Well, green transparent acrylic with a couple of yellow LEDs inside would do for making the edges "glow". If the edges were chamfered towards the front, and sanded to a dull finish, they'd even look quite "vector-y". Personally, I'd go for a 3-layer sandwich, black front and back, and green inbetween, lit from inside.Kamineko wrote:You think you've got it all figured out, ready to design and laser cut the wooden case and then it will be done, done, done!
And then somebody suggests 'Why don't you make the case out of green semi-transparent fluorescent-edged acrylic, in the shape of an asteroid?'
Smeg...!
Asteroids on Oscilloscope
Re: Asteroids on Oscilloscope
"massive retard™", "full of shit™", "tiny bitch™", "retarted™", "a bit of a one trick pony™", "petty™", "simple minded idealistic twat™", "miserable humourless fucker™"
Re: Asteroids on Oscilloscope
Do you have a CNC laser cutter? Cutting acrylic sheet by hand is horrible, then you have to flame the edges, then straighten out the melted bits, then polish out the marks from where you held it whilst unbending, then you have to resize the pieces to make them fit together again which brings you back to flaming the edges. And on the Monday morning the gasman came to call.
Re: Asteroids on Oscilloscope
Hey everybody, want to see a video? 
MATTSTEROIDS and the Ocelot Arcade System make their (second) debut at DoES Liverpool Maker Day!
DoES has two big laser cutters that are good for wood, most plastics and all kinds of other stuff. Making these kind of cases is absolutely their bag.
The multi-talented Ocelot Arcade Corporation creative dept. (my bro, who also did the music) has come up with a concept render to show how a fluorescent-edged acrylic case would look:

I 'unno. I liked the wooden case idea. I'm having trouble visualising that case with all the bits inside.
MATTSTEROIDS and the Ocelot Arcade System make their (second) debut at DoES Liverpool Maker Day!
DoES has two big laser cutters that are good for wood, most plastics and all kinds of other stuff. Making these kind of cases is absolutely their bag.
The multi-talented Ocelot Arcade Corporation creative dept. (my bro, who also did the music) has come up with a concept render to show how a fluorescent-edged acrylic case would look:

I 'unno. I liked the wooden case idea. I'm having trouble visualising that case with all the bits inside.
Re: Asteroids on Oscilloscope
you can use wood & plastic if you like - the laser will cut everything to exactly the right size. some people do layers, so flat top and bottom then thin outlines for the sides, like those rpi cases, bit wasteful mind
now that you've finished i can show you the easter egg mode on the HP 54622
http://kotaku.com/heres-how-to-play-tet ... 1560280553
now that you've finished i can show you the easter egg mode on the HP 54622
http://kotaku.com/heres-how-to-play-tet ... 1560280553
Re: Asteroids on Oscilloscope
Aye I've seen those pi cases at the club, striped layers and whatnot. I guess they're something simple you can make out of offcuts.
And I had no idea about that! I have to wonder though, if that thing is that new, it's probably a raster screen/displaying everything in a raster mode, and that would be cheating.
And I had no idea about that! I have to wonder though, if that thing is that new, it's probably a raster screen/displaying everything in a raster mode, and that would be cheating.
Re: Asteroids on Oscilloscope
Hollow out a real asteroid. From space.
Re: Asteroids on Oscilloscope
That idea rocks.
Edit - But y'know... aren't all rocks from space anyway? And if you'd watched the video, you'd know I hate rocks. (Apparently...?)
Edit - But y'know... aren't all rocks from space anyway? And if you'd watched the video, you'd know I hate rocks. (Apparently...?)
Re: Asteroids on Oscilloscope
yes, does look very rastery - the scopes got a floppy drive too
Re: Asteroids on Oscilloscope
Well that's OK. If it's got a floppy then it's still retro.
A lot of LCD scopes boot Windows, so by all rights they should let you play Minesweeper or Freecell on them. Would be amusing to see vector versions of those. What might be tricky, though, is the effect at the end of a game of Solitaire.
A lot of LCD scopes boot Windows, so by all rights they should let you play Minesweeper or Freecell on them. Would be amusing to see vector versions of those. What might be tricky, though, is the effect at the end of a game of Solitaire.
Re: Asteroids on Oscilloscope
The sudden realisation of how quickly the current time and your project deadline are converging?gumboot wrote:What might be tricky, though, is the effect at the end of a game of Solitaire.
Re: Asteroids on Oscilloscope
Custom cases custom cases everybody loves custom cases




Re: Asteroids on Oscilloscope
venusian wrote:The sudden realisation of how quickly the current time and your project deadline are converging?gumboot wrote:What might be tricky, though, is the effect at the end of a game of Solitaire.
Thinking about how you'd even do that made me wonder about Doom or Quake.
Probably not on a PIC.
[edit]
Actually maybe yes on a PIC. I think a huge proportion of the compute is replaced with empty space, so all you have to do is find intersections of walls.
Re: Asteroids on Oscilloscope
Lines and 3D are not the hard part. The hard part is coming up with a model for, and implementing, obscured line removal.
If I were super-cheating, I could port most of my MS-DOS fixed-point-based car sim (simulation and renderer) across by just changing the types around (MS-DOS 32-bit ints to dsPIC33 long ints).

