As near as I can tell, all the usenet groups in big8 and alt have roughly the same amount of spam per group. (There may be some specific groups that get extra spam, but all of the groups I've looked at are about the same.)
But how much you _notice_ it definitely depends on the group. In particular, it depends on the level of traffic the group normally carries. High-traffic groups drown out the spam by sheer volume. If you read a very popular group, the kind where new messages roll in almost faster than you can read them, the spam is such a low percantage of that huge bulk that it gets positively buried under all the off-topic threads.
On the other hand, if you spend a much smaller amount of time each day reading a dozen low-traffic groups, spam will dominate your usenet experience. For one thing, you get mostly the same spam messages in all groups, only with munging to prevent automatic duplicate suppression. In addition, if the group you are reading only gets two or three legit messages a week, five spam messages a day seems like a lot in comparison.
> It never ceases to amaze me that somewhere there is someone who is > glad to have received spam and buys something from it.
People are idiots. A _lot_ of otherwise-surprising things can be explained if you operate under the assumption that people are idiots.
I'm not saying _all_ people are idiots _all_ of the time, but a significant percentage of people are idiots (or behave in a manner that makes them indistinguishable from idiots) a significant percentage of the time.
> Somewhere someones eyes just light up when see that 5th "** Very > Important Message **" turn up in thier mailbox. I just can't grok > that. I would love to see a photo of some of these customers.
They would look, to the untrained eye, just like regular people.
> There is no good reason why its not illegal either.
We can only hope that spam law will eventually catch up with fax law, but I should note that where I work we receive a number of junk faxes every week, and people I work with have contemplated making purchases as a result.
> They restrict what telemarketers are legally allowed to do. They > can't keep calling you over & over with the same pitch,
They can, and they do. Although they usually wait long enough that you would have a hard time knowing it's the same company calling again, since you've had the same salespitch meanwhile from several _other_ telemarketing firms...
> but you can be spammed countless times.
This is mostly because of the fundamental ecconomic differences between telemarketing and sending spam: the telemarketing firm has to pay a human being to be on the line with you. They have to do this for every person they actually reach (albeit not for every attempted call). Spammers don't have that problem.
> I'm glad there are people out there making these spammers lives > hell. More power to them:) Hopefully someone will have the balls to > just start serial killing these spammers.
First off, I did confuse Verdana with Georgia, but it was early
in the morning...
> Fixed-width fonts mean the spacing between characters is equal.
Really? I thought it meant every character was a _different_
width... </sarcasm>
> there are no mostly-sans font types
Andale Mono is mostly sans-serif (meaning, most of the characters
don't have serifs). However, certain select characters have
serifs, in order to distinguish them from otherwise similar
characters. Compare the following chars: Il1!| O0 If any
two of them look the same, your font fails to adequately make
all the distinctions it should. This won't matter for the end
user doing nothing but word processing, but for many uses it is
totally unacceptable. (Granted, most of these uses are geeky,
but they still matter.) Fonts that accomodate this via select
serifs but omit serifs on most letters are mostly-sans.
Furthermore, most of the uses that require every character to
be distinct also absolutely require a fixed-width font. If
you don't understand why a fixed-width font might be needed,
then you don't understand source-code indentation. This is
why we need a good quality mostly-sans fixed-width font, similar
to Andale Mono. Alternately, we could use a serif fixed font
like Courier New, except that serifs are ugly for large blocks
of regular-sized text. Don't get me wrong, they're great for
headlines and stuff, but I do NOT want to look at sixty
consecutive lines of seriffed text. (Yes, I know the publishing
industry does it all the time, but that's one of the reasons
I prefer reading stuff on-screen to a physical book. The other
reason is of course the search feature. The only redeeming
feature of dead-tree books is that you can take them to the
bathtub or wherever, conveniently.) The reason many fixed
fonts are mostly-sans is because the need to make every char
distinct and the need to make them line up happen to both be
important for many of the same uses. You say fixed fonts
should only be used on screen, but that's preposterous. Any
text that needs to be fixed-width on the screen needs to be
fixed-width on paper, too. This includes almost all source
code, all ASCII art of more than one line, and a large amount
of what is posted to usenet, as well as a lot of the email
that geeks exchange with one another. (End-user email can
probably be shown in a proportional font without problems.)
One last thing: if you are about to post a reply advocating
tab stops for program indentation, please first read
this example that I posted to usenet some time back, explaining
why tab stop indentation is inadequate for many situations.
> Unfortunately most clients/browsers seem to go out of their > way to discourage self-signed certificates with error messages > that sound like "This certificate was self-signed.
Yes, and at that point the user's eyes glaze over and if he doesn't have a guru to call, he clicks any button at random. VERY few users would deign to read the entire message. The dialog probably has "Okay" and "Cancel", plus the close box on the window frame. Since "Okay" is the default button, it's highlighted, and hitting "Enter" will select it too, so there's probably at _least_ a one in three chance the user will hit "Okay". That's on the first try. What is more, if the desired result is not achieved the first time, most users will try again and hit a different button.
Translation: SSL certs only matter to people who care about security and privacy.
This is not helped any by the fact that older browsers used to display a dialog that looked basically identical to the users whenever any information was sent over an unencrypted socket -- for example, every time the user did a web search at an http site like Yahoo! Users who have been around for a few years have learned to just bop Okay whenever they see that dialog -- and they teach this behavior to the newer users.
So users who don't know anything about security or privacy (i.e., almost everyone) are fairly unlikely to be dissuaded from visiting a site just because the certificate is invalid. They're WAY more likely to skip a site because it uses a plugin that didn't come preinstalled, or takes too long to load during peak hours.
