You don't seem to have much experience with RF connectors. An RF connector that performs to 6 GHz, fits in the hinge, is durable under the torsional flexing that the hinge exerts on it, and is low-loss, plus the overhead in installing it and testing it, in combination with all the digital connectors, can add easily $20 to the retail price.
But I do agree with part of what you say. I have been speaking in terms of any individual maker implementing such a feature. If the whole industry standardized (like happened with USB or 1394), then an otherwise expensive connector can be made cheap by expensive production tooling. But that is not the direction the industry is moving.
If they really wanted to be clever, they could wipe the key after breaking it into 100 factors, any 20 of which would be enough to re-construct the key, then distribute that to 100 independent 3rd parties. That way, if they want to update the code, they could sign it only after scores of outsiders have reviewed it. It should be hard to sneak a back-door past that. If the government wants the code-signing key, they have to serve subpoenas to a large number of outsiders for the key factors without someone blabbing the fact and blowing the investigation's cover.
There is no GOOD reason to make the top half irreplacable by a reasonably skilled person
Not good enough maybe, but they do have have reasons. Connectorizing the hinge takes up space, adds cost, and degrades RF performance. Just guessing on the cost numbers here, but is it worth adding $20 to every unit (more if you go through contortions to recover the lost RF performance) to save $100 in repair costs on the 5% that will have cracked screens? Not unless you place a premium on quality.
But then, if the market were to put a premium on quality, the marketshare of Mac/Linux vs. Windows would be the reverse of what is actually is.
Get real RS232 from a real RS232 company, B&B. They have been doing RS232 gizmos of every sort since the 80's. PCMCIA, USB, parallel to RS232, multiple ports, long distance extenders, RS422, RS485, serial for PDA's, you name it. Not cheap though.
Is there something intrinsically magical about the screen hinge and graphics connection of a laptop that keeps them forever joined
Yes, and it will only get worse in the upcoming years. One of the many constraints in laptop design is routing the cables through the hinge. You have a back light and its control, and all the crazy data & clock lines (not analog video) for the LCD display. Now with WLAN you have co-axial cable, since since real-world experience has shown that locating the antenna up high is worth the cable losses. The trend is to put more stuff up there, like webcams, where the machine can see, and the microphone, further from those fans whose noise everyone is complaining about in posts here. And more antennas, for WWAN, TV, DVB, UWB, blah, blah, blah.
Not just Mac either. I can personally attest that it was quite easy for Amigans to browse the web and email by the time Bill Gates wrote The Road Ahead, which (before its redaction) barely mentioned the internet.
The biggest hurdle was finding an ISP that didn't suck (i.e. wasn't AOL, which refused to port its proprietary client to Amiga anyway) and wasn't long-distance. By the 90's, long-haul long distance was cheaper than short-haul, so my first ISP was 2,000 miles away in Silicon Valley. Web-browsing was an after-11pm affair, though in those days there was as much on Gopher as on the Web. My small-town environment didn't get a local ISP until 1994.
Well, the key exchange can be made secure on untrusted hardware, if you use one-shot keys or a challenge-response protocol. But that is really a case of extending the unsecured part of the network from the ethernet interface to the human-computer interface. You still have trusted hardware: the blinky-number keyfob, or simply a trusted sheet of paper with crypto-keys.
To make the whole session secure, all the displays and responses would have to be encrypted, possible in principle by directing all web browsing though an encrypting service, but rather cumbersome. You would have formulate encrypted codes ahead of time for the transactions you want to initiate, and/or print out your gibberish-looking bank transactions & take them home to decrypt with your WW2-style code book. Basically a return to card-punch batch programming.
Fair enough I suppose. The president isn't required to convene a joint session. But the OP was driving at denying the president an audience for the SotU. So in that sense, Congress is required to convene if the president calls for it. As for the "extraordinary occasion" bit, the threshold for convening Congress is much lower with the advent of mechanized travel and electronic communication. Washington isn't so far away anymore; even California reps can commute on weekends.