It would be the same up to the routines for drawing models and terrain and then it would put lines to the screen instead of calculating fragment edges for the MSDOS scan buffer. The Ocelot wouldn't have to go so far as to actually calculate these fragment edges, it would only have to calculate the endpoints.
(The fragments would have to be recalculated eventually, of course. The oscilloscope gimmick is basically a 4096x4096 1-bit pixel screen that has a lot of temporal persistence and some shiny special effects at saturation. The rendering has two parts: the game loop assembles a list of instructions for a LOGO-like bytecode every frame, while the cursor acts as a super-fast asynchronous 'coprocessor' eating this bytecode and spewing out coordinates to the DAC. The fragments would be recalculated over and over eventually in realtime constantly as the cursor shoots around drawing from its line list, but whatcha gonna do.)
If I'm reading the specs right, the dsPIC has single cycle multiply and divide, which makes it quite nice and shiny for this stuff. The double-buffered display instruction list is taking up 2 * 4096 bytes of my 16834 RAM though, so Quake -may- be pushing it a little.
The knack here would be to be able to draw an opaque triangle/convex polygon onto this LOGO-like bytecode instruction list (or alternatively some intermediate list of lines that are yet to be LOGO-ified that have more information and are easier to calculate against). Do that, and you've won the universe.
Massive improvements would be had if you could do that test against the convex polygons beyond triangles. (The naive method I think would test existing lines against each sub-triangle within a polygon, resulting in bits of sub-line being created and destroyed as the polygon gradually gobbles it up over each triangle. Yuck.)
If I were super-cheating, I could port most of my MS-DOS fixed-point-based car sim (simulation and renderer) across by just changing the types around (MS-DOS 32-bit ints to dsPIC33 long ints).

It would be the same up to the routines for drawing models and terrain and then it would put lines to the screen instead of calculating fragment edges for the MSDOS scan buffer. The Ocelot wouldn't have to go so far as to actually calculate these fragment edges, it would only have to calculate the endpoints.
(The fragments would have to be recalculated eventually, of course. The oscilloscope gimmick is basically a 4096x4096 1-bit pixel screen that has a lot of temporal persistence and some shiny special effects at saturation. The rendering has two parts: the game loop assembles a list of instructions for a LOGO-like bytecode every frame, while the cursor acts as a super-fast asynchronous 'coprocessor' eating this bytecode and spewing out coordinates to the DAC. The fragments would be recalculated over and over eventually in realtime constantly as the cursor shoots around drawing from its line list, but whatcha gonna do.)
If I'm reading the specs right, the dsPIC has single cycle multiply and divide, which makes it quite nice and shiny for this stuff. The double-buffered display instruction list is taking up 2 * 4096 bytes of my 16834 RAM though, so Quake -may- be pushing it a little.
The knack here would be to be able to draw an opaque triangle/convex polygon onto this LOGO-like bytecode instruction list (or alternatively some intermediate list of lines that are yet to be LOGO-ified that have more information and are easier to calculate against). Do that, and you've won the universe.
Massive improvements would be had if you could do that test against the convex polygons beyond triangles. (The naive method I think would test existing lines against each sub-triangle within a polygon, resulting in bits of sub-line being created and destroyed as the polygon gradually gobbles it up over each triangle. Yuck.)
Re: Asteroids on Oscilloscope
It turns out there aren't very many (if any) free programs on the internet that design laser-cutter patterns for non-cuboid 'notched box' project cases. Plenty of cuboids, one or two regular polygons, but no arbitrary polygons.
So I wrote one from scratch! http://casemaker.mrdictionary.net
Making non-cuboid cases means that you have to deal with the side-effects of arbitrary angles and lengths appearing anywhere in the prism footprint. My algorithm handles both convex and concave angles (so you could have an L shaped case for example). The base panel footprint needs to be adapted to accommodate the angles, the face panels need to be extended in width with the teeth of each changing depth depending on the angles where they meet. All of these depend on the material thickness in multiple ways.
Haven't been able to test it because I don't have a laser cutter in my bedroom. The SVGs it produces seem legit though.
After the biscuit tin proved to be way too small, I envisaged Ocelot having a flat case with a square-ish footprint and a control panel slightly angled back from the player. Making prism shaped cases is Case Maker's cup of tea, so I zooped in the coordinates to see what its solution would be. It surely looks like tilted-control-panel shaped case design to me. http://casemaker.mrdictionary.net/outpu ... 111427.svg
Wednesday was Maker Night at DoES Liverpool, so I placed my reputation on the line and proceeded to laser-cut directly from Case Maker's output without tweaking it.

