Linux Desktop Cohesion part 1

3 01 2010

I’m a big fan of Linux and also Windows, I work on a simple concept of using the right tools for the job. All the servers I work with are Linux based, but that’s because it provides the right environment and performance over Windows (though its a much narrower advantage now than ever before). If Windows turned out to be the right server for as particular deployment, in exactly the same way if a Sun box or iSeries did, then I would use it. Operating System Fundamentalism is pointless and does nothing to drive forward anything useful.

On the desktop it gets more complex. Windows works. OSx works. Linux really struggles. Why does Linux struggle? Well look at Windows and OSx, they do not provide a dozen different window managers, desktop managers, theme engines, windows controllers, widgets, gadgets, thingies, whatsits and godknowswhateles. They do one! I’m all for diversity and choice but how about some natural selection too? (or is that what’s happening and the Linux Desktop is breeding itself out of the gene pool?)

Under Windows or OSx pretty much every thing works in a similar way, looks the same at the application control level, uses the same mechanisms for interaction. One of Linux’s greatest features is the one thing that continually stops it making any real inroads into the desktop market. Saying Linux allows you to make it work exactly how you want it to is great, pointless, but great. Sadly though no one wants to spend a week building a desktop environment that works and doesn’t hurt their eyes. Even distributions that try to (and continue to) tidy up the mess only get so far and once you go past the very basic and very short list of distributed applications you’re lost in a sea of miss matched GUI’s and bad UX. I think a fair measure of acceptable levels of success would be to be able to move from word processor to email to IM to web to micro blogger without having to adapt to 5 different UI’s for the basic applications.

You would (sensibly) think that this mess should have been resolved by now, but in my opinion its getting worse. Its probably true to say that Gnome and KDE represent the two dominant desktops under Linux and they are both as bad as each other. I love them both but boy they make it hard to create a desktop environment that looks decent and works across all the applications required. And its not just the Linux side that continues to make a mess of it, so called ‘cross platform’ systems like Adobe AIR also make a hash of it, seriously how much of an ask is it to expect a application to use the standard frame buttons for min/max/close?

So how much of this is down to the development philosophy of ‘my way is better’, well possibly all of it. I’m not saying ‘your way’ isn’t better, but if it makes ‘your’ application behave differently then quite likely that’s going to mean it doesn’t get used as much as it would otherwise do.

The UI/UX battleground is a war torn landscape dotted with small victories lost in a mire of defeats.

Perhaps the increasing number of interface systems; desktop, web, mobile, netbook, is making for a problem that is simply too large, too out of anyone’s hands to ever be solved in a meaningful way. Gesture based multi touch can greatly simplify things and if you can think past your preconceived ideas of Windows Icons Mouse Pointer, can be intuitive and productive, but it does not lend itself easily to being a cross device method. We will see multi touch desktop based systems and even if their price makes them real they wont work as the keyboard is the primary interface and ‘touch’, multi or otherwise is a control method.

An early conclusion could be that the reason we have such a diverse ecosystem of interfaces is because no one has got it right yet, I would say that after Xerox got the basics and everyone stole/copied it progress has been very slow.

So, what do I want for my Linux desktop? Well I would like to be able to have a desktop environment where everything worked well, was easy to make it look like it is all part of the same thing (without a degree in astro-emerald-compiz-physics) and looked good out of the box. Until then I will still be using it as a desktop (alongside Windows of course) but I’ll be one where a thousand others will stick to Windows (or OSx) only.

Posted via email from Steve’s Blog





Google OS

23 11 2009

This week Google gave a preview of its upcoming operating system, ChromeOS. It also open sourced the project as ChromiumOS and the Chromium Browser (Google Chrome). Now then, apart from confusing the hell out of everyone with its naming (including me, so if my understanding is wrong, what did you expect?), this has been talked about as either the Microsoft Windows killer or a waste of time. Personally I think its neither.

I downloaded the source code and built it. I then installed it on my netbook. I thought I would spend a few hours playing with it but after 30 minutes there wasn’t anything left to play with. If you have been wondering why there seems to be a lack of screenshots on the interweb its because there is precious little to take a screenshot of.

Basically, Google’s OS is a completely stripped down linux variant (they are working with Canonical of Ubuntu fame), it boots very quickly (I got mine booting in something like 10 seconds) but thats because it doesn’t boot anything other than the bare minimum needed to run a very limited list of hardware and the Google browser, Chromium. Once running everything is in the ‘cloud’, its all on line, apps, data, the lot, its all on the net. No local storage.

The browser has matured into a ‘full’ front end to everything you want to do with your netbook, to be honest I don’t see anything really new here, in fact I’m not sure Google OS gives so much as it takes away, less is more, more or less.

I think the UI concept has some limitations, but these are offset by the netbook, you wont run 10 apps at the same time because your netbook doesn’t do that, its really 1 or 2 things at a time and the UI (not just Google’s but the others too) handles this quite well. In fact you could give Google an extra point here for having applets inside the bigger frame, such as Google Talk or a notepad.

Its all online, it will be interesting to see how well unreliable connections are handled, writing your presentation on your netbook on the train to work sounds great. Its something we have all done. But today when the train goes through a tunnel I don’t lose the connection to my hard drive, tomorrow with Google’s OS I would lose my connection to the cloud, without some clever caching and recovery at best I will have to wait for the connection to return, at worst, well lets see.

You can read more everywhere about it. I think its good, it shows what you can do if you focus on your purpose and nothing else. Is it going to kill Windows? No! Its not going to even raise more than a ‘oh yeah we have heard of it’ from the boys at Microsoft and why should it?

But, in a years time when they launch it with the hardware as well, remember that all they have let out in the wild is the open source project, well then I think it will be time to sit up and take notice. A proper light weight OS, no frills, on hardware built for the task. Think of the applications. Google’s web browser tablet, news reader, media player, cell phone, video phone… this is just the first showing of the first building block of Google’s OS. Only a fool would write it off based on this, lets see what Google does next.

So is Microsoft going to be looking at the next 12 months and wondering what the landscape for discrete devices might look like if Google makes its play? Yeah, I think so.

I read today that, shock and surprise, its likely that Google OS and Android (Google’s OS for cell phones and mobile devices – you can see the overlap) will converge, read here