You say you don't want to play childish games, but then immediately call for an investigation of a case where every salient fact has been well established for months. The leak was Richard Armitage. The prosecutor knew this before Libby was called to testify. Plame was not covert in the legal sense (though someone at the CIA called her "covert", apparently in some looser sense). Investigate what? Sounds like yet another meaningless Beltway political game.
We shall be hearing "I don't recall" a lot more now.
Libby's conviction was for testimony in front of a grand jury. You don't have an attorney present and are not permitted to check documentation, and you get no cross-examination of opposing testimony. But if what you say is inconsistent...Bam! Criminal liability.
The precedent of Libby's conviction makes the lesson clear. Whatever you do, never, ever co-operate with a grand jury investigation. Deny your memory, take the fifth, halt the proceedings to consult your attorney (the only way to see an attorney is to stop everything and exit, then come back in), answer every question minimally, and generally drag your feet on every point. In other words, the way to avoid charges of obstructing justice is to obstruct justice, but in a passive-aggressive way.
That by itself is adequate reason to pardon Libby. Not because Libby necessarily deserves a full pardon, but because this conviction screws the grand jury system up even more than it already is.
Re:Slackware... ironic that it's too much effort
on
Slackware 12.0 Released
·
· Score: 3, Informative
That isn't unique to Slackware by any measure. Even Debian, with its fat 14-CD install set, can't include everything. You are more likely to find what you need in a Debian package, but once you step outside the repository, it's tough going. With Slack's simpler layout and simpler package tools, at least you have a chance when rolling your own.
AFAICT, there is no silver bullet, only trade-offs. Even Windows will get hamstrung by dependencies if you stray too far from the mainstream, though commercial Windows software often packages tons of OS updates with it, solving your dependency problems but often breaking previous software installs. Debian's repository is a walled garden, and they try to mitigate the pain of the walls by making the garden really, really big. Slackware is all open, but that leaves you on your own sometimes. RPM-type distros are somewhere in-between. Choose distro most suited for your needs. If, like me, you want to do lots of oddball stuff, automated package managers create more problems than they solve. But if you can live comfortably without ever installing things outside the repository, then Debian is for you.
There is no constitutional requirement for the state of the union to be a speech.
From Article 2 Section 3
He shall from time to time give to the Congress Information of the State of the Union, and recommend to their Consideration such Measures as he shall judge necessary and expedient; he may, on extraordinary Occasions, convene both Houses...
Sounds to me like the president has the power to convene Congress for the purpose of giving a speech. George Washington thought so too; the timing & manner of delivering the State of the Union dates back to the Washington administration.
Parce mihi, "Spare me," everyone here remembers Latin grammar worse than I do. vir, plural viri, masculine, 2nd declension, "man" virus, 2nd declension neuter, "venom" or "toxicicity" bacterium, 2nd declension neuter, plural bacteria
We don't have any surviving attestations to a plural of virus, or AFAIK, any -us 2nd declension neuter. The meaning is by nature a collective noun: you don't hear "Potassiums" very often, do you? Or is it Potassia? My own sense is that the neuter-plural-ends-in-a wins out over -us-ending-takes-i-for-plural. After all, it's unus nauta not una nauta: grammatical gender wins out over consistent form. Cicero and Vergil aren't around anymore to ask, so that is the best we can do.
argue points to favour their bias like they're barracking for sports teams
The US has a fixed election cycle, & right now is probably the second-worst time in the cycle for intelligent political talk. We are 6 years into the maximum of 8 for a presidency, so the top talent is already starting to wander out to search for the next winning horse to back. It's the lame-duck effect: You can expect lots of execution mistakes over the next year & a half, no matter how fastidious Dubya is individually. And yet we are still far enough from the election that the wider population is not really paying attention yet. So the political talk-sphere is filled with players trying to force a watershed moment, hoping that, if they yell loud enough, they can license the OS for the new IBM PC while the other guys are out flying a private plane (so to speak). Less partisan & more thoughtful types are still silent, waiting to hear what the emerging candidates say and balance the arguments of the partisans.
