra per month.

The switch will take place shortly, and will allow me to upgrade to the new Movable Type 5.01 which has just been released. MT5 requires MySQL 5.0 (released several years ago) yet my current host is still offering MySQL 4.1, as well as a very old version of Apache. There’s also no support for persistent environments such as FastCGI which are key for getting the best performance out of Movable Type.

I’m aiming to have the migration to the new host complete by the 14th, which will be the 8th anniversary of the first post to this blog.

Firefox, Performance, and You

by

I’ve been blogging quite a bit about the beta versions of Firefox 3, which can be downloaded here, and how much better it is than Firefox 2. If anything, I’m as excited about this version as I was when version 1 came out, as it feels like the browser I have been using for 5 years has now grown up and many of the issues it has had in the past have been ironed out.

In this entry, I am going to go through the various changes that have been made in Firefox which improve the performance of the browser – the startup times, the time it takes to render web pages, and the general responsiveness of the program. Improving performance has been a major focus of the Firefox developers for this release, and the changes are notable.

Memory usage

Do a search for ‘firefox bloated‘ and you’ll get over 500 000 results, many of which complaining that Firefox is big and slow. Many put this down to the amount of memory it uses, which, in Firefox 2, appears to be more than its main competitors – Opera, Internet Explorer and Safari. (Note that, unless you’re using Windows Vista, it is difficult to measure memory usage of individual programs on Windows successfully)

For Firefox 3, reduces Firefox’s ‘memory footprint’ (the amount of memory it needs to run) was one of the main priorities. Improvements have been made both in reducing ‘memory leaks‘ – memory taken by Firefox that isn’t released back to the operating system when it is no longer needed – and memory fragmentation, which means that the memory it does have allocated is used more efficiently. The culmination of this is that, on Windows, Firefox 3 uses less memory while running than any other current web browser. It also released more memory back to the operating system after a period of inactivity, meaning that it’s no longer needed to close and restart the browser because it’s eaten half of your RAM.

How does this improve performance? If you have 16 GB of RAM, then, not much really. But if like most computers your machine has just enough RAM to perform, then it means more of Firefox can fit into RAM and less of it has to use ‘virtual memory’ – essentially a big file on your hard disk which acts as an overflow when all of your RAM is used up, but that is much slower than RAM. It will also allow more space for other programs in RAM, so you’ll be able to run more programs at the same time as your web browser.

Code optimisation

Several parts of Firefox have been re-written to fix bugs and simplify code, and in some cases, improve performance. There have been performance improvements in:

  • Page rendering – a marked performance increase that also means that Firefox 3 will pass the Acid2 test
  • Image drawing – a bug fix sorted out problems with image caching and a new JPEG decoding routine also means that images will render faster

Mac-specific improvements

Firefox 3 finally moves away from the really old QuickDraw API to the newer Cocoa API. As well as giving nicer widgets (scrollbars, buttons etc) on web pages, this also means the whole program should feel more snappy. This, combined with the new theme, means that Firefox 3 is the most Mac-friendly version to date.

Profile guided optimisation

Firefox is now, essentially, compiled twice – the first compile runs as usual, and then the compiler runs a series of tests which optimises the code by looking at what parts of the browser will be most used in a working environment. This has lead to performance improvements of at least 10%.

Because we’re now approaching the ‘release candidate’ stage, ahead of a final release probably some time in June, most of the performance changes have now been made, so if you want to see if it is an improvement for you, now is as good a time as any to try; any further changes will be to fix any major bugs and stability issues. All in all, the massive strides made in performance make Firefox a much bigger competitor in the web browser arena and may win back some of its earlier detractors.

openoffice.virginmedia.com

by
A screenshot of openoffice.virginmedia.com showing folders and files to download.

I know there’s at least one other person here who uses OpenOffice.org as their office suite and Virgin Media as their ISP, so you may be pleased to know that Virgin mirrors OpenOffice downloads at openoffice.virginmedia.com. If you’re a Virgin Media customer, then this is the fastest way of downloading it as it should require the least number of hops.

Virgin also mirror all downloads from Tucows and possibly some other sites so it may well be worth checking out before you download, especially if it’s a big download and you’re lucky enough to have Virgin’s 20 MB broadband offering, as you’ll be able to make the most of your bandwidth.

Akismet and Data Protection

by

In Britain, we have this law called the Data Protection Act, which dictates what companies and organisations can and can’t do with data about its customers, clients or employees. It basically puts a duty on organisations to ensure that a person’s data is kept private and cannot be compromised, that the person is aware if that data is being shared with third parties (and seek their permission if needed) and that the person is aware if data is to be processed in another jurisdiction, where data protection laws are not equivalent to this act. There’s more to it than that – Wikipedia goes into more detail and the full text of the act is here.

So how does this relate to Akismet, the spam-filtering web service that I conveniently mentioned in the title of this post? Well, I use Akismet on this site as a way of stopping spam (though to be honest it’s been largely redundant since I started using Comment Challenge). This site is based in the UK, and therefore falls under UK law, but Akismet is a US service, and right now every comment submitted (bar those from approved TypeKey and OpenID commenters) is being sent through it.

