I'm pleased to announce the arrival of Brian Ziming Johnson and Cody Zihong Johnson. They were born on July 14, both weighing around 7 pounds. Mom and both babies are doing well.
Full name is Scott Andrew Johnson; there are zillions of Scott Johnsons in the world, it seems. (Probably still zillions even if you include the middle name).
Also known as engineer_scotty (not the most original moniker, I know...but outside of a certain Star Trek character, nobody else seems to be using it. At any rate, my Program Manager at work has asked me why I multiply schedule estimates by a factor of four...)
Software engineer at Tektronix Inc since 1995, working in what is currently known as the Video Product Line. At work, I get to use (lucky me) Cee Plus Plus and Vx Works. That particular architecture will teach you Defensive Coding practices very quickly.
Have a MSEE, BSEE, BSCS, all from Oregon State University. CS degree was pursued as an afterthought. Since graduating, my career has been almost exclusively software.
Have a strong interest (outside of work) in programming language design (which is one reason I'm here at Wards Wiki). (Didn't study this topic in college; only taking one undergraduate course. Now I regret this...)
My latest PC (an Athlon 64), I'm sorry to report, comes pre-installed with Windows Xp. Course, I installed SUSE Linux Professsional 9.3 on it (a real 64-bit OS), and have booted Windows exactly once since - to make sure that when SUSE repartitioned my hard disk, it didn't scotch the Windows installation. (It didn't).
Also have never owned a PC with an Intel-made CPU in it; I've owned a long succession of AMD-made processors. From the original 40Mhz 386DX back in college, to the K6, to the Duron, to the Athlon, I've yet to regret it.
Married (wife Louise), three children (sons Alex, Brian, and Cody).
Enjoy good coffee, good beer, and bad wine. Good music and bad movies.
Like Wards Wiki because I learn something new every day... usually.
One of these days, I suppose I should put up a picture...
My preferred email address is Ee En Gee Eye En Ee Ee Are Ess Cee Oh Tee Tee Wye At Gee Em Ai Eye Ell Dot Cee Oh Em. I have gmail invites available, if anyone wants them let me know.
Also have installed Media Wiki at home as my own personal scratchpad... although my free time these days is scarce with twin babies in the house. :)
Scott, was Hostile Teacher entirely for me or am I just a paradigmatic example? LOL.
Actually, it's for both of you. No offense to either, but you and Costin have more in common than either of you probably care to admit. (That said, you were the author of Hostile Student IIRC, so there is an obvious parallel there). You two are probably two of the most intelligent persons on Wards Wiki (along with perhaps Doug Merritt and Jonathan Tang; if anyone else reads this don't be offended I didn't mention any other names here), as well as two of the stubbornest. Both of you are highly knowledgeable on a wide variety of subjects. Both of you are highly driven by ideology, and a firm sense of what is right and wrong. Neither of you suffer fools gladly.
And both of you have the tendency to think that the universe of fools includes (by definition) anyone who disagrees with you on certain topics. Which makes a debate with either of you an exercise in frustration, for those of us who want a good enlightened discussion of the issues but don't care to get dragged into a free-for-all. And which explains rather quickly why Bad Things happen when the two of you decide to tangle with each other.
Don't take any of the above as a flame; rather than take it as a Good Faith observation.
Yeah, there's lots of things that set off my hostility in a teacher mode. Failing to respect two out of three of the following almost always sets me off: my arguments as arguments, myself (posting anonymously), their own arguments. If they disrespect only one, I'm usually okay with that. But then there are people that manage to disrespect all three in a single paragraph.
Often because THEY take the Hostile Teacher posture. The world is their classroom, their intellectual stomping ground.
(And yes, Costin can call me an idiot and the antichrist all he wants, by signing he's showing more respect to me than most people do. Costin just doesn't respect anything else.) -- RK
Nobody thinks you're an idiot; Costin ought to know better. Irascible and obnoxious, perhaps. An asshole, certainly. :) But I suspect you take pride in those words. :)
Actually I don't recall calling Richard am idiot in a very long time, and if I did that long time ago, I should have known better. Richard gives me that permission because he knows I'm smarter than to exercise it, but then Richard is known for his cheap bravado. --Costin
Scott, aparently you don't have an email address on this page. But if you don't bother to expose this well kept secret to me, drop me an email and I'll promise I won't abuse it and afterward I'll make sure to secure erase the information from my computer and go under hypnosis to erase it from my brain as well. -- Costin Cozianu
Look up above. I just wrote it out in the Wiki Alphabet.... Ee En Gee Eye En Ee Ee Are Ess Cee Oh Tee Tee Wye At Gee Em Ai Eye Ell Dot Cee Oh Em -- Scott Johnson. And don't worry, the spammers already have it. :) Gee Mail is fairly good at filtering out most of the spam...
Scott I have seen you giving advice of Lambda The Ultimate as a good programming site. Do you have advice for a good application development / implementation / system integration site, preferably with a community keen on collaboration? -- dl
Unfortunately, I can't recommend any. Not that such doesn't exist (I'm sure it does), just that I don't particpate on an such forum that I can vouch for. -- Scott Johnson
It already exists; just use C++ and avoid the virtual keyword. (I also sent you email on the subject, Doug). -- Scott Johnson
Re: lambda-the-ultimate.org -- good for you, well said. You probably missed it, but perhaps 6 months ago, in talking to Pete about this sort of thing here, I suggested to him that interaction may add no power at all, but in any case couldn't add more power than whatever the TM interacted with. An obvious example is a TM talking to a quantum random number generator (very nearly impossible to ideally achieve, but skip that). The TM will gain in power very precisely -- it was unable to produce true randomness before, but now it can, and the mathematical implications of that are quite well known.
