Monday, June 21, 2004

Our Friend, the Meter

This evening, I learned that one meter equals 39.3700787 inches. While this may come as no surprise to some people, it was one to me - for years, I had mistakenly believed a meter was 39.77 inches, and now I know it's basically 39.37.

Of course, I'm not alone in my confusion. A bit of research on Google revealed quite a few different conversions from meters to inches. Here are some of them:
Once again, the correct answer is right around 39.37 inches. Remember that - it'll be on the quiz!
Updated 6/24: More than a few people have kindly pointed out two things:
  1. When rounding to 1 significant digit, 40 inches is actually "correct," as is 39 when rounding to two, and 39.4 when rounding to three. So the USDA and Navy, along with the quilting folks, are not technically "wrong." The H2WTech folks, however, list 1 meter as 39.40 inches - four significant digits - and are therefore still wrong, as are the folks at Fife Products who list it as 39.0 inches - three significant digits.
  2. More usefully, there is, of course, a precise conversion factor of 1 inch = 2.54 centimeters. That's the correct way to convert things - inches x 2.54 / 100 = meters; meters x 100 / 2.54 = inches. Fair enough. This does, though, make me wonder why numerous sites listing metric conversion factors include both the precise one for inches and centimeters and somewhat arbitrary ones for inches and meters.

Friday, June 18, 2004

Peek-a-boo Moray

While skindiving, I noticed a nice red pencil urchin by a head of coral. Swimming closer, I discovered that there was also a Puhi Paka (Yellowmargin Moray, Gymnothorax flavimarginatus) "guarding" that particular area. Given that morays tend to have quite a lot of needle-sharp teeth, I decided to not get too close. Canon PowerShot S20, Ewa-Marine D-MM enclosure, 3-4 meters depth. I took several shots and chose this one. At first, it looks like it's just a photo of some coral and underwater scenery... surprise, it's not!

Wednesday, June 16, 2004

iMac and PowerBook G5: Heat is not the problem.

 Apple got a lot of buzz for their Power Mac G5 in 2003, as did IBM for the PowerPC 970 chip inside it, made using a 130nm process in a shiny new highly automated fabrication facility. There was also a lot of buzz about the Xserve G5in early 2004, which took a while to actually start shipping due to difficulty getting enough functional PowerPC 970FX chips, made by IBM using a new 90nm process. Now there's buzz about the iMac G5, which had originally been planned for the "back to school" summer buying season, but is now expected to launch at the end of August and ship in mid-September. And of course there will be buzz about the inevitable PowerBook G5, at least until Apple announces and ships one.


Anyway, a 
lot of the buzz is along the lines of "G5 chips dissipate so much heat that it's hard to fit them into small spaces like 1U rackmount servers, iMacs and PowerBooks." Compared to G4 chips, this was certainly true of the original 130nm PowerPC 970. Heat dissipation is linked to power consumption, and 51 watts of "typical" power consumption per chip isn't too bad for a desktop processor, but it's not something you put in a laptop if you want anything in the way of battery life.

However, there's less certainty about the power consumption of the newer 90nm PowerPC 970FX. In Apple's recent earnings conference call, the company's chief financial officer, Peter Oppenheimer, said multiple times that the most critical issue affecting the introduction of G5 chips in, well, 
everything, is the need for IBM to get the kinks out of its 90nm process and actually start supplying enough working 970FX chips to Apple to meet its needs. He said this in response to questions about heat problems.

I'd just like to take this opportunity to say that the buzz about alleged heat problems is, well, a lot of hot air. :)

Currently, Apple's PowerBook G4 laptops use Motorola 7447A "G4" processors, running at speeds of 1.33 GHz (12" and 15" models) and 1.5 GHz (15" and 17" models). According to page 19 of a preliminary technical document (PDF) from 
Atmel, the 7447A running at 1.33 GHz has "typical" power consumption of 18 watts, and "maximum" consumption of 25 watts. At 1.42 GHz, those numbers rise to 21 and 30, respectively.

