I found that the hardest thing for me to learn over the past twenty-some years was web development. The whole concept of stateless request/response interactions was, for some reason, incredibly difficult to get my head around. Hopefully the books available today are better than what was out there at the turn of the century, but it's still a fairly significant paradigm shift.
the excellent and free IDE available from Microsoft
Boy, wish I'd gotten that. I'm stuck with Visual Studio 2010 "Premium", which sucks rocks compared to the Java IDEs I've used (Eclipse and NetBeans). Gotta love writing debug statements to a console that evaporates when the program dies. And I wouldn't be doing that, except VS randomly decides whether or not it'll honor breakpoints in the debugger.
Uh, maybe if you're a rice-picker or something. As far as writing code goes, there's a hell of a lot that's changed: Design: Object orientation. Use cases. UML instead of flow charts and DFDs. Coding practices: The whole paradigm of TDD. Security as a design center. Agile methods instead of waterfall. Technology: Relational databases. The whole concept of the web and n-tier systems. SOA and web services. Multi-core CPUs.
Modern PCs and phones look a lot like mainframes from the 1980s
Oh, yeah, I hate when my phone ABENDs with a S0C7 and I have to re-init JES before I can make a call...
The 911, generally accepted as one of the world's best-handling cars, has 40% of its weight behind the rear axle. If Porsche can figure out how to make that work, I don't think they'll have any problems.
Price isn't a primary design center for Porsche. They tend to sell everything they can produce (the 918 was originally suggested for a limited run of 1000, and they had 900 people put down deposits before they even announced that it had been approved for production). FWIW, they consulted with Toyota in the (early 90s?) and brought in some automation and streamlined their manufacturing and the price of 911s actually dropped. But they're also probably the only manufacturer in the world who can get away with offering a hardtop version of a convertible and charging more money for it.
What about using a parabolic mic to record conversations at an outdoor cafe? You could be sitting on a bench across the street and hear every word. Should that be legal? Or was Scott McNealy right?
Personally, I'm on the fence regarding this whole question. It's fine to keep an eye on police behavior, but what's to prevent someone from taping a traffic stop and using the recording to blackmail the driver? Imagine it's a teenager who stands to lose his driving privileges if his parents find out he got pulled over. Lots of cops pull kids over for speeding or running a stop sign, make them sweat a little, then let them go with a warning. No harm done, lesson learned, end of story. But if it gets recorded, then it isn't the end. Maybe we do need a few legal cases to get this sorted out.
The only real numbers I took are for repaginating a Word document I was working on. My work machine took ~34 seconds, my wife's machine took 30. Not a hard-core profiling job, but both machines were similarly loaded (no heavy background tasks). My work machine has 4G of memory, my wife's has 1.5G, both were using about 800M before I launched Word.
I was surprised my wife's machine seemed faster, as it has a 5400RPM drive and my work machine has a 7200. So I assumed the difference must have been due to CPU speed, since if any significant swapping had happened, the disk speed advantage would have kicked in. (The document I was working on was on a USB drive in both cases.)
When we moved, I put my old Sun workstation (single core 2.4GHz Opteron) in storage. It took longer to find a house than I expected, so I wound up building a quad-core box. When we finally got moved in, I was quite surprised to notice that my old box actually felt faster than my new one. I put it down to OpenSolaris being tuned for the Sun hardware, plus having a proper graphics card (nVidia Quadro vs the on-board Radeon 4200 on my quad-core box). So now I'm working on setting up my new box to use as a compile/compute server and doing all my editing/surfing/etc on my old one.
Yep, my wife's machine is a 5yo Gateway laptop with a 3GHz P4 and 1.5G of memory. For lots of stuff, it runs faster than my 2.2GHz dual-core machine at work. Lots of stuff is still single-threaded, and even though that's changing, there's often a critical path that can't be partitioned. Faster CPUs still == win much of the time.
I think part of it is the lack of raw stuff to work with. How many garages in your neighborhood have scraps of lumber or a few surplus bricks lying around? Or some bell wire and a 6V lantern battery? Or even a hammer? I remember running around from garage to garage with my friends, rounding up all the pieces and parts we needed to build a fort or a jump ramp or a hang glider or whatever it was we were building that time. "Mom, can we take a piece of rope from the clothesline, the old trellis, the wheelbarrow and the camping tarp? We have an idea!"
With an error rate that high, it doesn't even sound worthwhile to sync with anything less than millisecond accuracy. Even if you sync exactly, you can't maintain accuracy long enough to reliably base any high-precision timing tasks on the internal clock.
D-STAR is a digital mode, so it's pretty efficient in terms of (signaling) bandwidth. It's also used on high-frequency bands, which gives it good information bandwidth as well.
Do really think that there is a spare 3 inches on the back of the iPhone?
I never claimed there was a spare three inches on an iPhone. In fact, I said that the form factor would provide significant constraint on antenna design.
Did you model your antenna for any sort of yield analysis to meet the requirements?
All of my antennas were one-offs for personal use, therefore there were no yield requirements (beyond producing one functional example).