Re:What about my toaster.....
on
LinuXbox Boots
·
· Score: 3, Funny
Ugh, so complicated. If you have Emacs installed with toast-mode set to autoload, you can just M-x make-toast. The first time you do this, you'll probably want to do M-x customize-group toast-mode and set up the various toast-mode variables to your liking. The defaults are reasonable for making wheat toast, but if you keep your toaster loaded with multigrain bread as I do, you may find that you need to set toast-mode-toast-threshhold a little higher, because the bread toasts more slowly than ordinary wheat bread.
Someone a while back was working on an enhancement to read in the toast darkness from the toaster's sensors (if you have one of the more advanced models) and thus automatically make the necessary adjustments for different kinds of bread, but I think he ran into a problem where Emacs 20 didn't support something he needed and put it on the back burner. I don't know whether he ever resumed work on it after Emacs 21 came out, but I haven't heard anything about it. Anyway, you want to adjust how dark the bread is toasted according to taste anyway, so this enhancement is really only useful if you don't always stock the same kind of bread.
> The fact that you don't even know to refer to additive primaries > and subtractive primaries reveals your lack of understanding.
Either that, or I understand the additive properties of light and absorptive properties of pigment but didn't think it necessary to get too technical.
> If you really think that CMYK is nothing but a buzzword thrown > around by marketing people, then you have no idea what you're > talking about.
CMYK is a printing system. But saying that a piece of computer software "supports CMYK" is meaningless buzzword speach. The document still won't look the same on paper as it does on the screen, or vice versa.
> The point of being able to work in CMYK is that you get a > very close representation of the printed color on screen.
My turn to laugh. You have bought the marketing line.
> The limitations of a "luminous medium" in representing a > reflective medium are minimal
Even if we go from 32-bit to 64-bit colour, a monitor still can't accurately reproduce flat, non-glossy, non-luminous tones. Representing fluorescent inks (well, some of them) actually works much better than representing a surface with low reflectivity. I'm talking about substances that are not really dark, certainly nothing like black, but just dull looking. They reflect some light, with biases toward certain wavelengths, but mostly they absorb it across the whole spectrum. You can play with gamma correction all day and not get it right. Why do you think photographs of dark unvarnished wood look so lousy on your screen? Things like dark blue yarn and raisins reflect mostly ambient light, a mixture of assorted wavelengths from an entire range (albeit with biases toward certain subranges), that are not at all well modelled via RGB.
Metallic inks are an extreme case, but I was talking about more everyday stuff.
> I'm sure that $22k was for a real workstation, like an IBM zSeries > or an HP Visualize or a Sun Blade 1k/2k (Or U60/U80).
From the article, it seems they were HP-UX systems.
> I'm a sysadmin at a large company and I've got a Blade 1000 on my > desk (with Sun's 24" LCD + XVR-1000 video board, thankyouverymuch:) > Anyway, the LCD is somewhat excessive, but the workstation certainly > isn't. I'm constantly compiling code and doing testing on my desktop > -- I need a good, reliable piece of hardware that'll function under > stress.
A regular, non-overclocked PC will do fine under stress as long as it has adequate cooling. From your description of your job, you are probably in an air-conditioned building, so as long as you don't get a bargain-basement system with cheapo cooling fans that'll go out in three years, you'd probably be fine. Spend the extra twenty bucks for the fans with good bearings, if you're getting a new system and want it to last. The PC is so much cheaper than a heavy workstation that you can afford to get a really _nice_ PC and still save a good deal of money.
> The other thing that nobody mentioned is that that $22,000 > workstation will probably last 6 or 7 years. Not so with that cheap > PC.
What if it lasts half as long? Would it be worth one extra instance of copying everything over to a new system, halfway through the 7 years, to save $20,000? Anyway, my PC is now going on 5 years old, and it's not on its last legs yet. 6 or 7 years is a decent lifespan, but it's not _impressive_. (15 years, now that might impress me.)
> I had one developer who was still using his SPARCstation 10 until > less than a month ago when we replaced it with a spare Ultra 2. Why? > Because it still worked. All he used it for was basically an X > display via SSH into the development boxes....
Yeah, so? If all you're doing is running an X server and sshing into other systems to do your work, a Pentium 75 (current market value approximately the same as a good lunch, except in high-tech areas, where people will probably pay you to take it away if they still have anything so ancient) will do the job with half its cycles tied behind its back while whistling dixie, provided it has a halfway decent graphics card. For maximum productivity you'd probably want to upgrade it with a three-button mouse for around ten bucks, which would be a significant part of the purchase cost of the thing.
CMYK is a color model that only works on absorption media (such as pigment on paper). On a luminous medium (such as a CRT), things fundamentally don't (and can't) work that way. As good as Photoshop is, speaking of its having "support for CMYK" is marketroidese. All this means is that it can convert from RGB formats (which *must* be used on your CRT computer screen) to formats intended for printing. The conversion is necessarily lossy, because ink on paper cannot represent all of the same colours that the computer screen can (and vice versa). Unless you're using phosphorescent paint and viewing it under a blacklight, or some trick along those lines, you can't represent the brightness of the sky (for example) on paper. Similarly, your CRT can never show a truly _flat_ (as in nonglossy, nonluminous) color.
You can throw buzzwords like "CMYK" at this all day long, but an image will NEVER look the same on paper (no, not even on glossy paper, although that's closer) as it does on a CRT monitor, and that's a problem Photoshop can't solve.
LCDs (at their current level of tech) are even worse, because they show colors inconsistently. Perhaps some future technology will allow computers to display both luminous and flat colors on the same display...
