Well, you naturally think that way, so it's not surprising that a functional model seams easier.
The thing about an OO design is that object instances can have state (internal per-instance variables) that is implicitly available to member functions (this in C++). That's a big plus.
In your case, if you want a purely functional object, you can have that too! In C++ you'd declare and define a member function operator()().
This would pay off big if you had an object that represented a function with several parameters, some of which you'd want to hold fixed, while you varied the others. You'd use "set" member functions to set the fixed parameters, and the () operator to call the function with the remaining variable ones. This is a slam-dunk if you always want to fix the same parameters.
I can see you thinking, "Yeah, but I can use globals for that, or hard-coded constants." True enough, but then the nature of the function object is polluted by how you want to use it at this particular point in time. There is a price to be paid, of course, to reference the constants within the function proper by dereferencing this, but that's tantamount to deciding "Do I want to do this at compile time or run time?"
Now, C++ templates let you do some neat compile-time hackery to pick the pre-optimized version of a function object class with known parameters as compile time constants or global fixed parameters (static members so as to not pollute the global namespace). And, unfortunately, I don't have time to go into a discussion about type traits and functors, but you can represent functions as objects in C++, and it is espescially elegant syntactically, when those functions have fixed-paramters states.
As for the speech recognition work, the OO methodology came in handy when it came to phonemic graphs and quotient graphs of them divided by particular equivalence relations used for forward estimation functions driving an A* search (the forward estimate obtained by a full Viterbi search over the much smaller quotient graph) -- I could just tell the quotient graph, in it's current state, to digest the next acoustic input frame. Later, I'd ask it for a forward score from a given node at a particular point in time (which it would cache for me).
Functional programming is better suited to solving engineering problems.
Not really, but there is an element of truth to that mantra: most engineers traditionally learned to program in Fortran, and there were some damn good Fortran compilers/libraries to handle the types of matrix manipulations commonly encountered.
You don't need the overhead of OO design for your engineering programs. I would stick to C unless you are trying to do some kind of fancy GUI.
While abstract classes will incur the hit of an extra level of indirection in a function call, and exception handling in C++ can be expensive (as can multiple inheritance), these features only cost you if you use them.
Now, to answer the question at hand, I have helped to design and develop commercial speech recognition products using C++. Of course, there was plenty of C and hand-tuned assembler there as well.
A more concrete example would be the VFS in Linux, as well as the classes of network, block, and character device drivers -- while generally coded in C, they represent the notion of abstract base classes unifying a common interface to many different implementations -- there's no reason C++ could not have been used there.
Well, it's tough to followup in this "click this link for immediate gratification" world, without a "click if you agree" link.
And, if we make it easy to click a link to send a canned email to a representative, well, it's just too easy, now isn't it? Furthermore, there has to be accountability: Does the email actually represent the sentiments of the signer? Is the signer a constituent?
On the one hand, personal letters, that take time to write, have greater impact, because of the effort. On the other hand, a well-written position paper, with thousands of verifiable signatures can be equally powerful, if not more so.
Why not, then, a site which contains position papers, or sample letters to elected representatives on issues of the day, as well as the means to register, and obtain a digital certificate with which to sign such letters?
The site itself could be position-agnostic, merely providing the technology. Position papers could be submitted in a manner similar to slashdot features, with comments, and rework due to feedback, prior to a final version being posted. Or it could be a link farm to similar such papers/letters. One would register once to obtain a digital certificate (yes, that would identify one), and could then sign those papers with which he or she agreed. Papers with a certain number of signatures would then be sent to members of congress, with an emphasis on congress-critters who elicited the most signatures from their constituents. If there were sufficient funding, printed copies could be mailed, though the current status should be available on-line at any time for browsing.
As to your situation: your boss has just given you license to take extra paid breaks. If your boss doesn't like the thought of that, tell him he should come up with a better solution.
Well, I no longer work there for the very reason you mentioned.
Regarding the quoted suggestion: this wouldn't work: I was constrained to get a certain amount of work done in a particular interval of time which translated to a particular coding rate (and my code is quite dense, it it wasn't simply a matter of racking up useless LOCs). I could not simultaneously get the work done in the time required and produce code for review at a slower rate. Yes, it was very much a "do the impossible and if you can't we will fire you without cause" place. Like I said, I stopped working there (3-1/2 years without any time off or compensation was too much). I would have left sooner, but being on an H1-B visa made that difficult.
I recently bought a Sony 32" HDTV-ready set for US$1800 delivered from Crutchfield. Combined with a Sony Sat-HD100 terrestrial/satellite receiver (another $800, from American Satellite) and a Terk TV55 terrestrial HD/Standard TV antenna, it is great.
Couple points, though, that will save you $$$:
While there is a dearth of HDTV programming, there is plenty of DTV programming, even terrestrial. All satellite programming, for example, is digital. The immediate bonus is that all terrestrial digital channels look great, and a great reason to drop cable in favour of an antenna -- if you get enough local digital channels.
16:9 sets add about $1000 to the price, and most broadcast material you will watch is still 4:3. Get a big enough screen and live with letterboxing for the next several years. Sony makes a 36" version of the set I bought, and I would say that 32" is the bare minimum you should consider. Of course, there are bigger and cheaper projection sets, but I never liked them and the convergence problems they have. Your call.
Similarly, you can save money if you buy an HDTV-ready set instead of an HDTV set (the former lacks an HD tuner/decoder). This provides some flexibility in the choice of outboard tuner/decoder and combining such a set with a HD terrestrial/satellite reciver is a nobrainer. The total cost amounts to about the same, but the flexibility is important. RCA makes a 38" set with a built-in HD Satellite receiver, but they have had problems with early versions of that model, and I've heard people complain about the noisy fan (yes), in them. I have no opinion of my own about the RCS sets, but have generally been pleased with the Sony's I've owned.
Important feature #0: Make sure it displays 720p, and possibly 1080i. There are some cheap DTV sets (480i, 480p) that accept HD signals (720p, 1080i) and downsample them. Beware.
