The time you most need horsepower, on a traditional aircraft, is during takeoff. Lose the engine at altitude and you have some time to figure out where you're coming down. Lose it on takeoff and, well, you better think fast.
A Light Sport Aircraft is limited in the amount of horsepower it can produce, the max speed in level flight and the range. It doesn't say anything about takeoff performance.
Use a small, lightweight, relatively weak Internal Combustion Engine (ICE) with an electric motor/generator as an assist for takeoff. That way, you would have more power available for takeoff and you wouldn't be completely powerless if the ICE failed (doesn't happen often, but it's usually catastrophic when it does).
Charge up some batteries before flight. Possibly even put some flexible thin-film photovoltaics on the wings and park in the sun for a couple days before flying. It'll add to the cost of the plane, but reduce the operating cost.
Electric + ICE for takeoff. ICE for cruise. For landing, turn the electric motor into a generator, allowing you to use the prop as a dive brake. Big turboprop singles like the Pilatus PC-12 routinely do this (flatten out the pitch at low RPM to increase drag). This way, when you land, you recharge your batteries part of the way. And, if you have a missed approach and need to pull out, you go back to electric + ICE.
Voila! A hybrid aircraft. Improved takeoff performance and takeoff safety with a small engine and a limited cruise speed (dovetailing nicely with the Light Sport Aircraft requirements). The improved safety factor will only make it more attractive than a traditional, ICE-only aircraft. Might have to get a waiver on the aircraft weight, though. Since you aren't trying to cruise on electric power, you won't need terribly many batteries.
Mobile. Whether you're talking iPhone or Android, most of the browsers are based on WebKit.
I can't imagine why we'd be seeing a surge in WebKit-based browsers:-P
I can't wait for a decent implementation of FireFox on my Android phone, especially if it supports plug-ins and add-ons. I'm dying to be able to use AdBlock Plus and Aardvark on my cell. I have a definite cap on my wireless bandwidth, and it can get VERY expensive if I'm roaming. I once paid over $12 for the privilege of reading a Linux Magazine article, because I was across the border in Canada. Being able to kill the bandwidth-hogging banners and other crap would've been so nice . . . .
I live in a beautiful valley in a rural area in southern Missouri. I used to have an Alltel phone with tethering. I was getting EV-DO at home. When I roamed to other areas (San Antonio, TX and Minneapolis, MN), I could still get data, but it was 1X-CDMA (153 kb/s). It has never affected my phone bill. People visiting us, with AT&T or T-Mobile service, were unable to send or receive text messages, or make or receive calls, while they were here. It was Alltel or nothing. I had very few complaints with them.
I have, at home, a Cradlepoint MBR-1000 router connected to a Alltel USB-3G module (Pantech UM-150, to be more precise). As it is not roaming, I get full EV-DO Rev A. Data access for the module is about $60/month. If they have a bandwidth cap, they aren't enforcing it. We routinely suck down over 10 GB/month. I have an external antenna, on an external mount, which helps tremendously (a Wilson antenna, acquired from a local truck stop). Without the antenna, I'm lucky if I get 1 bar. With it, I get 4. The connection frequently flirts with 1 Mb/s inbound, with < 100 ms latency. DSL and Cable modem are not an option, at any price, so it is 3G, satellite (horrible latency) or dial-up for us.
I was very disappointed when Alltel was acquired by Verizon, with which I have had "negative experiences" in the past. I now have a Sprint smartphone (HTC Touch) with Phone as Modem (tethering). It has excellent 3G connectivity when I'm at work. When I go home, I have to roam to Verizon, so I'm back to 1X-CDMA speeds. The tethering is mostly used when I'm away from home, so this isn't such an issue. Roaming typically does NOT affect my phone bill. I haven't completed my 2-year contract on the data module, which is why Verizon is still getting my money on that count.
My dad is a truck driver. He uses AT&T. He has an external antenna on his rig, which gives him greater range. There are parts of Montana (mountains) and the Dakotas where he has trouble getting a signal. This is particularly true when passing through Native American Reservations in the western US. He doesn't have a data plan.
My advice: get a data module from Verizon (since their acquisition of Alltel, they do have the largest network), an MBR-1000 and a good external antenna. If you really want to get fancy, buy/build a directional antenna and get used to aiming it (it will be manual; I know of no system for automatically aiming a directional cell antenna; there is a market looking for a solution). There are places where you will not be able to get a signal, but a good antenna will shrink those areas. A signal booster will shrink it further. Verizon has roaming agreements with most other CDMA carriers, so you won't get screwed on your bill from roaming charges.
Oh, and stay the hell out of Canada. Nothing against Canada or Canadians (Vancouver Island was beautiful, and we met some very nice people on my last trip) but the data roaming rates in Canada are little more than legalized rape. I managed to rack up over $20 in roaming charges just reading one article in LinuxMag on my aformentioned HTC and briefly using Google Maps (with the GPS in the phone) to find our hotel.
If they respond to #1 with a blank stare or "huh?" I'd advise running for the door. If they are not using any kind of revision control software, and don't know what it is, the place is very poorly managed and will cause you some serious nightmares. All of the places where I've worked, which didn't use revision control, went out of business within a couple years.
Personally, I prefer Subversion, and I converted my prior employer to using it. They didn't understand why I wanted to use it, but they weren't averse to trying something new. After it saved our butts a couple times, they understood perfectly well why I wanted to use it and continue to use it after I left.
The job before that, I used StarTeam. Borland bought StarBase (the maker of StarTeam) while I was with that employer.
My current employer uses CVS. It has its warts, but it works.
For question #2, my current employer uses Bugzilla. It works reasonably well. My prior employer didn't have a bug-tracking system. The second or third web app they had me write was a bug-tracking system, not oriented toward software development but toward the larger company (hospitality). We ended up using a wiki for bug-tracking. StarTeam had a "Change Control" system built into it, so that employer had excellent bug-tracking software.
I've had only one employer who knew what CMMI was (question #3). They could've passed a level 3 certification if necessary. If you're going to develop for the DoD, you used to need a level 4 cert; not sure if that's the case any more.
For those who don't know, CMMI is merely a methodology for ensuring that:
coding standards are in place
development documentation is in place
the project is well-managed on, time, responsibilities and risks
there is a reasonable probability of delivering on-time
It can be tedious, consistent, or both. The better places are the latter, not so much the former.
In short, they should have good, solid answers for #1 and #2, while #3 is a nice-to-have.
Finally, a correct description of the problem. And me without mod points. Mod parent up!