While we're on the subject of Photoshop, I agree that Photoshop on Linux is a good thing. Photoshop is very entrenched in the publishing community, and for good reason; it's quality stuff. It also has a pricetag to match, so I surely hope Gimp continues to develop (as it has been doing great so far), for those of us with less expansive budgets. Photoshop may be (and probably is) better, but my take on the matter is that Gimp is _comparable_, which is a tremendous achievement. (I have a friend who does graphics work for a living; he works at Eisenbraun's, a publisher specialising particularly in ancient near-east stuff. He works with Photoshop a lot. He'd been trying out Gimp, and was in some ways (not all ways, but some) impressed with it, and had noted that it had some really nice features Photoshop 6 did not have. (He didn't specify which features.) Then he got the new Photoshop, and they had it, he said, "in spades"). That says to me that the two programs are in roughly the same league, a huge accomplishment. But people who already know Photoshop and have the budget for it will want to stick with it, rather than learn Gimp which, although it's free, is not substantially _better_ than Photoshop (at least, not at this time), surely not better enough to justify a non-programmer to switch.
To me, Photoshop on Linux is a great thing, because it's cross-platform technology, one more step toward separating the decision of what OS to use from the decision of what other software to use -- and THAT is a VERY good thing.
Seriously, IM is basically email with poorer clients, less standardisation, and an intrusive, pop-to-the-front notification feature. You can turn a regular mail service into essentially the same thing (sans compatibility with existing IM clients -- but you get compatibility with the even larger installed-base of standard email) just by adding a feature that pops incomming messages up in a window in front of whatever you were doing. (Whether this is a GOOD feature remains to be debated; I suspect it would be good in some corporate settings and maybe for people who don't get much mail and crave it, and merely annoying for most of us. But if the notification feature can be toggled at will, everybody wins.) Call it an "email pager" if you want.
IM, with its smaller, fragmented installed-base, really has very little to offer to the discerning communicator. Most of the other hyped features it boasts can be replaced with MIME, which is older and more standardised and supported by most major email clients these days.
Not that IM is likely to go away. If I understand the article, only the server has to pay license royalties, and companies like AOL and MS will, if they can't beat it in court (which they quite possibly can) will pay the royalties. But if the OSS community is worried about not having an equivalent... email is a pretty good answer, IMO.
Although it's high time the email protocols get upgraded to ensure that the sender can necessarily be reached or at least identified. That could have been built into SMTP in the first place, if it weren't for that nonsense about needing to support third-party relaying for historical reasons. We need SMTP2, backward-compatible with SMTP, but with measures to ensure that the sender has an account on the sending server and that the username associated with that account is disclosed in the transaction, and then SMTP3 can be compatible with SMTP2 but not with SMTP, closing the loop. (The middle-ground SMTP2 has no benefits in itself, since messages sent via SMTP can still be received without the requisite sender-account info, but it provides for a gradual migration time.)
> With the exceptions of klingon, esperanto, and elvish, no > language is planned
Aside from the fact that you left out Songo (probably the most important planned language, since it is the most important trade language in much of central Africa, with French and English coming in second and third), you are talking at cross-purposes with the original poster. You are talking about _language_, and while the OP used the _word_ language, he was really talking about writing systems, as was clear in context. Unlike languages, writing systems generally _are_ deliberately invented and planned. (Hop over to omniglot.com and browse around for a while, and count how many writing systems they indicate were deliberately developed.)
> On a computer with decent algorithms, a skilled radical > typist can easily type an order of magnitude faster than > a western typist--no "word" should take more than three > strokes.
This implies that the number of possible words in the language is not more than about P(30,3) + P(30,2) + P(30,1), and in practice (since some combinations will invariably be insensible) probably not more than P(30,3) (the number of permutations of 30 strokes taken 3 at a time), which comes to fewer than 25000 words and is probably generous at that.
As an English speaker, that sounds like an incredibly minute vocabulary to me. Perhaps this is due mainly to the expansive vocabulary of English, generally considered to exceed that of any other language, but still... 25000 is SO few, even the _smallest_ English lexicons, intended for gradeschool children, boast _substantially_ more entries than that.
If by "word" you really meant "syllable", then what you say makes more sense. Omniglot says there are about 1700 syllables possible in Mandarin (presumably this is since it has fewer consonants than English). This does not account for homophones (some are distinguishable to native speakers, Mandarin being a tonal language, and more importantly for our purposes they are written distinctly), but the 25000 figure has ample room in it for that. Still, some words can be numerous syllables long. One example given at omniglot is schizophrenia, at five syllables. 5x3=15; if a couple of syllables don't use all three strokes that's maybe 12 strokes. (Yes, I'm guesstimating.) Sure, it's a moderately complicated word and is 13 letters in English too, but that's pretty similar; as near as I can tell that makes the two languages essentially comparable; claims that Mandarin is significantly _faster_ to type would seem to be unwarranted.
What you fail to point out is that besides the traditional ideographic system (Zhongwen), Chinese is also written with assorted other systems, including at least one alphabetic system (which still looks very like chicken scratchings to the Western eye, but nevermind _that_).
It is trivial to demonstrate that alphabetic writing systems have significant advantages from a programming standpoint. However, it's also a well-regarded maxim in the (modern) computer industry that it's easier to make the programmers do hard things than it is to make the users do hard things. Which raises the question: what is the big advantage (to humans) of ideographic writing that justifies the extra programming effort? Other than "we've always done it that way", I mean.
As I understand it (and my understanding is limited here, so if there's an actual philologist or Chinese scholar about, please chime in), the primary difficulty with using Chinese alphabetic writing systems is that the pronunciation differs significantly from region to region (way moreso than with the various regional accents in the US), and using an alphabetic system requires a (relatively) standardised pronunciation, if people from various areas need to read it. Ideographic systems have more latitude in that regard, since the way a word is written is not tied nearly so directly to the exact details of how it is pronounced.
