Wave Fronts

Black Thursday for Engineering?

Hoping I'm wrong

  (found on the 'Net while searching for the
    real psalm. Adapted slightly.)

  ***********
  Psalm 6.111
  ***********

  The FlipFlop is my Synchronizer, I shall
   not want.
  He maketh me pipeline my whole circuit,
  He leads me to program more EPROMs,
  He restores my sanity.

  He guides me in paths of synchronicity
   for his clock's sake.
  And lo, though I walk through the valley
   of the shadow of CMOS,
  I will fear no intermediate voltages,
   for He is gated,
  His one or His zero,
   they comfort me.

  He prepares a haven for me
   in the presence of metastability,
  He annoints my design with blocks,
   my logic shall rationalize.
  Surely good reviews and no recalls will
   follow me all the days of this fiscal
   year, and I will create the right buzz
   forever.

----

Wishing all of you safe passage,
dear friends at Sun.

Posted at 12:00AM Jun 22, 2006 by microwaves in General  |  Comments[0]

Intelligent Multiboot Support for Windows, Linux and Unix

Multiboot selection should be possible with a single command line or GUI, trivial to accomplish locally or remotely, and generally available.

There ought to be a way to trivially utter to my computer:
Reboot with a boot menu choice like "3:5" where "3" means "first menu choice three" and "5" means "first menu choice three's fifth menu choice."

Why? I'm sitting here in the living room with a Powerbook, and my Sun Ultra-20 is sitting nine feet below me running Solaris 10. Should I have to get up, hike to the basement, do an init 6, and step through the above choices manually just so I can run a command that is only available to me through another operating environment choice?

Surely the keepers of Grub and its Windows multiboot equivalent can agree to about three bits of shared state to support this feature? Guys: how about lunch in Portland for a nice chat and agreement that a command like "init -- 3:5 6" will just do the right thing? Of course a remote desktop would need an icon/dialog set that you click a few times (no keyboarding, please). Do this and the world will thank you. And you can spend the lunch just getting to know each other!

This would serve well until virtualization becomes so easy and available that something like the following can be uttered:
as

for example:

javac | `as ddeal/FC4 tac` | less

Posted at 03:47AM Jun 18, 2006 by microwaves in Operating Environments  |  Comments[0]

Belated OpenSolaris Birthday Celebration

OpenSolaris is the next OE on my list to learn about.

I couldn't make Mozilla speak IRC at the office and couldn't take the time to go find something that worked. So I missed the celebration of the first birthday of OpenSolaris. Here's a progression cum evolutionary tribute.

This is an incomplete and not quite correctly ordered list of the operating environments I've spent a reasonable amount of time using or that I had a hand in writing:
Univac Exec VIII
Univac APL
Unnamed OE (I wrote this on the bare metal of a Data General Nova 1200 while we [Univ. of Alabama Huntsville Isolation Lab] were waiting for our Nova 2+software order to be filled. It used a decent quality 300 baud cassette tape drive of some sort and a very nice 300 baud thermal paper terminal)
Data General RTOS
Data General RDOS
Data General timeshare doc system
Data General AOS
Data General Easy (DG too stupid to ship this)
Texas Instruments DX10
Business Application Systems BASPort
(Wrote part of one VM, one native assembler, one native linker, a native debugger, and various other bits and pieces long forgotten. Other members of the BASPort team included Steve Goldman, Bob Leivian, Don Parce, George Franzen, and my name memory just collapsed. Forest Earl Gilmore managed us [RIP, dear friend].)
SWTP 6800 MIKBUG
SWTP 6809 UNIFLEX (+ Bob Uiterwyk's Basic: Bill's Altair Basic blew chunks)
MSDOS
DRDOS
Unnamed OEs inside Network Products BabyMux, BabyNet and Commponent products (I wrote these with Steve Schleimer and others)
Version 7
System III
Opus Unix
BSD
SVR4
Mach
Linux
Solaris

And now, at long last, OpenSolaris is well and truly in the world. Some times you have to be patient before you end up with something really, really good designed for the long haul. I'm looking forward to running OpenSolaris on an Ultra-20 in my basement as soon as I can. But the download's going to have to get easier.