So don't worry too much. It will get better as the election season heats up and the yellow-dog partisans are outed for what they are. Until election day draws really near. I said "second-worst" above. The worst time for political debate is a few days before the election. At that time, media hit-men will, in an effort gain an edge in a close race, release ads that appeal to fear, knowing there is not enough time for counter-arguements to diffuse out. Though Dems have blown this tactic the last couple of cycles (like the Repubs did in '96); due to lack of rank & file discipline, appeals to fear appeared too early, & the Repubs, taking advantage of faster news cycles, skewered them for it.
I think the reason for that, as far as demographics go, and leaving out the historical and regulatory factors (beyond the scope of this post), is that high population-density areas in the US typically contian low-income people, whereas the higher income groups--the ones who buy broadband--live in exurban McManisons. The European cities I have seen seem to have the reverse pattern, swanky condos in the city and prosaic block housing at the city edges.
There are exceptions both ways though, I wonder how good broadband service is in American urban boutique neighborhoods (like Bexley or Shaker Heights, OH), or in the Islamic slums of Europe. Information on that would test my little theory.
removing the democratically elected leader of Iran and replacing him with the Shah
This bit of leftist mythos is a pet peeve for me, probably because it had me fooled for awhile. That "democratically elected leader" was a Communist, and, in the manner of all Communists who come to power by election, promptly set about destroying the democratic institutions that brought him to power. At the risk of invoking Godwin's Law, even Hitler was duly & democratically elected chancellor by the Reichstag. Does that mean Eisenhower should have stopped at the Rhine?
If the lobbying groups don't buy influence, what exactly is it they are spending money on?
They don't have to buy much influence. When 533 people are controlling $2 trillion, you only have to steer a very small amount power to be worth spending a good-sized pile of money. Or you keep a lobbying firm on retainer for years, just to have them available to put in your side of the story on the once-in-a-century moment when it can make a difference. It's sorta like patents, they're usually not good for anything, but the megacorps cultivate piles as a form of insurance.
In the corporate world many things are printed and never read. I had a tech lead years ago that swore he put a photocopied page from a russian engineering textbook in every large report he ever submitted to management - never got asked about it.
The classic example of this gambit is the Write Only Memory. According to the Jargon File, somebody got annoyed with all the required apporovals where nothing was actually being checked, so he put out a totally illogical spec. Heh, it specifies a standard filament voltage of 6.3 volts AC (compatible with your tubes, y'know).
Re:But what about the battery?
on
The Zune Cometh
·
· Score: 1
I work (when I'm not blathering on./) at a place that makes components for cellphones, so I see many different models come & go. Sometimes a CDMA and GSM phone will be derived from a common platform, but that seems to be the case mostly for high-end, lower volume smartphones. I've seen Nokia go as far as to make a SIM card for a CDMA phone (don't know if that idea made it to production though). The big volume cheap-o's, however, cater less to consumers and more to the network operators, who want a phone that will help with their dropped call statistics, help promote money-making servies, and so on, so these phones will be purpose-designed for one type of network or another. If Cingular thinks some GSM phone has appeal, they don't want to see the Verizion dealer carrying the same thing in a CDMA version.
I suspect the main reason CDMA phones still have replaceable batteries is not so much commonality in phone design, but commonality in battery design. And to a lesser extent commonality in the manufacturing process, and connector design. CDMA phones may be about half the market in the US, but they are a much smaller fraction globally, so why futz around with separate supply contracts and process flows?
Re:But what about the battery?
on
The Zune Cometh
·
· Score: 1
I think you are missing one major reason cellphone batteries are replaceable: To remove SIM cards safely, the power must be disconnected. So the SIM slot is arranged so that the battery must be removed to change out the SIM.
I have seen cellphones (smallish Sony-Ericsson T68 or some such) that required a screwdriver (a Torx T-6 for extra annoyance) to change the batery (and SIM, unless it was a CDMA type without a SIM...been too long I can't remember). It was a pain in the butt.
The Gimp can also display (instructions) images to simulate the various forms of colorblindness (though in that case you would have to do a static screen capture of your stuff). For much more./ style prattling on developing for colorblindness, an ask./ item was posted on this topic a while back.
If you're really good, hack your xorg.conf to simulate colorblindness. But don't ask me how to do it.