IIRC, Dolby NR used compression. Basically, everything below a certain level was considered noise; the signal to be recorded was compressed and then shifted up. Consequently, if something was Dolby B-encoded, you had to use Dolby B decoding to get it to play back properly. If you didn't, you got something with a compressed dynamic range, but with the recording level jacked up. Different Dolby noise reduction types used different amount of compression, with different recording levels. Dolby S, IIRC, used a logarithmic compression/expansion instead of a linear one; that's why it was a relative late-comer to the game, and why such equipment was more expensive.
Audio compression, such as described by the parent, and as used by Dolby NR, was a manipulation of the original, analog signal. Data compression, which is what MP3 is, is a completely different animal. Yes, MP3 is a lossy codec, so some of the detail will be "lost;" they try to make sure it's things you are less likely to perceive, but different people have slightly different perceptions, so it won't be perfect. The usual rule of thumb is more bits-per-second equals more detail carried. I can tell the difference between 128 kb/s and 192 kb/s MP3's, which is why I usually go with 192. With the equipment I own, and my aging ears, I have a hard time distinguishing between 192 and 256, so there's not much point (for me) in going higher. If I had better quality equipment, it might be worth it.
A point made by the article was that audio compression is being used to squash the differences between "loud" and "soft," then the recording level was jacked up to push the "loud" parts as far as they can go without clipping (and, in some cases, with noticeable clipping). The waveform comparisons between U2's "with or without you" was an excellent illustration. Nuance has been lost, in favor of volume. Simply turning the volume down on your audio player (CD, MP3 or otherwise) will NOT restore the lost nuance.
You want nuance? Grab Boston's "Third Stage" album, in vinyl, and queue up "The Launch." You can barely hear the opening. If you crank up the volume to hear it clearly, you may damage your hearing later on. CD's have, theoretically, more dynamic range than mid-range turntables, and yet, the CD is mixed differently. There's a much smaller range between the "soft" and "loud" aspects of the song, on the CD (yes, I have them both and yes, I have compared them). CD's are being mixed with greater audio compression (less nuance) and MP3 data compression is only compounding the problem.
O.k. I'm seeing lots of comments on here about "trams don't need batteries because they're connected to electrical infrastructure" and "infrastructure for trams is too expensive." Y'all are missing the connection:
The reason infrastructure for trams is so expensive is BECAUSE they are constantly wired to an electrical supply system. If you don't have to electrify every last foot/meter of railway, the cost of the infrastructure drops, drastically. By making the trams run on battery, only the stations need to have electrical infrastructure. It leaves a station, runs on battery, gets most of its kinetic energy recovered through regen braking, and only has to "top off" the battery at the next station. And, with a 15 km range, it could actually hit several stations without electrical infrastucture, before it would need recharging.
Because of the lower speeds, and the reduced rolling resistance from metal wheels on metal rails, trams are much more efficient, in terms of energy expended/person/mile, than any other form of transport (bicycles and walking excepted). Making it better able to do the regen braking, and eliminating the need for most of electrical infrastructure will only make the cost of setup and operation more attractive.
In terms of energy content, 1 kg of H2 is about equal to 1 gallon of gasoline.
1 liter of Liquid H2 weighs about 71 grams. 14 liters of LH2 = 1 kg.
So, for each 3.785 liters (1 gallon) of gasoline you'd normally carry, you need 14 liters of LH2. That's 3.7x the volume for an equivalent amount of energy. Don't forget that gasoline can be stored in a tank stamped out of steel or aluminum, while LH2 required a double-walled container, with vacuum insulation in between the walls, made out of thicker materials which won't get brittle when exposed to extreme temperatures. The tank will be a lot heavier for any given volume.
3.6 kg H2 / 100 km = approx 3.6 gallons of gasoline / 100 km.
That's 3.6 gallons of gasoline / 62.15 miles, or 17.3 mpg.
That's pretty lame, even for a BMW.
As mentioned, a gallon of gasoline weighs about 6.25 pounds. Compare 14 liters of LH2 weighing 2.2 pounds. Gasoline weighs almost 3x as much, per unit of energy.
If you need light weight (an aircraft or spacecraft), you need something with more energy per pound. If you need minimal volume (a car or truck), you need more energy per liter/gallon/cubic foot.
When I first read about the Abaq (Byte Magazine, 1988 IIRC), I was excited. The original idea was something like a Mega ST functioning as a host, with Transputer modules connected to the DMA port on the back end. Then, it changed. You had a typical-looking computer case with a limited number of Transputer boards which plugged in. Do some digging around; you can find pictures.
Limited amount of expansion available. Normal-looking computer. Disappointing, really.
Another aspect of the whole Transputer thing is that INMOS (the company which developed the Transputer) had small things called TRAnsputer Modules or TRAMs. You could buy an ISA board (VL-BUS and PCI didn't exist, back then) which had an adaptor for the serial channel and a soft-configurable switch, then plug multiple TRAMs into it (think daughterboards). Assuming each TRAM had four serial channels on it, two of them were hard-wired and the other two went through the soft switch. In that fashion, you could create all manner of topologies without physically modifying the layout on the board. You could get a TRAM with:
a screaming fast number-cruncher and some RAM
a video adaptor with a certain amount of VRAM and a Transputer
one bit-plane for a graphics adaptor and a Transputer you could link multiple of these TRAMs to one which actually did the graphics output lots of raw processing power applied as a GPU
a 16-bit Transputer (T212) and a SCSI adaptor think about the kinds of parallel processing RAID-based systems you could build with THAT!
It's hard to even find information about these things, these days.
I was using a potential answer to this in 1990. I was working for a small company in the Provo/Orem area, called Computer System Architects, which was selling Transputer hardware. For those who haven't heard of Transputers, they were small, 16- or 32-bit processors, with a small amount of built-in RAM (not a cache; this was actually in the memory map and you could do small tasks on a Transputer without any external RAM), 2-4 high-speed serial channels (easily implemented with 4 wires) and a stack-based architecture. Adding megabytes of external RAM was easy, and it was embarrassingly easy to connect up networks of these things, even on one board (in a single ISA slot), and build cluster. An external card cage, in those days, could hold 20 slots, which would hold up to 80 Transputers, using our products.
I did some Assembly and some C, but the kicker language for this chip was called Occam II. Among other things, it used the indentation in the code to determine block structure. A quick example: PAR
step A
step B
step C
SEQ
step D
step E
In this example, steps A, B and C would all be executed in parallel with another task which ran step D then step E. If you had one Transputer in your machine, it would multi-task. If you had multiple CPU's available, it would spread the task across the CPU's.
It also has a basic construct called a Channel. These were very easy to set up and use. These were how the different tasks communicated with each other.
