Monty

Baby dinosaurs!

monty@xiph.org — Sat, 26 May 2012 21:58:44 GMT

The local mockingbirds have fired up the springtime brood-vats, and I was surprised to hear fledgelings begging for food yesterday.

I was pretty sure I could hear three of them, and finally managed to catch the whole brood in one place (otherwise, not a great picture).

This also afforded some opportunity to practice driving stick on the camera (handheld shooting a manual focus long telephoto)!

In general, the challenge is what you'd expect-- the lens is slow, long, and prone to serious purple fringing unless stopped down to F11 or so. Pretty much what you'd expect from 35-year-old optics. There's not much point to even trying to get off a shot slower than 1/200 or wider than F8 (PF wasn't really under complete control until F11). I got lots of feeding shots, all of them wanted to be much much faster-- they're mostly blurs of feathers.

Other than that.. yeah, it's not a bad lens for $50.

Oh! Right! Birds! These little guys have been running the parents ragged for days.

SO! FLUFFY!

Silhouette or Look Out Below!

"YO"

"Pizza's here!"

Beak full of wiggly caterpillars?

Obligatory hot-bug injection picture

stretching and preening

Good night...

equipment wanking: photography

monty@xiph.org — Wed, 21 Mar 2012 16:21:34 GMT

Now that we have an unnaturally early spring^H^H^H^H^H^Hsummer here in New England, I finally had a few patches of green grass against which to try white balancing the new-ish D7000 with an IR filter in place.

Persistence payed off... it will do it!

Sadly, the Sigma zoom shows a huge honking hotspot in most shots. I shall have to try the primes and ancient Tamron zoombeast next.

Looks like the kingfishers are back.

monty@xiph.org — Sun, 11 Mar 2012 03:22:13 GMT

Alternate photo title: "Winter? What winter? Wait, that was winter?"

Alternate alternate photo title: 600mm is just not enough

Why 24-bit/192kHz music downloads make no sense

monty@xiph.org — Mon, 05 Mar 2012 18:29:50 GMT

(by Monty and the Xiph.Org community)

Articles last month revealed that musician Neil Young and Apple's Steve Jobs discussed offering digital music downloads of 'uncompromised studio quality'. Much of the press and user commentary was particularly enthusiastic about the prospect of uncompressed 24 bit 192kHz downloads. 24/192 featured prominently in my own conversations with Mr. Young's group several months ago.

Unfortunately, there is no point to distributing music in 24-bit/192kHz format. Its playback fidelity is slightly inferior to 16/44.1 or 16/48, and it takes up 6 times the space.

If you just said 'Whaa?', you may want to read the whole article.

It's fairly long... but hearing, perception and fidelity are complicated topics. Shysters and charlatans exploit that nuance (and misunderstanding) to bilk unsuspecting consumers of their money, all the while convincing them they're paying for 'quality'.

Anyway, happy reading and comments welcome!

A Denton pic

monty@xiph.org — Wed, 22 Feb 2012 13:55:37 GMT

This is a pic from last summer that I had meant to post then. I ran across it again a few nights ago, and didn't want to forget a second time.

XiphQT components, MAC OS X and 64 bit iTunes

monty@xiph.org — Sat, 21 Jan 2012 08:24:40 GMT

Camilla forwarded a necessary tip for installing the XiphQT components on a 64 bit Mac OS X so that it works with iTunes. This is a reasonably well known tip, but it wasn't in our FAQ or installation instructions (well it is now as of about ten minutes ago) so I'm passing it along now too...

I upgraded to Lion, and my ogg files stopped being able to play in iTunes (silently). Here's how to make it go:

"show in finder" your iTunes binary (either navigate to the Applications folder, or right/control click on it in the dock, and choose "show in finder")
right/control click on iTunes in the finder, and select "Get Info"
Under General, check the box marked "Open in 32-bit mode"

You should put the above on something linked from: http://www.xiph.org/quicktime/download.html I paraphrased it from roaringapps.com.
If XiphQT can be rebuilt in 64 bit mode, and that shipped that way to Lion users, that would also be a good solution.

That last comment is actually a bit of an embarrassment for us at the moment; neither the XiphQT builds nor code have been updated since 2009 or so, despite multiple releases, fundamental improvements and new features in the Xiph codecs since. There are actually more recent beta builds of updated Mac OS X and Win32 XiphQT components than never got bumped to the official XiphQT download page, but even these builds are from mid 2009.