I realize you think you're an uber-genius just talking down to some simpleton, but next time you might want to try actually understanding what someone else has written before spazzing out about it.
I liked where it said "The candidate will be expected to performance radiation performance measurements". If the spell checker doesn't flag it, it must be OK!
Pffft. I've designed (FSVO) antennas with no more than an HP-41 and a copy of the ARRL Antenna Handbook. Sure, if you're designing an antenna from scratch and you've got some odd-wad requirements you might have to get esoteric on it, but cell phone antenna design is a fairly well-researched field, and the design's going to be largely dictated by form factor considerations. No, you're not going to get somebody from Marketing to design it, but it's not like you're going to be spending weeks with a copy of Mathematica just to get a handle on things.
Um, not sure I see the connection between failing to keep proper records and being a victim of a violent assault (except possibly the admonishment to CYA applies to both). If you have to hand in timesheets, you *always* keep a copy. If they're on-line you make sure either you get a copy/confirmation e-mailed to you or take a screen cap before you submit it (hint: ctrl-PrntScr copies the current window to the clipboard, paste it into Paint and print it from there). I first learned to do this when I was consulting (my timesheet to my employer had to be accompanied by the timesheet I submitted to the client), but now I do it whenever I have to deal with timesheets. It's saved my 4$$ a couple of times.
But what if "two highly-placed officials within the Vatican" announced that Judgement Day was imminent? Don't you think Solarcaine would rush out tubes of "Rapture-Ready" sunblock (extra-large, of course, since you'll need full-body coverage)?
Any serious security policy will ensure tha there is no "give your boss, or employer, or anybody else the passwords you are custodian off".
So, you're on call 24/7/365? Can't go anywhere that you can't get a VPN connection, because you and only you have the critical system passwords? Sounds more like a serious *insecurity* policy to me.
Doesn't matter. You won't get as much of a speed advantage, and you might not be able to take advantage of newer capabilities, but the code will still run. AIUI, you simply generate a system image for whatever hardware you want, and VM loads and runs it, translating stuff on the fly.
And, as a lot of people found out during Y2K, if your system's had any binary patches applied, even if you have the source, you don't have the source.
Slashdot Mirror
The web & tiered systems: not much new here.
Well, there we have it, then.
I found that the hardest thing for me to learn over the past twenty-some years was web development. The whole concept of stateless request/response interactions was, for some reason, incredibly difficult to get my head around. Hopefully the books available today are better than what was out there at the turn of the century, but it's still a fairly significant paradigm shift.
the excellent and free IDE available from Microsoft
Boy, wish I'd gotten that. I'm stuck with Visual Studio 2010 "Premium", which sucks rocks compared to the Java IDEs I've used (Eclipse and NetBeans). Gotta love writing debug statements to a console that evaporates when the program dies. And I wouldn't be doing that, except VS randomly decides whether or not it'll honor breakpoints in the debugger.
I'm sure the Petzold book's nice, though.
Almost nothing has changed between 1980 and 2010
Uh, maybe if you're a rice-picker or something. As far as writing code goes, there's a hell of a lot that's changed:
Design: Object orientation. Use cases. UML instead of flow charts and DFDs.
Coding practices: The whole paradigm of TDD. Security as a design center. Agile methods instead of waterfall.
Technology: Relational databases. The whole concept of the web and n-tier systems. SOA and web services. Multi-core CPUs.
Modern PCs and phones look a lot like mainframes from the 1980s
Oh, yeah, I hate when my phone ABENDs with a S0C7 and I have to re-init JES before I can make a call...
The 911, generally accepted as one of the world's best-handling cars, has 40% of its weight behind the rear axle. If Porsche can figure out how to make that work, I don't think they'll have any problems.
Price isn't a primary design center for Porsche. They tend to sell everything they can produce (the 918 was originally suggested for a limited run of 1000, and they had 900 people put down deposits before they even announced that it had been approved for production). FWIW, they consulted with Toyota in the (early 90s?) and brought in some automation and streamlined their manufacturing and the price of 911s actually dropped. But they're also probably the only manufacturer in the world who can get away with offering a hardtop version of a convertible and charging more money for it.
you don't want to have everyone drawing hundreds of kilowatts straight from the grid.
Boy, talk about your rolling brownouts... <rimshot/>
12:1 compression? Sounds a little fishy to me. Most cars of that era ran about 8:1, and the high-performance engines only ran a bit over 10:1.
What about using a parabolic mic to record conversations at an outdoor cafe? You could be sitting on a bench across the street and hear every word. Should that be legal? Or was Scott McNealy right?
Personally, I'm on the fence regarding this whole question. It's fine to keep an eye on police behavior, but what's to prevent someone from taping a traffic stop and using the recording to blackmail the driver? Imagine it's a teenager who stands to lose his driving privileges if his parents find out he got pulled over. Lots of cops pull kids over for speeding or running a stop sign, make them sweat a little, then let them go with a warning. No harm done, lesson learned, end of story. But if it gets recorded, then it isn't the end. Maybe we do need a few legal cases to get this sorted out.