So:
Happy birthday to you,
Happy birthday to you,
Happy birthday OpenSolaris,
Happy birthday to you,
And many morrrrrrrrrrrre!
21

Posted at 11:04PM Jun 15, 2006 by microwaves in Operating Environments  |  Comments[1]

Software Defined Radio Use Case

Smarter elimination of SSB adjacent signal interference

I've been a licensed radio amateur for 11 years and I'd only had one interest as far as application and that was HF radio contesting. HF is short for "high frequency" and covers about 1-30mhz with 10 amateur bands of frequencies within this range. HF contests involve trying to make more contacts in a fixed amount of time where each contact is short to very short, the jargon equivalent to "I'm Pete and my favorite color is violet." It's actually like "F5XY 59 5" spoken very quickly or "F5XY 5NN 5" by morse code ("N" is contester short hand for "9" as dah dit is shorter than dah dah dah dah dit).

But after moving house and deciding against another tower and "serious antenna system" I realized I needed a new interest. And I need a reason to be in the "radio room" (the back of a detached storage building) as this is where my daughers electronic tinker toy set is in position to keep all the bits and pieces from going to never land. I've settled on exploring something called a "software defined radio" and the rest of this is about my first application goal.

I recently took part in the ARRL phone(voice) "Sweepstakes" in which you can make a lot of contacts if you can send and receive signals well and can stand to pick out other voices in a cacaphony of noise. Really, a major phone contest is just weird. I wish I had an mpeg to link to to give a sense of what it's like. Basically you're soliticing contacts, saying effectively "please let me tell you my favorite color" while somebody just below you in terms of frequency and somebody else just above are doing the same, hammering your ears with interference relative to your goal.

With single side band (what we use for "phone" contests) it's like an old style telephone party line: all voices can be heard at the same time, but shifted up or down in tone. This is very different from AM and FM in which one signal steps on or completely displaces another. So with side band you've got somebody soliciting contacts below you but their audio is shifted up or down by the difference in radio frequency of their signal relative to yours and also your listening frequency. The net effect is that it's kind of like trying to carry on short conversations while Donald Duck and Tweety Bird are yammering in your ears. And sometimes the folks responding to their solicitations are also audible. In some cases the soliciting signal is louder than their responders, other times its visa versa or about the same loudness. The loudness is proportional to signal strength, ignoring automatic volume control factors.

So I started wondering if it might be possible to apply a computer to a situation where even state of the art amateur band receivers can't get the job done. That is, if I can hear and recognize the lower pitch and higher pitched Donald Duck and Tweety bird interference, couldn't a computer do it too? Computers with DSP (digital signal processing) are extremely effective for locating and notching out a single tone, such as is frequently caused by a continuous carrier produced by another station. So one moment you've got a killer tone in your ear and the next, after you push the "heterodyne filter button" it vanishes.

So why can't this scheme be generalized to cause Donald Duck to vanish from my ears? Even if I could just get the level down a little it would be great.

So this is my first goal: to try to arrange for "alien" audio components within my passband to be attenuated while leaving "friendly" audio tuned to my frequency alone. I have no idea if this is an impossibility or if it has been done already.

Unfortunately, the GNU Radio Universal Software Radio Peripheral I bought to explore this area seems to be defective. I haven't had the time to discuss this with its maker. I'm hoping it's something simple, or maybe something that is at least repairable, like a cracked solder joint.

And, in the meantime, I've broken through self-consciousness with the fumble-bumble way I do morse code (heavily depending on computer assistance), and I find CW contests to be so much easier than phone, I might put off this SSB-research for some time. And in a CW contest you can eat and drink while participating, something that's very hard to do when you're making a lot of contacts per unit time for hours on end.

Posted at 09:00PM Jan 06, 2006 by microwaves in Radio  |  Comments[2]

Thoughts about God's Debris

A few weeks back something led me to Scott Adam's free book God's Debris ^[1], and I gave it a quick read. I mentioned it to my best friend Steve at work and then forgot about it. I just read the preface and first pages of the story again, and this blog is about my first "discussion point."