Meanwhile, the XServe G5 uses (and the iMac G5 will presumably use) IBM PowerPC 970FX "G5" processors, running at speeds of 2.0 GHz in the XServe. According to the PowerPC 970 & 970FX page of 
IBM's own PowerPC Quick Reference Guide, the 970FX running at 1.4 GHz has "typical" power consumption of 12.3 watts. At 2.0 GHz, its "typical" consumption is 24.5 watts.

So... a 1.4GHz G5 uses, say, 40% less power than a 1.33GHz G4. And a G4 pushed to 1.5 GHz - standard in the 17" PowerBook and optional in the 15" one - most likely uses just about as much power as a 970FX G5 at 2.0 GHz, and almost certainly more than a 970FX G5 at anything below 2.0 GHz.

So, once more - it's 
not about heat. If the internal design of the new iMac leads to heat problems with a G5, those heat problems would be worse with a G4 at the same clock speed! It's about Apple just not being able to get enough 970FX chips out of IBM's new fab due to the quality-control issues that tend to affect any new smaller process in a relatively new facility. Once the floodgates are opened and working 970FX chips are available to Apple in larger quantities, there's very little standing in the way of getting them into PowerBooks... and, for that matter, anything else Apple wants to put them into.

Do I think we'll see the PowerBook G5 in 2004? No. Mid-2005 is my guess, though I'd be most pleasantly surprised if they were announced at MacWorld San Francisco in January of 2005. And with the iMac G5 out of the way this August/September, Apple will have to have 
something big to announce then, won't they?

That VIS place I keep mumbling about

I took this photo just after sunset, from the top of a cinder cone on Mauna Kea. This is the Visitor Information Station at the Onizuka Center for International Astronomy, located at 9,300 feet elevation. Yes, it's on (more than a bit of) a hill.

Inside the building are a whole bunch of exhibits, computers, a video projection area with seating, and of course a gift shop that sells all kinds of nifty stuff.

Outside on the patio and in the parking lot, a half-dozen telescopes are set up for the evening stargazing program. They are, from left to right:

  • Orion 8-inch newtonian reflector on a dobsonian mount.
  • Orion 4.5-inch newtonian reflector on a dobsonian mount (behind the 8-inch).
  • Orion 6-inch newtonian reflector on a dobsonian mount.
  • Meade 16-inch LX200GPS cassegrain reflector on an alt-azimuth fork mount.
  • Celestron 14-inch cassegrain reflector on a Losmandy equatorial mount.
  • Televue 102mm apochromatic refractor on an Astro-Physics equatorial mount.
The Meade, Celestron and Televue all have motor drives that move them slowly to follow stars moving across the sky. The Meade and Televue also have "goto" computer controls with star catalogs, so celestial objects can be easily and automatically targeted. Not seen here is another Celestron, 12-inch if I recall, that's used for solar viewing during the day. That one has a secondary scope for viewing solar eruptions and a spectrograph mounted on it, as well.

Yes, I know how to use 'em (and how to get 'em out, set 'em up and put 'em away). Oh, the yellow thing on the rock wall is a case for eyepieces and adapter rings.

Saturday, June 12, 2004

Subtle Honu

During a quick dip in the ocean the other day, two Honu (Green Turtle, Chelonia mydas) swam past me. I squeezed off a bunch of shots, but only one caught a turtle far enough from the camera for its entire body to be in the frame - they get close, sometimes! Canon PowerShot S20, Ewa-Marine D-MM enclosure, 2-3 meters depth.

Friday, June 11, 2004

On a clear night, you can see... a lake of fire?

I've mentioned before that I volunteer at the Visitor Information Station of the Onizuka Center for International Astronomy. To call Onizuka a "world-class" facility would not be exaggerating, at all - more than ten nations on 5 continents are involved in it, and the 12 observatories it supports include four of the ten largest optical telescopes in the world, as well as the largest dedicated infrared and submillimeter-band telescopes. The telescopes are among the highest terrestrial telescopes. A 1998 article in Outside magazine by Richard Panek gives some idea of just how extreme this all is. 

Even the VIS is pretty extreme - there aren't a lot of places in Hawaii you'll see tour buses and vans discharging hordes of tourists... in arctic-style parkas! On a clear night, the temperature drops to about 40 F (that's under 5 C), and there might be a good stiff breeze to contend with as well. But a clear night at 9,300 feet, with a dark sky above and basically zero light pollution is pretty incredible.