It was not difficult to spawn thousands of tasks, each one doing a relatively small part of an overall task, with full communication and synchronization. Again, if you had multiple CPU's available, it would spread the tasks across them. A board with multiple Transputers was usually doing ray-tracing or rendering Mandelbrot fractals as a demo anytime we went to a trade or tech show. They could knock it down to one processor, and things got done relatively quickly. Then, they'd kick in 4 or 16 CPU's and blow people's minds.
This was in 1990. A 386DX-33 was high-end, back then. The Transputer didn't run DOS or Windows, so it didn't survive in the market of the time. That was a shame; I benchmarked a variety of them, then ran identical benchmarks on various other machines as technology marched on. A T805 running 30 MHz (the top end Transputer I ever got to play with) blasted through mixed integer/floating-point calculations about as fast a 486DX2-66 (which didn't come on the market for another couple years). There was an occasion where I had 16 of those T805's sitting my machine. You'd need a Pentium II to be able to match that occasion. It was well over a decade later that the P-II became available.
Cool tech, but the programming tools were what allowed you to really use the parallelization. It was typical to achieve over 95% linear speedup (i.e. 20 CPU's gave real-world 19x performance); sometimes we went over 99%. Most Intel SMP machines are lucky if they give 80% linear speedup (4 CPU's = 3.2x total performance).
She was originally married to a German weapons supplier. Consequently, she knew about things like tanks and torpedoes.
Came up with what we now call frequency-hopping spread spectrum technology, trying to make a torpedo which could be directed after launch, but couldn't be jammed.
Reasonably good actress. Brainy as all hell. Drop-dead gorgeous.
Now THERE'S a Geek Girl rolemodel who simply needs better publicity.
Computer Sysstem Architects, for whom I was working, was selling the Inmos Occam development kit and Logical Systems C, both of which allowed you to exploit the parallelism.
I was already familiar with C before I started there, and I learned Occam and Transputer Assembly Language on the job. I'd never even HEARD of a stack-based architecture before, so I was in for a real eye-opener.
What's truly sad is that, these days, some geeks would just port Linux to it and it would have a thriving market. Back then, it didn't run DOS or Windows 3.0, so it didn't sell.
Also, in my most humble opinion, it is immoral to write software unless you have some Paul van Dyk, Tiesto, or at least some Ministry of Sound in the CD/MP3 player, anything else would just be WEIRD.
Not familiar with those. I usually work best with Eric Johnson and Blues Saraceno. Few, if any, vocals to distract the mind, variation in tempos (kicking adrenaline all the time doesn't work, long-term), fairly complex and somewhat repetitive.
Tried a variety of other stuff. This just happens to be what works for me.
Paul Graham is somewhat known for saying that web-based applications (what they're calling Software-as-a-Service) is going to take the market away from installed applications, for some very simple reasons:
Installed apps only run on the machines where they're installed. Web-based apps run from just about any Internet-connected machine, which is much more flexible. I mean, if my desktop machine is having problems, I can go to the local library, or a friend's house, and I can hit GMail from there. Access to my data is much more flexible.
How hard is it to push software updates/upgrades to multiple machines? Granted some of the newer management tools are making it easier if you're managing a network of MS Win machines, and most Unix gurus can update a roomful of Unix boxes in short order. But, if you're dealing with a web-based app, you have the potential to roll out multiple updates and upgrades in a single day, especially bug fixes. I've seen this in my own career. I was developing a number of web-based apps for our company Intranet. Someone in HR complained that they were inputting the same data quite often (address of the office where someone was employed), and I was able to throw together something which accelerated the process. There was less than 30 minutes between her whining about an issue and her using the updated system. Let's see you do that with an installed app.
Web-based apps are usually running in data centers, which tend to have better data backup and disaster recovery features than most users. How many users run backups as religiously as a data center? I don't know any. Consequently, if I have a system problem, the data is on their server. If they have a system problem, they are more likely to be able to restore my data than I would be.
Most users simply want to be able to just use the system. Most of them don't want to mess with the settings, run hotfixes, install patches, install new software. They just want to sit down and have it work. So long as your network connection is reliable, web-based apps provide that much easier than installed apps. Yes, I realize there's a potential problem right there. For most people, though, that's not a problem, especially if they're working in an enterprise environment.
It's great for you that you like your Clie, but I don't own one and am not about to run out and spend, what, over $200, or heck, even $100 to be able to purchase a book.
I picked up an SL-10 on eBay a while back. I spent < $20, including shipping. It runs on AAA batteries (I have NiMH rechargables and a charger) and has a good screen and a working scroll-wheel. If you want the absolute latest and greatest, you can drop $100+ if your really want to. Otherwise, you can get your feet wet for less than the cost of most hardbound books (typically $24.95 or more).
I tend to prefer Plucker as my e-Book reader. Anything which is posted on the 'Net can become portable content. It comes in real handy when I'm waiting for the wife or a kid to get through with a dental appointment, not to mention the fact that it also trips an alarm a week before my wife's birthday (so I can remember to get her something) and a week before the anniversary (same reason). My point is that it is ALSO an eBook reader, but it's useful for other things as well. The last time I filled out a job application, all my previous job information was in my handheld (a Palm IIIXE, at the time). The last time I filled out a lease application, all my previous address information was there, too. Did I mention that it was handy? Does that cover the "more than just an eBook reader?"
I replace the batteries every couple weeks (rotate in a different pair of rechargables), but this thing uses so little power that I can literally spend hours reading on it without the batteries dying on me. I believe I've had that happen a grand total of once in the last few years that I've been using Palm-compatible machines.
I also have an account on Safari, so I tend to spend a lot of time reading from a digital display. I guess I'm just more comfortable with that than most people. I just wish there was a way to put my Safari bookshelf on my Clie.
Where I work, most of the IT guys (myself included) run around with USB sticks attached to themselves (hanging around the neck, attached to a belt loop, etc.). Our main support guy has a Linux distro on one of them, and can boot desktop machines off the silly thing; comes in real handy when someone has REALLY hosed up their WinXP machine and he has to try to rebuild it without completely wiping their drive and losing their data. Each of us have a "personal" one which has.mp3's, etc. on them. In my case it's an old 128 MB Sandisk Cruzer. I got it free when we ordered a bunch of hardware from someplace. It's getting harder to buy something that small, these days. Even that little thing can easily haul 100 MB of files around.
Quite a few employees have iPods or other small, personal media players, with capacities that dwarf my Cruzer.