We don't have any high-powered Mac OS hackers in the core Xiph group at the moment. I have some relatively insignificant amount of experience coding for Mac OS X and Quicktime, but I've been hoping for a volunteer with more chops. Any takers?

Ghost Update: Demo 4

monty@xiph.org — Sat, 21 Jan 2012 07:19:29 GMT

Turns out I missed blogging about the latest Ghost update... back in November...

Ghost Demo4 is up on the demo list showing the sinusoidal extractor doing some very early sinusoidal tracking frame to frame, and a very early example of the analysis performing real sinusoidal/non-sinusoidal audio splitting. Pictures and interactive listening, oh my!

It looks like I'll be putting a month or two into transOgg before getting back to Ghost work (and demo 5). The work that went into demo4 raised a number of questions I'm not sure how to approach answering yet, so I'm going to let that percolate for a bit.

Planet Xiph posts now featured on Xiph.Org front page

monty@xiph.org — Thu, 19 Jan 2012 14:33:54 GMT

I made a quick change to the Xiph.Org front page that a few people have suggested now over the past few years.

The top few blog posts aggregated by Planet Xiph now appear as a five-item teaser list near the top of the Xiph.Org home page. The idea is both to get some more live content on the front page as well as to draw more attention to both the Planet and our developer community.

Comments and feedback welcome!

Upgrading the Infrastructure

monty@xiph.org — Sun, 19 Jun 2011 12:55:07 GMT

Last weekend I was having some trouble reliably handling thin acrylic stock, specifically putting two .220" wide .015" deep slots into .660" wide strips of .030" acrylic. This morning was more successful:

The router leaves a rather pretty moire pattern as I'm feeding by hand so my feed rate isn't perfectly steady.

Doing so required a better sawbox. (In case you were wondering, the Jeapordy clue is "Super Star Destroyer")

...and New router jig.

FTC "Patent Hold-Up" Workshop

monty@xiph.org — Wed, 15 Jun 2011 09:36:15 GMT

Quite a few otherwise interested people may not have heard that the FTC (Federal Trade Commission) is holding a panel and workshop next week concerning how patent trolls are abusing standards body processes. This is our field, and we didn't find out about it until end of last week.

Regardless, Xiph.Org has assembled an official comment document, and will be represented in person by at least Dr. Tim Terriberry and possibly a few other core members (I won't be there).

If you're interested in software patents, some of the US Government's thinking on the issue, and participating in the process, have a look at the above two links. Also, feel free to distribute our comments far and wide. It's somewhat more gripping than the usual, dry "Percy Q. Business Leader Advises the Federal Goverment".

Death By Graphs (a new Ghost update/demo)

monty@xiph.org — Sun, 05 Jun 2011 03:28:34 GMT

I'd mentioned in the previous update that we're (Xiph is) using a chirp estimation algorithm that we published back in 2007, and that the original paper has precious little space to devote to describing in detail how the algorithm actually performed. One of the upshots of not having done extensive characterization tests of our own algorithm was that it has already surprised me a few times this year (in both good and bad ways).

Therefore, Ghost Update 20110604 concerns itself with describing and graphing algorithm behavior in mind-numbing detail.

Death! By! Graphs!

monty@xiph.org — Sat, 04 Jun 2011 00:39:30 GMT

monty@xiph.org — Thu, 26 May 2011 00:14:51 GMT

Pleasant surprise: Vinyl tubing is surprisingly easy to heat form. Put a spring in it, heat to 165F for ten minutes, cool slowly, remove spring. It retained the bend perfectly with no kinking.

I have found...

monty@xiph.org — Thu, 26 May 2011 00:08:41 GMT

...the ignition I was looking for. Surprisingly difficult to track down. Not scarce, just one of those things where the search terms were hard to come by.

lubricating models and styrene safe oil

monty@xiph.org — Tue, 24 May 2011 21:03:03 GMT

I've recently started a model project (souping up a Visible V8!) and one of the first things that came up is... with what do you lubricate it? Ask a chemist "what lubricants attack polystyrene?" and the answer is something close to "Oh, wow, just about everything."

The Visible V8 manual advises using vegetable oil, which is a great choice only if you want to sell a brand new model every six weeks. Polystyrene slowly dissolves in virtually all natural oils, and vegetable oil also polymerizes into a solid gunk in the presence of oxygen.