That (as I understand it) is why it's worth the effort for someone to do the (substantial) extra work to create full software support for the various non-alphabetic Asian writing systems, of which Chinese (Zhongwen) is probably foremost. (The primary Japanese writing system, as I understand it, is a syllabary, which should be easier to support than ideographs, at least in theory. Hangul ("Korean"), it seems, uses an alphabet.)
> Reading Ideograms IS FASTER THAN READING ALPHABET BASED SCRIPTS.
This is at least partly because most alphabetic writing systems pad most words out with fluff that makes _learning_ to read them easier. Vowells as distinct characters (as opposed to diacritical marks as in Hebrew), spaces, punctuation, capitals. These all help a society increase its literacy rate (well, capitals are disputable), because they ease learning, but they are unnecessary for a person skilled in the language (and skilled in the writing system and practiced at reading without spaces and distinct vowells) and make the writing take up more space, making reading less efficient.
Anyway, speed of reading is less important to a computer interface than speed of data entry. The latter is always a good deal slower, and thus is the bottleneck. Yes, I know the writing system was not designed for computers, but we were talking (originally, at any rate) about why Chinese is harder for computers to support properly.
> The key to Descent was the fact that you could simultaneously > move in three directions with control.
Exactly. Descent was truly 6DOF. The premise was that you were either in an asteroid with very little gravity, or your ship's computers took care of automatically detecting and compensating for gravity, so all you had to do was fly. The Pyro GX was *way* easier to fly than any flight sim I've ever seen, and it had *amazing* maneuverability. This made for good gameplay: once you remapped your controls (the defaults were horrible) and learned them, you could forget entirely about the flight controls; your Pyro GX became like an extension of your body, and you could concentrate just on rooting out and blasting the droids without getting shot to pieces, finding your way around the mines, and accomplishing your objectives (mainly, blowing up the place and getting out alive).
Longer than that. Since the mid 1990s at least. I don't know when the red trex logo was first introduced, though; that may have been more recent.
Sun + Linux == Desktop ?
on
Linuxworld Fun
·
· Score: 2, Informative
> "What you will see from Sun is a lot more attention paid to > Linux on the desktop, because there is a lot more growth there > than anyone is willing to suggest," said Jonathan Schwartz, > executive vice president for Sun's software group.
This is in direct contrast to IBM's approach, and IMO it makes sense for Sun, because it hurts a certain competitor with a very large market share more than it hurts Sun. Think about it: Sun doesn't want to commoditise the server market if they have any brains, because that's where they make their money. But they *do* want to commoditise the desktop market, because that will prevent anyone from leveraging control of the desktop market (since no one entity can control a commoditised market) to push Sun (along with other competitors) out of the server market.
This is Sun being smart. *And* it's something the Linux community really needs badly: a major desktop OEM.
Now, granted, this is highly speculative, since the product they're unveiling right now is a low-end server. But I would very much like to see Sun (or any major OEM -- sorry, WallMart doesn't count as a major OEM) unveil an affordable Linux-based desktop system.
It's different for IBM, because they make a lot of money on the consulting and support end of the business, so that if the server becomes a commodity, it doesn't hurt them really. Sun has a bottom line in the server market to worry about, but they can better afford to commoditise the desktop, since that's a natural complement of the server.
> Why compare with Windows? The interesting thing about Windows > is how long it takes to erase.
Not very long if you use fdisk.
Seriously, though, with Windows 9x the (very significant) amount of pain it took to install was relevant, because it needed to be periodically _re_installed. With NT, it really doesn't matter so much, because almost everyone is going to be working with an OEM install.
> Having reinstalled Windows many times, I know what makes > it better for me: automatic hardware detection.
Automatic hardware detection if you don't mind having no working ethernet card and 640x480x8bpp graphics at 60Hz (i.e., much visual pain). The majority of systems I've installed Windows on, I've had to visit at LEAST two manufacturer websites to get working drivers (please, don't talk to me about the drivers on the CD that the manufacturer ships with the hardware, grumble), and the video driver is almost always one of the two, meaning you have to spend five or ten minutes trying to navigate a corporate website (heaven help you if it's one of the major OEMs that make you jump through six hoops just to FIND the downloads section) using a painfully bad display.
I've never had that problem with Mandrake. Just tell it what video mode to use, and it _does_ it, no flak about needing a $#@! driver. Granted, I've only installed Mandrake on about four systems (verses Windows on lots more systems than that, sometimes numerous times on the _same_ system), but the ethernet and video have in my experience always been correctly detected automagically.
Now, granted, soundcards (especially onboard ones) and software modems are more of a problem under Linux. But to say that Windows does "automatic hardware detection" is a pretty generous assessment. It TRIES to do automatic hardware detection... _sometimes_ it actually gets it right. On the whole, Windows is more likely to detect your soundcard and modem, and Linux is more likely to (correctly) detect your graphics card and NIC. (Okay, so Windows always _detects_ them, if by that you mean detects that they exist, but often it reports that it has detected "unknown hardware" and asks you where to find the driver... I don't consider that form of detection to be successful.)
Still, for easy-to-install, no OS can beat the BeOS. BeOS has some problems (like, the company that created it became insolvent, for example), but it had the ease of installation thing *nailed*. Way ahead of its time on that front, BeOS was. Shame about what happened to the company.
> You might not get paid much flipping burgers but at least you > won't be asked to work a 7 day week and you can actually take > a lunch break or even, gasp, a holiday!
