Making the Processing IDE play nicely with xmonad

If you’ve had problems when running the Processing IDE, or indeed any other Java Swing applications, under the xmonad window manager on Linux, here’s the cure – just a simple edit to the ‘processing’ executable (which is really just a shell script). Open it in you favourite text editor, then select the following lines:

# JARs required from JDK (anywhere in/below the JDK home directory)
JDKLIBS="rt.jar tools.jar"

And paste the following in their place:

#Make us work with xmonad
export AWT_TOOLKIT=MToolkit

# JARs required from JDK (anywhere in/below the JDK home directory)
JDKLIBS="rt.jar tools.jar"

You should also be able to add the export line to your .xinitrc or your .bash_profile, the former being more sensible if you never use X applications over a remote login. For some reason, however, Processing seems to require the variable to be set every time it is run, hence we export it from the startup script.

Setting up the Processing environment on Slackware64 13.0

[Fairly useless post - removed until I figure out some better solutions...]

More new toys…

I’ve always wanted an L1N64. With GPGPU computing finally coming of age with the ubiquitization of OpenCL, now would be a brilliant time to happen upon a deal on such a board, with support for upto 4 graphics cards, wouldn’t it?

Well, I guess I’m very fortunate! For a little over £30 (inc. P&P), a lovely eBayer has handed me an ‘untested’ L1N64 SLI-WS complete with 2 Athlon 64 FX-70s and a smattering of accessories. Worst case scenario, I thought, would be if everything was broken, in which case I could at the very least use one of the CPU backplates on my Arima board.

Seeing as the L1N64 was mainly an overclocking platform, I was a little afraid I’d find overvolted-to-death CPUs and numerous burned MOSFETs – basically, a complete lack of magic smoke. Imagine my glee when I found no visible defects on the board, apart from plenty of dust. I apologize in advance for the terrible-quality photos:

As you can (perhaps not) see, the board is revision 1.01G, with a soldered BIOS chip, which may need rectifying with my planned move to quad-cores… Apart from that, all seems well, until we remove the CPUs…

Ah. Bent pins. Bent pins galore. And a cat hair. Believe it or not, this actually made me happy. Not because I love cats, but because none of the pins are actually broken, and with a bit of scalpel-aided provocation, they should fall neatly back into line. This also suggests that the system in which the board and CPUs were used probably did not suffer instant, inexplicable stop-working-ness, but rather simply wasn’t put together well in the first place.

In conclusion, then, I have in my hands a board I’ve always wanted, but never thought I could actually afford, and it might even work! On the other hand, I don’t have the time to repair it. Nor do I have a spare case, or power supply, except back at home in another city. I guess this project will have to wait until the summer, then. In the meantime, I’ve removed one of the CPU backplates and mounted it onto the Arima board – that’s a project which shouldn’t take too long to complete at all, and hopefully is going to end up being pretty awesome in its own right.

The eagle-eyed amongst you will notice that this means I now have 3 Socket F systems in various stages of completion. That’s a potential 20 cores and 7 graphics cards of number-crunching power. Since joining WCG, I haven’t really been devoting any machines to the effort. I hope that will change soon as soon as these systems are up and running.

Lost in Translation

This is a difficult movie to review. It begins with nothing, ends with nothing, and leaves you unsure of quite what happened inbetween. It is a film we can all relate to, filled with moments we can all relate to; but it leaves the viewer feeling somewhat alien and detached. This is storytelling at its truest – one journey, two people, and the things that might have been. Lost in Translation is a love story, but there are no lovers – only love itself: a formless, loosely-defined yet rigidly enforced emotional construct which fails to create because it can’t bear to destroy. The kind of love that builds respect; the kind of respect that builds walls; the kind of walls that halt love.