The Internet is full of completely ludicrous advice on the subject of lubricants and styrene. And, of course, vendors don't want to tell you what's in their products or what their criteria is for declaring a particular lubricant 'compatible' with a given polymer. MSDSs are useful for establishing that several lube vendors are probably, to put it bluntly, lying through their teeth about compatibility.

So, what to do? We test.

I've partly immersed two expanded styrene foam peanuts (the very thin skin of the expanded bubbles should show damage well) and one strip of transparent 'crystal' styrene to a number of different lubes and liquids.

In this test, which I expect to leave running for a few years, we have:

Contestant 1 (above): Soybean oil. Sort of a control, because it should be solid useless gunk surrounding trashed styrene in the space of about two months.

Contestant 2 (above): 3-in-1 (black label). Another 'oil' that is usually the worst possible choice for lubricating anything. I expect it to be only marginally better than the soybean oil.

Contestant 3 (above): 3-in-1 (blue label). This is the more serious '3-in-1' that contains a paraffinic oil actually worth using and less of the magic secret crap that makes the black label stuff such a bloody menace. That said, it's a natural group I oil and so should also attack the styrene long-term.

Contestant 4 (above): white mineral oil. Highly refined/purified Group I/II natural oil, also expected to attack styrene long term though more slowly.

Contestant 5 (above): Synco "Super Lube" synthetic lightweight oil without PTFE. This is explicitly listed by the manufacturer as having 'excellent' compatibility with polystyrene. However the MSDS indicates that it contains a substantial fraction of normal mineral oil (approximately 80/20 mix), which means that it is neither fully synthetic, nor is is compatible with polystyrene. We'll see what the test determines.

Contestant 6 (above): NyOil II (Nye Lubricants #123S). This is a product of Nye Lubricants that has legendary status as a super-lubricant on the Pinewood Derby circuit. The MSDS lists the contents as 'proprietary' and 'PAO base' with no additional information. Also, Nye lists the compatibility of the PAO base with polystyrene as 'fair' where other manufacturers list it as 'excellent'.

Contestant 7 (above): Nye Lubricants #269. This oil is recommended for plastics and appears to be ~ 100% PAO. Perhaps just a slightly thinner version of NyOil II? No way to know.

Contestant 8 (above): Labelle #108 plastic-safe oil. Another oil that is legendary as the only thing safe / worth using within a specific modeling community, this time the model train people.

Contestant 9 (above): Amsoil 100% synthetic 0W20. In the interests of trying out something relatively cheap-in-bulk, I've added the only motor oil still explicitly claimed to be entirely Group IV (PAO, diester, etc) synthetic. For the record, 'fully synthetic' motor oils are allowed to be up to 30% Group-III non-synthetic in the US, so other 'synthetic' motor oils are not necessarily substitutes. I believe in Europe, anything claiming to be synthetic does actually have to be 100% group IV+ synthetic.

Contestant 10 (above): 1.5wt Airsoft Silicone oil. Very thin. Silicone is probably the only thing 100% safe forever on polystyrene, so it might seem like a no-brainer as it's safe and cheap. That said, it actually kinda sucks as a lubricant, it dissolves silicone gasketing material, and it tends to contaminate workshops. The problem with silicone oil is nothing sticks to it, aerosolized droplets get everywhere, and it doesn't evaporate so it stays put in inadvertant places forever-- if you use silicone lubricants in a workshop, you can probably forget about trying to paint or powdercoat with reliable results in said workshop ever again.

Contestant 11 (above): Feser One (not an oil). This is actually a non-conductive fluid sold for watercooling loops in computers. I have a use for it (it's the right color for antifreeze :-) and want to make sure none of the additives will hurt the styrene.

Contestant 12 (above): Thermochill Ec6 (not an oil). See 11 above.

Contestant 13 (above): WD40. Sort of an anti-control-- this should be an utter disaster with styrene. In fact, I think it was already dissolving the foam peanut by the time I took the picture. In any case, WD40 is not a lubricant and is in this test only as a warning to others.

...and several compartments left empty as controls, plus one filled with regular water.

...In six months to a year, it will be time to check!

A Ghost update! For... last month!

monty@xiph.org — Wed, 27 Apr 2011 17:27:47 GMT

Not actually last month, but pretty close at this point.

I never publically released the my previous Ghost update delivered internally to Red Hat at the beginning of the month because I hadn't finished some of the diagrams I wanted to do for the last section on chirp coding. Well, the diagrams are done! Here's the latest Ghost demo update, just in time for the next one to almost be due!

