Cogs can think.

Wow, a blog post!

Google, please fix this!

I find it rather pathetic that the only entry on Planet GNOME so far regarding KDE 4.2 is some guy complaining about the release announcement using Flash for the screencasts. From a project that embraces such a controversial technology like Mono.

Jealous much?

So, like, GNOME was on its way to irrelevance, simply because they had no plan for the future, no direction, no leadership, no vision, no nothing.

Then came GUADEC, and at GUADEC, they came up with a vision for GNOME 3.0. And it’s all about tabs.

This basically means no more GNOME for me. Tabs suck balls in just about every implementation except settings panels. Tabs are bad because they constrain you. Tabbing is all the shizzle in web browsers, but all it does is stop you from having differently sized windows, having websites side by side - and to make it all even worse, tabbed programs introduce a new place to manage windows: the application window itself. So, users have to think about where to switch to a certain window - do we switch using the panel, or via an application? Wait, we have to switch to the application via the panel first, and then switch to the particular window we want inside the application? And what about closing documents versus closing windows? What about having 15 scientific .pdf’s loaded in a tabbed Evince? Can I still read the tabs, or are they shortened to only the first few letters of the filename?

What do you mean, pointless clicks?

We have been trying for ages now to move away from an application-centric world, towards a document-centric world, and Mac OS X is doing really, really well in that regard (Quick Look!), and GNOME itself was not doing bad either. By focussing efforts on tabbed applications, all that work has been in vain. They are setting the clock back, I don’t know, 15 years?

I’m happy that GNOME has a vision, but sadly, it’s one step forward, three billion steps back. I mean, vertical damn tabs? Why don’t you start eating babies while you’re at it?

There’s a geek soap opera going on over at Planet GNOME, with the center of attention being Jeff Waugh and his (non?) work for GNOME. Since I really don’t give a rat’s ass about the people behind the software I use (really, I don’t. I care about the product or art, not the programmer or artist), and I only had one run-in with Waugh, I really cannot comment on this stuff. That run-in was perfectly resolved, and he didn’t come across me in a way that resembles whatever Cumming is claiming.

In the end, this is all useless political blabber. GNOME is falling apart. As I have said many times before on this blog as well as on OSNews, GNOME is dead. Dead, dead, dead. If you, as a major software project in a fast-changing world, do not have a well thought-out plan for the future, then you are dead. Cold, harsh, but reality. People have flamed Eugenia and I for saying it, but if a strong leader does not step up soon, a leader that dares to make the hard decisions needed to set a viable plan for the future, than GNOME is doomed to insignificance. It won’t be able to adapt. And we all know what happens to things that do not adapt.

KDE had the guts to make the tough decisions, and despite delays and setbacks, they are slowly but surely seeing the fruit of their labour. I may not like everything coming out of KDE 4, but at least they are trying. And I commend them for it.

I do want KDE 4.0 to be released after December 31st, 2007, though. I want to win my bet with Aaron.

It’s finally here, Dooce’s redesign. It looks distinctive, and luckily, she didn’t go all pattern and colour happy. I only see few colours, and barely a pattern. Great stuff. The buttons are a bit… Nineties, but hey, as long as she keeps on writing the way she does I don’t really give a rat’s ass about her blog’s looks. Congratulations to Heather and her coding slave husband Jon.

As I said before, for all I care, she just dumps a text file on the net.

My own redesign is on an indefinite hold. I am way too busy with real life and Grow, and I simply don’t have the time left to dive into Cogs Can Think. v4. It will happen one of these months, but for now, you’ll have to do with v3. I mean, it still looks pretty goddamn nice if you ask me.

Arrogance is a virtue.

About seven months ago, Eugenia wrote on her blog:

Don’t you think that this looks sweet? The statusbar/toolbar font is -2 points smaller than the default font size (minimum size is 8pt). We filed a bug report on GTK+ over a year ago about this but no one seems to care, even if it makes the windows look so much better (applications like Baobab that now comes in Gnome 2.18+ by default would greatly benefit from it because it has a lot of toolbar text). So far in my Gnome desktop I had to disable the toolbar text completely, but with these changes I would leave it on. BeOS and Mac OS X’s toolbar font is also smaller than the rest of the fonts and it’s details like these that make these UIs look “cleaner”. The devil is in the details.

Eugenia and I regularly disagree, but on this one, I agree wholeheartedly with her. Let me explain.

A window is a user interface element comprised of several different areas. From top to bottom, a standard window is made up out of the window title, menubar, toolbar, actual content area, and a statusbar. See the below schematic representation.

Furthermore, a window may contain loads of other areas, such as an address bar, or additional menubars like the bookmarks toolbar in many web browsers. In addition, each window has widgets, such as scrollbars window manipulation widgets (close, minimise, maximise, etc.). Lastly, the content area itself can be divided up into different areas, but you can forget that for the moment.

All these elements of a window need to be differentiated. You see, users need to be able to instantly recognise where each of the standard window areas are, so that he can quickly familiarise himself with said window. You can achieve differentiation in a lot of different ways - by using colours, separating horizontal lines, font differentiation (both typeface as well as font style), those sorts of things.

The challenge, of course, is to strike a perfect balance between easy differentiation on one side, and a clean appearance on the other. If you use all of the differentiation possibilities I just mentioned, you’ll end up with a very messy and cluttered window - achieving exactly the opposite of what you are aiming for. However, if you disregard all of these features, you will end up with, yes, a very clean window - but also a window that is very hard to navigate because it is very difficult to see where one area ends, and the other starts.

Consequently, I’ve been following the KDE4 maturation process with great interest. I have been very eager to see how the KDE guys would balance the scale between easy differentiation, and clean looks - especially taking into account KDE’s history of, well, dumping widgets all over the place. And sadly enough, only a few months before the final release of KDE 4.0, this is what KDE4 looks like.

