If they're checking the software design for inconsistencies, then they're too late. What is needed is some way to formally specify user requirements, so that they can be checked for completeness and consistency. Use cases are nice, but they're not sufficiently rigorous to capture absolutely all the requirements. I know there have been some schemes tossed around for requirements validation, but none that I've seen have really been general-purpose enough for the average project.
I've got one of their standard 104-key models, and it's quite nice. Not quite as nice as my Avant Stellar (the reborn Northgate OmniKey), but certainly worth the money. I tend to spend more money on my peripherals, simply because I keep them longer than I keep computers.
It's just too bad that Unicomp's Linux keyboard (with the control key to the left of the A, where $DIETY intended) isn't available with USB. They do custom orders, though, hopefully it wouldn't be too expensive to get that layout.
nobody is able to find out you're tuning to that station, unless they actually hear you
Not quite true, a receiver must be tuned (and therefore have an oscillator resonating at) the desired frequency. If you've got a highly-sensitive receiver, you can determine pretty accurately the frequency a given radio is tuned to. You can't really cover much of an area, but if someone wants to know what station you're tuned to, and they can get close to you (say within a hundred feet or so), then they can tell what you're listening to. Google up "TV detector vans" for more details.
the transmitter needs to have an huge (like mounted on a house huge) antenna to transmit large distances
You're partially right. To get the greatest distance, you'll want to use the lower frequencies (say 80 meters). The lower the frequency, the longer the antenna needs to be (to transmit efficiently). The center of the 80M band is ~3.75MHz, so you'd need an antenna that's (468 / 3.75 = 124.8) ~125 feet long. If you're talking a standard dipole (and we are), then each leg is going to be 62.5 feet long. Each leg is just a hunk of wire, and can easily be rolled up. A complete 80M antenna with a 100W transmitter can be stuffed in a small backpack and hiked up into the hills somewhere (the higher the better). Then the antenna can be unrolled and set up almost anywhere (use a slingshot to hurl the center section over/into a tall tree and fan out the legs, or drape it over a wooden fence or even some bushes). Just because it's a long antenna doesn't mean it's really physically *big*. And there are plenty of other, smaller (and less efficient) antennas that would probably get the job done.
Or you could just get one of these . Doesn't fit in a wall socket, but it's a full-fledged thin client (doesn't require a "terminal server" [how quaint!] environment). Also doesn't come with keyboard/mouse/monitor (supports DVI or VGA). Not sure what the advantage of either of these is, but I guess if you need them, they're nice to have. Actually, the one I listed supports smart cards, so it's handy if you move around a lot (your session gets switched to whichever node you plug your card into, kind of like using VNC). For $250, you might be able to make a case for thin clients, but for $400, it's a bit tougher.
Does he seriously think it's a good idea to have somebody who's new to the language writing production code in that language?
I don't.
Worked for me. I was hired for a position almost identical to this: established company with a VB-based product (three, actually) needed a developer to maintain and enhance existing systems and design and build new ones. I had no VB experience, but I had about five years of experience in the field (pharmaceutical R&D). I also had written production code in three or four different languages previously. I picked up VB (16-bit VB4 at the time) and ran with it with no problems (although I did write the 16-to-32-bit VB translator in Perl). People with experience in the application domain will probably be more valuable to the team than a language guru who doesn't have any insight into the problem he's being asked to solve.
We had some gradual students who were collating and uploading data from remote sensors (basically a Z-80 box with a modem and a dozen or so A/D converter hooked to various weather instruments). They would have to use tip(1)[0] to call the sensors, then download the files and reset the sensors. And if they didn't do this at the same time every day, then they'd have to dick with the data file to get the readings to line up with previous captures. So I hacked tip(1) to accept scripting commands from a file (fairly easy, it already had built-in commands, it just couldn't read them from a file), wrote a cron job to kick off a shell script, and *boom*: two freshly-unemployed grad students! [1] Not as cool as wiring up a current-loop-BAUDOT-to-RS232-ASCII converter and stuffing NWS reports directly into a newsgroup, but that one wasn't mine....
[0] Why yes, this was the early 80s -- how could you tell?:P [1] Not really, they were actually being paid to analyze the data, not collect it.
Yes -- that's their problem. Face it, no matter how elegant or wickedly clever your algorithm, or how well you manipulate the intrinsics of whatever language you're using...it all gets executed as assembler in the end. And unless you know how assembly language works, you can't make it work for you. Once you learn it, it's like scales fall from your eyes, and those seriously cryptic error messages (the ones that show up just when you think you've finally got your program finished) suddenly make sense.