I think you are right on all those points, but I think it is easier for people to understand when described thusly: First-past-the-post forces coalition-building out of the legislature and into the two major parties. Successful canidiates in FPTP must already have a coalition behind them to get a full majority in whatever district they represent. American primary elections often have the multipolar flavor found in elections for European proportional legislatures. You get the such-and-such "caucus" or the whatchmacallit "council", groups organized to get certain kinds of candidates nominated, playing the ideologically pure roles found in European-style political parties.
And the GP is ignoring the UK (merely the birthplace of modern democracy), whose elections are FPTP.
As for my own view, I think that proportional representation is fine on a smaller scale, but it seems to me that, because candidates are insulated from the voters (who vote for parties), as the size of the nations (and the parties) scales up, corrupt (or ideologically extreme) individuals can entrench themselves into the party machinery and get away with much more mischief. FPTP scales better. The US is, after all, much larger than Germany.
I wouldn't say Unix. What baffles me about the 80's is why companies like Phoenix & AMI, having sucessfully clean-room cloned IBM's BIOS, didn't take the obvious next step and clone MS-DOS (obvious in hindsight maybe). I suppose part of it was they had to clone BIOS because IBM wouldn't sell or license, whereas DOS was (falsely) proffered as an open platform. There certainly was a strong institutional memory in the industry to not trust IBM to be open about anything. But 8 years dragged by before DR-DOS appeared on the scene.
With a variety of DOS providers all implementing a de facto industry standard, the following evolution of desktop OS'es would have been vastly different. My best guess is that it would have followed a path similar to the one trod by Unix a decade earlier (and the web a decade later): gradually growing features in a modular fashion and deprecating superannuated cruft all according to a vague sort of consensual best-practices doctrine. What actually happened was a bunch of comapanies (like Phar-Lap for example) made DOS extenders to use more than 640k, provide GUI toolkits, and so on, the start of a unix-like evolution. But all of these extenders depended on MS-DOS. When Win 3.0 (whose DOS box tolerated extenders rather poorly) came bundled with PC's, the multipolar competitive evironment crashed and MS monoculture set in. We have been stuck with it ever since. People in the OS-enhancement aftermarket fled for the hills: Nobody in their right mind would spend money to develop a Win 3.0 extender.
I would guess that the Amtrak system works due to self-selection bias. Their error rates improved as people like me gave up, and the only people left using it are those whose mode of speaking matches the recognition algorithms' assumptions.
They used to have an excellent DTMF system, but the new system doesn't even give the option of falling back to something that actually works. For me, Amtrak doesn't have a phone number anymore; if I can't get a web browser or call an agent, I have to find another mode.
I don't know if it is the tambre of my voice, or that I find machine voice cognitively unnatural. When I speak, I form sentences, and listen for cues the other party understands the concept I am attempting to convey. These machines make me feel like I'm talking to a tree, or in the case of Julie (which gives phony conversational feedback) like I'm talking to a sociopath. I get the urge launch into a Euro-style multilingual greeting....Dicisne linguam Latinae? At least DTMF systems are honest about the formalistic nature of the interaction.
Slashdot Mirror
You don't seem to have much experience with RF connectors. An RF connector that performs to 6 GHz, fits in the hinge, is durable under the torsional flexing that the hinge exerts on it, and is low-loss, plus the overhead in installing it and testing it, in combination with all the digital connectors, can add easily $20 to the retail price.
But I do agree with part of what you say. I have been speaking in terms of any individual maker implementing such a feature. If the whole industry standardized (like happened with USB or 1394), then an otherwise expensive connector can be made cheap by expensive production tooling. But that is not the direction the industry is moving.
If they really wanted to be clever, they could wipe the key after breaking it into 100 factors, any 20 of which would be enough to re-construct the key, then distribute that to 100 independent 3rd parties. That way, if they want to update the code, they could sign it only after scores of outsiders have reviewed it. It should be hard to sneak a back-door past that. If the government wants the code-signing key, they have to serve subpoenas to a large number of outsiders for the key factors without someone blabbing the fact and blowing the investigation's cover.