If we wanted to, I'm sure we could slurp a large amount of data and walk off with it. More than a few people have pointed out, though, that it would be unethical. For most people, that's enough of a reason not to do it. Probability of getting burned for doing so isn't really the motivating factor. Most people are ethical enough, without needing any kind of threats hanging over their heads.
On the other hand, my wife applied, at one point, for a position with a defense contractor. She wasn't allowed to bring any kind of personal media player, CD's, etc. into the premises. If she had a camera cellphone, she wouldn't be allowed to bring it in, either. A regular cellphone was allowed, but she couldn't turn it on or take/make calls inside the building; she'd have to be outside on break. She couldn't even bring a personal CD player into the place (no recording capability, at all). She had to go through a metal detector any time she entered the building; good luck sneaking an electronic device past that thing.
It all depends on the environment. Obviously, some places are "locked down" more than others.
Santa Monica enacted a local ordinance to that effect. Within 24 hours, all of the banks directed their ATM's to only dispense cash to customers of that particular bank. In short, you couldn't go to a Bank of America ATM and get money from an Commerce Bank account.
Dr. Steven Hawking was getting a tour of the Paramount set for Star Trek: TNG. It was as a result of this tour, and his professed appreciation for Trek, which resulted in his cameo appearance in one of the later episodes (and no, I'm not a big enough Trek geek to remember the title).
Anyhow, at one point during his tour, they were working their way through Engineering, looking at the main warp core, and he signalled they should stop. He tapped on his communicator for a moment or two, and indicated:
I agree with you on this one. In my book, GMail has done a lot more than just up the ante on disk space.
I use my GMail account as my primary account. I've got three POP3 accounts at various places, plus Yahoo and Hotmail, but GMail is where my "legit" e-mail goes. The Hotmail account gets handed out anywhere I figure SpamBots are searching, so it usually has 200+ messages, all of them crap. Fitting place for it, if you ask me.
We use Outlook at work (company-mandated). I've got Thunderbird loaded on my machine at home (decent RSS aggregation). And yet, I'd still rather use GMail for e-mail functionality. The message threading is simply too useful.
When e-mail clients can match that functionality, allowing me to follow a thread of e-mail messages as easily as following a thread on Usenet, I'll find a reason to switch. When I can manually reorganize message threads (so that messages with different subjects are included), I'll be even happier. You guys at Google listening?
At first, I thought "labeling" messages was a nuisance. Then, I started using it, and promptly found the older, "folders" method to be constraining.
The fact that I don't have to wait for the whole screen to redraw every time I expand a message is simply icing on the cake.
These aren't just "show-off" features, which look nice on the "features" list but never get used. There are plenty of practical, useful and heavily-used features here. The fact that no one else has copied them (yet) doesn't mean they're a failed trend. It just means the rest of the world hasn't caught up, yet.
The record high price for gasoline (set in August, 1981), adjusted for inflation, works out to $3.08 / gallon in today's money. Stations in the Kansas City area BROKE that record, this afternoon. We do not, by any stretch, have the highest prices in the nation.
Many of our refineries are in the Gulf Coast region, and shut down and/or damaged by Hurricane Katrina. There's the choke point in the supply / demand equation. The price of crude has hit records, as well, but Uncle Sam is releasing some of the Strategic Petroleum Reserve, to try to keep that from going too high.
I have my Gentoo machine set to update itself every three days at about 2 a.m. It has recompiled X at least twice, GLIBC three times, the kernel at least twice, at least one round on Firefox and, recently, OpenOffice. All told, the system averages about five-six hours compile time each week. All of this happens at some ungodly hour of the morning when I'm out cold.
If a tree falls in a forest, and no one hears, does it really make a noise? If a computer recompiles major components on a regular basis, but no one is trying to use it during that timeframe, is it really an annoyance?
Setting it up, on the other hand, can be SO DAMMED ANNOYING. It took me a week to finally get mplayer and ALSA to play nice (I'd never used ALSA before, so there was a learning curve). Luckily, you don't have to recompile the entire 2.6 kernel every time you want to add/remove another module (moved from 2.4 to 2.6 in the process; another learning curve). I'm currently getting some funky behavior with lbxproxy which I still haven't figured out (Debian, which I used to use, still used XFree86, while Gentoo is X.org; yet another learning curve).
Once it's set up, though, it ROCKS! With mplayer on Debian, the system wasn't able to render DivX videos full-screen without dropping frames. GMPlayer on Gentoo can handle it, without dropped frames, at about 70% CPU load. Enlightenment hustles along smoothly and beautifully, even at 1600x1200. There is definitely something to be said for optimized object code.
I'm not sure I'd want to put Gentoo on a server, but it has replaced Debian as my desktop distro of choice. Cron is your friend, even more so with Gentoo.
I still can't get the scriptmonkeys around here to grasp the notion of structured programming, let alone OO or functional.
Here, here! I'm the senior developer around here, with a B.S. in CompSci. The CIS majors around here can play with a point-and-drool GUI developer tool and put together a decent-looking web page with some functionality on the back-end. When the requirements get complicated, they end up re-using something I wrote months ago, or they're in my cubicle, going "how would you tackle this?"
CIS majors are carpenters; they can use standardized tools and standardized components and put together someting which works. Carpenters can do wonderful work, but they're largely limited in their tools and materials. CompSci majors are architects; they can use the standardized stuff, or they can go off the deep end, develop their own components, and leave the carpenters going "whoa! how did you do that?!!!"
I had Scheme, as well as C/C++ and Java, in college. I've also done a significant amount of Perl development in my career. My ASP and.Net coding for my current employer reflects my experience. I can build complex data structures in ASP (using VBScript, of all things!), built on Arrays (a.k.a. lists) and Dictionaries (a.k.a. hashes) which leave them with their jaws in their laps. Then, when someone comes up and says "that's great, but can we add this . . . and this, . . . oh, and this," I add a few more lines to my data struture and they've got what they want. This isn't the way I design EVERYTHING, but I'm used to the users in this place, and how they ALWAYS want to add "just one more feature (or two . . . or three . . . or)."
The CIS majors have to go back to their tools and re-write parts of their programs. If they're maintaining something I wrote, they can usually figure out what needs to be tweaked to add stuff. They just don't, typically, write their apps the way I write mine.
If they trained on VB, they have a hard time dealing with C# (which is very similar to Java, C++ and, to a lesser degree, C). Perl might as well be greek. Scheme doesn't even look, to their eyes, like programming. Most of them can't figure out why anyone would want to mess with JavaScript. They can usually hack some HTML, but they'd rather use DreamWeaver or FrontPage (shudder!), instead. Oh, and don't even THINK about tackling.Net without Visual Studio.
I use vim for all of the above, from VBScript, to SQL, to C#, HTML and JavaScript.