Important feature #1: progressive scan component video inputs -- at least two sets (one for sat receiver, one for DVD player). I don't know of any HD sets that don't have this, but it is important.
Important feature #2: a line doubler. This takes interlaced material (like from an analog broadcast, or source) and makes it progressive (i.e. 480i becomes 480p). The result is a sharper-looking picture. Line doublers vary in quality and poor ones can have difficulty with motion. A bonus is that if you have a DVD player with interlaced component output, instead of progressive, the set can "sharpen" (figuratively) the picture.
Important feature #3: On 4:3 sets, make sure that the set actually displays 720 or 1080 lines of resolution on letterboxed material, instead of downsampling to the area between the black bands. This feature goes by various names, and works by cutting the amplitude of the vertical drive to get the letterbox aspect ratio instead of downsampling. Of course, the shadow mask will be the limiting factor in actual resolution.
DirectTV has only 3 HD channels on one of their non-main satellites, so unless you subscribe to HBO, a sports package (I think), or like to watch the demo loop, you won't find much HD material (yet). But, because it is on another satellite, you will need two LNBs and, in most of the U.S., an 18" x 24" elliptical dish with four coax cables (two from each dual-LNB), or wo dishes. Spanish programming (Para Todos) is on a third satellite and requires an additional single-LNB (for a total of three on the dish). The point of all this is that if you get an HDTV or HDTV-ready set and DirectTV (Dish competes with them in the U.S.A., and there are comparable services elsewhere in the world), spring for the twin dual-LNB 18"x24" dish instead of the standard 18" round one -- you won't want to have to redo installation later that way (while the extra LNB and elliptical dish add about $100 to the cost, initial installation is usually free, while a retrofit will probably cost that $100).
All totaled (set, sat rx, dish) I must have spent about $2800. So far (three months later) I am pleased.
Oh, if you do get a satellite system as well, you will have to learn all about multiswitches (satisfied customer plug: Hometech has 5x8 Trunkline multiswitches for about $160).
I didn't grow up in the 50s, but rather the 60s: I was born in 1961, and started school (kindergarten) in 1966. So, from about 1968 to 1974 I spent a lot of free time building things: first with Lego (tm) blocks, then, by the time I was 11 or 12, small electric projects: you know, a battery, light bulbs, and switches, with the battery hidden away in a corner of the room, lights in all sorts of places, and an array of switches in one "control panel".
By the time I was 12 or 13 I discovered high-voltage (having mistaken a potted transformer for a relay and getting a zap off the secondary as I tried to energize the primary "relay coil" with a battery). I quickly built all sorts of high-voltage circuits. I knew enough about AC, DC, and step-up transformers to be dangerous, though it would still be a while before I'd think to charge up old capacitors salvaged from TV sets via half-wave rectifiers (I was too cheap to get 4 diodes for a full-wave bridge). About this time I got an TI SR-52 calculator to help with practical calculations related to my other interests: calculus, physics (espescially general relativity, which, though I had saved up to buy a college level text, was at the limit of my ability to comprehend), and programming. In 1974, I finally had access to a timesharing computer: an HP 2000, accessed via a teletype and acoustic coupled modem in the "terminal room" in my high school. This was heady stuff.
In those days, building anything electrical, or electronic was "something". Oh sure, you could get kits, but I wanted to design my own stuff, and the hand-me-down parts I scrounged from my father proved handy. The fact was that anything technical was rare. LCDs didn't exist and LEDs were a novelty.
We graduated from a black and white TV to colour around 1967, and cable TV shortly thereafter. I remember the first UHF stations, and the difficulty to receive them. My father had a "Super-8" movie camera, and projector for making home movies. I remember him having to send 35mm and movie film away for processing and waiting a couple of weeks to get it back. There were no VCRs, walkmans, or answering machines in the late 1960s. Eight track tape decks were a big thing.
Things started changing slowly in the early to mid-1970s.
"We" had been to the moon, and it was clear that change was afoot. I got a mono cassette deck for Christmas, 1972, though I still coveted my father's open real 2 track 7-1/2 ips reel to reel deck. Open reel decks were to be prefered for "serious" recording for some time after that, though 15ips and stereo. During high school, I had progressed to designing radio-controlled devices and anoying the heck out of neighbours by remote control (remotely-exploded firecrackers in the flower beds, anyone?). Getting time on the HP2000 was a major priority, especially when we got an upgrade to a 300 baud modem and a DECwriter. The big thing for geeks to do then was design multi-terminal spacewar games (text-based)... in BASIC... with files for inter-process communication. Then, in 1975, something big happened: The MITS Altair.
I never got an Altair, but my father "worked with someone who knew someone who knew someone who worked for someone who got one for his business", seriously. By this time, I had progressed to logic circuits (designing my own state-machine based gizmos, and had multi-channel remote control down to a tee. I got a chance to see the Altair and play with it. I ended up writing much of the initial code for an "invoice" program for the owner of the machine. And what a beauty it was! 16K RAM, and a CRT terminal (about $3000), glowing nicely with blue lettering. BASIC took 8 minutes to load from cassette tape (one of my first projects was to build a second cassette interface), and the invoicing program about 30 seconds. And the printer! Sure, it was upper case only, but it printed at 1200 Baud! (well, it accepted data that fast, most of the time). During lunch, when I didn't have school, I got to play with that computer.
This brings us to about 1979, when I started on my B.Comp.Sc. degree program. Notice anything? From 1965 to 1979 there were very few innovative things! Colour TV, calculators, bulky video cameras, and finally the first VHS and Beta VCRs. Computers were a real expensive hobby item, only for garage tinkerers, though the Apple ][ was starting to make an impact, as well as this thing called CP/M, which brought order to hardware abstraction. I turned 18 in 1979, the legal start of adulthood where I lived. So, during the course of my childhood, very little change occured. And I thought, at the time, what modern times and things are being invented! After all, the big things for my parents' generation were cars, planes, electricity (with it, phones, radio, and TV) -- I had seen as many new and cool things during my childhood as they had seen during their whole lives.