WebM Community Cross-License announced

monty@xiph.org — Wed, 27 Apr 2011 13:09:19 GMT

In case folks have missed it (or worse, read about about it on Ars Technica)...

The WebM folks have finally finished up their work on the WebM Community Cross-License project and announced the license launch. This is a FOSS defensive license/pool similar to what a couple other groups are trying out (and similar to the defensive patent license that Xiph is already using for our parts of Opus within the IETF).

The basic idea of the cross-license is:

"Everyone is free to use any known or unknown WebM patents. Unless you sue over patents related to WebM. In that case, we all agree to yank your license."

In short, it's sort of a NATO for FOSS patents; a free license with an agreed-upon mutual defense clause that tries to enforce everyone playing nice. This strategy is not a new idea, but it's interesting that several different FOSS groups, Xiph and WebM included, are finally trying the idea for real in practice.

Welcome to Somerville, turkey.

monty@xiph.org — Mon, 25 Apr 2011 14:30:49 GMT

Opus (CELT) demo page 2: Listening tests

monty@xiph.org — Fri, 15 Apr 2011 03:24:11 GMT

Greg Maxwell has just posted a nice second 'demo' page for Opus. It mostly covers the recent listening testing done by volunteers at Hydrogen Audio. Pretty colors and interactive listening/comaprison scripting!

For those of you new to Opus, it's the FOSS/RF codec we're working on within the IETF codec working group. Opus is a collaborative hybrid speech / high-fidelity audio codec built using primarily Xiph's CELT codec and the Skype SILK voice codec as inputs. That makes Opus similar in some ways to what MPEG is trying to achieve with USAC (Unified Speech and Audio Coding), though Opus is also ultra-low latency, so it looks like we're considerably ahead of MPEG here.

George has the best hair.

monty@xiph.org — Fri, 01 Apr 2011 04:05:46 GMT

A total eclipse of the George allows us to see his corona.

The big hair keeps getting bigger

monty@xiph.org — Mon, 31 Jan 2011 03:09:24 GMT

Several other photos were better composition-/execution-wise, but there was just something very Denton about this one.

monty@xiph.org — Mon, 31 Jan 2011 02:53:25 GMT

George.... hard at work...

(Yeah, it's out of focus.)

monty@xiph.org — Mon, 31 Jan 2011 02:38:58 GMT

Yep, looks like an infestation of xenomorphs all right... (they *did* pop out of Camilla so to speak).

X61T rebuild pics, and some rhetorical musing at the end....

monty@xiph.org — Tue, 04 Jan 2011 08:07:20 GMT

...I had promised a few days ago I would post the X61T disassembly and reassembly pics... :-)

I picked up a nice little X61T off eBay assuming ahead of time that I'd replace a number of unavoidably worn parts to make it effectively new. Thinkpad or no, some things in a modern notebook unavoidably wear out. The difference is that in most Thinkpads they're easily replaced. The backlight is an exception, though I have some experience replacing backlights...

Anyway, I bought a once top-of-the-line 1.8GHz model and a 1.6Ghz 'beater' that I planned to cannibalize for parts as necessary. The beater actually turned out also to be in good condition (it was broken, but only in trivial ways), so I'm rebuilding it now too. I'll sell it back off on eBay, there's no sense wasting a perfectly good Thinkpad.

Starting at the beginning... Open up the Bible to Genesis, remove the battery and hard drive, pop the stylus out (if you don't, you'll snap the pen holder off its mount when removing the palmrest... the hardware service manual neglects to mention this) and remove every screw from the bottom.

The hardware manual has you pop the hinge cover caps off next, but in retrospect it would have been much easier to do this step after unbolting the hinge after removing and opening the lid...

The hard part about getting the hinge cap off is the catches on both sides inside it... if you try it, you'll see what I mean... pop one side, flip the lid to get to the other side, the first side pops back in.

Pop off the smaller second piece of the hinge cover.

Keyboard slides forward, then hinges up from the bottom.

Don't forget to actually disconnect the keyboard cable.

Pop the palmrest off.

...and then the keyboard bezel...

Naked Thinkpad!

Remove bluetooth daughtercard/cable, disconnect WLAN cables and remove the daughtercard, and pull the LCD extender board.

Uncrew the remaining four screws holding on the hinge and behead the machine.