This week, I spent three evenings at the VIS. Wednesday and Friday started out cloudy, cleared up off and on, and then toward the end of the stargazing period at 10 PM, cleared up entirely. Thursday, though... was clear. Insanely clear. So clear that I could see the city lights of Hilo, 26 miles away at sea level. So clear that I could see the red of Kilauea's Pu'u O'o vent erupting in Hawaii Volcanoes National Park 34 miles away and a mile lower, and even the red of lava flowing across the coastal plain and into the ocean, another 5 miles beyond it.

Toss in a visible pass by the Hubble Space Telescope, other satellites, meteors, two comets - C/2001 Q4 (NEAT) and C/2002 T7 (LINEAR) - and of course a view of Polaris, the Southern Cross, and quite a bit of stuff in between... and you've got a pretty spectacular experience, albeit a cold one.

Thursday, June 10, 2004

Vanquished: the NetCam

Remember the strong-willed camera I was ranting about before? Well, it wound up working!

Working, that is, as long as the person accessing it is on, or VPN'ed into, its local LAN.

Great. Just... great.

Now I know what brand of web camera not to buy.

Tuesday, June 1, 2004

When Good Marketers Go Bad.

 The folks at Apple have a reputation for cranking out well-designed, well-integrated products, and generally marketing them in a way that's pleasing to the eye and all that.


An example of this integration is the Apple Display Connector, which carries a DVI signal, USB channel and power from my 
Power Mac to my Apple Cinema display. There's only one cable coming out of the display, which cuts down nicely on clutter.

Like all of us, though, Apple marketers sometimes have a bad day. And when that happens, look out!

Several of my acquaintances and colleagues have been discussing the fact that 
PowerBooks include a mini-DVI port, and an assortment of adapters to connect that port to various things, but not an adapter to connect it to any of Apple's line of flat-panel LCD displays, which all use ADC.

Apple 
does offer a $99 DVI to ADC adapter, which takes DVI and USB from a computer (be it a PowerBook, the secondary DVI port on a Power Mac, or indeed any computer, Apple or not, with a DVI port) and power from an AC adapter, and combines them into an ADC cable.

When one purchases a PowerBook on-line, a display and adapter are presented as a bundled time-of-purchase option. However, if one visits 
Apple's online store to purchase an Apple Display "after the fact" - for use as the second display on a Power Mac, as an external display on a PowerBook, or as a display on a DVI-equipped computer from another manufacturer - the information presented is less than helpful. "Double your desktop area," the site says. "The Apple DVI to ADC adapter makes adding a second flat-panel display simple. Now only $99." No mention of it being necessary if you're not using the panel as a second display.

It gets worse, though. Select one of Apple's three display models, and thirteen "Accessories" are presented. The all-important DVI to ADC adapter is among them, yet for some reason it's listed 
fifth, while the top-listed item is a $19 VGA adapter that doesn't work with ADC displays, and is primarily intended for systems that don't even have DVI ports. In fact, nine of the thirteen accessories listed are of absolutely no relevance or usefulness to an ADC display. (If you're curious, the other three relevant ones are two color calibration tools and a cleaning kit). This is confusing at best.

So what's the cure? Well, the first step is to stop making the problem worse with all those irrelevant "accessories." There should only be four things listed, and the DVI to ADC converter must be listed 
first.

Secondly and more importantly, the DVI to ADC converter needs to become a "default" for people purchasing Apple's flat-panel ADC displays. This can be done by having it "checked" by default in the accessories listing (much as a 56K modem is "checked" by default when one buys a Power Mac, but can be unchecked, resulting in a $29 price cut)... or by simply including one in the box with 
every ADC display Apple sells. After all, anyone who's using an ADC display as anything other than the primary display on a Power Mac is going to need a DVI to ADC adapter, whether Apple communicates that clearly to them or not.

So, yes, Apple's marketers had a bad day when it came time to sell ADC displays. Let's hope they have a good day soon, and untangle this mess.

Why I'm leaving Twitter.

I've stuck it out and continued participating on Twitter while Elon Musk has run it into the ground, made it progressively more inhospit...