There is no GOOD reason to make the top half irreplacable by a reasonably skilled person
Not good enough maybe, but they do have have reasons. Connectorizing the hinge takes up space, adds cost, and degrades RF performance. Just guessing on the cost numbers here, but is it worth adding $20 to every unit (more if you go through contortions to recover the lost RF performance) to save $100 in repair costs on the 5% that will have cracked screens? Not unless you place a premium on quality.
But then, if the market were to put a premium on quality, the marketshare of Mac/Linux vs. Windows would be the reverse of what is actually is.
Get real RS232 from a real RS232 company, B&B. They have been doing RS232 gizmos of every sort since the 80's. PCMCIA, USB, parallel to RS232, multiple ports, long distance extenders, RS422, RS485, serial for PDA's, you name it. Not cheap though.
Is there something intrinsically magical about the screen hinge and graphics connection of a laptop that keeps them forever joined
Yes, and it will only get worse in the upcoming years. One of the many constraints in laptop design is routing the cables through the hinge. You have a back light and its control, and all the crazy data & clock lines (not analog video) for the LCD display. Now with WLAN you have co-axial cable, since since real-world experience has shown that locating the antenna up high is worth the cable losses. The trend is to put more stuff up there, like webcams, where the machine can see, and the microphone, further from those fans whose noise everyone is complaining about in posts here. And more antennas, for WWAN, TV, DVB, UWB, blah, blah, blah.
Hey, I paid 99 cents to make that my ring tone. No other phone in the room would ring like that.
third party basic compiler
Blitz Basic. AMOS was another one, though I think it came a bit later.
Not just Mac either. I can personally attest that it was quite easy for Amigans to browse the web and email by the time Bill Gates wrote The Road Ahead, which (before its redaction) barely mentioned the internet.
The biggest hurdle was finding an ISP that didn't suck (i.e. wasn't AOL, which refused to port its proprietary client to Amiga anyway) and wasn't long-distance. By the 90's, long-haul long distance was cheaper than short-haul, so my first ISP was 2,000 miles away in Silicon Valley. Web-browsing was an after-11pm affair, though in those days there was as much on Gopher as on the Web. My small-town environment didn't get a local ISP until 1994.
Well, the key exchange can be made secure on untrusted hardware, if you use one-shot keys or a challenge-response protocol. But that is really a case of extending the unsecured part of the network from the ethernet interface to the human-computer interface. You still have trusted hardware: the blinky-number keyfob, or simply a trusted sheet of paper with crypto-keys.
To make the whole session secure, all the displays and responses would have to be encrypted, possible in principle by directing all web browsing though an encrypting service, but rather cumbersome. You would have formulate encrypted codes ahead of time for the transactions you want to initiate, and/or print out your gibberish-looking bank transactions & take them home to decrypt with your WW2-style code book. Basically a return to card-punch batch programming.
Fair enough I suppose. The president isn't required to convene a joint session. But the OP was driving at denying the president an audience for the SotU. So in that sense, Congress is required to convene if the president calls for it. As for the "extraordinary occasion" bit, the threshold for convening Congress is much lower with the advent of mechanized travel and electronic communication. Washington isn't so far away anymore; even California reps can commute on weekends.
You say you don't want to play childish games, but then immediately call for an investigation of a case where every salient fact has been well established for months. The leak was Richard Armitage. The prosecutor knew this before Libby was called to testify. Plame was not covert in the legal sense (though someone at the CIA called her "covert", apparently in some looser sense). Investigate what? Sounds like yet another meaningless Beltway political game.
We shall be hearing "I don't recall" a lot more now.
Libby's conviction was for testimony in front of a grand jury. You don't have an attorney present and are not permitted to check documentation, and you get no cross-examination of opposing testimony. But if what you say is inconsistent...Bam! Criminal liability.
The precedent of Libby's conviction makes the lesson clear. Whatever you do, never, ever co-operate with a grand jury investigation. Deny your memory, take the fifth, halt the proceedings to consult your attorney (the only way to see an attorney is to stop everything and exit, then come back in), answer every question minimally, and generally drag your feet on every point. In other words, the way to avoid charges of obstructing justice is to obstruct justice, but in a passive-aggressive way.
