the corner office : tech blog

I don't know how I stumbled across it but I read a post on Mark Shuttleworth's blog about tooltips on the panel being removed in the next version of Ubuntu. There's been a bit of an outcry on a related bug and elsewhere.

Like many users I use the tooltips all the time - hover over the system monitor, hover over the time to check the date, etc etc. It's a bit tenuous saying 'less is more' when it means that something that was afforded by a simple hover now requires multiple clicks (especially on something as sluggish as Gnome, but that's another story...) Less has now become more (effort).

It's just annoying, and it happens often with free software (and for all I know, proprietary software) where someone makes a decision where removing a useful feature is done in a way that seems arbitrary or ill-considered. There was once a feature in Eclipse which allowed you to disable the 'Insert' key so that you didn't accidentally go into overwrite mode, a mode used (I suspect) by about 3 human beings on the planet and is nothing but a headache for every other developer out there. It was removed a few versions ago and when someone opened a bug to say 'hey, what's happened to this feature?' a developer replied saying, in effect, 'I asked around my colleagues and nobody used it, so we took it out, now run along'. I remember it well because it was one in a line of 'changes' that left me generally thinking that Eclipse sucks, no matter how many other cool features it's added since.

Missing tooltips and disabled insert keys aren't the kind of thing you're going to switch distributions or operating systems or IDEs over, but it leaves you feeling done in, and when enough frustrations add up, you do switch. Switching software isn't always a rational, coldly-considered decision, it's often just someone saying 'Sod this, I've had enough', when they'll barely remember all of the things that've gotten them so annoyed in the first place.

The moral of the story, really, and it applies everywhere, is that people are a lot unhappier when you take something away from them than if you'd never given it to them in the first place.

For some reason Google's not indexing my tech blog. It gets the main page and knows about a few older permalinks, but the majority of pages aren't indexed. I'm not sure why - the blog isn't any different to my main blog, which is indexed properly, and I can see Googlebot visiting unindexed pages.

This is a bit frustrating because the whole point of having half my tech blog content is that it's easier to have Google find things for me when I need it than it would be to do a search locally.

Sitemaps seem to be the answer. I haven't touched my blog software in ages, but I think it might be time to get tinkering.

gitvsmercurial.com

All fine and well, but which to choose? I've used CVS for my personal projects for years, and been planning to move to svn for ages. Is it worth going with svn? I figure if I'm going to make the jump, I may as well make it a big jump. The choice between Git and Mercurial seems to boil down to 'fast, finicky and poorly documented' versus 'slow but predictable and well-documented'. I can see benefit to both, which makes me think that a worthwhile approach would be to see what works best with Eclipse (since that's what I do most of my coding in.)

StackOverflow rocks, as always.

Upgrading to Karmic Koala has been a pain. The version of GTK+ used by Koala has exposed a bug in how SWT uses GTK+, apparently. It manifests in dialog buttons sometimes not working - you hover the mouse over a button but clicking doesn't work.

Anyhoo, to fix, you can export GDK_NATIVE_WINDOWS=1 (in a startup script, say), to get the proper behaviour back. Apparently to be fixed in 3.5.2, no idea when that's due.

(via)

Firefox 3.5 has blurry fonts for me, apparently it doesn't obey Gnome font settings. From Fix Firefox fonts in Ubuntu 9.10 Karmic:

sudo rm /etc/fonts/conf.d/10*
sudo dpkg-reconfigure fontconfig

Cameron Purdy blogs infrequently but when he does it's usually good:

Large organizations often exist (both in aggregate and at any observable level of division) primarily to continue to exist and can best be described as a social form of Brownian motion

More: Eight Theses

Try 10 OSes you've never heard of

Trading Shares in Milliseconds

This year's been an interesting mix but also with frustrations.

Workwise, I'll not say much (as usual) except that it's been good. Another year of server-side heavy lifting and low-down nitty gritty stuff, which is exactly what I love doing. I also got to spend a chunk of the year doing C++, which I always enjoy.

On the home front, it's been a bit frustrating. My secret pet project fizzled out due to lack of time (mainly due to my starting studying again) and what I might call 'feasibility' prototypes which suggested that I was wasting my time. I learned a few things along the way, but for the most part, it's just mothballed code now.

I've also fiddled with GUIs this year, playing with everything from wxPython to C++ wxWidgets to Swing. I'm sorry to say I've achieved proficiency with none of them, and wasted an inordinate amount of time vascillating between them. Too damned complicated. I don't want to be a guru desktop app developer, I just want to be able to knock up a functional interface when I need it. I've never considerd my needs exotic; I'm sure it should be easier, and I wish it was, but it just isn't. Part of me wonders whether I'd be as frustrated by things not quite working how I want them to, if I was using .Net. I've fired up Monodevelop a few times and thought 'hmmm'.

Why not .Net then? Mainly, I'll admit, because it's too close to Java. The irony (and main problem) is that despite repeated promises to myself that I won't use Java at home while it's my day job, it's still easiest for me to knock out quick pieces of code in it. With a young family and studies and the like, tinkering time is always at a premium. Since I'm almost always chasing the clock (and oftentimes too far past bedtime), the temptation to revert to Java when I just want to get things done is too great.

So, looking back at my goals for 2009, it's a mixed bag. I did the C++ thing (but could always do more), I'm almost permanently in Linux (except for a brief stint this year playing Eve Online). I'm still doing Java at home, still not yet done the Unixey stuff, and I've sort-of played with Python.

For 2010? For now I'lljust say:

• on the work front, more of the same
• no more Java at home, no matter how convenient it is

That's a little open-ended, and it's that way for a reason. I've got exams in February, and I'm going to give myself the next month to think about what I'd really like to do hobby-wise this year. I want to do something achievable, I want to pick a project. I'm just not sure what, yet.

Compared to a wxGrid, it's altogether too complicated adding rows and selecting cells in a JTable. I still don't fully understand cell selection and focusing and it grates me when what's supposed to take 5 minutes of Googling for the answer to add a feature to a simple GUI app turns into a drawn-out schlep with false starts and dead-ends and head-scratching.

JTable comboboxes are keyboard-friendly, though, and there's more sample code and discussion for Swing functionality than there is for wxWidgets. Breaks my heart to admit it, but it's true.