My daughter is now 8-1/2 years old. Since 1993 she has seen the advent of the internet (email, www, on-line shopping and payment), cell phones becoming ubiquitous, DVDs, hand-held electronic games, satellite TV. Most importantly, I grew up in an analog world and her's is most certainly digital. My son probably won't even remember the time of the dial-up ISP: we dropped it in favour of DSL when he was 14 months old -- a whole connectivity paradigm shift in half his sister's childhood.
When people lament that their kids' lives are on some kind of techno-amphetamine induced fast-forward, with little time for imaginative play, I wonder if its just that their world changes so much faster than that of their parents', that keeping up and absorbing all of it results in a short span of attention to any one thing, lest something else be missed. There is no time to imagine: when you return to reality, it will have changed so much, you won't recognize it.
I do not expect this trend to continue indefinately. Moore's Law and Quantum Physics will collide at some point. Until that is reached however, we will progress at breakneck speed through a socio-economic upheaval that can probably be compared to the industrial revolution. Entire industries will disappear, though not without resistance. New companies will fight to dominate in a digital, information-rich world. Our kids will be caught up in this turmoil, and the usual effects that social upheaval causes, fighting for rights that were impossible to curtail a decade ago, that the point in their lives where they barely realize that such rights are important.
The world is in the middle of a technological transformation that will result in a new one, without breakneck change, or at least a change of an implementative and not architectural nature. Once that happens, kids will once again be able to take a breather and let their imagination run wild in that frontier of new possibilities.
I have in the past managed 25000 lines of code per month, two months running. But this involved 16 hour days, 6 days a week.
The code proved damn robust, but I could not maintain that level of production much longer. Then again, it really was about 50k lines of code and was done at the end of those two hellish months, so there was no more to do. I suppose if I had a greater vision, and a solid architecture, work could have progressed.
Oh, it was a mix of C and 80x86 assembler.
The point is, that, yes, 5000 lines per week is doable, but something I doubt I could maintain for a year.
And this is the same government that wants to flaunt the constitution that gives it any sense of legitimacy?
Give me one good reason why one should not take up arms against such a monstrous institution.
Never mind... I can think of two:
1) This was probably dreamed up by some power-hungry flunky and does not reflect on the government as a whole. As much as seeing myself write those words makes me want to retch, even the government deserves the benefit of the doubt.
2) Taking the law into one's own hands is an act of vigilanteeism: never a good idea unless there is widespread support for one's position (and the mechanics and institutions of the supposed law corrupt).
So, I exercise the restraint that all law abiding people do, but remain ever watchful. It is a painful lesson that one can't simultaneously love the principles of the constitution and the institution empowered to defend it.
/me gets out the heater (man, it's cold in this thread).
They are in fact a kind of commmunist activity, embedded in a capitalist framework, but recognisably communist (rather than, say, Socialist, even).
And that's exactly where I disagree, because the motivation is generally to scratch a personal itch, and not for some greater good inconsistent with the scratching of that itch.
The problem I have with your definitions is that anything not directly capitalist, i.e. invested for the purpose of obtaining greater investable capital, is presumed (by you) to be non-capitalist.
To wit, I would not use the term communist to describe open-source software development, but rather cooperative, because it is an investment with a personal payoff expectation. Rather capitalist, if you ask me.
A capitalist system is certainly open to such cooperation, but capitalist activities can not flourish to any great extent within a communist system, lest the needs of the one start to displace his obligations to the many. In a capitalist system, if the individual wishes to frivolously waste his capital, or altruistically give it to others, this does not undermine the system itself. Yet, if one undertakes capitalist investment activity within a communist system, the system is threatened.
I would consider capitalism a more flexible system, then, though I will admit that many situations can arise which are localy optimal (in a temporal sense), that are undesirable and the subject of attack by communist detractors of capitalism. Excessive perpetuation of artificial scarcity is one such cause of difficulty.
Yes, this has been fun. Oh, and thanks for not ripping me up over mixing economics (capitalism vs. communism) and politics (individualism vs. socialism). I was being a bit imprecise there even though socialism and communism do tend to go hand in hand.
Sorry, I find your points somewhat rambling, but I will try to respond.
First, my friend was not using the 99% forced unemployment as an example of socialism -- he was arguing that in a socialist society, where the unemployable are supported, it is not morally wrong to deny them work because of incompetence, and might be more efficient, overall.
Socialism is about the most good for the community, not how to make one person more productive. The most good for a community actually involves having most people feeling productive. When people do not feel productive, they get bored. When they get bored, they start mucking around. When people start mucking around, the bell shaped curve idicates that a certain percentage will turn to crime. The more the people who are bored, the longer the tail on the curve...
How can you define what is "good for the community", if not in terms of what is good for individual members? Further, "feeling productive" does not put food on the table -- being productive does. And when people do not see the fruits of their labour as benifiting them directly, they tend to be less productive. The horse requires the carrot, in other words.
Socialism is not about raping the productive individual, but rather, the individual not raping the society.
Ah, but "raping society" is invariably defined as "not sharing the fruits of your labour" often to a degree decided by others. I'd define "raping society" as living off the fruits of others' labour, without their consent. Let me provide an example: my father died because he could not save up enough money to travel to the U.S. for a lifesaving operation -- his taxes having gone to pay for a social healthcare system that had collapsed to the point of not being technically capable of providing the required surgery (which, admittedly has only a 70% recovery rate). He certainly earned enough money over his life to pay for that operation, but, alas, it was taken and given to others. Should we debate whether society is better off because, perhaps, someone else's life was saved?, or even more lives were saved? Socialism, as practiced, in Canada, reached the point where the state decided who lived and who died (there being a celebrated case of someone dying of a broken leg because they had to wait in line too long for care). Note: I am not bitter because of my father's death -- I was a libertarian for many years before that event, but I do note that Canada's brand of socialism certainly accelerated it, and surely the deaths of others.
To counter your argument, just because someone is born a cripple, is that reason to make them lve on the street?