Yeah, I had to disagree with that one too. I had a SCSI ZIP drive, and it was fast enough (on a 450MHz P-II mind you) that I could work on it directly (as opposed to just copying files to/from). 100MB was plenty of space for my projects, and it made it easy to swap them with my cow-orkers.
Hmmm, that reminds me, I'd better back up all those old disks before I decomission my old PC and can't access them anymore....
In about 20 years I expect a metric shit-ton of blackmail material will be available for our future up-and-coming politicians.
Nope, there'll be an Act of Congress declaring all such material "inadmissable" for the purposes of political discussion. You won't be able to go back farther than seven years when dredging for dirt on a candidate. Any non-felony convictions will be covered, as will any blogs, comments, web postings or letters to the editor. Everyone gets the "I learned from my mistakes" get-out-of-hassle-free card, not just conservative right-wingers.
You can't really have a remote antanna, per se, because you still have to generate the power (not just the signal) to send it. What I have seen is computer-controlled rigs that have software the present the complete control panel on-screen, and can route the audio through the sound card. So what you'd have is a complete transceiver/antenna setup somewhere, then access it remotely across the internet. One such set-up demonstrated at a local ham radio club meeting even had a little diagram showing which way the beam antenna was pointing, overlaid on a map of the world which showed what countries were along the beam's path.
It's especially nice if you have a vacation home in a place where you can put up a Big Honkin' Antenna (tm), then access it from your desk at home. I'd like to do it just so I could operate during lunch here at the office, but I like old tube rigs, and they weren't making computer-controlled transceivers back in the 1960s...
Yeah, right. Not even the Dells in this price range have separate graphics cards. Good luck finding a 2Ghz dual-core laptop for a grand or so without integrated graphics. Hell, good luck finding a 2GHz dual-core laptop in this price range, period. To get the equivalent from Dell, you have to get a high-end Inspiron, and that'll set you back more than $1500. Of course, with that you'll get a 17" screen, which is cool, but if you want to stick close to the $1200 price point, you'll have to settle for a 1.66GHz Core Duo.
I was the lead developer on a team that converted from VB to Java. For us, it made sense, as we wanted to web-enable our app and VB's web support at that time (VB5, back around the turn of the century) was too simplistic for us. It was a big challenge, and lots of hassle, but we got it done and it's been a pretty successful app in its space.
I can think of several reasons why your IT manager wants to make this change, and none of them are good. For us, it made send to make the move to the OOP Web world, because that's where we wanted to take our software. I really don't see the upside moving from C++ to Java. It sounds like offshore Java developers are cheaper, and that's the big incentive.
Bullshit. I've owned a Porsche for the last twelve years, and they take no more maintenance than any other car I've owned. Well, unless you count the extra work I do on it to get it ready for track days, but I'd have to do the same work on anything else I took to the track.
And twelve years and over 100K miles later, it looks a hell of a lot better than my last car; in fact, it looks a hell of a lot better than most of the cars I pass every day.
The fact that your buddy is a fair-weather driver says more about him than the type of car he drives.
it is just clear that Scientific computing is not the forte os MacOSX
Well, now that SGI's gone tits up, they might try pushing harder into the scientific visualization market. I know lots of physical chemists that would love to have some OpenGL love for their MOL files...
Inventors of what is arguably the best pointing device known to man, the roller mouse (PDF warning). No space wasted on track pads and sissy "palm rests" (you're not supposed to be resting your wrists while typing!), no need for desk space for a mouse or trackball, just a small roller and a couple of buttons. I couldn't believe it when Tandy (who bought GRiD Systems) dropped this from their lineup.
OSX does poor [sic] with database products which leads me to believe raw disk io sucks
From what I understand, the biggest problem with OSX is the interrupt overhead. There are so many layers (disk driver to device I/O subsystem to Mach to BSD I/O to BSD kernel to application) that performance suffers. It looks like every release improves this, but no matter how much you shine that Jetta it'll never be a Porsche.
They get to watch their power meters run backwards!
Heh. Most of the time, the power co just dumps the power into a big load (read: resistor), because it's not worth the trouble trying to phase-shift the consumer-generated power to sync it to the neighborhood supply. Especially given the fact that the frequency can vary with load (e.g., during the summer, your power frequency may drop to 55-56Hz at 5:30 when everyone's running the A/C and the stove while making dinner). At least, that's what I was told when I was considering a large UPS/generator I was offered cheap in Feb of 2000...
Slashdot Mirror
If they're checking the software design for inconsistencies, then they're too late. What is needed is some way to formally specify user requirements, so that they can be checked for completeness and consistency. Use cases are nice, but they're not sufficiently rigorous to capture absolutely all the requirements. I know there have been some schemes tossed around for requirements validation, but none that I've seen have really been general-purpose enough for the average project.