The film follows the lives of two very different people, temporarily stranded in Tokyo for reasons of compromise and searching for an escape. The acting is wonderfully subtle, not ‘real’ in the traditional sense – gritty, raw and vigorous – rather, calm, restrained and angsty. You can sense the naivety of the characters, you can feel their weaknesses – their endearing, attractive failings. Yet they are not weak, nor are they flawed – they are merely human. And as they stumble their way through Tokyo, we see them struggling to let go of that humanity, frightened of chasing the contentment awaiting in the distance, for fear of losing it along the way. They would rather not find out that they love each other, though they already know it. And they cannot hide – they have seen each other naked, as it were. There is nothing left to reveal, so they run, struggling to break away lest they fall into happiness. Lost together in a rainforest of concrete, they seek respite: in monasteries and saunas; alcohol and karaoke, but to no avail. Each distance brings them nearer. Every mistake perfects their feelings for one another. The harshest, most clumsy words turn into expressions of longing and unfaithfulness veils desire.

It is a destructive superposition. There are slight highs, slight lows – moments of multiplication and resonance, and moments of subtraction and dischord. Lost in Translation has the wonderful effect of merging all life’s actions into one long, meaningless wave of emotion – a continuous ebb and flow which is bound to follow its own path, no matter our intervention – like a juggernaut jackknifed across the freeway. This is shoegazing translated into film – filled with the deepest of emotion, but never providing opportunity for threads to unravel, for one feeling to become distinct from another. If this is a comedy, then its message is far from humorous; if it is a black comedy, then it denies its own normality. Lost in Translation cannot be pinned down because it refuses to define itself, for that would be too inhuman. Instead, it remains stubbornly as it is, and as it always was – in a word, beautiful.

New toys…

new toys + leftovers = fun

More soon… Need a 3.5″ pitch backplate before I can use the stock cooler you see in the picture.

Five Minutes of Love

I set up a shell script. chmod +x it. Type the command, and get prepared to be patient. But there’s no need – Love is ready for me. 21fps. I click into ‘low-end mode’. It’s a shame I can’t change the resolution. But Love is beautiful at 2048×1536, and without the awkward swaying of the trees, and with the roughness, as opposed to wobblyness, of the lines, Love is more beautiful in low-end mode. I enter my details. “ahmad” – I take care typing my name. There may be no going back.

The game begins. It’s quiet, disconcertingly so. “Hellooo?”, I type. “Anybody here??” Nothing. My Name is ‘Dredel.’ Why is my name Dredel? I think back – no, even I’m not that careless. I begin to move. The landscape shimmers under the setting sun. Glowing power lines zap across the sky in the distance. I follow them towards the mountain: up some stairs; over a river; though a forest.

I reach a clearing in the trees. A lone obelisk stands, backlit by the sun – vibrant, just like the bullets. Oh no, bullets! I sway this way and that, the mouse lag limiting my agility. A character appears – running, shooting, backflipping, reeling, ecstatic. “A player”, the game explains to me. The font could be better. “If that was a player,” I think to myself, “Why was she shooting me?.” I’ve already assumed she’s female. This is the internet, after all. I run after her. “Finally! someone is online”, exclaims Sindra. The name sounds feminine – maybe I was right. I begin typing back, before realizing I’m mute. My hand clumsily scrambles back to the mouse, dashing for the middle button. ‘Click’. “My first time here,” I reply. “Me too, kinda”, she says, before explaining how she didn’t have an activation code yesterday. But three euros is a tempting price. We’re back at the obelisk. “Why is my name Dredel?”, I enquire, as if she will know. “It’s Random…” – (s)he does, and I am wrong on both counts. “…did you build the obelisk?” I glance downwards for a second, fumbling for the mouse. I begin typing. But there are no more shimmering trees. There’s no more piercing sunlight. The game has crashed.

Sindra: If you’re reading this – 9pm, by the obelisk. I’ll shoot first this time.

Raytracing is dead. Long live raytracing.

Delayed post – this should have been submitted months ago

Reading about CausticGraphics’ CausticOne card today got me wondering.  While I can’t help but be impressed by the technology (LOOK – SODIMMs!! aren’t they cuuute!), I’m pretty sure this won’t last for long – And it’s not because of Larrabee, nor even CUDA.  It’s because the ‘proprietary-algorithms-on-a-proprietary-chip-that’s-on-a-proprietary-card’ model is long dead.  Remember 3DFX?  I do.  In fact, I’m surprised that anyone with a background at nVidia ended up in CausticGraphics – after all, it was nVidia who killed off Glide, and with it, Voodoo.