Hate to burst your bubble, but McDs won't _ask_ you to work ten hour days eight days a week, they'll just _schedule_ you for it if that's what they feel like, and the _only_ thing you can do about it is quit. The worst they ever scheduled me was from midnight to 1pm Sunday-Monday and Monday-Tuesday (so I could do overnight cleaning and then work grill area (kitchen) for breakfast and lunch) then 5am-1pm (breakfast and lunch) Wed. through Saturday. All the 1pm times turned into 1:30 or 2, of course. I had to do that three weeks running when they were getting ready for corporate inspections. There were some other people working shifts like 9am-1am around the same time. None of us had any choice, other than quitting.
As for taking a holiday, the scheduling manager literally started taking the request book home and "forgetting" to bring it in (for weeks on end), because too many people were requesting time off. Turn in a request on a piece of ordinary paper? Nope, they'd throw that away, because of the disorganisation it would create, lots of pieces of paper everyplace -- couldn't have that. You want time off, it's got to be in the request book... which isn't here today. What I can't figure out in retrospect is why I put up with that for as long as I did. If I'd had any sense, the first time I couldn't put in a request for time off, I'd have put in notice instead.
The one thing is, there's always another burger joint you can go to and put in an application, and most of them are pretty much always hiring. But I for one am quite glad to have the job I have now. (I have the privilege of being "the computer guy" at the local public library. I work in an air-conditioned building, and we're CLOSED on Sundays. It's much nicer.)
The thing is, my job at McDs wasn't one of the really bad ones. If I have to, I'll work fast food again (though I don't relish the thought), but I'll NEVER work fast food _management_. I've seen the manager's job, and I do NOT want it.
It seems a lot of people are under the impression that the library's computer remembers everything you ever checked out. While it is theoretically possible that there may be libraries whose automation systems do that, it is certainly not usual. I work at a library, and our vendor (Gaylord) produces two of the major catalog systems on the market, Galaxy (which we use) and Polaris (which is newer and less, erhm, mature). Neither provides even the _option_ of storing this information.
The library _does_ of course know what books you _currently_ have checked out. That's sort of necessary for them to be able to hold you accountable if you fail to return the item. They may even be able to check a book that recently came back and see who just had it out, but that information is not stored forever, either. (On our system, it's stored either for three days, or until someone else checks the item out and returns it, whichever is sooner. There is no way to look it up on a per-person basis, not even with the report-generation facilities.)
So, if you are worried that having checked out a book on bomb making a couple of years back for a report will make you a suspect when the next terrorist attack rolls around, set your mind at ease.
Furthermore, it is in many states (including Ohio) illegal for a public library to disclose to anyone outside the library your personal information (such as what you have out or what your phone number is) except in certain special situations, such as at the request of a parent of a minor patron, or a court order.
So, to summarise the risk, the feds could, with a court order, find out what you _currently_ have out, and your address and such. Actually, I'd be more concerned about J. Random Criminal (or someone who decides to hold a grudge for some reason) walking up to an unattended circ terminal while the librarian on duty is off helping a patron in the stacks (this happens quite a bit at smaller libraries) and quickly looking up your address, or charging you fines, or whatever. Very little computer knowledge would be required to do this, because library computer systems are designed for librarians, many of whom are not geeks.
Perhaps the most interesting insight I have to offer here is that librarians tend WAY further toward the privacy-nut view on this issue than the typical citizen. A significant number of patrons would prefer (some of them strongly, to the point of being quite annoyed at our refusal) that we retain a complete list of every item they have ever checked out, in order to be able to inform them whether they've already read a given book, which books we have by a given author that they have not read, and so on. Our suggestions that they retain such a list themselves fall on deaf ears. They don't want to be troubled with that. They want the convenience. (I personally am appalled that anyone could take the trouble to read an entire book and then not remember the plot (or the major points, or whatever), to say nothing of not even remembering whether they've read it or not, but apparently I am nearly alone in this view. Anytime I state it, people look at like I've just announced I'm from Mars.) If there are libraries that do retain such information, I'm quite sure it's because they caved in to patron demand.
> It has nothing to do with addiction to lifestyle, it has to > do with that technology being necessary for most people's > LIVES, since so few people have any alternate means of > transportation.
Malarke. The previous poster was being stupid, yes, but your argument is just as lame. Lots of people get by without a car all the time, with no discernible ill effects.
> The majority of the population would have a hard time > getting to and from a grocery store, typically located > several miles from their home.
This is nonsense. Nearly half of the population of North America live in communities not more than a couple of miles from one end of town to the other, usually with not one but _several_ groceries. Most of the _rest_ of the population live in cities, where things are even closer. Only the most extreme rural populations would be unable, on pain of starvation, to walk to the grocery, and most of those are near (or on) a farm. Quite aside from that, location of domicile is part of lifestyle, and if we're altering our transportation habits we would presumably alter that as well in many cases, to say nothing of most people needing to change jobs...
It's not life and death; it _is_ lifestyle. That said, it's preposterous to suggest that it might ever be in any way appropriate to ask every person in the world to alter his lifestyle so you can conduct an experiment in climatology.
The distinction between that and which is irrelevant, however. (The relevant distinction is between sentient beings (such as "someone") and insentient items (such as "something").) Granted, we all knew what was meant. Nevertheless, get your criticism right, or don't criticise.
The best way to have fun in a computer lab is with fractal sound. I find a simple one-dimensional version of the ever-popular "plasma clouds" fractal, using the range of frequencies a PC speaker (yes, the old pre-soundcard kind) emits for the values at each point, works great. Get about three PCs in the lab doing this at once, and there WILL be excitement if there's anybody else in the lab.