I've got one of their standard 104-key models, and it's quite nice. Not quite as nice as my Avant Stellar (the reborn Northgate OmniKey), but certainly worth the money. I tend to spend more money on my peripherals, simply because I keep them longer than I keep computers.
It's just too bad that Unicomp's Linux keyboard (with the control key to the left of the A, where $DIETY intended) isn't available with USB. They do custom orders, though, hopefully it wouldn't be too expensive to get that layout.
nobody is able to find out you're tuning to that station, unless they actually hear you
Not quite true, a receiver must be tuned (and therefore have an oscillator resonating at) the desired frequency. If you've got a highly-sensitive receiver, you can determine pretty accurately the frequency a given radio is tuned to. You can't really cover much of an area, but if someone wants to know what station you're tuned to, and they can get close to you (say within a hundred feet or so), then they can tell what you're listening to. Google up "TV detector vans" for more details.
the transmitter needs to have an huge (like mounted on a house huge) antenna to transmit large distances
You're partially right. To get the greatest distance, you'll want to use the lower frequencies (say 80 meters). The lower the frequency, the longer the antenna needs to be (to transmit efficiently). The center of the 80M band is ~3.75MHz, so you'd need an antenna that's (468 / 3.75 = 124.8) ~125 feet long. If you're talking a standard dipole (and we are), then each leg is going to be 62.5 feet long. Each leg is just a hunk of wire, and can easily be rolled up. A complete 80M antenna with a 100W transmitter can be stuffed in a small backpack and hiked up into the hills somewhere (the higher the better). Then the antenna can be unrolled and set up almost anywhere (use a slingshot to hurl the center section over/into a tall tree and fan out the legs, or drape it over a wooden fence or even some bushes). Just because it's a long antenna doesn't mean it's really physically *big*. And there are plenty of other, smaller (and less efficient) antennas that would probably get the job done.
the government derives its mandate from the people, so they do answer to us
You really need to educate yourself...
Or you could just get one of these . Doesn't fit in a wall socket, but it's a full-fledged thin client (doesn't require a "terminal server" [how quaint!] environment). Also doesn't come with keyboard/mouse/monitor (supports DVI or VGA). Not sure what the advantage of either of these is, but I guess if you need them, they're nice to have. Actually, the one I listed supports smart cards, so it's handy if you move around a lot (your session gets switched to whichever node you plug your card into, kind of like using VNC). For $250, you might be able to make a case for thin clients, but for $400, it's a bit tougher.
Does he seriously think it's a good idea to have somebody who's new to the language writing production code in that language?
I don't.
Worked for me. I was hired for a position almost identical to this: established company with a VB-based product (three, actually) needed a developer to maintain and enhance existing systems and design and build new ones. I had no VB experience, but I had about five years of experience in the field (pharmaceutical R&D). I also had written production code in three or four different languages previously. I picked up VB (16-bit VB4 at the time) and ran with it with no problems (although I did write the 16-to-32-bit VB translator in Perl). People with experience in the application domain will probably be more valuable to the team than a language guru who doesn't have any insight into the problem he's being asked to solve.
We had some gradual students who were collating and uploading data from remote sensors (basically a Z-80 box with a modem and a dozen or so A/D converter hooked to various weather instruments). They would have to use tip(1)[0] to call the sensors, then download the files and reset the sensors. And if they didn't do this at the same time every day, then they'd have to dick with the data file to get the readings to line up with previous captures. So I hacked tip(1) to accept scripting commands from a file (fairly easy, it already had built-in commands, it just couldn't read them from a file), wrote a cron job to kick off a shell script, and *boom*: two freshly-unemployed grad students! [1] Not as cool as wiring up a current-loop-BAUDOT-to-RS232-ASCII converter and stuffing NWS reports directly into a newsgroup, but that one wasn't mine....
:P
[0] Why yes, this was the early 80s -- how could you tell?
[1] Not really, they were actually being paid to analyze the data, not collect it.
The calls are coming from inside the house!
Yes -- that's their problem. Face it, no matter how elegant or wickedly clever your algorithm, or how well you manipulate the intrinsics of whatever language you're using...it all gets executed as assembler in the end. And unless you know how assembly language works, you can't make it work for you. Once you learn it, it's like scales fall from your eyes, and those seriously cryptic error messages (the ones that show up just when you think you've finally got your program finished) suddenly make sense.
Yeah, I had to disagree with that one too. I had a SCSI ZIP drive, and it was fast enough (on a 450MHz P-II mind you) that I could work on it directly (as opposed to just copying files to/from). 100MB was plenty of space for my projects, and it made it easy to swap them with my cow-orkers.