They can live wherever they can afford, or from what charity is available to them. Certainly I am not responsible for their state of affairs? Should others have to feed me because I am disadvantaged in that I am not a stellar athlete? I have also met many so called "cripples" that were quite capable of supporting themselves at least as well as I support myself.
Using your text, I suspect that the answer would be euthanasia, but, as I recall, that is illegal in America....
Why kill someone if they do not want to die? I do not understand.
Anyway, to make this vaguely on topic, if you brush your teeth, as dentists recommend, twice a day, three minutes a time, that is roughly 3 hours a month just wasted. Better kill all those people with clean teeth too.
Now, I really don't understand the relevance of this.
If you drink one coffee a day, assuming 1 minute to walk to the machine, get the coffe etc, then you are wasting a 1/2 hour per month. Better kill those people who like coffee as well. They are just robbing you of your riches!
Oh, I see, you're equating inefficiency in others with a lack of productivity in the productive. That wasn't my friend's point. Such people have a neutral effect (unless the coffe machine crowd is so noisy as to be a distraction). His point was that many workers actively impede productivity. A better example would be a policy (and yes, this is an exagerated example) where one had to change all 1's in a file to 2's, by hand, and being forbidden to use sed to automate the process because (a) not everyone knows sed, and (b) it makes other's "look bad".
"We" is the invarying productive component of society, in this case (beholden to all of society, according to socialists).
The general idea is if you pay people to not work and mess up or slow down what productive people do, those productive people would be so much more efficient that they could earn enough to pay the non-workers.
IOW: Factory employs 1000 people to make widgets. Do the investors care if they can make the same widgets by paying the same total wages to 10 really efficient people who automate most everything? Probably not: those 10 people are worth 100 times the average. So pay each of those people 100 times as much, and tax them at a 99% rate to "pay off" the other 990 to not have to work.
Of course, why should some work while others get a free ride just because it is possible? So those others don't make our working lives miserable? To some extent this may actually be worthwhile, but not as a general principle.
I can cite a personal example here: I produce pretty good code with below average defect rates at above average coding rates. (After 25 years, you'd think that I'd better!) In many shops, we had formal review processes. I was asked to reduce my output because it was taking too much of others' time to keep up reviewing it. This, despite taking on the coding efforts of many of those other reviewers in the process. The right thing to do, IMHO, would be to relax review requirements where there was a history of low defect rates. But, PHBs being what they are...
What a stinky way to crack encryption. But, I suppose someone has to do the dirty jobs.
OffTopic: In Czech, "shit" is "hovono" (phonetically), and "to suck" is "tzUtzit" [hard U]. The machine that cleans out septic tanks, therefore, is colloquialy called a "hovono-tzUtz".
A socialist friend of mine used that argument to justify a 90-90% unemployment rate: if we could just feed, clothe, and house people who messed up the work we did, we'd be more than 20 times more efficient with the same effort, and could easily afford the astronomical taxes to support the welfare state that would result.
Being a libertarian, I said a few disparaging words upon hearing this suggestion, but I have to admit, he had a point.
Of course the proper counter is, having increased my efficiency 2000% or more, why shouldn't I reap the benefits? Because you didn't stop me? Is that a threat? There are other ways of getting threatening freeloaders out of the way...
I suppose the files that are GPL with "Red Hat" in them are O.K., but not all the content is GPL. In particular, many of the images (and the Red Hat "hat" logo) are copyright (and possibly trademarked, like the "hat") Red Hat.
While at Teradyne, I helped produce a Red Hat derived O/S for a product we were building, and these were serious issues (The derived O/S? "Teradyne GNU/Linux" -- we had less trouble with GNU than with Red Hat).
Think Yaz's (Yazoo in the U.K.) "Upstairs at Eric's" (recorded in E.C. Radcliffe's studio), and "You and Me Both". These combined Alyson "Alf" Moyet's voice with Vince Clarke's (from Depeche Mode) synth-pop sound (rumour has it that she was a club act looking to sing lead for a "rootsy blues outfit" and he said, "I can be that").
For it's time, it was dance-able club music, but with meaningful lyrics, and a lead that could actually sing, despite her, what?, 300 pound bulk?
Sadly, Ms. Moyet considers it the worst stuff she's done. Me, I think the exact opposite.
And what's wrong with waking up and smelling the coffee?
At some point, you're beholden to others: you have to earn a living, insure against tragic loss, etc. But, you save a little for retirement, reach the point of not needing insurance (estate tax issues aside), and generally become independent of those to whom you were beholden in the past (employers, insurers, etc.)
You now have time to reflect: did they treat you fairly? Did the relationship over time appear equitable, or was one side ("them") excessively leveraging your vulnerabilities against you? Do other people in the same situation think so as well?
If the answer is yes, now that the "ties that bind" are broken, as it were, maybe it is time to voice your opinions of the injustice that was perpetuated, and try to end it for people now in the same position as you were.
I keep reading and hearing of absurd recording contracts, with no chance for legal review, and I can't help but think, "who would bend over so far for a recording contract?" Someone naive, vulnerable, and desperate, that's who.
As much as I am a libertarian, and think that whatever the market will bear is fair, I do not particularly like participants in a free market that (a) leverage their counterpart's vulnerabilities to their advantage, and more importantly (b) fight tooth and nail to prevent their counterparts from seeking alternatives. It's like the baker saying, "Oh, you're hungry? Well, the bread costs twice as much today as yesterday." Accepting it might be fair, but it certainly isn't nice, and I, for one, prefer to do business with nice people.
The problem with your approach is that it completely ignores the non-investment benefits that capitalism provides. I prefer to view those benefits as natural within a capitalist system. This does not mean that they do not fit in other systems, and some would argue that socialist systems more directly address consumptive needs, but the question, of course, is do they do so more effectively? Clearly I think not.
Witness that too often, capitalists are criticized for not being "sensitive" to comsumptive needs, as if they are somehow non-capitalist, and so ignored.