That by itself is adequate reason to pardon Libby. Not because Libby necessarily deserves a full pardon, but because this conviction screws the grand jury system up even more than it already is.That isn't unique to Slackware by any measure. Even Debian, with its fat 14-CD install set, can't include everything. You are more likely to find what you need in a Debian package, but once you step outside the repository, it's tough going. With Slack's simpler layout and simpler package tools, at least you have a chance when rolling your own.
AFAICT, there is no silver bullet, only trade-offs. Even Windows will get hamstrung by dependencies if you stray too far from the mainstream, though commercial Windows software often packages tons of OS updates with it, solving your dependency problems but often breaking previous software installs. Debian's repository is a walled garden, and they try to mitigate the pain of the walls by making the garden really, really big. Slackware is all open, but that leaves you on your own sometimes. RPM-type distros are somewhere in-between. Choose distro most suited for your needs. If, like me, you want to do lots of oddball stuff, automated package managers create more problems than they solve. But if you can live comfortably without ever installing things outside the repository, then Debian is for you.
There is no constitutional requirement for the state of the union to be a speech.
From Article 2 Section 3
Sounds to me like the president has the power to convene Congress for the purpose of giving a speech. George Washington thought so too; the timing & manner of delivering the State of the Union dates back to the Washington administration.
Parce mihi, "Spare me," everyone here remembers Latin grammar worse than I do.
vir, plural viri, masculine, 2nd declension, "man"
virus, 2nd declension neuter, "venom" or "toxicicity"
bacterium, 2nd declension neuter, plural bacteria
We don't have any surviving attestations to a plural of virus, or AFAIK, any -us 2nd declension neuter. The meaning is by nature a collective noun: you don't hear "Potassiums" very often, do you? Or is it Potassia? My own sense is that the neuter-plural-ends-in-a wins out over -us-ending-takes-i-for-plural. After all, it's unus nauta not una nauta: grammatical gender wins out over consistent form. Cicero and Vergil aren't around anymore to ask, so that is the best we can do.
And yes, virii, is just awful, delicto summo.
The US has a fixed election cycle, & right now is probably the second-worst time in the cycle for intelligent political talk. We are 6 years into the maximum of 8 for a presidency, so the top talent is already starting to wander out to search for the next winning horse to back. It's the lame-duck effect: You can expect lots of execution mistakes over the next year & a half, no matter how fastidious Dubya is individually. And yet we are still far enough from the election that the wider population is not really paying attention yet. So the political talk-sphere is filled with players trying to force a watershed moment, hoping that, if they yell loud enough, they can license the OS for the new IBM PC while the other guys are out flying a private plane (so to speak). Less partisan & more thoughtful types are still silent, waiting to hear what the emerging candidates say and balance the arguments of the partisans.
So don't worry too much. It will get better as the election season heats up and the yellow-dog partisans are outed for what they are. Until election day draws really near. I said "second-worst" above. The worst time for political debate is a few days before the election. At that time, media hit-men will, in an effort gain an edge in a close race, release ads that appeal to fear, knowing there is not enough time for counter-arguements to diffuse out. Though Dems have blown this tactic the last couple of cycles (like the Repubs did in '96); due to lack of rank & file discipline, appeals to fear appeared too early, & the Repubs, taking advantage of faster news cycles, skewered them for it.
There are exceptions both ways though, I wonder how good broadband service is in American urban boutique neighborhoods (like Bexley or Shaker Heights, OH), or in the Islamic slums of Europe. Information on that would test my little theory.
This bit of leftist mythos is a pet peeve for me, probably because it had me fooled for awhile. That "democratically elected leader" was a Communist, and, in the manner of all Communists who come to power by election, promptly set about destroying the democratic institutions that brought him to power. At the risk of invoking Godwin's Law, even Hitler was duly & democratically elected chancellor by the Reichstag. Does that mean Eisenhower should have stopped at the Rhine?
If the lobbying groups don't buy influence, what exactly is it they are spending money on?
They don't have to buy much influence. When 533 people are controlling $2 trillion, you only have to steer a very small amount power to be worth spending a good-sized pile of money. Or you keep a lobbying firm on retainer for years, just to have them available to put in your side of the story on the once-in-a-century moment when it can make a difference. It's sorta like patents, they're usually not good for anything, but the megacorps cultivate piles as a form of insurance.