The only real numbers I took are for repaginating a Word document I was working on. My work machine took ~34 seconds, my wife's machine took 30. Not a hard-core profiling job, but both machines were similarly loaded (no heavy background tasks). My work machine has 4G of memory, my wife's has 1.5G, both were using about 800M before I launched Word.
I was surprised my wife's machine seemed faster, as it has a 5400RPM drive and my work machine has a 7200. So I assumed the difference must have been due to CPU speed, since if any significant swapping had happened, the disk speed advantage would have kicked in. (The document I was working on was on a USB drive in both cases.)
When we moved, I put my old Sun workstation (single core 2.4GHz Opteron) in storage. It took longer to find a house than I expected, so I wound up building a quad-core box. When we finally got moved in, I was quite surprised to notice that my old box actually felt faster than my new one. I put it down to OpenSolaris being tuned for the Sun hardware, plus having a proper graphics card (nVidia Quadro vs the on-board Radeon 4200 on my quad-core box). So now I'm working on setting up my new box to use as a compile/compute server and doing all my editing/surfing/etc on my old one.
Yep, my wife's machine is a 5yo Gateway laptop with a 3GHz P4 and 1.5G of memory. For lots of stuff, it runs faster than my 2.2GHz dual-core machine at work. Lots of stuff is still single-threaded, and even though that's changing, there's often a critical path that can't be partitioned. Faster CPUs still == win much of the time.
More like you install a new stereo and now the car won't start.
I think part of it is the lack of raw stuff to work with. How many garages in your neighborhood have scraps of lumber or a few surplus bricks lying around? Or some bell wire and a 6V lantern battery? Or even a hammer? I remember running around from garage to garage with my friends, rounding up all the pieces and parts we needed to build a fort or a jump ramp or a hang glider or whatever it was we were building that time. "Mom, can we take a piece of rope from the clothesline, the old trellis, the wheelbarrow and the camping tarp? We have an idea!"
Hammering gets the wrinkles out. Nothing screams "amateur" like wrinkles in your duct tape...
With an error rate that high, it doesn't even sound worthwhile to sync with anything less than millisecond accuracy. Even if you sync exactly, you can't maintain accuracy long enough to reliably base any high-precision timing tasks on the internal clock.
D-STAR is a digital mode, so it's pretty efficient in terms of (signaling) bandwidth. It's also used on high-frequency bands, which gives it good information bandwidth as well.
I think SSL violates the "no ciphers" prohibition of Part 97.113[0].
[0] "Part 97" refers to CFR (US Code of Federal Regulations) Title 47, which governs telecommunications. Part 97 covers the Amateur Radio Service.
Do really think that there is a spare 3 inches on the back of the iPhone?
I never claimed there was a spare three inches on an iPhone. In fact, I said that the form factor would provide significant constraint on antenna design.
Did you model your antenna for any sort of yield analysis to meet the requirements?
All of my antennas were one-offs for personal use, therefore there were no yield requirements (beyond producing one functional example).
I realize you think you're an uber-genius just talking down to some simpleton, but next time you might want to try actually understanding what someone else has written before spazzing out about it.
I liked where it said "The candidate will be expected to performance radiation performance measurements". If the spell checker doesn't flag it, it must be OK!
Pffft. I've designed (FSVO) antennas with no more than an HP-41 and a copy of the ARRL Antenna Handbook. Sure, if you're designing an antenna from scratch and you've got some odd-wad requirements you might have to get esoteric on it, but cell phone antenna design is a fairly well-researched field, and the design's going to be largely dictated by form factor considerations. No, you're not going to get somebody from Marketing to design it, but it's not like you're going to be spending weeks with a copy of Mathematica just to get a handle on things.
Um, not sure I see the connection between failing to keep proper records and being a victim of a violent assault (except possibly the admonishment to CYA applies to both). If you have to hand in timesheets, you *always* keep a copy. If they're on-line you make sure either you get a copy/confirmation e-mailed to you or take a screen cap before you submit it (hint: ctrl-PrntScr copies the current window to the clipboard, paste it into Paint and print it from there). I first learned to do this when I was consulting (my timesheet to my employer had to be accompanied by the timesheet I submitted to the client), but now I do it whenever I have to deal with timesheets. It's saved my 4$$ a couple of times.
But what if "two highly-placed officials within the Vatican" announced that Judgement Day was imminent? Don't you think Solarcaine would rush out tubes of "Rapture-Ready" sunblock (extra-large, of course, since you'll need full-body coverage)?
Any serious security policy will ensure tha there is no "give your boss, or employer, or anybody else the passwords you are custodian off".
So, you're on call 24/7/365? Can't go anywhere that you can't get a VPN connection, because you and only you have the critical system passwords? Sounds more like a serious *insecurity* policy to me.
Doesn't matter. You won't get as much of a speed advantage, and you might not be able to take advantage of newer capabilities, but the code will still run. AIUI, you simply generate a system image for whatever hardware you want, and VM loads and runs it, translating stuff on the fly.
And, as a lot of people found out during Y2K, if your system's had any binary patches applied, even if you have the source, you don't have the source.