The nine bolts holding the lid together are covered by a small square cover, two rubber bumpers and six adhesive disks. An X-acto knife is perfect for prying them off. Keep a finger over 'em, they tend to pop free all of a sudden.

Remove the nine screws holding the lid together. The lid is also held closed by snaps along all four edges, as well as an adhesive strip along the bottom of the screen. It's easiest to begin popping the snaps along the bottom seam, then along the right side. Pry the bezel away and out from the lid and the snaps should pop easily. work up the right side, then about halfway along the top. Gently pry the adhesive strip away from the LCD panel along the bottom. At this point the remaining latches along the left and top left of the screen should release easily.

Lid open, hooray! Disconnect the fingerprint reader cable and set the bezel aside.

Slide the cables through the hinge. The thick cable should go first-- it's a tight fit but it slides through without difficulty.

Remove the two screws holding in the button/LED board, the one screw holding in the backlight inverter, the ground cable screw behind the LED board, and the four screws holding in the LCD panel. Flip the panel, cables and two small boards out onto something soft.

Untape the cables from the LCD and disconnect the two connectors. Set the cables and boards aside. We need to pull the LCD panel completely apart to get to the backlight.

A few notes about backlights...

A CCFL backlight has a relatively short/finite lifetime, usually about 5000-10000 on hours until it's down to 50% brightness. Before a CCFL fails completely, it's typical for it to start very pink or red when it first fires before becoming more white as it warms up.

A bad inverter looks different. Either the lamp stops firing at all, or is so dim the screen is barely visible in a dim room. The failure is usually sudden. Although some eBay sellers claim otherwise, in my experience a failed inverter isn't very common, whereas every old notebook has a worn CCFL.

Anyway, onto peeling the panel apart... Remove the adhesive plastic/metal shield over the LCD driver, unscrew the driver board, disconnect the resistive touchscreen cable from the digitizer board.

Cut the thin tape holding the metal frame along the bottom...

Pop apart the snap lugs holding the metal retaining frame onto the panel.

Flip the panel out of the metal frame.

The frame includes the resistive touch layer and glass protection plate. Cover the frame (to keep dust and fingerprints out) and set aside someplace safe.

The LCD panel is exposed now. It's not actually all that fragile, but it'll shatter if flexed too hard or dropped. Also, any fingerprints and dust you don't get on it now you won't have to remove later.

One thing about the LCD panel is very fragile; beware where the wide flex cables fasten to the glass along the top. If the cable pulls apart from the panel, game over, screen destroyed. Do not, under any circumstances handle the screen by the cable and by extension don't grab the controller board either.

Next, flip the LCD panel panel itself out of the frame onto something soft. cover and set aside somewhere safe.

The diffuser/polarizer is exposed now. It's the one annoyingly fragile part of the screen in the sense that if it gets dirty (eg, big fat fingerprint), any attempt to clean it will probably leave some slightly visible scratches. So, don't touch the diffuser for any reason. Also, don't try to blow dust off of it (spit will probably leave visible spots too), use canned air instead.

The layers of diffuser/polarizer plastic are held in place by two small clips. Remove the clips.

Flip the diffuser/polarizer layers all out together onto a soft, clean cloth. Cover and set aside.

The next layer is a clear waveguide with a microdot texture on the back. It's fairly tough and somewhat flexible. Alcohol and other solvents will cause it to craze or fracture though.. if for some reason you get a fingerprint on it, use windex. or just don't get fingerprints on it to start with.

Pop the top corners of the waveguide free of the flexible plastic frame by flexing the corners of the frame down. Cover with a soft cloth and set aside.

The lowest layer is a floppy white plastic backing. Leave it in place for now. Be aware that it will accidentally kink/crease more easily than it appears so don't be too rough with it. Under the plastic backing, behind the frame, is the Wacom digitizer.

Flip the frame over, free the flexible Wacom panel from the clips along the top of the frame. It's then held in place by adhesive tape along the sides and bottom. The Wacom is tough and not easily damaged. Free the left side first, then the bottom and finally the right side. Set it aside.

Peel off the remaining thin tape along the bottom of the to expose the backlight bracket and free the white plastic backing.

Remove the screws and clips holding in the backlight.

Flip the frame over and free the backlight wire that runs along the face.

Peel the backlight out. There's one more piece of adhesive tape that is holding the backlight in, so work it free slowly from one side to allow the tape to slowly release.