There is no typeface differentiation. No font style differentiation. No colour differentiation (except for the content area). No colour differentiation. I could live with all that, were it not for the fact that it also lacks… Separating lines. Titlebar, menubar, toolbar - they are on big blob of white. Sure, themes can be changed and all that, but as has been repeated often on the ‘net, defaults matter. And if this is the default, it’s simply a fcuking mess. They put “clean” atop their list of priorities, but ended up with something so clean, it’s close to unusable.

To prove my point, I added a few separating lines between the window areas, and see how much it cleaned up already, by using just a few 1pix lines! Clickety-click for full-size.

This is Amarok 1.4.x, the current tree.

On OSNews, for the past few days, some discussions have centered around Amarok being ‘the best music player hands down’. I beg to differ, and based on this main window’s screenshot alone, I can point to various major flaws in this application that will prevent me from ever using it. Please note that these are just my personal concerns (that’s why they call it a personal weblog, boys and girls), and they do not reflect the opinions of my employer.

In the top left corner of the window, you see the Music/Lyrics/Artist tabs. This indicates that this row (yes, even modern graphical user interfaces can be divided up into textual rows) is a tab bar. Great, but, then, why are there file/navigation buttons on the same row, only a few pixels east?

The reason for this is clear: the Amarok developers are trying to cram so much information into the main window, they were forced to split the window up in two sections: a contextual section (left) and the actual section that matters to this kind of application, the playlist/buttons (right). You could argue that the play/pause/stop/etc. buttons in the right section ought to be on top (seeing they are the most vital buttons for a music player) but alas, I’ll let that one pass.

Let’s focus on the left section. The Amarok developers were so hell bent on cramming as much information as possible into this limited space, that they were not only forced to add a vertical scrollbar (and sometimes, a horizontal one too), but also not one, not two, but three (!) tabs.

The above leads to this ridiculous situation where you have two completely different types of sections crammed into one window, where rows switch their function (tab to button), simply because they wanted to cram way too much (pointless, in my book) information into a single window. The end result is that the actual part that matters (playlist, play/pause/stop/etc. buttons) is now demoted to that side of a window that receives the least focus (the right side). On top of that, as said (can’t let it pass by, I’m sorry), the most important buttons (play/pause/stop/etc.) are now rendered somewhere at the bottom right, far away from the focal area of a window (which is the top-left).

The ever-growing hunger for more functionality and information forced the Amarok developers to take even more drastic measures. The left section of the main window needed to function not only as a three-tabbed contextual tab (read that aloud five times if the ridiculousness doesn’t sink in immediately), but also as a devices tab. And a Magnatune tab. And a collection tab. And a files tab. And a playlists tab.

And in order to cram all that information and functionality into one single window, they did what makes Amarok, to this very day, the most ridiculous application ever written, UI wise: vertically text labeled tabs, with normal horizontally oriented icons. This is wrong on so many different levels, it’s just not funny any more. Whoever thought of that brilliant idea ought to never be allowed to “design” a graphical user interface, ever again.

Amarok is not the only audio player that suffers from functionality and information creep. Windows Media Player, iTunes, they all suck major balls because they all try to present their users with so much goddamn pointless information it almost makes my head spin. Every part of Amarok except for the parts that matter are just screaming “Look at me! Over here! I got boobs!”.

That’s why I refuse to take anyone seriously who says “Amarok is just about the cleanest as it gets UI wise”.

My dream operating system would have BeOS’ kernel, responsiveness, and soul. Mac OS X’ attention to detail and polishedness. Windows’ industry support. Linux’ price tag. VMS’ stability. OpenBSD’s security. My nightmare operating system would have the Linux kernel. Windows’ attention to detail and polishedness. SkyOS’ industry support. Vista’s pricetag. Windows 98’s stability. BeOS’ security.

Just saying.

I switched my Ubuntu installation over to Kubuntu. I see more future in KDE4 than I see in GNOME (as I have said many times before). If I don’t see someone from the GNOME community stand up to bring GNOME to the next level, I don’t see why I should invest time in using it and reporting bugs on it. As a user, I expect my platform of choice to have (what I consider) a viable future, and at this point, I just don’t see a viable future in GNOME (and no, some bug fixes and minor changes to the 2.x tree do not count as a viable future).

Sure, KDE3 lacks polish here and there, but at least KDE has set out a path for the future. And it has Kopete and Konqueror, which kick GAIM Pidgin and Firefox’ asses any day.

Warning: rant ahead.

I don’t give a rat’s ass what all the anti-MS idiots say. Surface is a truly innovative product, and I applaud Microsoft for it. They had the guts, back in 2001, to devote money and manpower to this idea, and now, 6 years later, it is paying off. Surface looks great, opens up a whole slew of possibilities, and, as far as I can see it, is truly a Pandora’s box of opportunities. I’ve seen use cases flash before my eyes like every other minute of the past two days.

Surface is just one of those things the OSS community will never come up with. Let’s face it; Linux, BSD, most of the other OSS projects, they are all followers. They are in it to regurgitate what companies like Apple and Microsoft serve them. The OSS community is supposed to be so great, right? Then why is it that they never seem to be able to come with something truly new, something groundbreaking, something that will make people all around the world go: “wow…!” I showed Surface to some of my computer-illetrate friends, and they were baffled. They all saw new use cases in front of them… Just like I did.

Fact remains, as much as I like the OSS world (I try to use the OSS equiv. when it is at least as good as or better than the closed-source counterparts), they are followers, not trendsetters. So, sure, there is enough to bash Microsoft and Apple about, but at least they have the guts to come up with truly new ideas.

In 2010, we’ll see a marginally different, poor rip-off of Surface. The joy of waiting.