In the corporate world many things are printed and never read. I had a tech lead years ago that swore he put a photocopied page from a russian engineering textbook in every large report he ever submitted to management - never got asked about it.
The classic example of this gambit is the Write Only Memory. According to the Jargon File, somebody got annoyed with all the required apporovals where nothing was actually being checked, so he put out a totally illogical spec. Heh, it specifies a standard filament voltage of 6.3 volts AC (compatible with your tubes, y'know).
I work (when I'm not blathering on ./) at a place that makes components for cellphones, so I see many different models come & go. Sometimes a CDMA and GSM phone will be derived from a common platform, but that seems to be the case mostly for high-end, lower volume smartphones. I've seen Nokia go as far as to make a SIM card for a CDMA phone (don't know if that idea made it to production though). The big volume cheap-o's, however, cater less to consumers and more to the network operators, who want a phone that will help with their dropped call statistics, help promote money-making servies, and so on, so these phones will be purpose-designed for one type of network or another. If Cingular thinks some GSM phone has appeal, they don't want to see the Verizion dealer carrying the same thing in a CDMA version.
I suspect the main reason CDMA phones still have replaceable batteries is not so much commonality in phone design, but commonality in battery design. And to a lesser extent commonality in the manufacturing process, and connector design. CDMA phones may be about half the market in the US, but they are a much smaller fraction globally, so why futz around with separate supply contracts and process flows?
I think you are missing one major reason cellphone batteries are replaceable: To remove SIM cards safely, the power must be disconnected. So the SIM slot is arranged so that the battery must be removed to change out the SIM.
I have seen cellphones (smallish Sony-Ericsson T68 or some such) that required a screwdriver (a Torx T-6 for extra annoyance) to change the batery (and SIM, unless it was a CDMA type without a SIM...been too long I can't remember). It was a pain in the butt.
The Gimp can also display (instructions) images to simulate the various forms of colorblindness (though in that case you would have to do a static screen capture of your stuff). For much more ./ style prattling on developing for colorblindness, an ask ./ item was posted on this topic a while back.
If you're really good, hack your xorg.conf to simulate colorblindness. But don't ask me how to do it.
And the GP is ignoring the UK (merely the birthplace of modern democracy), whose elections are FPTP.
As for my own view, I think that proportional representation is fine on a smaller scale, but it seems to me that, because candidates are insulated from the voters (who vote for parties), as the size of the nations (and the parties) scales up, corrupt (or ideologically extreme) individuals can entrench themselves into the party machinery and get away with much more mischief. FPTP scales better. The US is, after all, much larger than Germany.
With a variety of DOS providers all implementing a de facto industry standard, the following evolution of desktop OS'es would have been vastly different. My best guess is that it would have followed a path similar to the one trod by Unix a decade earlier (and the web a decade later): gradually growing features in a modular fashion and deprecating superannuated cruft all according to a vague sort of consensual best-practices doctrine. What actually happened was a bunch of comapanies (like Phar-Lap for example) made DOS extenders to use more than 640k, provide GUI toolkits, and so on, the start of a unix-like evolution. But all of these extenders depended on MS-DOS. When Win 3.0 (whose DOS box tolerated extenders rather poorly) came bundled with PC's, the multipolar competitive evironment crashed and MS monoculture set in. We have been stuck with it ever since. People in the OS-enhancement aftermarket fled for the hills: Nobody in their right mind would spend money to develop a Win 3.0 extender.
They used to have an excellent DTMF system, but the new system doesn't even give the option of falling back to something that actually works. For me, Amtrak doesn't have a phone number anymore; if I can't get a web browser or call an agent, I have to find another mode.
I don't know if it is the tambre of my voice, or that I find machine voice cognitively unnatural. When I speak, I form sentences, and listen for cues the other party understands the concept I am attempting to convey. These machines make me feel like I'm talking to a tree, or in the case of Julie (which gives phony conversational feedback) like I'm talking to a sociopath. I get the urge launch into a Euro-style multilingual greeting....Dicisne linguam Latinae? At least DTMF systems are honest about the formalistic nature of the interaction.