Note also, that, while non-invenstment activities are consistent with, and part of, capitalist endevours, they are not the primary activity of capitalist systens, and by themselves, or with little choise in invenstment, do not a capitalist system make.
So, no the serf, or slave, is not a capitalist, per se, but this does not mean that he does not undertake activities harmonious with such a system.
Capitalism persists when there is a preponderance of activities indicative of an investment of capital, clearly more than merely consumptive ones.
Slashdot Mirror
The thing about an OO design is that object instances can have state (internal per-instance variables) that is implicitly available to member functions (this in C++). That's a big plus.
In your case, if you want a purely functional object, you can have that too! In C++ you'd declare and define a member function operator()(). This would pay off big if you had an object that represented a function with several parameters, some of which you'd want to hold fixed, while you varied the others. You'd use "set" member functions to set the fixed parameters, and the () operator to call the function with the remaining variable ones. This is a slam-dunk if you always want to fix the same parameters.
I can see you thinking, "Yeah, but I can use globals for that, or hard-coded constants." True enough, but then the nature of the function object is polluted by how you want to use it at this particular point in time. There is a price to be paid, of course, to reference the constants within the function proper by dereferencing this, but that's tantamount to deciding "Do I want to do this at compile time or run time?"
Now, C++ templates let you do some neat compile-time hackery to pick the pre-optimized version of a function object class with known parameters as compile time constants or global fixed parameters (static members so as to not pollute the global namespace). And, unfortunately, I don't have time to go into a discussion about type traits and functors, but you can represent functions as objects in C++, and it is espescially elegant syntactically, when those functions have fixed-paramters states.
As for the speech recognition work, the OO methodology came in handy when it came to phonemic graphs and quotient graphs of them divided by particular equivalence relations used for forward estimation functions driving an A* search (the forward estimate obtained by a full Viterbi search over the much smaller quotient graph) -- I could just tell the quotient graph, in it's current state, to digest the next acoustic input frame. Later, I'd ask it for a forward score from a given node at a particular point in time (which it would cache for me).
Not really, but there is an element of truth to that mantra: most engineers traditionally learned to program in Fortran, and there were some damn good Fortran compilers/libraries to handle the types of matrix manipulations commonly encountered.
You don't need the overhead of OO design for your engineering programs. I would stick to C unless you are trying to do some kind of fancy GUI.
While abstract classes will incur the hit of an extra level of indirection in a function call, and exception handling in C++ can be expensive (as can multiple inheritance), these features only cost you if you use them.
Now, to answer the question at hand, I have helped to design and develop commercial speech recognition products using C++. Of course, there was plenty of C and hand-tuned assembler there as well.
A more concrete example would be the VFS in Linux, as well as the classes of network, block, and character device drivers -- while generally coded in C, they represent the notion of abstract base classes unifying a common interface to many different implementations -- there's no reason C++ could not have been used there.
If the idea caught on, an email, with thousands of verifiable signatures should be just as good.
And, if we make it easy to click a link to send a canned email to a representative, well, it's just too easy, now isn't it? Furthermore, there has to be accountability: Does the email actually represent the sentiments of the signer? Is the signer a constituent?
On the one hand, personal letters, that take time to write, have greater impact, because of the effort. On the other hand, a well-written position paper, with thousands of verifiable signatures can be equally powerful, if not more so.
Why not, then, a site which contains position papers, or sample letters to elected representatives on issues of the day, as well as the means to register, and obtain a digital certificate with which to sign such letters?
The site itself could be position-agnostic, merely providing the technology. Position papers could be submitted in a manner similar to slashdot features, with comments, and rework due to feedback, prior to a final version being posted. Or it could be a link farm to similar such papers/letters. One would register once to obtain a digital certificate (yes, that would identify one), and could then sign those papers with which he or she agreed. Papers with a certain number of signatures would then be sent to members of congress, with an emphasis on congress-critters who elicited the most signatures from their constituents. If there were sufficient funding, printed copies could be mailed, though the current status should be available on-line at any time for browsing.
Well, I no longer work there for the very reason you mentioned.
Regarding the quoted suggestion: this wouldn't work: I was constrained to get a certain amount of work done in a particular interval of time which translated to a particular coding rate (and my code is quite dense, it it wasn't simply a matter of racking up useless LOCs). I could not simultaneously get the work done in the time required and produce code for review at a slower rate. Yes, it was very much a "do the impossible and if you can't we will fire you without cause" place. Like I said, I stopped working there (3-1/2 years without any time off or compensation was too much). I would have left sooner, but being on an H1-B visa made that difficult.
IIRC, their reject pile would pass Q/A for a Nikon Nikor lens, which is hardly junk.
Couple points, though, that will save you $$$:
While there is a dearth of HDTV programming, there is plenty of DTV programming, even terrestrial. All satellite programming, for example, is digital. The immediate bonus is that all terrestrial digital channels look great, and a great reason to drop cable in favour of an antenna -- if you get enough local digital channels.
16:9 sets add about $1000 to the price, and most broadcast material you will watch is still 4:3. Get a big enough screen and live with letterboxing for the next several years. Sony makes a 36" version of the set I bought, and I would say that 32" is the bare minimum you should consider. Of course, there are bigger and cheaper projection sets, but I never liked them and the convergence problems they have. Your call.
Similarly, you can save money if you buy an HDTV-ready set instead of an HDTV set (the former lacks an HD tuner/decoder). This provides some flexibility in the choice of outboard tuner/decoder and combining such a set with a HD terrestrial/satellite reciver is a nobrainer. The total cost amounts to about the same, but the flexibility is important. RCA makes a 38" set with a built-in HD Satellite receiver, but they have had problems with early versions of that model, and I've heard people complain about the noisy fan (yes), in them. I have no opinion of my own about the RCS sets, but have generally been pleased with the Sony's I've owned.
Important feature #0: Make sure it displays 720p, and possibly 1080i. There are some cheap DTV sets (480i, 480p) that accept HD signals (720p, 1080i) and downsample them. Beware.