So what does that piece of parchment (the degree in CompSci) do? Not much. What difference is made by the exposure to:
a variety of programming languages
a variety of programming styles
very different programming paradigms (imperative vs functional)
data structures and theory
that comes with EARNING that piece of paper? All the difference in the world.
Programming is arguably a craft, and occasionally art. When I can open a file I haven't touched in months, tweak a couple lines of code and add the new functionality the clients want (in minutes, not hours), I consider that application to be a piece of art. To do that, the code has to be clean, comprehensible and pretty robust. In short, it has to be easy to maintain. If things aren't neatly laid out, and the code isn't very clean, it's NOT easy to maintain. If adding a couple lines causes stuff to break, it is DEFINITELY NOT easy to maintain.
Not everything I create is art. In fact, some of my earlier stuff is, in hindsight, crap. But, as time goes by, and my experience increases, the art-vs-crap ratio keeps improving. How does a programmer increase this ratio? The same way an artist does: exposure (to others' works and re-exposure to his/her own), practice, reflection (this is how you learn to recognize crap) and feedback.
Slashdot Mirror
The time you most need horsepower, on a traditional aircraft, is during takeoff. Lose the engine at altitude and you have some time to figure out where you're coming down. Lose it on takeoff and, well, you better think fast.
A Light Sport Aircraft is limited in the amount of horsepower it can produce, the max speed in level flight and the range. It doesn't say anything about takeoff performance.
Use a small, lightweight, relatively weak Internal Combustion Engine (ICE) with an electric motor/generator as an assist for takeoff. That way, you would have more power available for takeoff and you wouldn't be completely powerless if the ICE failed (doesn't happen often, but it's usually catastrophic when it does).
Charge up some batteries before flight. Possibly even put some flexible thin-film photovoltaics on the wings and park in the sun for a couple days before flying. It'll add to the cost of the plane, but reduce the operating cost.
Electric + ICE for takeoff. ICE for cruise. For landing, turn the electric motor into a generator, allowing you to use the prop as a dive brake. Big turboprop singles like the Pilatus PC-12 routinely do this (flatten out the pitch at low RPM to increase drag). This way, when you land, you recharge your batteries part of the way. And, if you have a missed approach and need to pull out, you go back to electric + ICE.
Voila! A hybrid aircraft. Improved takeoff performance and takeoff safety with a small engine and a limited cruise speed (dovetailing nicely with the Light Sport Aircraft requirements). The improved safety factor will only make it more attractive than a traditional, ICE-only aircraft. Might have to get a waiver on the aircraft weight, though. Since you aren't trying to cruise on electric power, you won't need terribly many batteries.
Gee, what's the fastest growing platform, today?
:-P
Mobile. Whether you're talking iPhone or Android, most of the browsers are based on WebKit.
I can't imagine why we'd be seeing a surge in WebKit-based browsers
I can't wait for a decent implementation of FireFox on my Android phone, especially if it supports plug-ins and add-ons. I'm dying to be able to use AdBlock Plus and Aardvark on my cell. I have a definite cap on my wireless bandwidth, and it can get VERY expensive if I'm roaming. I once paid over $12 for the privilege of reading a Linux Magazine article, because I was across the border in Canada. Being able to kill the bandwidth-hogging banners and other crap would've been so nice . . . .
I live in a beautiful valley in a rural area in southern Missouri. I used to have an Alltel phone with tethering. I was getting EV-DO at home. When I roamed to other areas (San Antonio, TX and Minneapolis, MN), I could still get data, but it was 1X-CDMA (153 kb/s). It has never affected my phone bill. People visiting us, with AT&T or T-Mobile service, were unable to send or receive text messages, or make or receive calls, while they were here. It was Alltel or nothing. I had very few complaints with them.
I have, at home, a Cradlepoint MBR-1000 router connected to a Alltel USB-3G module (Pantech UM-150, to be more precise). As it is not roaming, I get full EV-DO Rev A. Data access for the module is about $60/month. If they have a bandwidth cap, they aren't enforcing it. We routinely suck down over 10 GB/month. I have an external antenna, on an external mount, which helps tremendously (a Wilson antenna, acquired from a local truck stop). Without the antenna, I'm lucky if I get 1 bar. With it, I get 4. The connection frequently flirts with 1 Mb/s inbound, with < 100 ms latency. DSL and Cable modem are not an option, at any price, so it is 3G, satellite (horrible latency) or dial-up for us.
I was very disappointed when Alltel was acquired by Verizon, with which I have had "negative experiences" in the past. I now have a Sprint smartphone (HTC Touch) with Phone as Modem (tethering). It has excellent 3G connectivity when I'm at work. When I go home, I have to roam to Verizon, so I'm back to 1X-CDMA speeds. The tethering is mostly used when I'm away from home, so this isn't such an issue. Roaming typically does NOT affect my phone bill. I haven't completed my 2-year contract on the data module, which is why Verizon is still getting my money on that count.
My dad is a truck driver. He uses AT&T. He has an external antenna on his rig, which gives him greater range. There are parts of Montana (mountains) and the Dakotas where he has trouble getting a signal. This is particularly true when passing through Native American Reservations in the western US. He doesn't have a data plan.
My advice: get a data module from Verizon (since their acquisition of Alltel, they do have the largest network), an MBR-1000 and a good external antenna. If you really want to get fancy, buy/build a directional antenna and get used to aiming it (it will be manual; I know of no system for automatically aiming a directional cell antenna; there is a market looking for a solution). There are places where you will not be able to get a signal, but a good antenna will shrink those areas. A signal booster will shrink it further. Verizon has roaming agreements with most other CDMA carriers, so you won't get screwed on your bill from roaming charges.
Oh, and stay the hell out of Canada. Nothing against Canada or Canadians (Vancouver Island was beautiful, and we met some very nice people on my last trip) but the data roaming rates in Canada are little more than legalized rape. I managed to rack up over $20 in roaming charges just reading one article in LinuxMag on my aformentioned HTC and briefly using Google Maps (with the GPS in the phone) to find our hotel.
If they respond to #1 with a blank stare or "huh?" I'd advise running for the door. If they are not using any kind of revision control software, and don't know what it is, the place is very poorly managed and will cause you some serious nightmares. All of the places where I've worked, which didn't use revision control, went out of business within a couple years.
Personally, I prefer Subversion, and I converted my prior employer to using it. They didn't understand why I wanted to use it, but they weren't averse to trying something new. After it saved our butts a couple times, they understood perfectly well why I wanted to use it and continue to use it after I left.
The job before that, I used StarTeam. Borland bought StarBase (the maker of StarTeam) while I was with that employer.
