Apple, why did you not ship me a 64-bit compiler, and for god's sake, a 64-bit perl with my 64-bit desktop? Do you know how embarrassing this is?
thunder% perl schwartz.pl.txt words
perl(29599) malloc: *** mmap(size=99618816) failed (error code=12)
*** error: can't allocate region
*** set a breakpoint in malloc_error_break to debug
Out of memory!
thunder%
Oh, oops, looks like you've got a thirty-two bit perl, Mr. Avriette, here's 2gb of memory. Go run along now and play, just don't try to ask for 2.00001gb, or your ass is getting canned. It's like speeding or something. Try to do too much and, zam, the gumballs come on, a guy in a hat comes up to you and gives you some garbage about what you were doing not being safe when, clearly, you're both still there (so, it's not unsafe, right?), but you're not allowed to continue along 270 at 140mph. Even though you're fully capable of it. It's a sixty-four bit Xeon for crissakes. Gr. It's the perl equivalent of an unrestricted autobahn.
And, just so nobody gets on my case, yeah, I did. I downloaded the source to build my own 64-bit perl, and discovered that Apple didn't give me a 64-bit gcc, so in order to do this simple task, I'd have to bootstrap a 64-bit compiler, which is traumatic in and of itself, and then I'd have to go along and build my perl.
Apple, sometimes you make me very sad.
07 April, 2009
06 April, 2009
Data Architecture
I love seeing the title Data Architect. See, it helps if one considers "data" as a big gelatinous mass of stuff from which we want things. So for one to be an "architect" of data, one has to grab hold of data as it is produced, and stack it into neat little piles from which our "things" can be extracted later, easily.
Unfortunately, we see Database Administrators using this title more and more because, well, I think the term "Architect" makes people feel very proud of themselves and all their myriad achievements. Ironically, using the title indicates a sad misunderstanding of what data, and databases, really are. A Database Architect, or even Schema Architect, is somebody who decides how data should be organized, once they are presented with that big gelatinous mass of stuff. This is where people use tools like perl, ruby, and even (cue twilight zone noises) awk, sed, cut, tr, and so on (ok, stop with the soundtrack) to format the stuff into the schema which the architect has designed. This is a good process because it means I can get my things out.
I personally love getting data in that giant mass of goo and forming it into something useful, gathering metrics about it so I know how much storage I need, and how my database is going to perform, and so on, but I'm kind of a perv in that regard. What I really wanted to get at, though, was that there's another side, and a title that might be more appropriate than "Data Architect" – Data Archaeologist or Data Archivist. But, please, people, stop tacking "architect" onto your title in places that puff up your business card. It's kind of like kids with Honda Accords and loud mufflers.
"...what's that? I can't hear you over the sound of how much data I'm architecting!"
Unfortunately, we see Database Administrators using this title more and more because, well, I think the term "Architect" makes people feel very proud of themselves and all their myriad achievements. Ironically, using the title indicates a sad misunderstanding of what data, and databases, really are. A Database Architect, or even Schema Architect, is somebody who decides how data should be organized, once they are presented with that big gelatinous mass of stuff. This is where people use tools like perl, ruby, and even (cue twilight zone noises) awk, sed, cut, tr, and so on (ok, stop with the soundtrack) to format the stuff into the schema which the architect has designed. This is a good process because it means I can get my things out.
I personally love getting data in that giant mass of goo and forming it into something useful, gathering metrics about it so I know how much storage I need, and how my database is going to perform, and so on, but I'm kind of a perv in that regard. What I really wanted to get at, though, was that there's another side, and a title that might be more appropriate than "Data Architect" – Data Archaeologist or Data Archivist. But, please, people, stop tacking "architect" onto your title in places that puff up your business card. It's kind of like kids with Honda Accords and loud mufflers.
"...what's that? I can't hear you over the sound of how much data I'm architecting!"
05 April, 2009
Complaining, loudly, from the monkey's head.
Once upon the time, at the foot of a great a mountain, there was a town where the people known as happy folk lived. Their very existence a mystery a world, obscured as it was by great clouds. Here they played out their peaceful lives, innocent of the litany of violence and excess that was growing in the world below. To live in harmony, with the spirit of the mountain called "Monkey" was enough.
The one day, a man came along and plugged his amplifier into the mountain. A strange buzzing began emanating from the mountain, sometimes camouflaged in the distorted amplifiers of other musicians who sought distortions in their amplifiers, to get the "tortured 303" or "overloaded 808" sound onto their wax. But without actually testing the amplifier independently of the mountain, the happy folk had no way of knowing that their very own amplifier was suffering from poorly made capacitors and a fundamentally poor ground channel. Poinz has a lot to say about this, but all I can say is ignorance is bliss, and I will only quote a very small bit of what he has to say here:
Well, I have serious ground hum in my amp, and I have serious lack of quality in my capacitors. The good news is I am looking for a headphone amp, so I needn't worry about anything too medieval and too chaise du électricité. The bad news is, I'm probably screwed if I want good ground and capacitance.
For the first, right now I have two choices: the wall, which is what most people use, except I am using either my computer or my iPod for audio, and as such I use an amplifier in the middle. The amp itself gets its ground from either a) the wall, as when it is at home, b) USB (when it is mobile) or c) AAA-sized batteries. You may be surprised to learn that the cleanest ground I've gotten so far is from the batteries. I think this is because my building was wired by monkeys, and nobody figured the preamp outs on the MacPro for audiophile quality (you'd be using digital-optical, not 1/8", right?), and really slacked on the ground there.