Important feature #1: progressive scan component video inputs -- at least two sets (one for sat receiver, one for DVD player). I don't know of any HD sets that don't have this, but it is important.
Important feature #2: a line doubler. This takes interlaced material (like from an analog broadcast, or source) and makes it progressive (i.e. 480i becomes 480p). The result is a sharper-looking picture. Line doublers vary in quality and poor ones can have difficulty with motion. A bonus is that if you have a DVD player with interlaced component output, instead of progressive, the set can "sharpen" (figuratively) the picture.
Important feature #3: On 4:3 sets, make sure that the set actually displays 720 or 1080 lines of resolution on letterboxed material, instead of downsampling to the area between the black bands. This feature goes by various names, and works by cutting the amplitude of the vertical drive to get the letterbox aspect ratio instead of downsampling. Of course, the shadow mask will be the limiting factor in actual resolution.
DirectTV has only 3 HD channels on one of their non-main satellites, so unless you subscribe to HBO, a sports package (I think), or like to watch the demo loop, you won't find much HD material (yet). But, because it is on another satellite, you will need two LNBs and, in most of the U.S., an 18" x 24" elliptical dish with four coax cables (two from each dual-LNB), or wo dishes. Spanish programming (Para Todos) is on a third satellite and requires an additional single-LNB (for a total of three on the dish). The point of all this is that if you get an HDTV or HDTV-ready set and DirectTV (Dish competes with them in the U.S.A., and there are comparable services elsewhere in the world), spring for the twin dual-LNB 18"x24" dish instead of the standard 18" round one -- you won't want to have to redo installation later that way (while the extra LNB and elliptical dish add about $100 to the cost, initial installation is usually free, while a retrofit will probably cost that $100).
All totaled (set, sat rx, dish) I must have spent about $2800. So far (three months later) I am pleased.
Oh, if you do get a satellite system as well, you will have to learn all about multiswitches (satisfied customer plug: Hometech has 5x8 Trunkline multiswitches for about $160).
By the time I was 12 or 13 I discovered high-voltage (having mistaken a potted transformer for a relay and getting a zap off the secondary as I tried to energize the primary "relay coil" with a battery). I quickly built all sorts of high-voltage circuits. I knew enough about AC, DC, and step-up transformers to be dangerous, though it would still be a while before I'd think to charge up old capacitors salvaged from TV sets via half-wave rectifiers (I was too cheap to get 4 diodes for a full-wave bridge). About this time I got an TI SR-52 calculator to help with practical calculations related to my other interests: calculus, physics (espescially general relativity, which, though I had saved up to buy a college level text, was at the limit of my ability to comprehend), and programming. In 1974, I finally had access to a timesharing computer: an HP 2000, accessed via a teletype and acoustic coupled modem in the "terminal room" in my high school. This was heady stuff.
In those days, building anything electrical, or electronic was "something". Oh sure, you could get kits, but I wanted to design my own stuff, and the hand-me-down parts I scrounged from my father proved handy. The fact was that anything technical was rare. LCDs didn't exist and LEDs were a novelty.
We graduated from a black and white TV to colour around 1967, and cable TV shortly thereafter. I remember the first UHF stations, and the difficulty to receive them. My father had a "Super-8" movie camera, and projector for making home movies. I remember him having to send 35mm and movie film away for processing and waiting a couple of weeks to get it back. There were no VCRs, walkmans, or answering machines in the late 1960s. Eight track tape decks were a big thing.
Things started changing slowly in the early to mid-1970s. "We" had been to the moon, and it was clear that change was afoot. I got a mono cassette deck for Christmas, 1972, though I still coveted my father's open real 2 track 7-1/2 ips reel to reel deck. Open reel decks were to be prefered for "serious" recording for some time after that, though 15ips and stereo. During high school, I had progressed to designing radio-controlled devices and anoying the heck out of neighbours by remote control (remotely-exploded firecrackers in the flower beds, anyone?). Getting time on the HP2000 was a major priority, especially when we got an upgrade to a 300 baud modem and a DECwriter. The big thing for geeks to do then was design multi-terminal spacewar games (text-based)... in BASIC... with files for inter-process communication. Then, in 1975, something big happened: The MITS Altair.
I never got an Altair, but my father "worked with someone who knew someone who knew someone who worked for someone who got one for his business", seriously. By this time, I had progressed to logic circuits (designing my own state-machine based gizmos, and had multi-channel remote control down to a tee. I got a chance to see the Altair and play with it. I ended up writing much of the initial code for an "invoice" program for the owner of the machine. And what a beauty it was! 16K RAM, and a CRT terminal (about $3000), glowing nicely with blue lettering. BASIC took 8 minutes to load from cassette tape (one of my first projects was to build a second cassette interface), and the invoicing program about 30 seconds. And the printer! Sure, it was upper case only, but it printed at 1200 Baud! (well, it accepted data that fast, most of the time). During lunch, when I didn't have school, I got to play with that computer.
This brings us to about 1979, when I started on my B.Comp.Sc. degree program. Notice anything? From 1965 to 1979 there were very few innovative things! Colour TV, calculators, bulky video cameras, and finally the first VHS and Beta VCRs. Computers were a real expensive hobby item, only for garage tinkerers, though the Apple ][ was starting to make an impact, as well as this thing called CP/M, which brought order to hardware abstraction. I turned 18 in 1979, the legal start of adulthood where I lived. So, during the course of my childhood, very little change occured. And I thought, at the time, what modern times and things are being invented! After all, the big things for my parents' generation were cars, planes, electricity (with it, phones, radio, and TV) -- I had seen as many new and cool things during my childhood as they had seen during their whole lives.
My daughter is now 8-1/2 years old. Since 1993 she has seen the advent of the internet (email, www, on-line shopping and payment), cell phones becoming ubiquitous, DVDs, hand-held electronic games, satellite TV. Most importantly, I grew up in an analog world and her's is most certainly digital. My son probably won't even remember the time of the dial-up ISP: we dropped it in favour of DSL when he was 14 months old -- a whole connectivity paradigm shift in half his sister's childhood.
