Reviving a slow XP machine

Lately, my trusty XP desktop has been grinding to a virtual halt. Things were getting so bad that I’d decided to trash the lot, and do a fresh install. But having made the decision, I got brave enough to do some radical / drastic spring-cleaning – sure, what did it matter if I borked it? I was going to reinstall anyway! – and it’s actually been enough to bring the thing back to life. Not quite as zippy as a box-fresh installation, but certainly back to acceptability.
A few tips I’d like to share from the experience:

  • There really aren’t more than a handful of software applications you really, really need. I made a list of barely 20 things I absolutely had to have if I did a fresh install (and 6 of those were web browsers).
  • CCleaner is a marvellous piece of freeware, which will wipe your Windows machine clean of all sorts of stuff that simply doesn’t need to be there. It also includes an ‘advanced’ (?) Registry Cleaner, which identified and sorted out hundreds (literally) of issues for me.
  • One particular area of agony was the time it took for Windows to draw up the list of ‘All Programs’ under the Start button. As a workaround, I started using Launchy – which effectively puts a ‘command prompt’ inside Windows. Press Alt-Space, start typing the name of what you want, and it’ll present a list of suggestions. Pretty quickly, you’ll find the right programs are coming up when you type the first letter. It’s a remarkable timesaver, and so efficient I’m still using it, even after reviving the machine.
  • On similar lines is the experimental Firefox addon, Ubiquity. As with Launchy, a ‘hotkey’ brings up a command line interface, where you type pre-determined commands to do online things. Some of the reviews have been glowing; personally, although it’s certainly clever, I didn’t find myself using it much. So as part of my spring-clean, I uninstalled it… and was amazed at the speed increase I saw in Firefox as a result. I can’t say for certain that Ubiquity was slowing things down; but there is mention of a known bug in the current version, which may have been causing my problems. I’ll be back (probably).
  • There are so many good uses for a service like Dropbox: an online storage / backup / syncing service, with Windows, Mac and Linux clients, plus a pretty web interface. 2GB is free, 50GB costs $99/year. It’s very tempting.

And finally, one that won’t be at all contentious…

  • Linux is ready, but the IT industry isn’t. I was beginning to think it’s time to finally ditch Windows. I’ve been tinkering with Linux on and off for 10 years; and there’s no doubt now that Ubuntu, Fedora or various other distributions are now up to scratch. Was this the moment to go Linux-first, with Windows as backup – be it dual-boot, or ‘virtual machine’? In the end, I bottled it. Partly because I’d revived the Windows installation, but mainly because there were going to be issues with many, perhaps most of my peripherals.I was able to find long, tortuous workarounds to make most (but importantly, not all) of my kit mostly work; but that’s not the point. I don’t have the time or energy to put up with lengthy, intricate installation processes, or accept limited functionality, when we know it would have been plug-and-play under Windows. True, there is a part of me that would fancy the challenge; and I dare say, for most hardcore Linux types, that challenge is (the main?) part of the attraction. But it’s too much to ask of mere mortals.Don’t get me wrong; I’m very pro-Linux in principle, and I actively want to make the switch – probably to Ubuntu, since you asked. But I need people like (quick glance round the room) Dell, Apple, Garmin and (ha!) Microsoft to catch up. Make your kit work with Linux, guys, out of the box – and we’ve got ourselves a revolution. In the meantime, I’ll have to satisfy myself with occasional playing in Sun’s VirtualBox.

Long Term Support for WordPress

Oops. I wrote this piece yesterday, wishing that WordPress offered Long Term Support for occasional releases, along the lines of Ubuntu. I then get a comment from Mr WordPress himself, Matt Mullenweg, telling me that there actually is a long-term supported release. Here it is for the record…

The official policy from Team WordPress about software upgrades, as described by Matt Mullenweg last month, is pretty straightforward: when we release a new version, you should upgrade. Like, immediately. But when you’re dealing with the corporate world, where you deliver a project and effectively walk away, it isn’t quite so simple… and I’d personally welcome a Long Term Support approach along the lines of Ubuntu.
WordPress was built for bloggers: technically literate self-publishers, with some grasp at least of what’s involved in running a website. But as I’ve documented here countless times, and as my continuing mortgage payments demonstrate, comms professionals with no particular IT skills find its convenience, flexibility and simplicity (not to mention the price) equally appealing.
But the chink in the armour, if you like, is WordPress updates. Corporate projects tend to come with lists of requirements, which push well beyond normal blog-based sites. Normally, these requirements are achievable using plugins or a bit of custom code. But as Matt acknowledges, when an upgrade comes, there’s no guarantee that a particular plugin will work. And even worse, given that most plugins are offered up by volunteers, there’s no guarantee that the plugin will be updated accordingly.
I’m afraid Matt’s assertion that ‘having a secure site is much more important than the functionality of a single plugin’ won’t really stand up in the corporate context. You’ll ultimately face a decision between a site which might be at risk, but does everything you want; or (to put it provocatively) an under-performing site which still won’t be 100% secure anyway, because nothing ever is. And I’m afraid most marketing or communications people will choose the former.
There’s also the issue of the high-visibility upgrade notifications in the more recent WordPress releases. Whilst these are fantastic for those of us who run our own server setups, and aren’t scared of the upgrade process, I’ve had several phone calls from clients who are seeing this warning, and panicking (I’d say) unnecessarily. And I can’t honestly promise them that ‘hey, just do it, nothing can possibly go wrong.’
There is a compromise solution here: and that’s the model of Ubuntu‘s Long Term Support releases.
There’s a new version of Ubuntu’s Linux package every six months, with a promise to offer product support (ie minor fixes) for at least 18 months. But some of these are designated as having Long Term Support: these come with a promise of three years’ worth of fixes for the desktop version, and five years’ worth for server versions. It doesn’t mean that you’ll never have to do a major upgrade. But it’s a guarantee that the fundamentals won’t change for a considerable period – long enough to put the IT manager’s mind at rest.
That’s the kind of commitment I’d value as a WordPress ‘developer’. I want to present pitches to clients based on guarantees, not probability. And I’ve seen specific examples of excellent WordPress plugins, perfectly secure and stable in their own rights, which suddenly become obsolete because something changes in the next major WordPress release. Looking back, the changes are almost certainly for the better overall; but not if I’ve built a particular function around a particular plugin which no longer works.
The v2.5 release of WordPress takes it to a new level of maturity. A policy of LTS releases, ideally via simple ‘overwrite this file with that file’ patching, would signal the product’s readiness to be taken most seriously in corporate environments. And it would make an already strong proposition almost undeniable.

… and here’s the info about the ‘legacy 2.0 branch’ which is almost exactly what I was asking for. Now, I consider myself fairly well versed in the ways of WordPress, but I’d never even heard of this, and various Google searches yielded nothing.
I guess my only response would be that the description of the legacy branch needs to be rethought. The word ‘legacy’ (to me anyway) sounds negative; the idea of ‘Long Term Support’ sounds positive.
My thanks to Matt for correcting me, without making me sound like a total idiot.