log of Logfish.net

The OpenMoko mainstream development branch has decided to go with Enlightenment as their main driver behind their GUI toolkit. More precisely, they are developing a new collection of libraries specially for mobile devices. Om2009 is going to be the first release which used Paroli with an FSO back-end.

Choosing FSO is a great thing to do and because almost all the distributions are now using it (or planning to) the bug reports and development effort has finally found a central point to converge.

Trying Om2009, I can not say the same for the Enlightenment developments. As I was fearing, the interface has been created in what I would call a classical Enlightenment way: looks are more important then usability. This classical approach that made me stop using Enlightenment has now hit the OpenMoko phone in much the same way: everything is white on black (ask yourself why school boards are not), the difference between a label and a button is not there, when you click things happen immediately even if it involves calling people, there is no lining around anything, things are hidden behind slide-able other things. Things like sliding are cool, but they should only be used when it is actually needed. Something that needs more confirmation then a button, like deleting files, but not for hiding other functionality.

I think all these flaws are overly obvious and in a year or two they may even be solved, if only the people who are behind the interface think this stuff matters. This got me thinking and I decided to post a community question: Why Enlightenment? Of course, I did not criticise the result, it probably took allot of work, but I can't understand the logic behind going with something that I have always found less-usable.

The responses I write here have not been posted on the community list, because they are all inflammatory. They are an elaboration on my opinion, and a reply to all the thing I thought where just down right stupid.

Here are some of the responses I got:
Why not?
The best thing you can do when making a decision is back up the resulting choice with a confident: we had nothing better to do, so we decided to just do it. Not really a good one I think, but evidently they had to much time left over. Two people even found it necessary to quote somebody else saying it. I think they thought that would make it "true" for some reason? Appart from that, they also abuse the quote as the quote was

You see things; and you say, 'Why?'
But I dream things that never were; and I say, "Why not?"
George Bernard Shaw,

Which is talking about dreaming about something that has not been created yet, but Om2009 is here (albeit in beta form). So all in all, the people saying why not where unable to contribute anything reasonable to the discussion.

Getting new improvements quickly (instead of waiting years for it or it never happening) is bad?
This was a reply to: why start a new toolkit instead of extending an existing toolkit. A fallacy of many questions, although telling the author this, he replied with: not really. Not being sure about which parts of any of his where real and which where just lies and fallacies, I'm left guessing at every retort he typed.

Scrolling with a touchscreen can not be done without a new widget toolkit
This is one of the weirdest replies I got, but the author insists on it being true and that I should trust him on this. Still, anybody with a scroll edge on their touchpad knows that it can work. So I still don't understand it, but as the author started writing in capitals, continuing on with that thread would end in certain flaming.

Cairo uses floating point operations for it's rendering, the OpenMoko has no FP unit
This is a good point, but I can't help but think that Cairo should have an AVOID_TP_OPERATIONS compiler flag which helps weed out as much of these operations as possible. Matter of opinion I guess.

Up-to-date. It's under constant development, and getting better by the day.
I know of no open source development project that does not have the same. And if there is the time to change stuff but the main development team doesn't apply the patches or stalled the changes, forking something will help. We have seen it with Compiz and other projects. The only project I know of that does not develop is tex, but that two has newcomers fixing that.

Finger scrolling - it works by default. If I know that app is written in Elementary - I know that it's finger-friendly.
A really good point in itself. But if you consider that any Elementary app is written to work on the phone, then it doesn't seem that different from a Qt app written for a phone. You could just say "If I know that app is written in QT for the phone - I know that it's finger-friendly", so I don't see it as a convincing point, but still.

It can look much cooler then anything else
Replying that I would rather have usability then looks would be besides the point of the thread: if that is why they chose Enlightenment, then that is the answer to the question. Along with this remark came a small movie of still unfinisched work by one of the developers:
http://www.rasterman.com/files/wp2.avi

This movie is not to show the development on the OpenMoko, but just to show how cool Enlightenment is. However, looking at this movie, I immediately see my usability fears confirmed. 1: A moving desktop background? Cool but have you ever heard of the motion after effect? 2: Why are the images popping up on the edge of the viewport? The whole point of a viewport is that is it like a virtual space which I can only see a part of. Who cares that the system takes does lazy loading in the background, if I can see the effect of it then that should be considered a bug, not a feature. 3: If you want to preview a background, why not put it on the desktop immediately? The whole point of the preview is to see if it goes well with any panel or window border I have previously chosen.

All of these are so obvious, that they must have been chosen to be counter intuitive, instead of being overlooked.

In the end the best points where made by Leonti Bielski, who I would like to mention as he/she wrote the best reply to the thread starting question. Finally also, hail to the FSO developers as with their hard work I'll probably have a stable and usable phone within the upcoming year which also allows most distributions to stabilise. I'll see what happens, I'm probably going to run a Debian derived distribution in the end simply because I like Debian. Hopefully Ubuntu is going to but in here and set up another great Debian specialization.

As this is a rant, I can't help but end with: fuck anybody who disagrees and if you can't see the fallacies you put out then you are to proud or stupid to admit to them!

;)