When people lament that their kids' lives are on some kind of techno-amphetamine induced fast-forward, with little time for imaginative play, I wonder if its just that their world changes so much faster than that of their parents', that keeping up and absorbing all of it results in a short span of attention to any one thing, lest something else be missed. There is no time to imagine: when you return to reality, it will have changed so much, you won't recognize it.
I do not expect this trend to continue indefinately. Moore's Law and Quantum Physics will collide at some point. Until that is reached however, we will progress at breakneck speed through a socio-economic upheaval that can probably be compared to the industrial revolution. Entire industries will disappear, though not without resistance. New companies will fight to dominate in a digital, information-rich world. Our kids will be caught up in this turmoil, and the usual effects that social upheaval causes, fighting for rights that were impossible to curtail a decade ago, that the point in their lives where they barely realize that such rights are important.
The world is in the middle of a technological transformation that will result in a new one, without breakneck change, or at least a change of an implementative and not architectural nature. Once that happens, kids will once again be able to take a breather and let their imagination run wild in that frontier of new possibilities.
The code proved damn robust, but I could not maintain that level of production much longer. Then again, it really was about 50k lines of code and was done at the end of those two hellish months, so there was no more to do. I suppose if I had a greater vision, and a solid architecture, work could have progressed.
Oh, it was a mix of C and 80x86 assembler.
The point is, that, yes, 5000 lines per week is doable, but something I doubt I could maintain for a year.
Kudos.
Give me one good reason why one should not take up arms against such a monstrous institution.
Never mind... I can think of two:
1) This was probably dreamed up by some power-hungry flunky and does not reflect on the government as a whole. As much as seeing myself write those words makes me want to retch, even the government deserves the benefit of the doubt.
2) Taking the law into one's own hands is an act of vigilanteeism: never a good idea unless there is widespread support for one's position (and the mechanics and institutions of the supposed law corrupt).
So, I exercise the restraint that all law abiding people do, but remain ever watchful. It is a painful lesson that one can't simultaneously love the principles of the constitution and the institution empowered to defend it.
...except when Joe Bloggs puts up a website, he is publishing it publicly.
refactor as a result of learning from your mistakes and redundancies;
and try to minimize the busy parts (where all developers have a hand) when things change (like lists of unique symbols, numbers, etc.)
They are in fact a kind of commmunist activity, embedded in a capitalist framework, but recognisably communist (rather than, say, Socialist, even).
And that's exactly where I disagree, because the motivation is generally to scratch a personal itch, and not for some greater good inconsistent with the scratching of that itch.
The problem I have with your definitions is that anything not directly capitalist, i.e. invested for the purpose of obtaining greater investable capital, is presumed (by you) to be non-capitalist.
To wit, I would not use the term communist to describe open-source software development, but rather cooperative, because it is an investment with a personal payoff expectation. Rather capitalist, if you ask me.
A capitalist system is certainly open to such cooperation, but capitalist activities can not flourish to any great extent within a communist system, lest the needs of the one start to displace his obligations to the many. In a capitalist system, if the individual wishes to frivolously waste his capital, or altruistically give it to others, this does not undermine the system itself. Yet, if one undertakes capitalist investment activity within a communist system, the system is threatened.
I would consider capitalism a more flexible system, then, though I will admit that many situations can arise which are localy optimal (in a temporal sense), that are undesirable and the subject of attack by communist detractors of capitalism. Excessive perpetuation of artificial scarcity is one such cause of difficulty.
Yes, this has been fun. Oh, and thanks for not ripping me up over mixing economics (capitalism vs. communism) and politics (individualism vs. socialism). I was being a bit imprecise there even though socialism and communism do tend to go hand in hand.
First, my friend was not using the 99% forced unemployment as an example of socialism -- he was arguing that in a socialist society, where the unemployable are supported, it is not morally wrong to deny them work because of incompetence, and might be more efficient, overall.
Socialism is about the most good for the community, not how to make one person more productive. The most good for a community actually involves having most people feeling productive. When people do not feel productive, they get bored. When they get bored, they start mucking around. When people start mucking around, the bell shaped curve idicates that a certain percentage will turn to crime. The more the people who are bored, the longer the tail on the curve...
How can you define what is "good for the community", if not in terms of what is good for individual members? Further, "feeling productive" does not put food on the table -- being productive does. And when people do not see the fruits of their labour as benifiting them directly, they tend to be less productive. The horse requires the carrot, in other words.
Socialism is not about raping the productive individual, but rather, the individual not raping the society.
Ah, but "raping society" is invariably defined as "not sharing the fruits of your labour" often to a degree decided by others. I'd define "raping society" as living off the fruits of others' labour, without their consent. Let me provide an example: my father died because he could not save up enough money to travel to the U.S. for a lifesaving operation -- his taxes having gone to pay for a social healthcare system that had collapsed to the point of not being technically capable of providing the required surgery (which, admittedly has only a 70% recovery rate). He certainly earned enough money over his life to pay for that operation, but, alas, it was taken and given to others. Should we debate whether society is better off because, perhaps, someone else's life was saved?, or even more lives were saved? Socialism, as practiced, in Canada, reached the point where the state decided who lived and who died (there being a celebrated case of someone dying of a broken leg because they had to wait in line too long for care). Note: I am not bitter because of my father's death -- I was a libertarian for many years before that event, but I do note that Canada's brand of socialism certainly accelerated it, and surely the deaths of others.
To counter your argument, just because someone is born a cripple, is that reason to make them lve on the street?
They can live wherever they can afford, or from what charity is available to them. Certainly I am not responsible for their state of affairs? Should others have to feed me because I am disadvantaged in that I am not a stellar athlete? I have also met many so called "cripples" that were quite capable of supporting themselves at least as well as I support myself.
Using your text, I suspect that the answer would be euthanasia, but, as I recall, that is illegal in America....
Why kill someone if they do not want to die? I do not understand.