My current employer uses CVS. It has its warts, but it works.
For question #2, my current employer uses Bugzilla. It works reasonably well. My prior employer didn't have a bug-tracking system. The second or third web app they had me write was a bug-tracking system, not oriented toward software development but toward the larger company (hospitality). We ended up using a wiki for bug-tracking. StarTeam had a "Change Control" system built into it, so that employer had excellent bug-tracking software.
I've had only one employer who knew what CMMI was (question #3). They could've passed a level 3 certification if necessary. If you're going to develop for the DoD, you used to need a level 4 cert; not sure if that's the case any more.
For those who don't know, CMMI is merely a methodology for ensuring that:
It can be tedious, consistent, or both. The better places are the latter, not so much the former.
In short, they should have good, solid answers for #1 and #2, while #3 is a nice-to-have.
Finally, a correct description of the problem. And me without mod points. Mod parent up!
IIRC, Dolby NR used compression. Basically, everything below a certain level was considered noise; the signal to be recorded was compressed and then shifted up. Consequently, if something was Dolby B-encoded, you had to use Dolby B decoding to get it to play back properly. If you didn't, you got something with a compressed dynamic range, but with the recording level jacked up. Different Dolby noise reduction types used different amount of compression, with different recording levels. Dolby S, IIRC, used a logarithmic compression/expansion instead of a linear one; that's why it was a relative late-comer to the game, and why such equipment was more expensive.
Audio compression, such as described by the parent, and as used by Dolby NR, was a manipulation of the original, analog signal. Data compression, which is what MP3 is, is a completely different animal. Yes, MP3 is a lossy codec, so some of the detail will be "lost;" they try to make sure it's things you are less likely to perceive, but different people have slightly different perceptions, so it won't be perfect. The usual rule of thumb is more bits-per-second equals more detail carried. I can tell the difference between 128 kb/s and 192 kb/s MP3's, which is why I usually go with 192. With the equipment I own, and my aging ears, I have a hard time distinguishing between 192 and 256, so there's not much point (for me) in going higher. If I had better quality equipment, it might be worth it.
A point made by the article was that audio compression is being used to squash the differences between "loud" and "soft," then the recording level was jacked up to push the "loud" parts as far as they can go without clipping (and, in some cases, with noticeable clipping). The waveform comparisons between U2's "with or without you" was an excellent illustration. Nuance has been lost, in favor of volume. Simply turning the volume down on your audio player (CD, MP3 or otherwise) will NOT restore the lost nuance.
You want nuance? Grab Boston's "Third Stage" album, in vinyl, and queue up "The Launch." You can barely hear the opening. If you crank up the volume to hear it clearly, you may damage your hearing later on. CD's have, theoretically, more dynamic range than mid-range turntables, and yet, the CD is mixed differently. There's a much smaller range between the "soft" and "loud" aspects of the song, on the CD (yes, I have them both and yes, I have compared them). CD's are being mixed with greater audio compression (less nuance) and MP3 data compression is only compounding the problem.
O.k. I'm seeing lots of comments on here about "trams don't need batteries because they're connected to electrical infrastructure" and "infrastructure for trams is too expensive." Y'all are missing the connection:
The reason infrastructure for trams is so expensive is BECAUSE they are constantly wired to an electrical supply system. If you don't have to electrify every last foot/meter of railway, the cost of the infrastructure drops, drastically. By making the trams run on battery, only the stations need to have electrical infrastructure. It leaves a station, runs on battery, gets most of its kinetic energy recovered through regen braking, and only has to "top off" the battery at the next station. And, with a 15 km range, it could actually hit several stations without electrical infrastucture, before it would need recharging.
Because of the lower speeds, and the reduced rolling resistance from metal wheels on metal rails, trams are much more efficient, in terms of energy expended/person/mile, than any other form of transport (bicycles and walking excepted). Making it better able to do the regen braking, and eliminating the need for most of electrical infrastructure will only make the cost of setup and operation more attractive.
In terms of energy content, 1 kg of H2 is about equal to 1 gallon of gasoline.
1 liter of Liquid H2 weighs about 71 grams. 14 liters of LH2 = 1 kg.
So, for each 3.785 liters (1 gallon) of gasoline you'd normally carry, you need 14 liters of LH2. That's 3.7x the volume for an equivalent amount of energy. Don't forget that gasoline can be stored in a tank stamped out of steel or aluminum, while LH2 required a double-walled container, with vacuum insulation in between the walls, made out of thicker materials which won't get brittle when exposed to extreme temperatures. The tank will be a lot heavier for any given volume.
3.6 kg H2 / 100 km = approx 3.6 gallons of gasoline / 100 km.
That's 3.6 gallons of gasoline / 62.15 miles, or 17.3 mpg.
That's pretty lame, even for a BMW.
As mentioned, a gallon of gasoline weighs about 6.25 pounds. Compare 14 liters of LH2 weighing 2.2 pounds. Gasoline weighs almost 3x as much, per unit of energy.
If you need light weight (an aircraft or spacecraft), you need something with more energy per pound. If you need minimal volume (a car or truck), you need more energy per liter/gallon/cubic foot.
Limited amount of expansion available. Normal-looking computer. Disappointing, really.
Another aspect of the whole Transputer thing is that INMOS (the company which developed the Transputer) had small things called TRAnsputer Modules or TRAMs. You could buy an ISA board (VL-BUS and PCI didn't exist, back then) which had an adaptor for the serial channel and a soft-configurable switch, then plug multiple TRAMs into it (think daughterboards). Assuming each TRAM had four serial channels on it, two of them were hard-wired and the other two went through the soft switch. In that fashion, you could create all manner of topologies without physically modifying the layout on the board. You could get a TRAM with:
you could link multiple of these TRAMs to one which actually did the graphics output
lots of raw processing power applied as a GPU
think about the kinds of parallel processing RAID-based systems you could build with THAT!
It's hard to even find information about these things, these days.
I was using a potential answer to this in 1990. I was working for a small company in the Provo/Orem area, called Computer System Architects, which was selling Transputer hardware. For those who haven't heard of Transputers, they were small, 16- or 32-bit processors, with a small amount of built-in RAM (not a cache; this was actually in the memory map and you could do small tasks on a Transputer without any external RAM), 2-4 high-speed serial channels (easily implemented with 4 wires) and a stack-based architecture. Adding megabytes of external RAM was easy, and it was embarrassingly easy to connect up networks of these things, even on one board (in a single ISA slot), and build cluster. An external card cage, in those days, could hold 20 slots, which would hold up to 80 Transputers, using our products.