That's pretty extreme. You don't need to actually _kill_ anyone. Just shocking their systems will do. My family tends to stay out of food I cook, because they know I have a tendency to include ingredients they don't like, such as beets, greens, or peppers. I know a guy who in college had a roomie who was getting into the food in the kitchen without even saying anything. This annoyed my friend, because he'd go to make a recipe and be missing something. So they had a little talk... then (and this was NOT planned), one day, while my friend was doing homework, they guy picked up the package of dried habaneros that my friend happened to have around and asked, "can I have some". "Yeah, but be careful, don't use too much". By which, they guy thought he meant too many and figured he needed them for a recipe -- only that wasn't what he meant, as you know if you've ever cooked with habanero. A moment later there was a horrible noise and the faucet turned on...
Now, he didn't do that to the guy on purpose, but you _could_ do something like that, and it would keep them out of your food without killing them per se.
The good PR, I mean. MS has been a little short on good PR lately, so this is a good thing for them. They could stand another couple of PR boosts, but this is at least something. Really, to get their PR up to a decent level, they could stand to do something that seems truly magnanimous, like announce an across-the-board discount (say, ten percent) on all of their products sold during the month of September, or release the source code for Notepad and offer a prize (say, a free copy of a compiler or something) for the best enhancement, or announce that IE7 will ship with Sun's JRE, or release a patch for a bug that hasn't been exploited yet, or something. Really, Microsoft needs a new PR manager badly. But this is _something_.
Slashdot Mirror
As near as I can tell, all the usenet groups in big8 and alt have
roughly the same amount of spam per group. (There may be some
specific groups that get extra spam, but all of the groups I've
looked at are about the same.)
But how much you _notice_ it definitely depends on the group.
In particular, it depends on the level of traffic the group
normally carries. High-traffic groups drown out the spam by
sheer volume. If you read a very popular group, the kind
where new messages roll in almost faster than you can read
them, the spam is such a low percantage of that huge bulk
that it gets positively buried under all the off-topic threads.
On the other hand, if you spend a much smaller amount of time
each day reading a dozen low-traffic groups, spam will dominate
your usenet experience. For one thing, you get mostly the
same spam messages in all groups, only with munging to prevent
automatic duplicate suppression. In addition, if the group
you are reading only gets two or three legit messages a week,
five spam messages a day seems like a lot in comparison.
> It never ceases to amaze me that somewhere there is someone who is
:) Hopefully someone will have the balls to
> glad to have received spam and buys something from it.
People are idiots. A _lot_ of otherwise-surprising things can be
explained if you operate under the assumption that people are idiots.
I'm not saying _all_ people are idiots _all_ of the time, but a
significant percentage of people are idiots (or behave in a manner
that makes them indistinguishable from idiots) a significant
percentage of the time.
> Somewhere someones eyes just light up when see that 5th "** Very
> Important Message **" turn up in thier mailbox. I just can't grok
> that. I would love to see a photo of some of these customers.
They would look, to the untrained eye, just like regular people.
> There is no good reason why its not illegal either.
We can only hope that spam law will eventually catch up with fax law,
but I should note that where I work we receive a number of junk faxes
every week, and people I work with have contemplated making purchases
as a result.
> They restrict what telemarketers are legally allowed to do. They
> can't keep calling you over & over with the same pitch,
They can, and they do. Although they usually wait long enough that
you would have a hard time knowing it's the same company calling
again, since you've had the same salespitch meanwhile from several
_other_ telemarketing firms...
> but you can be spammed countless times.
This is mostly because of the fundamental ecconomic differences
between telemarketing and sending spam: the telemarketing firm has to
pay a human being to be on the line with you. They have to do this
for every person they actually reach (albeit not for every attempted
call). Spammers don't have that problem.
> I'm glad there are people out there making these spammers lives
> hell. More power to them
> just start serial killing these spammers.
There are two problems with that approach:
1. It's wrong.
2. Death's too good for them.
First off, I did confuse Verdana with Georgia, but it was early in the morning...
> Fixed-width fonts mean the spacing between characters is equal.
Really? I thought it meant every character was a _different_ width... </sarcasm>
> there are no mostly-sans font types
Andale Mono is mostly sans-serif (meaning, most of the characters don't have serifs). However, certain select characters have serifs, in order to distinguish them from otherwise similar characters. Compare the following chars: Il1!| O0 If any two of them look the same, your font fails to adequately make all the distinctions it should. This won't matter for the end user doing nothing but word processing, but for many uses it is totally unacceptable. (Granted, most of these uses are geeky, but they still matter.) Fonts that accomodate this via select serifs but omit serifs on most letters are mostly-sans.
Furthermore, most of the uses that require every character to be distinct also absolutely require a fixed-width font. If you don't understand why a fixed-width font might be needed, then you don't understand source-code indentation. This is why we need a good quality mostly-sans fixed-width font, similar to Andale Mono. Alternately, we could use a serif fixed font like Courier New, except that serifs are ugly for large blocks of regular-sized text. Don't get me wrong, they're great for headlines and stuff, but I do NOT want to look at sixty consecutive lines of seriffed text. (Yes, I know the publishing industry does it all the time, but that's one of the reasons I prefer reading stuff on-screen to a physical book. The other reason is of course the search feature. The only redeeming feature of dead-tree books is that you can take them to the bathtub or wherever, conveniently.) The reason many fixed fonts are mostly-sans is because the need to make every char distinct and the need to make them line up happen to both be important for many of the same uses. You say fixed fonts should only be used on screen, but that's preposterous. Any text that needs to be fixed-width on the screen needs to be fixed-width on paper, too. This includes almost all source code, all ASCII art of more than one line, and a large amount of what is posted to usenet, as well as a lot of the email that geeks exchange with one another. (End-user email can probably be shown in a proportional font without problems.)