God damn, I love being smart. Being smart is awesome.
The ocelot logo on the top of the box is engraved into the surface of the wood. (A shallow area-filling burn from a fast moving laser sweeping over the shape.) The image is from animalgalleries.org. (http://www.animalgalleries.org/cache/Wi ... 0_h399.jpg) I used Threshold in GIMP to make the image two colour, directly imported it into the laser-cutter application and overlaid it onto the panel cut outline. Didn't need to vector trace it or anything. Super pleased with that result.
There was a snafu here when the laser cutter kept jumping out of alignment during the job. I was told the thing is getting a bit on in years and it doesn't like long engraving jobs; it will occasionally throw the laser arm out of calibration (more likely during large engraves), ruining whatever you're doing. If you were to look at the underside of the lid, you'll see the previous attempt with the poor kitty's face crudely splinched half way across.
Let's see what happens when I mount the instrumentation onto this frame and place the boards inside.

Looking good!

Looking super good! =^_^=.
Smells like burny wood...
This victory is brought to you entirely by the majesty of electronic calipers. Electronic calipers are -amazing-. You want to measure something? DONE. You want to measure the interior of an aperture? DONE. I love things that exemplarily serve a well-defined purpose. These calipers? These calipers are GENIUS incarnate. I would never have been able to design and cut the new OAS front panel without them. I want to give a precision engineered high-five to their inventor.
I've been up for a while.
So I wrote one from scratch! http://casemaker.mrdictionary.net
Making non-cuboid cases means that you have to deal with the side-effects of arbitrary angles and lengths appearing anywhere in the prism footprint. My algorithm handles both convex and concave angles (so you could have an L shaped case for example). The base panel footprint needs to be adapted to accommodate the angles, the face panels need to be extended in width with the teeth of each changing depth depending on the angles where they meet. All of these depend on the material thickness in multiple ways.
Haven't been able to test it because I don't have a laser cutter in my bedroom. The SVGs it produces seem legit though.
After the biscuit tin proved to be way too small, I envisaged Ocelot having a flat case with a square-ish footprint and a control panel slightly angled back from the player. Making prism shaped cases is Case Maker's cup of tea, so I zooped in the coordinates to see what its solution would be. It surely looks like tilted-control-panel shaped case design to me. http://casemaker.mrdictionary.net/outpu ... 111427.svg
Wednesday was Maker Night at DoES Liverpool, so I placed my reputation on the line and proceeded to laser-cut directly from Case Maker's output without tweaking it.

God damn, I love being smart. Being smart is awesome.
The ocelot logo on the top of the box is engraved into the surface of the wood. (A shallow area-filling burn from a fast moving laser sweeping over the shape.) The image is from animalgalleries.org. (http://www.animalgalleries.org/cache/Wi ... 0_h399.jpg) I used Threshold in GIMP to make the image two colour, directly imported it into the laser-cutter application and overlaid it onto the panel cut outline. Didn't need to vector trace it or anything. Super pleased with that result.
There was a snafu here when the laser cutter kept jumping out of alignment during the job. I was told the thing is getting a bit on in years and it doesn't like long engraving jobs; it will occasionally throw the laser arm out of calibration (more likely during large engraves), ruining whatever you're doing. If you were to look at the underside of the lid, you'll see the previous attempt with the poor kitty's face crudely splinched half way across.
Let's see what happens when I mount the instrumentation onto this frame and place the boards inside.

Looking good!

Looking super good! =^_^=.
Smells like burny wood...
This victory is brought to you entirely by the majesty of electronic calipers. Electronic calipers are -amazing-. You want to measure something? DONE. You want to measure the interior of an aperture? DONE. I love things that exemplarily serve a well-defined purpose. These calipers? These calipers are GENIUS incarnate. I would never have been able to design and cut the new OAS front panel without them. I want to give a precision engineered high-five to their inventor.
I've been up for a while.
Re: Asteroids on Oscilloscope
That is utterly awesome.