I did some Assembly and some C, but the kicker language for this chip was called Occam II. Among other things, it used the indentation in the code to determine block structure. A quick example:
PAR
step A
step B
step C
SEQ
step D
step E
In this example, steps A, B and C would all be executed in parallel with another task which ran step D then step E. If you had one Transputer in your machine, it would multi-task. If you had multiple CPU's available, it would spread the task across the CPU's.
It also has a basic construct called a Channel. These were very easy to set up and use. These were how the different tasks communicated with each other.
It was not difficult to spawn thousands of tasks, each one doing a relatively small part of an overall task, with full communication and synchronization. Again, if you had multiple CPU's available, it would spread the tasks across them. A board with multiple Transputers was usually doing ray-tracing or rendering Mandelbrot fractals as a demo anytime we went to a trade or tech show. They could knock it down to one processor, and things got done relatively quickly. Then, they'd kick in 4 or 16 CPU's and blow people's minds.
This was in 1990. A 386DX-33 was high-end, back then. The Transputer didn't run DOS or Windows, so it didn't survive in the market of the time. That was a shame; I benchmarked a variety of them, then ran identical benchmarks on various other machines as technology marched on. A T805 running 30 MHz (the top end Transputer I ever got to play with) blasted through mixed integer/floating-point calculations about as fast a 486DX2-66 (which didn't come on the market for another couple years). There was an occasion where I had 16 of those T805's sitting my machine. You'd need a Pentium II to be able to match that occasion. It was well over a decade later that the P-II became available.
Cool tech, but the programming tools were what allowed you to really use the parallelization. It was typical to achieve over 95% linear speedup (i.e. 20 CPU's gave real-world 19x performance); sometimes we went over 99%. Most Intel SMP machines are lucky if they give 80% linear speedup (4 CPU's = 3.2x total performance).
She was originally married to a German weapons supplier. Consequently, she knew about things like tanks and torpedoes.
Came up with what we now call frequency-hopping spread spectrum technology, trying to make a torpedo which could be directed after launch, but couldn't be jammed.
Reasonably good actress. Brainy as all hell. Drop-dead gorgeous.
Now THERE'S a Geek Girl rolemodel who simply needs better publicity.
Computer Sysstem Architects, for whom I was working, was selling the Inmos Occam development kit and Logical Systems C, both of which allowed you to exploit the parallelism.
I was already familiar with C before I started there, and I learned Occam and Transputer Assembly Language on the job. I'd never even HEARD of a stack-based architecture before, so I was in for a real eye-opener.
What's truly sad is that, these days, some geeks would just port Linux to it and it would have a thriving market. Back then, it didn't run DOS or Windows 3.0, so it didn't sell.
Not familiar with those. I usually work best with Eric Johnson and Blues Saraceno. Few, if any, vocals to distract the mind, variation in tempos (kicking adrenaline all the time doesn't work, long-term), fairly complex and somewhat repetitive.
Tried a variety of other stuff. This just happens to be what works for me.
Most users simply want to be able to just use the system. Most of them don't want to mess with the settings, run hotfixes, install patches, install new software. They just want to sit down and have it work. So long as your network connection is reliable, web-based apps provide that much easier than installed apps. Yes, I realize there's a potential problem right there. For most people, though, that's not a problem, especially if they're working in an enterprise environment.
O.k. I'm sitting in my small cubicle. Is she attractive?
Oh, THAT definition of "eaten." Nevermind.
I picked up an SL-10 on eBay a while back. I spent < $20, including shipping. It runs on AAA batteries (I have NiMH rechargables and a charger) and has a good screen and a working scroll-wheel. If you want the absolute latest and greatest, you can drop $100+ if your really want to. Otherwise, you can get your feet wet for less than the cost of most hardbound books (typically $24.95 or more).
I tend to prefer Plucker as my e-Book reader. Anything which is posted on the 'Net can become portable content. It comes in real handy when I'm waiting for the wife or a kid to get through with a dental appointment, not to mention the fact that it also trips an alarm a week before my wife's birthday (so I can remember to get her something) and a week before the anniversary (same reason). My point is that it is ALSO an eBook reader, but it's useful for other things as well. The last time I filled out a job application, all my previous job information was in my handheld (a Palm IIIXE, at the time). The last time I filled out a lease application, all my previous address information was there, too. Did I mention that it was handy? Does that cover the "more than just an eBook reader?"
I replace the batteries every couple weeks (rotate in a different pair of rechargables), but this thing uses so little power that I can literally spend hours reading on it without the batteries dying on me. I believe I've had that happen a grand total of once in the last few years that I've been using Palm-compatible machines.
I also have an account on Safari, so I tend to spend a lot of time reading from a digital display. I guess I'm just more comfortable with that than most people. I just wish there was a way to put my Safari bookshelf on my Clie.
Where I work, most of the IT guys (myself included) run around with USB sticks attached to themselves (hanging around the neck, attached to a belt loop, etc.). Our main support guy has a Linux distro on one of them, and can boot desktop machines off the silly thing; comes in real handy when someone has REALLY hosed up their WinXP machine and he has to try to rebuild it without completely wiping their drive and losing their data. Each of us have a "personal" one which has .mp3's, etc. on them. In my case it's an old 128 MB Sandisk Cruzer. I got it free when we ordered a bunch of hardware from someplace. It's getting harder to buy something that small, these days. Even that little thing can easily haul 100 MB of files around.
Quite a few employees have iPods or other small, personal media players, with capacities that dwarf my Cruzer.
If we wanted to, I'm sure we could slurp a large amount of data and walk off with it. More than a few people have pointed out, though, that it would be unethical. For most people, that's enough of a reason not to do it. Probability of getting burned for doing so isn't really the motivating factor. Most people are ethical enough, without needing any kind of threats hanging over their heads.
On the other hand, my wife applied, at one point, for a position with a defense contractor. She wasn't allowed to bring any kind of personal media player, CD's, etc. into the premises. If she had a camera cellphone, she wouldn't be allowed to bring it in, either. A regular cellphone was allowed, but she couldn't turn it on or take/make calls inside the building; she'd have to be outside on break. She couldn't even bring a personal CD player into the place (no recording capability, at all). She had to go through a metal detector any time she entered the building; good luck sneaking an electronic device past that thing.
It all depends on the environment. Obviously, some places are "locked down" more than others.
Santa Monica enacted a local ordinance to that effect. Within 24 hours, all of the banks directed their ATM's to only dispense cash to customers of that particular bank. In short, you couldn't go to a Bank of America ATM and get money from an Commerce Bank account.
Reference
Basically, they legislated what the banks could and couldn't charge, so the banks reduced services.