One last thing: if you are about to post a reply advocating tab stops for program indentation, please first read this example that I posted to usenet some time back, explaining why tab stop indentation is inadequate for many situations.
Bitmap fonts were great, back in the day when everyone used
the same screen resolution. These days, however, we really
need scalable fonts.
> ...but Georgia is the serif font, and Verdana is the sans serif
Err, I wrote the article early in the morning. Sorry about
that. I do understand what serifs are.
> Unfortunately most clients/browsers seem to go out of their
> way to discourage self-signed certificates with error messages
> that sound like "This certificate was self-signed.
Yes, and at that point the user's eyes glaze over and if
he doesn't have a guru to call, he clicks any button at
random. VERY few users would deign to read the entire
message. The dialog probably has "Okay" and "Cancel",
plus the close box on the window frame. Since "Okay" is
the default button, it's highlighted, and hitting "Enter"
will select it too, so there's probably at _least_ a one
in three chance the user will hit "Okay". That's on the
first try. What is more, if the desired result is not
achieved the first time, most users will try again and
hit a different button.
Translation: SSL certs only matter to people who care
about security and privacy.
This is not helped any by the fact that older browsers
used to display a dialog that looked basically identical
to the users whenever any information was sent over an
unencrypted socket -- for example, every time the user
did a web search at an http site like Yahoo! Users who
have been around for a few years have learned to just
bop Okay whenever they see that dialog -- and they teach
this behavior to the newer users.
So users who don't know anything about security or privacy
(i.e., almost everyone) are fairly unlikely to be dissuaded
from visiting a site just because the certificate is invalid.
They're WAY more likely to skip a site because it uses a
plugin that didn't come preinstalled, or takes too long to
load during peak hours.
$ \rm -rf *toast*
$ mkdir toast
$ chmod soft-eatable-noblack toast
Ugh, so complicated. If you have Emacs installed with
toast-mode set to autoload, you can just M-x make-toast.
The first time you do this, you'll probably want to do
M-x customize-group toast-mode and set up the various
toast-mode variables to your liking. The defaults are
reasonable for making wheat toast, but if you keep your
toaster loaded with multigrain bread as I do, you may
find that you need to set toast-mode-toast-threshhold
a little higher, because the bread toasts more slowly
than ordinary wheat bread.
Someone a while back was working on an enhancement to read
in the toast darkness from the toaster's sensors (if you have
one of the more advanced models) and thus automatically make
the necessary adjustments for different kinds of bread, but I
think he ran into a problem where Emacs 20 didn't support
something he needed and put it on the back burner. I don't
know whether he ever resumed work on it after Emacs 21 came
out, but I haven't heard anything about it. Anyway, you
want to adjust how dark the bread is toasted according to
taste anyway, so this enhancement is really only useful if
you don't always stock the same kind of bread.
> The fact that you don't even know to refer to additive primaries
> and subtractive primaries reveals your lack of understanding.
Either that, or I understand the additive properties of light
and absorptive properties of pigment but didn't think it
necessary to get too technical.
> If you really think that CMYK is nothing but a buzzword thrown
> around by marketing people, then you have no idea what you're
> talking about.
CMYK is a printing system. But saying that a piece of
computer software "supports CMYK" is meaningless buzzword
speach. The document still won't look the same on paper
as it does on the screen, or vice versa.
> The point of being able to work in CMYK is that you get a
> very close representation of the printed color on screen.
My turn to laugh. You have bought the marketing line.
> The limitations of a "luminous medium" in representing a
> reflective medium are minimal
Even if we go from 32-bit to 64-bit colour, a monitor still
can't accurately reproduce flat, non-glossy, non-luminous
tones. Representing fluorescent inks (well, some of them)
actually works much better than representing a surface with
low reflectivity. I'm talking about substances that are
not really dark, certainly nothing like black, but just dull
looking. They reflect some light, with biases toward certain
wavelengths, but mostly they absorb it across the whole
spectrum. You can play with gamma correction all day and not
get it right. Why do you think photographs of dark unvarnished
wood look so lousy on your screen? Things like dark blue yarn
and raisins reflect mostly ambient light, a mixture of assorted
wavelengths from an entire range (albeit with biases toward
certain subranges), that are not at all well modelled via RGB.
Metallic inks are an extreme case, but I was talking about
more everyday stuff.
> I'm sure that $22k was for a real workstation, like an IBM zSeries
:)
> or an HP Visualize or a Sun Blade 1k/2k (Or U60/U80).
From the article, it seems they were HP-UX systems.
> I'm a sysadmin at a large company and I've got a Blade 1000 on my
> desk (with Sun's 24" LCD + XVR-1000 video board, thankyouverymuch
> Anyway, the LCD is somewhat excessive, but the workstation certainly
> isn't. I'm constantly compiling code and doing testing on my desktop
> -- I need a good, reliable piece of hardware that'll function under
> stress.
A regular, non-overclocked PC will do fine under stress as long as it
has adequate cooling. From your description of your job, you are
probably in an air-conditioned building, so as long as you don't get a
bargain-basement system with cheapo cooling fans that'll go out in
three years, you'd probably be fine. Spend the extra twenty bucks
for the fans with good bearings, if you're getting a new system and
want it to last. The PC is so much cheaper than a heavy workstation
that you can afford to get a really _nice_ PC and still save a good
deal of money.
> The other thing that nobody mentioned is that that $22,000
> workstation will probably last 6 or 7 years. Not so with that cheap
> PC.
What if it lasts half as long? Would it be worth one extra instance
of copying everything over to a new system, halfway through the 7
years, to save $20,000? Anyway, my PC is now going on 5 years old,
and it's not on its last legs yet. 6 or 7 years is a decent lifespan,
but it's not _impressive_. (15 years, now that might impress me.)