For the second (if you've lost track already, we're at capacitors), good caps are both expensive and dangerous. The better they are, the bigger they get, the hotter they get, and the better they are at killing you. So I'm kind of peeved that it's not likely that I'll be able to get capacitors of any decent breeding in a portable amp – such as the amp that has so tragically failed me, the Headroom Total Bithead. I think, with my new soldering table and everything, I may actually start doing the DIY-DAC-PIMETA amp thingie and finding a portable case for it. Or, maybe, find one that's good enough to be used both on my desk and in my pocket on the metro. Maybe I'll even have a special discharge probe I can use as a taser from my caps. :(
If you love me, and you've already paid my rent, and you've already bought me that pair of Sennheisers (the reason I'm complaining is I just "tuned up" Sandy's Shure's any my Ety ER4's, and I'm hearing noise in my amp, and I am a sad panda), and you've also already bought me that amplifier for my desktop, please just buy that portable amp that's built out of better components than .. well, I'm not going to insult; they got me onto audio-hi-fi, and I now know what ground and capacitance hum is because of them (well, and Poinz), and for that I can't throw stones. But, please, buy it for me, so I don't have to build my own amplifier.
Do it for America. Do it for 9/11. Do it for Obama. He'd want me to have a better amp. You know he would.
The one day, a man came along and plugged his amplifier into the mountain. A strange buzzing began emanating from the mountain, sometimes camouflaged in the distorted amplifiers of other musicians who sought distortions in their amplifiers, to get the "tortured 303" or "overloaded 808" sound onto their wax. But without actually testing the amplifier independently of the mountain, the happy folk had no way of knowing that their very own amplifier was suffering from poorly made capacitors and a fundamentally poor ground channel. Poinz has a lot to say about this, but all I can say is ignorance is bliss, and I will only quote a very small bit of what he has to say here:
Two things are probably occurring to you right about now. First, that the proper functioning of all this single-ended system depends very heavily on the net resistance of the ground reference being precisely zero ohms (so that no voltages are developed in this leg of the circuit); and that this is maybe not precisely the case. The second is that, the preceding being true, this grounding deal is starting to look like a real can of worms. Right on, bro or sis; that queasy feeling you're having about the grounding of your piece is bringing you onto common ground with people who have many years of schooling in this discipline. There's also a kicker here; in that it's important, both for operator safety and to inhibit the ingress of RF nasties into the circuit, to have the box that it's mounted in grounded as well; this now becoming also a part of the ground reference.
...
The ground bus technique is a very close electrical approximation of star grounding that avoids the million miles of ground wires everywhere. The ground reference is a nice fat (I use 12 gauge Romex conductor) wire that runs the length or width of the chassis, to which individual circuit and PS grounds can be brought at convenient points. This is then brought to chassis ground as before; usually starting at this point. I start mine at the rear chassis wall, attaching it to the bus that picks up all the ground sides of the input RCA's. The output transformer secondary grounds and the (also fat) wire to the chassis, which is attached to an output tranny mount, attach right here also. The bus runs straight forward through the chassis, ending up at a terminal strip that's mounted to a front screw of one of the input tubes; about an inch from the front wall of the chassis. All the power supply grounds are gathered at the ground terminal of the dual 100µF PS cap, and then brought to the bus by a piece of 16 gauge stranded (a cutoff of the line cord). I tried positioning this major pickup experimentally, and could not measure or discern any difference at any reasonable position. Try it yourself. The ground bus should not be brought to chassis at more than that one point. This creates a circuit (loop, remember) called a ground loop, which can develop considerably more current (and noise) than the single-ground bus. That is, the bus (between the two chassis ground points) forms one leg of the circuit, and the chassis (between the two bus ground points) the other. A great deal of the troubleshooting in pro-sound and recording studio situations involves finding and eliminating ground loops.
Well, I have serious ground hum in my amp, and I have serious lack of quality in my capacitors. The good news is I am looking for a headphone amp, so I needn't worry about anything too medieval and too chaise du électricité. The bad news is, I'm probably screwed if I want good ground and capacitance.
For the first, right now I have two choices: the wall, which is what most people use, except I am using either my computer or my iPod for audio, and as such I use an amplifier in the middle. The amp itself gets its ground from either a) the wall, as when it is at home, b) USB (when it is mobile) or c) AAA-sized batteries. You may be surprised to learn that the cleanest ground I've gotten so far is from the batteries. I think this is because my building was wired by monkeys, and nobody figured the preamp outs on the MacPro for audiophile quality (you'd be using digital-optical, not 1/8", right?), and really slacked on the ground there.
For the second (if you've lost track already, we're at capacitors), good caps are both expensive and dangerous. The better they are, the bigger they get, the hotter they get, and the better they are at killing you. So I'm kind of peeved that it's not likely that I'll be able to get capacitors of any decent breeding in a portable amp – such as the amp that has so tragically failed me, the Headroom Total Bithead. I think, with my new soldering table and everything, I may actually start doing the DIY-DAC-PIMETA amp thingie and finding a portable case for it. Or, maybe, find one that's good enough to be used both on my desk and in my pocket on the metro. Maybe I'll even have a special discharge probe I can use as a taser from my caps. :(
If you love me, and you've already paid my rent, and you've already bought me that pair of Sennheisers (the reason I'm complaining is I just "tuned up" Sandy's Shure's any my Ety ER4's, and I'm hearing noise in my amp, and I am a sad panda), and you've also already bought me that amplifier for my desktop, please just buy that portable amp that's built out of better components than .. well, I'm not going to insult; they got me onto audio-hi-fi, and I now know what ground and capacitance hum is because of them (well, and Poinz), and for that I can't throw stones. But, please, buy it for me, so I don't have to build my own amplifier.
Do it for America. Do it for 9/11. Do it for Obama. He'd want me to have a better amp. You know he would.