About a year ago, I'd spent many weeks reading and pondering the first half of a highly speculative physics book by Frank Tipler (Physics of Immortality, the second half of which is for real scientists and mostly impenetrable). Steve lent it to me and when I brought it back from a trip to Denmark, dog-eared and with some notes added to the back, I was eager to discuss it with him systematically. We didn't do that but instead spent an hour or two dancing around some of Tipler's major points and agreeing he'd gone over the top and that his earlier and less popular (style) book written with with John Barrow (The Anthropic Cosmological Principle) was more enjoyable, albeit slower reading (unfathomable math and physics concepts interlaced with understandable text rather than the sectional approach of POI). So I guess we did discuss it, just not the way I expected to.

Anyway, Adams suggests readers discuss God's Debris to compare notes about fallacious remarks made by the book's characters to "try to figure out what's wrong with the simplest explanations." At second glance, I was struck by Adam's mention of the tension between "most believable" and "true" characterizations of "smart sounding answers" and his observation that while the skeptic's creed says the simplest explanation is usually right his experience has been that the simplest explanation is usually dead wrong. This matches my experience and in my business is perfectly typified by the fact that sometimes I can't properly explain anything technical to management using the number of words they allot me for the task. So while in a recent "Dilbert" strip in which a full day's productivity was defined as "eight slides", I've been working on just three short bullets of one slide for a statement of work to be proposed for the next major version of Sun's Java SE implementation code named Dolphin. But behind those three short bullets is a lot of thinking and verbal spewage into online notebooks. In fact, I could write several pages about each bullet while admitting that I'd just scratched the surface of each topic. I guess one of the Java Gods^tm could write bullets that don't cause management to respond with "what do you mean by that, exactly?"^[2], but I still have a long way to go with slide composition.

So I started to read God's Debris again, and before I could get to any heavy points to ponder, I read this opening remark by the main character, a small package delivery truck driver:

"My story begins on a day I delivered to a place I'd never been. That's usually a fun challenge. There's a certain satisfaction when you find a new place without using the map. Rookies use maps."

Bzzzt. My BS meter pegs as I mentally add the missing sentence "Of course, there's a certain dissatisfaction when you find you misjudged the location, are on the wrong side of a city with serious traffic, and realize you'll be delivering packages well after dark and explaining to customers what went wrong."

Surely an experienced guy^[3] integrates the familiarity of the address to drive to with a hundred other factors and is likely to use a navigation tool part of the time with the "no assist challenge" approach being stochastic. So the character's remark seems aimed at leading rookies to the wrong side of town rather than capturing the real difference(s) between a rookie and an experienced delivery guy.

OK, I guess that was practice for spotting the BS disguised as profundity that Adams has cleverly crafted into this book. This item seems more trite the longer I consider it while hopefully the opposite will be true of the hidden rhinestones. But I'm also beginning to see this book as a tool to help address one of Adams's obvious concerns that is captured by the combination life observation and book title When Did Ignorance Become a Point of View?

And I realize I don't have time to read this book three times and can't be bothered to write down chapter and verse. So I'd better do this second pass the old fashioned way with hard copy and a highlighter while I look for somebody to discuss it with. Hmm. Is this why some folks break down and buy an Adobe Postscript/PDF tool? If it makes creation of a marked up version of the book trivial, it would surely beat using up fractional trees and real yellow ink. But maybe the PDF can't be modified, and I can carry the paper and highlighter in my pockets while my Powerbook is too large and delicate. And this makes me recall the utter disappointment of my Zaurus and what it actually does vs what I'd hoped it would do for me by now. Sigh. With luck I can think of something more cheerful to write about next time.

[1] Warning: Adams cautions readers that this book is not for young minds and recommends against reading it if you're under 14. I personally take this warning very seriously, having had a very bad experience with existential overload that took me many years to adapt to. If you're too young and reading this, please don't race over and read Adams's book because he specifically warns against it. With luck it would just go over your head, but it might also create so much confusion that some part of your life is seriously disturbed while you get yourself sorted out again. There is likely plenty of time to make your head spin like a top once all the parts are firmly and prop erly in place for the spinning. As the saying goes "you've been warned."

[2] I don't know how the Java Gods^tm view themselves but I can't get past the sense sometimes that my feet are playdoh just pretending to be clay, and my head frequently feels as though it's full of tapioca pudding.