Anyway, to make this vaguely on topic, if you brush your teeth, as dentists recommend, twice a day, three minutes a time, that is roughly 3 hours a month just wasted. Better kill all those people with clean teeth too.
Now, I really don't understand the relevance of this.
If you drink one coffee a day, assuming 1 minute to walk to the machine, get the coffe etc, then you are wasting a 1/2 hour per month. Better kill those people who like coffee as well. They are just robbing you of your riches!
Oh, I see, you're equating inefficiency in others with a lack of productivity in the productive. That wasn't my friend's point. Such people have a neutral effect (unless the coffe machine crowd is so noisy as to be a distraction). His point was that many workers actively impede productivity. A better example would be a policy (and yes, this is an exagerated example) where one had to change all 1's in a file to 2's, by hand, and being forbidden to use sed to automate the process because (a) not everyone knows sed, and (b) it makes other's "look bad".
"We" is the invarying productive component of society, in this case (beholden to all of society, according to socialists).
The general idea is if you pay people to not work and mess up or slow down what productive people do, those productive people would be so much more efficient that they could earn enough to pay the non-workers.
IOW: Factory employs 1000 people to make widgets. Do the investors care if they can make the same widgets by paying the same total wages to 10 really efficient people who automate most everything? Probably not: those 10 people are worth 100 times the average. So pay each of those people 100 times as much, and tax them at a 99% rate to "pay off" the other 990 to not have to work.
Of course, why should some work while others get a free ride just because it is possible? So those others don't make our working lives miserable? To some extent this may actually be worthwhile, but not as a general principle.
I can cite a personal example here: I produce pretty good code with below average defect rates at above average coding rates. (After 25 years, you'd think that I'd better!) In many shops, we had formal review processes. I was asked to reduce my output because it was taking too much of others' time to keep up reviewing it. This, despite taking on the coding efforts of many of those other reviewers in the process. The right thing to do, IMHO, would be to relax review requirements where there was a history of low defect rates. But, PHBs being what they are...
What a stinky way to crack encryption. But, I suppose someone has to do the dirty jobs.
OffTopic: In Czech, "shit" is "hovono" (phonetically), and "to suck" is "tzUtzit" [hard U]. The machine that cleans out septic tanks, therefore, is colloquialy called a "hovono-tzUtz".
Being a libertarian, I said a few disparaging words upon hearing this suggestion, but I have to admit, he had a point.
Of course the proper counter is, having increased my efficiency 2000% or more, why shouldn't I reap the benefits? Because you didn't stop me? Is that a threat? There are other ways of getting threatening freeloaders out of the way...
I tried. Oh how I tried! But it was not ready for prime time then, and I didn't have the time to hack it to make it work.
I suppose the files that are GPL with "Red Hat" in them are O.K., but not all the content is GPL. In particular, many of the images (and the Red Hat "hat" logo) are copyright (and possibly trademarked, like the "hat") Red Hat.
While at Teradyne, I helped produce a Red Hat derived O/S for a product we were building, and these were serious issues (The derived O/S? "Teradyne GNU/Linux" -- we had less trouble with GNU than with Red Hat).
If archives were available, say on CD- or DVD-ROM, that would be nice. Even for $$$, naturally.
Now, Charlotte Church can SING!
Heh, tell that to the dung beatle.
Seriously, though, I do acknowledge your point, but do not have the time to discuss it now.
Think Yaz's (Yazoo in the U.K.) "Upstairs at Eric's" (recorded in E.C. Radcliffe's studio), and "You and Me Both". These combined Alyson "Alf" Moyet's voice with Vince Clarke's (from Depeche Mode) synth-pop sound (rumour has it that she was a club act looking to sing lead for a "rootsy blues outfit" and he said, "I can be that").
For it's time, it was dance-able club music, but with meaningful lyrics, and a lead that could actually sing, despite her, what?, 300 pound bulk?
Sadly, Ms. Moyet considers it the worst stuff she's done. Me, I think the exact opposite.
At some point, you're beholden to others: you have to earn a living, insure against tragic loss, etc. But, you save a little for retirement, reach the point of not needing insurance (estate tax issues aside), and generally become independent of those to whom you were beholden in the past (employers, insurers, etc.)
You now have time to reflect: did they treat you fairly? Did the relationship over time appear equitable, or was one side ("them") excessively leveraging your vulnerabilities against you? Do other people in the same situation think so as well?
If the answer is yes, now that the "ties that bind" are broken, as it were, maybe it is time to voice your opinions of the injustice that was perpetuated, and try to end it for people now in the same position as you were.
I keep reading and hearing of absurd recording contracts, with no chance for legal review, and I can't help but think, "who would bend over so far for a recording contract?" Someone naive, vulnerable, and desperate, that's who.
As much as I am a libertarian, and think that whatever the market will bear is fair, I do not particularly like participants in a free market that (a) leverage their counterpart's vulnerabilities to their advantage, and more importantly (b) fight tooth and nail to prevent their counterparts from seeking alternatives. It's like the baker saying, "Oh, you're hungry? Well, the bread costs twice as much today as yesterday." Accepting it might be fair, but it certainly isn't nice, and I, for one, prefer to do business with nice people.
The problem with your approach is that it completely ignores the non-investment benefits that capitalism provides. I prefer to view those benefits as natural within a capitalist system. This does not mean that they do not fit in other systems, and some would argue that socialist systems more directly address consumptive needs, but the question, of course, is do they do so more effectively? Clearly I think not.
Witness that too often, capitalists are criticized for not being "sensitive" to comsumptive needs, as if they are somehow non-capitalist, and so ignored.
Note also, that, while non-invenstment activities are consistent with, and part of, capitalist endevours, they are not the primary activity of capitalist systens, and by themselves, or with little choise in invenstment, do not a capitalist system make.
So, no the serf, or slave, is not a capitalist, per se, but this does not mean that he does not undertake activities harmonious with such a system.
Capitalism persists when there is a preponderance of activities indicative of an investment of capital, clearly more than merely consumptive ones.