Green brain-twisted names

Poetry; Green Knights

I've been reading Sir Gawain and the Green Knight, written by an unknown poet in medieval England. J.R.R. Tolkien translated it to a (*somewhat* :-) more modern English.

I actually didn't realize it was a poem when I got it. I've never enjoyed any poetry before (well, apart from stuff like Where The Sidewalk Ends ;), but I have been trying to. Fortunately, I'm enjoying the story of Sir Gawain and The Green Knight (I enjoy knight stories in general), but I must say that the form is *really* distracting. I do get distracted when reading prose, but when I'm reading poetry, I get distracted about once every page. My current thinking is "Gah, this would be so much better in prose", but I will try to get used to it.

Brain I/O

I recently read a very exciting article about Brain Input. http://www.popsci.com/popsci/medicine/article/0,12543,576464-1,00.html. The idea is to stick things in your brain that listen to firing neurons and send the data off to a processor. Stuff like "typing" words with your brain is obviously quite possible, from what the article says. I'm not very excited about the fact that it requires brain surgery: not because I fear its danger, but because I doubt I'll ever be able to afford it. Fortunately there are alternative mechanisms being researched to receive signals from your brain that don't require surgery. But anyway, that's not what I want to talk about.

Let's assume we have a working brain-input device that lets you "think" commands and send them to a computer. What I'm interested in is integrating this with brain *output*. Here's my anecdote: I was reading Sir Gawain and The Green Knight last night in bed. I don't know some of the archaic vocabulary used. If I had a brain I/O device hooked up to some tiny computer that can talk to my wireless network, I could have thought: "Look up 'gules'". After a half-second of delay, my response would come back from dict.org: "poetic; that which is red". I could have either 'heard' it if my brain-output device is hooked up to the audio-sensing bits of my brain, or maybe have it overlaid on my vision, or even have the words just *pop* in to my consciousness. I guess that the former two are more likely. :-)

Of course, something like contact-lens-like video devices and tiny wireless ear-pieces would be tolerable, but they're still probably going to be a pain.

Anyway, this would be so cool! An augmented brain! One's "memory" would become MUCH more robust (think google and dict.org). But be warned!

For you Sci-Fi readers, something very similar is introduced in Larry Niven's/Steve Barnes' "Saturn's Race". Humans would have their brains augmented so that they could store memories in a secondary, reliable storage inside their brain. But one of the characters who had such an augmentation got his destroyed; he became insane afterwards. I guess that some sort of reliance on such an augmentation might have negative effects, but that doesn't turn me off of the idea.

Stuff

Sorry, still no more Blendering. I've been doing some more hacking on Twisted Names and fixing random Twisted bugs, though. Twisted Names is actually becoming a reasonable replacement for Bind, I think (ignoring the fact that it already *has* replaced Bind on a couple of semi-critical servers).

Everyone go watch the MARIO MOVIES, because they are really awesome and profound and emotional.

* Part 1

* Part 2

* Part 3

They require Flash.

MANIC

My brain is slooooow.

I did a lot of work on Twisted Names today, in order for it to usefully be an "automatic" secondary name server for a domain, by doing AXFR queries to the "primary" server. lookupZone (The method for doing AXFR queries) was effectively totally broken. Fortunately, DJB came to the rescue with his excellent explanation of how the AXFR protocol works, which is totally un/misdocumented in the DNS RFC.

I was trying to sleep, but I suddenly realized that extrusion is the answer to my face-modeling woes. I need to add another section to my Learning to Model journal with some of the tricks I've learned, soon. I'll add screenshots of all my horrible attempts at modeling a face, and hopefully a shot of a *good* face modeled with extrusion :-)

*Four*-D

I've updated my 3d learning journal with stuff about animation and realistic (human) body modelling (Get it? Four D? Animation is about changing a model over *time*??).

Oh, and I solved a problem I've been having for *months* with my Nvidia graphics card. NVidia's Linux package unconditionally installs a set of "Thread-Local Storage"-using drivers in /usr/lib/tls. There *should* be a condition for installing these, however: on a Linux 2.4 kernel, they cause problems in many GL-using programs. On Linux 2.6, they're absolutely necessary. The Debian installer for the drivers does the right thing, but I hadn't been using that because it's usually out of date. I installed it after someone suggested it might fix my problem, and voila - the installer clarifies the situation (and asks you whether you _really_ want the TLS drivers or not).

So, now I have a working Soya3d, which means I can be productive and write fancy graphical software. However, unfortunately, I also have a working Rune, which means I can now be distracted for hours playing on Low Caliber Club's server...

Learning to Model part 1

This is the first part in my series about my experiences learning to model with the excellent program Blender.

I read castle tutorials at http://www.blender3d.org/Education/

Had lots of problems with spurious faces: learned about UV Face Select mode on my own, figured out how to select faces, then go into edit mode, then {x} -> delete face (UV face select mode is not strictly necessary: just select all vertices that are part of face)

Also had problems with *non*-existent faces: that's easy: select vertices then {shift+f} (fill). Actually I think this might have been explained in the tutorial, but my problem popped up somewhere where the tut-writer wasn't expecting it to happen...

I read the User Interface, The Blender Windows, and Navigating in 3d Space tutorials at the same location, and I should have read them before the castle tutorials (or maybe in parallel with them). :-)

I decided to start a model of a Man for a game I want to write.