Though I haven’t yet got a privacy policy on this site – it’s something I’ve been working on now and again for some time – this does bring up some privacy implications. Without Akismet, the privacy policy would say something like this:

Upon submission of a comment, the details provided (name, email address, URL and comment), along with your IP address, will be stored in a database. Your comment will also be displayed publicly on this web site.

Any comments you have made can be removed at any time, by contacting the site owner and requesting their modification/removal.

With Akismet brought into the equation, we have to add the following:

The details you submit will also be sent to the Akismet service, for the purpose of identifying possible spam comments. Akismet is based in the United States of America and falls under the laws of the State of California. The details submitted will not be stored, unless the comment is marked as a ‘false positive’ (a legitimate comment which is automatically identified as spam) in which case it may be stored for some time for diagnostic purposes.

For more details, please consult the Akismet Privacy Policy.

Now I’m not a lawyer and my experience with data protection mostly comes from a university module that I took recently, so this is certainly not legal advice. But it’s something that I hadn’t thought about until reading an email from the Six Apart Professionals Network this morning.

Akismet is done by the WordPress guys and I’m sure they’re trustworthy, and it’s also not entirely fair to pick them out as this could be any other web service – it just happened to be the topic of conversation at the time. But while a few bloggers using it isn’t going to cause much of a kerfuffle, a big organisation could land themselves in hot water if they’re not totally upfront about what is happening to their users’ data.

1000 reasons why IE is better than Firefox

by

It’s in French, but here’s 1000 reasons why IE is better than Firefox. Except that there’s only actually 14 reasons, so it’s 986 short. And some aren’t really reasons. In fact, it’s a pretty pointless piece of flame-bait designed to stir up the Firefox community.

In any case, here are those reasons translated into English, with the help of the Google Translator and my French A-level (feel free to correct me if I get any of these wrong):

  1. Internet Explorer is integrated with Windows XP so you can surf the internet without having to decide which browser to use.
  2. Internet Explorer is accessible for everyone.
  3. IE can display all web pages – not all of them work in Firefox.
  4. IE requires hardly any extensions, whereas Firefox has a plethora.
  5. IE does not ship with tabs, but they can be added if you wish.
  6. IE does not have an RSS reader, which is a fad. You can just visit the site every day if you need to.
  7. IE is less secure than Firefox, but patches are released regularly.
  8. IE is closely linked to Windows.
  9. IE does not pass the Acid2 test like Safari does.
  10. You need to use IE to access Windows Update – even though Microsoft released a Firefox plugin for Windows Genuine Advantage, they will never open Windows Update to Firefox.
  11. IE doesn’t crash when visiting online banking sites, Firefox does.
  12. IE7 will adopt the standards of 2006 and include innovations of Microsoft and others, such as RSS, SSL3 and better security.
  13. IE does not respect W3C standards in their entirety.
  14. IE is the ‘de-facto’ standard that developers work to.

There you go. Pretty lame reasons, in my opinion, and some of them are reasons not to use IE as far as I’m concerned.

[Via Glazblog]

RSS on the go

by

Reading this comment by Dave gave me an idea for a cool feature that I’d like to see some desktop RSS aggregators like FeedDemon to adopt: remote access.

Desktop aggregators are great as they have a nice interface and can work independent of a web browser. But unlike web-based services like Bloglines, they’re not easy to use on the move – if you’re not using your machine then it can be difficult to keep track of what you’ve read. You could find an aggregator that fits on a USB keychain but if you’re in an internet cafe that won’t let you plug those in then you’re stuck.

Remote access, to some extent, solves this. It builds on a feature implemented in P2P clients like eMule and Shareaza, which allows you to administer the client via a web interface in addition to using the desktop interface. This works by the client including a small web server and opening a port on your computer which you can then access from other machines (providing your computer is switched on and has an active internet connection). So, say I have Shareaza running on this computer with remote access enabled – I can then go into university, find a computer, type in my laptop’s IP address and port number for Shareaza’s remote access feature into a web browser, and then login. Once there, I’d be able to see how my transfers were progressing and even do searches and start new transfers through my laptop. And when I got home, those transfers may well have finished. That’s all currently possible.

Now, imagine FeedDemon had this feature. I could go into university, find a computer, type in my laptop’s IP address and port number for FeedDemon’s remote access feature in a web browser, and then login. Once there, I could read the all the feeds I was subscribed to, update channels and add new feeds. And when I got home, the feeds I had read would show up as being read.

I don’t believe any aggregators support that yet but it would be an excellent feature to add, since it would combine the convinience of a desktop aggregator with the ability to read feeds anywhere that a web-based aggregator has.

I’ll Lazyweb this post, just so that it can attract attention.

GMail Invitation

by

Update: (24th Jan 2006) I have no more invitations left. Thanks for your interest.

Hanni dropped me a GMail invitation earlier in the week, and although I haven’t really used my account, I now have an invitation of my own to hand out to someone. So, if you want one, be the first person to comment on this post. Make sure you include a real email address that I can actually contact you with, otherwise you won’t get one.

Creative Commons License
Except where otherwise noted, the content on this site is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.