[3] I was born in Michigan where "guy" is a unisex term; sorry if this bothers you.

Posted at 01:00PM Dec 31, 2005 by microwaves in Books  |  Comments[1]

Intro

"My Name, is of No Importance"

As a kid I learned about interference and reinforcement of wave fronts and more recently I read Frank Tipler's "Anthropic Cosmological Principle" with its references to wave functions as universal expressions. Then the other day I came across a DVD of the BBC TV series "Hitchhiker's Guide to the Galaxy" with its single, hilarious number "42" as the answer to, well, you know, the big question. So I expect that what I write here will, in a microscopic way, reinforce or inhibit what's going on in various areas while contributing to the really long computation that planet earth may be carrying out if Douglas Adams turns out to have been a historian...

At Sun I fix bugs.

Legal notice:
|°373 47 $0p3® Ð07 µ$
Is my personal email address in a dialect of l33t.

Posted at 12:00AM Oct 20, 2005 by microwaves in General  |  Comments[0]

Home communication

A first installment about my home communication setup

My home equipment related to communication looks mostly like the diagram below. It's "mostly" because since making this graphic there have been a number of changes.

Everything but the FM transmitters and Linksys storage box (NSLU2) has been running for some time, with the latter two items still "in development." The two adventures so far have involved trying to use the Motorola WE800G in pairs bridges and tuning the FM transmitters.

The WE800G is similar to the Linksys WET11 and Dlink DWL-810 (the three specific models I've had direct experience with). I'm convinced he majority of users put a bridge like this in place as a means of connecting a wired ethernet port to a wireless network. That is, the bridge connects a wired port to a wireless access point or bridge. I'm pretty sure a miniscule minority of users put a pair of bridges in place as "wireless CAT5 cable" to connect two networks or network segments togetgher. I think this is true because both the WET11 and WE800G seem to be extremely unreliable for the latter application. I'd been warned about the WET11 and didn't waste much time on them, but the much more recent WE800G design and an initial successful two day trial led me to believe they would be solid. I was wrong and after buying a third to try to eliminate failed hardware as a factor and after countless experiments I gave up on them and redesigned my network so I didn't need a pair of bridges. Although they don't show on my network map I have a pair of the DWL-810 bridges but I haven't been tempted to see if they are reliable in 'wireless CAT5" mode. I should mention these three bridges are all cheap to very cheap and I never found a moderately priced alternative (say in the $250 range) that offered prospects for rock solid performance. The prices seemed to go from $50-150 to $500 and up. Taking hassle factor into account I probably should have payed the bigger bucks.

The FM transmitters (a British Veronica model) are wonderful except they weren't designed with current receivers in mind. It's straight forward to tune them to an unused portion of the band, but it's beyond tedious to tune them precisely enough that a receiver with digital tuning receives their signal properly. Tuning is by a compression type variable capacitor that is beyond touchy. So the two old GE radios in the house and the clock radio work fine, but the home stereo and walkman type radios are useless. Even supposing I can make the tuning tool I need, the frequency stability is questionable. Veronica is absolutely straight up about this, saying 25khz is what you get. So I have to tweak analog tuned receivers from time to time to get them back in line with the transmit frequency. If I'm lucky and can get the tuning just right after allowing temperature to stabilize it might stay within the capture range of a receiver for the long term. But to go along with an exaggerated T-handle tuning device I may need something akin to a crystal oven for the transmitter to try to control the temperature and make retuning less frequent. If this sounds like a hassle you're appreciating the situation.

The other factor is that I want coverage around the house. The Veronica claim is 100 meters range and I've confirmed half of that range with transmitter and receiver on the same level, but most of the house gear is in the basement while most of the living is on the main floor. Now in most parts of the world it would be a trivial issue to just run a shielded cable from the sound source up to the transmitter. However here in the southeast USA lightning storms are a chronic issue and the property here is on relatively high ground. So I have to weigh the risk of blowing up the critical bits and pieces of house equipment for the sake of getting the FM antenna up to a proper position. Actually, increasing the chances of preserving equipment with lightning storms is a whole other story.

Posted at 06:54PM Oct 08, 2005 by microwaves in Home Nets  |  Comments[0]