Mostly based on knowledge of castle tutorials, but I did learn a couple things on my own / with IRC help.

Faces can only exist for up to 4 vertices. This explains why some things that I expected to work don't. For example, I had a cube like this:

  (Front View)
     ____
    |    |
    |    |
    |____|

And I wanted to break the top face in two:

  (Front View)

    |\  /|
    | \/ |
    |____|

Now remember, this was a cube: I wanted to take the top *face* and split it, from a single rectangle, into two adjacent rectangles.

  (Top View)

     ____          ____
    |    |        |    |
    |    |  to:   |----|
    |____|        |____|

I selected the four vertices representing the top face. After playing around I *thought* edit->faces->"convert to quads" would work, but it didn't; "convert to triangles" did something that was *almost* right, but it converted it into two adjacent triangles, making the "split" run diagonally between vertices.

  (Top View)
   ___
  |\  |
  | \ |
  |__\|

Eventually I found edit->edges->Loop subdivide. It did something that *looked* correct, but was confusing: it split my other faces too! Certainly I didn't want that. Long story short, I eventually found out (via a kind soul on #blender on irc.freenode.net) that faces can only exist for *four* vertices at a time. Splitting only the top face would require the *side* faces have more than four vertices (or, perhaps more intuitively, more than four edges).

  (Front View)
   Impossible!
   5 vertices      Correct
    |\  /|          |\ /|
    | \/ |    to:   | | |
    |____|          |_|_|

I got bored with my crappy Man and decided to learn a little about animation: I read the animation tutorials at

* http://10secondclub.org/users/juicy/mid_tute105.html
* http://10secondclub.org/users/juicy/mid_tute106.html
* http://10secondclub.org/users/juicy/mid_tute107.html

No problems here, but they're pretty simple. Oh -- one UI-change nit. I couldn't find where the "SCREEN button at the top of the window" was. However, I figured out how to get the same effect by 1) splitting the window vertically (rightclick the divider above the 3d window, select "split"), then setting the window on the right to the IPO curve editor view. BTW, zooming in this view is *trippy*. :-)

DeeDeeDee

I decided to keep a journal of my experience learning [1]Blender and

[2]Soya3d. So far I've learned these parts of Blender: Basic

modelling, basic texturing, trivial animation (no moving parts of an

object, just moving/rotating/scaling entire objects). See more!

[3]HERE.

I'll make changes to that page and post blog updates when I do.

References

1. http://www.blender3d.org/

2. http://oomadness.nekeme.net/en/soya/

3. http://radix.twistedmatrix.com/Texts/How%20I%20Learned%203d%20modelling%20and%20programming/

Blogpost

SimCapitalism Now!

On [1]TerraNova (my favorite group blog, which is about virtual

worlds), there's a [2]post about a [3]marketing agency that bought a

good chunk of "land" in a virtual world called [4]Second Life

(actually, it was one of their employees doing it semi-independantly;

however, the company's name *will* be involved). Second Life is really

the most amazing VW that exists, as far as I'm concerned and as far as

I can tell. But that's not relevant at the moment. The company that

made SL "sold" some of their land (with a monthly upkeep) to the guy,

for $1200 up front and $200 / month. Yes, USD, not virtual currency.

I offhandedly commented (in an IRC channel about virtual worlds that I

hang out it) that this was interesting. [5]Dash didn't agree (that it

was interesting :). He and I debated the issue for about half an hour,

and while I didn't find his typically negativist position useful as

such, the conversation that resulted got me thinking. Basically I was

arguing that it was an interesting precedent; no other

non-game-related commercial entities have shown interest in doing

anything in a virtual world. Now, dash and I want to build a massively

distributed virtual world: something rivaling the web itself. I

explained that this precedent would be useful to show to other

companies in order to get them to set up virtual worlds connected to

the DVW (= Distributed Virtual World) network.

But I wondered: Why do I *want* a general-purpose, distributed virtual

world? I *do* want to develop games, of course: but why do I want it

to be general-purpose? The only reason I can think of is "the web

sucks, we need something better." But then, I'm not *really* that

interested in smashing the web that much. It's something that really

ought to happen eventually, but I'm not very passionate about it any

more. I guess I'll have to give up on my grandiose world domination

plots and just make good games :-(

Bloggin'

Probably gonna switch to [6]MT. Working on "integrating" into [7]Plonk

now.

Gamin'

I'm kinda learning [8]Blender. I want to make a game like Secret of

Mana! One that's massively multiplayer and stuff! Yeah right.

Whatever. I'll probably rip/modify [9]Egoboo's models, because they're

cute and simple. Of course the game will be written in [10]Python,

probably using the [11]Soya3d graphics engine (the link isn't working

as of this post, hopefully it'll be up soon). I have to figure out why

I'm getting a "ImportError: libGL.so.1: cannot handle TLS data" when I

try to import soya, though. :-(

References

1. http://terranova.blogs.com/

2. http://terranova.blogs.com/terra_nova/2004/02/2ls_virtual_lan.html

3. http://www.riversrunred.com/

4. http://secondlife.com/

5. http://ghostwheel.ddts.net/

6. http://movabletype.org/

7. http://radix.twistedmatrix.com/Code/Plonk

8. http://www.blender.org/

9. http://egoboo.sourceforge.net/

10. http://python.org/

11. http://oomadness.tuxfamily.org/en/soya/