Pull the CCFL tube out of its bracket. It's held in place by three small silicone O-rings along its length that are squeezed into place in the bracket.

Pull down the silicone boot that covers the solder joint at each end of the CCFL tube. Don't damage or tear the boot, it's preventing a kilovolt from arcing to ground.

Desolder the tube and remove/save the three O-rings. Dispose of the old tube properly (it contains mercury, dispose like any other fluorescent lamp).

Put the silicone O-rings onto the new tube and solder it to the wires. Note that the pink and red leads are set 90 degrees to each other.

Trim the lead tails, reseat the boots and set the tube back into the bracket.

There's a trick here; the bracket springs closed enough once the tube is first removed that getting the tube back in is difficult. The metal edges are sharp and tend to slice the O rings. The shaft of a name-brand Q-tip(TM) cotton swab is a nice diameter for running through the bracket a few times to spread the edges open just enough to allow getting the tube back in but not so much that the bracket no longer seats snugly against the waveguide.

If you end up spreading the bracket open too far and it does not lightly grip the waveguide evenly with no gaps, you'll end up with uneven very noticable bright spots along the bottom of the screen. If that happens, it's easy enough to gently pinch it more closed.

Reassembly is the inverse of disassembly with the caveat that we leave reattaching the Wacom until the panel is reassembled.

Put the backlight, backing plastic, and waveguide back into the frame. Use some thin stiff plastic (a strip of polyester overhead projector transparency sheet is perfect) as a 'shoehorn' to guide the clear waveguide back into the backlight reflector bracket. The shoehorn makes reinserting the waveguide much easier and eliminates any possibility of damaging or peeling away the thin reflective film laminated to the inside of the bracket.

A few pieces of dust trapped between the backing and waveguide will be invisible. If there's a lot, blow it out with canned air. Protip: a humidifier not only lessens chances of a static spark damaging a circuit, it will also prevent static buildup on the backing/waveguide from proactively sucking dust out of the air.

Drop the diffuser lateys back into the frame. The black keys along one side fit under the black strip in the white frame. Replace the metal clips that hold the diffuser in place.

Place the panel back into the frame. Any dust between the diffuser layers and panel will be visible. Inspect for and flush any dust with canned air. Replace the metal frame by gently inserting the side of the white frame with the flat LCD flex cables, then dropping the metal frame down over the panel and snapping the lugs back together. When properly aligned, the snap-together process will go easily. If it seems like the metal frame just won't lower all the way into its lugs, don't force it! The LCD itself might not be seated properly and forcing will shatter it.

Flip the panel and replace the tape we removed from along the bottom with some fresh polyester tape. Trim to fit, of course.

Reattach the Wacom digitizer. Align it along the top of the white frame by fitting the top edge under the small white clips, then align the sides and press into place.

Bolt the LCD controller board back into place, replace the adhesive plastic shield, reinsert the touchscreen flex cable into the digitizer board, and reattach the LCD cable assembly and boards.

Bolt everything back into the lid.

Attach a new hinge (the old one was still solid, but had picked up some play), reattach the fingerprint reader cable, snap the bezel back into place, replace the nine screws and return the cover discs, bumpers and square to their rightful places. The lid is done!

To replace the fan, we need to remove the motherboard. To free the motherboard, remove the modem card and detach its cable, remove the video shield, pop the disk drive card subcard off the mobo, remove the lid switch and speaker and remove the two screws on either side of the card cage. The IBM manual also leaves up till now one screw to remove from the bottom.

And the motherboard hinges right out.

Disconnect the fan cable, loosen the three spring-loaded CPU heatsink screws and detach the remaining screw on the GPU heatsink. The fan will pop right off.

Reassembly is the reverse of assembly...

Am I this vain? Yes, I am this vain.

Hooray! All done!

...but does it blend? Yes! It blends!

How much difference does that new backlight make?

And now, for the final absolutely necessary step...

As I mentioned, I have a second machine that I was going to use for parts then didn't. Mechanically and electrically, it's in just as good shape as the one in these pics, but it's considerably more scuffed up. No dents, no cracks, but all the edges are worn through the paint down to metal and the top and bottom honestly look like they suffered alot of sliding on something gravel-like. It's not pretty.

I'm currently seriously considering stripping the paint and polishing the bare magnesium up to shiny...

Heh.

monty@xiph.org — Fri, 31 Dec 2010 06:17:06 GMT