You see, the only reason  ART VPS has survived this long is that they have established industry presence and have proven the quality of their algorithms in comparison to other, proprietary, software-based RenderMan-compatible solutions.  In fact, even they seem to have given up on putting a raytracer on a card (PURE) and are now seem to be concentrating on a more software-centric approach.

I’m not saying we need to get caught up in the CPU-GPU bubble into which Intel seem to be naively trapped.  The problem is that some people’s optimistic faith is severely misplaced – an OpenGL-based API in no way implies any form of open standards-compliance.  The world of raytracing is moving to more and more universal platforms – ART only did well because they were RenderMan-compliant

Still, I’d love to pick up a CausticOne on ebay in a few years time…

An open letter to Always Innovating, or why I won’t be buying the Touchbook

Dear Sir/Madam,

It is with great interest that I have followed, over the preceding months, the development of the Always Innovating Touchbook. It is for me, without doubt, the most exciting thing to happen in the world of portable computing since Apple’s Newton. Why, then, you might be wondering, will I not be buying one?

I refer you to a phrase with which one is greeted with after following the ‘open source‘ link on your website:

Always Innovating believes that open hardware will stand out in the coming years as open software has done recently…

Seems good so far. However, the passage continues,

…Based on the Beagleboard project, we are providing the design of the Touch Book to parties that show strong commitment to the open source movement. Please follow this link for more information.

This is where, for me, alarm bells begin to ring. What, exactly, is a party that shows strong commitment to the open source movement? Are you in some way implying that, even though your design is heavily based on BeagleBoard, an open-source hardware project, that you are going to be selective about contributing your changes back to the open-source community?

Following the link mentioned above, we are taken to a page which states,

The Always Innovating strongly believes and supports both the open source movement and the recent open hardware efforts. At the same time, we have also been a little bit disappointed by the behavior of some companies who fail to discern open source and one-way self service!

As we will see upon further delving into the text on this page, the above statement is nothing but hypocrisy:

The Touch Book brings a lot of innovation and we feel unconfortable to just throw away all our efforts and let some companies just copy-paste our design for free.

How, in that case, can you claim to support the Open-Source Hardware movement? Anyone can ‘copy-paste’ an Ubuntu CD, and profit off of it, ‘for free’. That is the meaning of freedom in the context of Open Source. Does any one of the thousands of contributors to the project, feel ‘uncomfortable’ about this? Of course not. Instead, they realise that anyone modifying, fixing and re-distibuting Ubuntu can only be a good thing for Ubuntu, at the very least because it promotes Ubuntu and Free/Open Source Software, but also because the mechanics of an Open-Source ecosystem dictate that any positive contribution will either be re-adopted upstream, or form part of a fork, improving quality, diversity and choice. It is in this context that the Touchbook (at least its electrical design) can be considered a fork of BeagleBoard, and in this context that I believe the actions of Always Innovating are in direct violation of the principles of Open Source.

At the same time, this project wouldn’t have been possible without the contribution of thousands of free developers or companies who honestly participate to this echosystem and give back to the community.

I’ll assume you meant ‘ecosystem’. If you really do care for these free developers, just how many of them are likely to get access to you designs under your current programme? Is someone who provided a vital bug report to one of the hundreds of open-source software products used in the Touchbook’s OS, deemed worthy of gaining access to your designs? Is an electrical engineer who has no provable contribution to open-source software or hardware, but a moral and ethical feeling of responsibility to honour the terms of an open-source license, likely to be able to view circuit schematics and hence improve the touchbook? Is the 19-year-old product design student, who knows not a thing about programming, but has a healthy interest in the mechanical design of computers, and a strong will to contribute design ideas to open-source hardware products, ever likely to see a CAD drawing of the Touchbook? Moreover, is anyone who does acquire designs from you allowed to redistribute them?