Hmmm, that reminds me, I'd better back up all those old disks before I decomission my old PC and can't access them anymore....
In about 20 years I expect a metric shit-ton of blackmail material will be available for our future up-and-coming politicians.
Nope, there'll be an Act of Congress declaring all such material "inadmissable" for the purposes of political discussion. You won't be able to go back farther than seven years when dredging for dirt on a candidate. Any non-felony convictions will be covered, as will any blogs, comments, web postings or letters to the editor. Everyone gets the "I learned from my mistakes" get-out-of-hassle-free card, not just conservative right-wingers.
You can't really have a remote antanna, per se, because you still have to generate the power (not just the signal) to send it. What I have seen is computer-controlled rigs that have software the present the complete control panel on-screen, and can route the audio through the sound card. So what you'd have is a complete transceiver/antenna setup somewhere, then access it remotely across the internet. One such set-up demonstrated at a local ham radio club meeting even had a little diagram showing which way the beam antenna was pointing, overlaid on a map of the world which showed what countries were along the beam's path.
It's especially nice if you have a vacation home in a place where you can put up a Big Honkin' Antenna (tm), then access it from your desk at home. I'd like to do it just so I could operate during lunch here at the office, but I like old tube rigs, and they weren't making computer-controlled transceivers back in the 1960s...
writes things that don't always make logical sense
Ahem.
change at the drop of a hate
Stolen, repainted and redistributed worldwide by the time you read this....
Yeah, right. Not even the Dells in this price range have separate graphics cards. Good luck finding a 2Ghz dual-core laptop for a grand or so without integrated graphics. Hell, good luck finding a 2GHz dual-core laptop in this price range, period. To get the equivalent from Dell, you have to get a high-end Inspiron, and that'll set you back more than $1500. Of course, with that you'll get a 17" screen, which is cool, but if you want to stick close to the $1200 price point, you'll have to settle for a 1.66GHz Core Duo.
I was the lead developer on a team that converted from VB to Java. For us, it made sense, as we wanted to web-enable our app and VB's web support at that time (VB5, back around the turn of the century) was too simplistic for us. It was a big challenge, and lots of hassle, but we got it done and it's been a pretty successful app in its space.
I can think of several reasons why your IT manager wants to make this change, and none of them are good. For us, it made send to make the move to the OOP Web world, because that's where we wanted to take our software. I really don't see the upside moving from C++ to Java. It sounds like offshore Java developers are cheaper, and that's the big incentive.
Bullshit. I've owned a Porsche for the last twelve years, and they take no more maintenance than any other car I've owned. Well, unless you count the extra work I do on it to get it ready for track days, but I'd have to do the same work on anything else I took to the track.
And twelve years and over 100K miles later, it looks a hell of a lot better than my last car; in fact, it looks a hell of a lot better than most of the cars I pass every day.
The fact that your buddy is a fair-weather driver says more about him than the type of car he drives.
it is just clear that Scientific computing is not the forte os MacOSX
Well, now that SGI's gone tits up, they might try pushing harder into the scientific visualization market. I know lots of physical chemists that would love to have some OpenGL love for their MOL files...
Apparently, neither do the millions of people buying iPods...
Do we have reason to believe they didn't train a satellite on each of them?
Yes, because that would imply that we had satellites in position to look within our own borders — and we know the government wouldn't do that!
GRiD Systems
Inventors of what is arguably the best pointing device known to man, the roller mouse (PDF warning). No space wasted on track pads and sissy "palm rests" (you're not supposed to be resting your wrists while typing!), no need for desk space for a mouse or trackball, just a small roller and a couple of buttons. I couldn't believe it when Tandy (who bought GRiD Systems) dropped this from their lineup.
OSX does poor [sic] with database products which leads me to believe raw disk io sucks
From what I understand, the biggest problem with OSX is the interrupt overhead. There are so many layers (disk driver to device I/O subsystem to Mach to BSD I/O to BSD kernel to application) that performance suffers. It looks like every release improves this, but no matter how much you shine that Jetta it'll never be a Porsche.
....since they don't have as much trouble getting those records...
They get to watch their power meters run backwards!
Heh. Most of the time, the power co just dumps the power into a big load (read: resistor), because it's not worth the trouble trying to phase-shift the consumer-generated power to sync it to the neighborhood supply. Especially given the fact that the frequency can vary with load (e.g., during the summer, your power frequency may drop to 55-56Hz at 5:30 when everyone's running the A/C and the stove while making dinner). At least, that's what I was told when I was considering a large UPS/generator I was offered cheap in Feb of 2000...