There's no reason, as you said, however, to think that all of a sudden anything goes, with interaction. Not at all.
LtU irritates me somewhat. Conversation there is often interesting, and the people smart, but as with this thread, all too often they're being smart in dumb ways -- or perhaps I should say, being simultaneously very smart but also very very wrong. -- Doug Merritt
Speaking of outsider perspective, the adage "can't see the forest for the trees" occurred to me. Especially so since you can't see a forest from inside it. And then I had a hunch and google confirmed that's one of those traits associated with autism .... -- rk
To truly appreciate a forest, one must view it from many different perspectives. From outside (especially from a distant vantage point, or from the air), where one may observe its scope and breadth. And, from the inside - where one may observe the details which are essential to the forest's being, details which are not discernable from outside. And not just from one place inside, but from many - complex and interesting systems, such as forests (and nontrivial software systems) are not homogenous entities which can be fully described and understood from a single observation.
Sometimes, a single vantage point is useful for some purposes; and in many cases - as you have pointed out - one doesn't have the luxury, due to opportunity cost and such - of performing a detailed examination of something before rendering a judgment or decision. But the cursory examination has its own cost - the cost of a greater probability of error.
Well, this analogy is proving unexpectedly fruitful. In biology there are many specialties. There are those who study the population dynamics of individual species. Then there are those who study ecosystems. And then there are those on the margins of biology who do biological analysis of satellite images. And there are so very many forests to study. And not just forests (boreal, coniferous, and broadleaf) but mountain ranges, swamps, bogs, tundras, plains, savannahs and deserts, without even going into littorals, coral reefs, and kelp forests. -- rk
Here's a question for you, does the write penalty for RAID 5 apply only to small writes or to large writes as well? In RAID 5 you have to read back the corresponding blocks on all the disks in order to write a single block. Now, if you write a segment large enough that it overwrites all of those blocks then there should be no write penalty. IF an implementation of RAID 5 is intelligent enough. But are they?
I ask for several reasons. Because if the write penalty does not apply then it means that RAID 5 is the hands down choice under a certain condition. That condition is that you use a logging filesystem. The LFS ensures that writes are always big so the penalty never applies. As a result, you don't have to sacrifice 50% of your storage for redundancy and you get all of the benefits of striping for free. Since LFSes have to be kept about 50% free to be operational, you sacrifice storage anyways, but if you go the LFS route you get a whole bunch of other things for free; like versioning.
The other reason is that it would vindicate my conviction that ReiserFS is an overcomplexified piece of evilly designed crap. ReiserFS can't feed large segments to the hard disk because it writes in place. I always knew that was idiotic and now it turns out to be a deal killer. ReiserFS is the epitome of Unix software. -- RK
Unfortunately, I'm the wrong person to ask both questions. I'm not that familiar with the dirty details of RAID - I know the basic principles, but for me to answer the question, I'd have to go do a bunch of digging. Nor do I know the details of how ReiserFS works... though keep in mind that it has to work with any reasonable hard disk configuration.
As for ReiserFS being the epitome of Unix software - not even close. It's long been controversial even with Linux; being as ambitious and Unix-architecture-stretching as it is. X11 is far more the epitome of Unix. Good and bad. Even Costin hates it. :) Now put that in yer pipe and smoke it! :)
FWIW, I'm running ext3; mainly because that's what SuSE uses as a default.
Would you say that Unix is stagnant or that its development is going in the wrong direction?
Do you have any suggestions where this issue should be moved to or what wiki name to put it under? It would be much appreciated.
Unix covers a lot of ground, and a lot of different systems (I'm using it here to refer to anything which is POSIX-ish; not just systems which derive somehow from the AT&T codebase). That includes all branded UNIX, BSD, Linux, Darwin Os, Be Os, and quite a few other things. It's hard to make a general claim about a group of systems this diverse. I'll pick on Linux, 'cause I know it best.
Linux is "stagnant" in that it isn't all that theoretically interesting. The Linux developer's community (which includes Linus Torvalds obviously; though how much he himself can bring about changes in Linux is questionable) is primarily interested in evolving a stable platform in a way which brings incremental improvement to, but without severely disrupting, a rather large installed base. Along the way, they desire to encourage migration from the Windows platform, an architecture which is dissimilar to Unix--but not radically so. They seem to be succeeding at this quest; according to their own goals they are moving in the right direction.
By the goal of advancing the "state of the art"? They don't advance it much, other than in incorporating implementation improvements to existing designs, as well as "sucking in" new technologies that have been proven out in research applications. Linus (and Linux) does encourage some experimentation - and Linux has been used as a research vehicle for many things - but if you want to study radical new OS architectures, Linux is not the place to do it.
What wiki name? Depends on the spin you want to put on it. A pro-Unix name might be Unix Isa Stable Platform; an anti-Unix spin might be Unix Is Obsolete. You're creative enough Richard that I'm sure you can come up with a suitable (and inflammatory) name without my assistance. :)
I'm sorry for being unclear. The stagnation or not of Unix relates to the representiveness (damn, that's not even a word) of X Window versus ReiserFS. But this is only a minor concern to me. Of more concern is the write penalty in RAID 5. -- rk
I believe you meant to say "representilliousablefull". :-) -- Doug
See original on c2.com