You have seen from the examples cited above that, with an Open-Source hardware project, it is extremely difficult for one to establish precedent for support of Open-Source, because most of the Open-Source movement until this point has been focussed on software. The simple fact is that a large number of Open-Source supporters are not software developers.

Lack of precedent in Open Hardware, of course, also affects companies like you, and I do understand that. However, instead of playing God and choosing who is and is not worthy of seeing your designs, would it not have been wiser to invest in the development of a true, Open-Source compliant hardware license, which would protect your rights aswell of those of the rest of the open-source community? Even without hardware equivalents of the GPL and BSD licenses, how can commercially-supported projects like Arduino, OpenMoko and, of course, BeagleBoard survive without having to resort to a policy like your own?

Supporters of Free/Open Source Software have made sure that no-one should be able to redistribute compiled code, wihout making the source code to their modifications easily-available. In the same way, should supporters of Free/Open Source Hardware not make sure that no-one should be allowed to distribute physical products based on open designs, without also sharing the relevant ‘source code’ (eg. hardware schematics, CAD files, part lists, design documents)? In other words, should all of these documents not be freely accessible to ayone who purchases a Touchbook?

To claim otherwise, as you do, fundementally opposes the Open Source movement. I understand that the stance you are taking may not be in violation of BeagleBoard’s licensing terms. However, it would seem to some that, in placing arbitary contraints on access to your designs, your so-called ‘support’ of open-source hardware is merely a masquerade; a marketing gimmick, employed only for the gain of your own profits. Open-source is not a one-way street, and Always Innovating cannot claim to be a supporter of Open-Source Hardware while keeping its revisions to the BeagleBoad framework to itself, and only releasing even its own designs to those whom it deems worthy.

It is therefore with disappointment that I am ubable to buy a Touchbook, until ether truly open and free access to the designs is established, or Always Innovating ceases to claim that the Touchbook is open hardware.

Yours faithfully,

A concerned supporter of the Open-Source Hardware movement

Using the Chromium Linux beta build on Slackware 12.2

Having previously tried (and failed) to build Chromium on my Slackware system, the news that a beta build is now available for Linux was a welcome surprise.  However, it is a little more tricky to get running than you’d expect, so here’s how to do it:

1. Download a package containing the required NSS libraries from here. This package was built using Road_map’s instructions on this forum thread and a slackbuild script by (who else?) Eric. Install it using the standard installpkg command; It shouldn’t be replacing anything…
2. Use the following commands (as root or with sudo) to set up Google’s weird, weird symlinks. Does Ubuntu use .so.*d files or something?!
   

   ln -s /usr/lib/libplds4.so /usr/lib/libplds4.so.1d
   ln -s /usr/lib/libnss3.so /usr/lib/libnss3.so.1d
   ln -s /usr/lib/libsmime3.so /usr/lib/libsmime3.so.1d
   ln -s /usr/lib/libssl3.so /usr/lib/libssl3.so.1d
   ln -s /usr/lib/libnssutil3.so /usr/lib/libnssutil3.so.1d
   ln -s /usr/lib/libnspr4.so /usr/lib/libnspr4.so.0d
   ln -s /usr/lib/libplc4.so /usr/lib/libplc4.so.0d
   ln -s /usr/lib/libplds4.so /usr/lib/libplds4.so.0d

3. Download the latest Chrome Linux beta tarball (well, zip file..) from here.
4. Extract the ‘chrome’ executable to somewhere in your path.
5. Pester me if it doesn’t work, or if I did something wrong.  I usually do..

For those interested, having followed the steps above, it should also be possible to build Chromium under Slackware. The poor documentation regarding ‘hammer’ put me off, even though I did find it in the end. Having gotten Chromium working, I decided not to bother with using about 15 Chrome-specific utilities just to set up the build system…

Eskil

Ever since I played with Loq Airou, I’ve been a fan of Eskil Steenberg.  Now this may just happen to be the most astute rambling on technology I’ll read in months…