Dr. Steven Hawking was getting a tour of the Paramount set for Star Trek: TNG. It was as a result of this tour, and his professed appreciation for Trek, which resulted in his cameo appearance in one of the later episodes (and no, I'm not a big enough Trek geek to remember the title).
Anyhow, at one point during his tour, they were working their way through Engineering, looking at the main warp core, and he signalled they should stop. He tapped on his communicator for a moment or two, and indicated:
"I'm working on one of these."
I agree with you on this one. In my book, GMail has done a lot more than just up the ante on disk space.
I use my GMail account as my primary account. I've got three POP3 accounts at various places, plus Yahoo and Hotmail, but GMail is where my "legit" e-mail goes. The Hotmail account gets handed out anywhere I figure SpamBots are searching, so it usually has 200+ messages, all of them crap. Fitting place for it, if you ask me.
We use Outlook at work (company-mandated). I've got Thunderbird loaded on my machine at home (decent RSS aggregation). And yet, I'd still rather use GMail for e-mail functionality. The message threading is simply too useful.
When e-mail clients can match that functionality, allowing me to follow a thread of e-mail messages as easily as following a thread on Usenet, I'll find a reason to switch. When I can manually reorganize message threads (so that messages with different subjects are included), I'll be even happier. You guys at Google listening?
At first, I thought "labeling" messages was a nuisance. Then, I started using it, and promptly found the older, "folders" method to be constraining.
The fact that I don't have to wait for the whole screen to redraw every time I expand a message is simply icing on the cake.
These aren't just "show-off" features, which look nice on the "features" list but never get used. There are plenty of practical, useful and heavily-used features here. The fact that no one else has copied them (yet) doesn't mean they're a failed trend. It just means the rest of the world hasn't caught up, yet.
The Cluetrain Manifesto says it best. Look, especially, at #74.
Non Sequitur Comic: The Too Much Information Age
In a few weeks? Try now!
The record high price for gasoline (set in August, 1981), adjusted for inflation, works out to $3.08 / gallon in today's money. Stations in the Kansas City area BROKE that record, this afternoon. We do not, by any stretch, have the highest prices in the nation.
Many of our refineries are in the Gulf Coast region, and shut down and/or damaged by Hurricane Katrina. There's the choke point in the supply / demand equation. The price of crude has hit records, as well, but Uncle Sam is releasing some of the Strategic Petroleum Reserve, to try to keep that from going too high.
I have my Gentoo machine set to update itself every three days at about 2 a.m. It has recompiled X at least twice, GLIBC three times, the kernel at least twice, at least one round on Firefox and, recently, OpenOffice. All told, the system averages about five-six hours compile time each week. All of this happens at some ungodly hour of the morning when I'm out cold.
If a tree falls in a forest, and no one hears, does it really make a noise? If a computer recompiles major components on a regular basis, but no one is trying to use it during that timeframe, is it really an annoyance?
Setting it up, on the other hand, can be SO DAMMED ANNOYING. It took me a week to finally get mplayer and ALSA to play nice (I'd never used ALSA before, so there was a learning curve). Luckily, you don't have to recompile the entire 2.6 kernel every time you want to add/remove another module (moved from 2.4 to 2.6 in the process; another learning curve). I'm currently getting some funky behavior with lbxproxy which I still haven't figured out (Debian, which I used to use, still used XFree86, while Gentoo is X.org; yet another learning curve).
Once it's set up, though, it ROCKS! With mplayer on Debian, the system wasn't able to render DivX videos full-screen without dropping frames. GMPlayer on Gentoo can handle it, without dropped frames, at about 70% CPU load. Enlightenment hustles along smoothly and beautifully, even at 1600x1200. There is definitely something to be said for optimized object code.
I'm not sure I'd want to put Gentoo on a server, but it has replaced Debian as my desktop distro of choice. Cron is your friend, even more so with Gentoo.
Here, here! I'm the senior developer around here, with a B.S. in CompSci. The CIS majors around here can play with a point-and-drool GUI developer tool and put together a decent-looking web page with some functionality on the back-end. When the requirements get complicated, they end up re-using something I wrote months ago, or they're in my cubicle, going "how would you tackle this?"
CIS majors are carpenters; they can use standardized tools and standardized components and put together someting which works. Carpenters can do wonderful work, but they're largely limited in their tools and materials. CompSci majors are architects; they can use the standardized stuff, or they can go off the deep end, develop their own components, and leave the carpenters going "whoa! how did you do that?!!!"
I had Scheme, as well as C/C++ and Java, in college. I've also done a significant amount of Perl development in my career. My ASP and
The CIS majors have to go back to their tools and re-write parts of their programs. If they're maintaining something I wrote, they can usually figure out what needs to be tweaked to add stuff. They just don't, typically, write their apps the way I write mine.
If they trained on VB, they have a hard time dealing with C# (which is very similar to Java, C++ and, to a lesser degree, C). Perl might as well be greek. Scheme doesn't even look, to their eyes, like programming. Most of them can't figure out why anyone would want to mess with JavaScript. They can usually hack some HTML, but they'd rather use DreamWeaver or FrontPage (shudder!), instead. Oh, and don't even THINK about tackling
I use vim for all of the above, from VBScript, to SQL, to C#, HTML and JavaScript.
So what does that piece of parchment (the degree in CompSci) do? Not much. What difference is made by the exposure to:
- a variety of programming languages
- a variety of programming styles
- very different programming paradigms (imperative vs functional)
- data structures and theory
that comes with EARNING that piece of paper? All the difference in the world.Programming is arguably a craft, and occasionally art. When I can open a file I haven't touched in months, tweak a couple lines of code and add the new functionality the clients want (in minutes, not hours), I consider that application to be a piece of art. To do that, the code has to be clean, comprehensible and pretty robust. In short, it has to be easy to maintain. If things aren't neatly laid out, and the code isn't very clean, it's NOT easy to maintain. If adding a couple lines causes stuff to break, it is DEFINITELY NOT easy to maintain.
Not everything I create is art. In fact, some of my earlier stuff is, in hindsight, crap. But, as time goes by, and my experience increases, the art-vs-crap ratio keeps improving. How does a programmer increase this ratio? The same way an artist does: exposure (to others' works and re-exposure to his/her own), practice, reflection (this is how you learn to recognize crap) and feedback.
Yes, that definitely qualifies as art. Technical, lethal, graceful and screaming fast, but still art.
http://www.bnr-art.com/aeronaut/valkyrie.htm
http://www.unrealaircraft.com/classics/xb70.php
http://www.labiker.org/xb70.html