> I had one developer who was still using his SPARCstation 10 until
> less than a month ago when we replaced it with a spare Ultra 2. Why?
> Because it still worked. All he used it for was basically an X
> display via SSH into the development boxes....
Yeah, so? If all you're doing is running an X server and sshing into
other systems to do your work, a Pentium 75 (current market value
approximately the same as a good lunch, except in high-tech areas,
where people will probably pay you to take it away if they still have
anything so ancient) will do the job with half its cycles tied behind
its back while whistling dixie, provided it has a halfway decent
graphics card. For maximum productivity you'd probably want to
upgrade it with a three-button mouse for around ten bucks, which
would be a significant part of the purchase cost of the thing.
> CMYK is a color model [...]
CMYK is a color model that only works on absorption media
(such as pigment on paper). On a luminous medium (such
as a CRT), things fundamentally don't (and can't) work that
way. As good as Photoshop is, speaking of its having "support
for CMYK" is marketroidese. All this means is that it can
convert from RGB formats (which *must* be used on your CRT
computer screen) to formats intended for printing. The
conversion is necessarily lossy, because ink on paper cannot
represent all of the same colours that the computer screen
can (and vice versa). Unless you're using phosphorescent
paint and viewing it under a blacklight, or some trick along
those lines, you can't represent the brightness of the sky
(for example) on paper. Similarly, your CRT can never show
a truly _flat_ (as in nonglossy, nonluminous) color.
You can throw buzzwords like "CMYK" at this all day long, but
an image will NEVER look the same on paper (no, not even on
glossy paper, although that's closer) as it does on a CRT
monitor, and that's a problem Photoshop can't solve.
LCDs (at their current level of tech) are even worse, because
they show colors inconsistently. Perhaps some future technology
will allow computers to display both luminous and flat colors on
the same display...
While we're on the subject of Photoshop, I agree that Photoshop
on Linux is a good thing. Photoshop is very entrenched in the
publishing community, and for good reason; it's quality stuff.
It also has a pricetag to match, so I surely hope Gimp continues
to develop (as it has been doing great so far), for those of us
with less expansive budgets. Photoshop may be (and probably is)
better, but my take on the matter is that Gimp is _comparable_,
which is a tremendous achievement. (I have a friend who does
graphics work for a living; he works at Eisenbraun's, a publisher
specialising particularly in ancient near-east stuff. He works
with Photoshop a lot. He'd been trying out Gimp, and was in
some ways (not all ways, but some) impressed with it, and had
noted that it had some really nice features Photoshop 6 did not
have. (He didn't specify which features.) Then he got the new
Photoshop, and they had it, he said, "in spades"). That says
to me that the two programs are in roughly the same league, a
huge accomplishment. But people who already know Photoshop and
have the budget for it will want to stick with it, rather than
learn Gimp which, although it's free, is not substantially
_better_ than Photoshop (at least, not at this time), surely
not better enough to justify a non-programmer to switch.
To me, Photoshop on Linux is a great thing, because it's
cross-platform technology, one more step toward separating
the decision of what OS to use from the decision of what
other software to use -- and THAT is a VERY good thing.
Seriously, IM is basically email with poorer clients, less
standardisation, and an intrusive, pop-to-the-front notification
feature. You can turn a regular mail service into essentially
the same thing (sans compatibility with existing IM clients --
but you get compatibility with the even larger installed-base
of standard email) just by adding a feature that pops incomming
messages up in a window in front of whatever you were doing.
(Whether this is a GOOD feature remains to be debated; I suspect
it would be good in some corporate settings and maybe for people
who don't get much mail and crave it, and merely annoying for
most of us. But if the notification feature can be toggled at
will, everybody wins.) Call it an "email pager" if you want.
IM, with its smaller, fragmented installed-base, really has
very little to offer to the discerning communicator. Most of
the other hyped features it boasts can be replaced with MIME,
which is older and more standardised and supported by most
major email clients these days.
Not that IM is likely to go away. If I understand the article,
only the server has to pay license royalties, and companies like
AOL and MS will, if they can't beat it in court (which they
quite possibly can) will pay the royalties. But if the OSS
community is worried about not having an equivalent... email
is a pretty good answer, IMO.
Although it's high time the email protocols get upgraded to
ensure that the sender can necessarily be reached or at least
identified. That could have been built into SMTP in the first
place, if it weren't for that nonsense about needing to support
third-party relaying for historical reasons. We need SMTP2,
backward-compatible with SMTP, but with measures to ensure that
the sender has an account on the sending server and that the
username associated with that account is disclosed in the
transaction, and then SMTP3 can be compatible with SMTP2 but
not with SMTP, closing the loop. (The middle-ground SMTP2 has
no benefits in itself, since messages sent via SMTP can still
be received without the requisite sender-account info, but it
provides for a gradual migration time.)
Hey, we have POP3, why not SMTP3?
> With the exceptions of klingon, esperanto, and elvish, no
> language is planned
Aside from the fact that you left out Songo (probably the most
important planned language, since it is the most important trade
language in much of central Africa, with French and English coming
in second and third), you are talking at cross-purposes with the
original poster. You are talking about _language_, and while the
OP used the _word_ language, he was really talking about writing
systems, as was clear in context. Unlike languages, writing
systems generally _are_ deliberately invented and planned. (Hop
over to omniglot.com and browse around for a while, and count how
many writing systems they indicate were deliberately developed.)
> On a computer with decent algorithms, a skilled radical
> typist can easily type an order of magnitude faster than
> a western typist--no "word" should take more than three
> strokes.
This implies that the number of possible words in the language
is not more than about P(30,3) + P(30,2) + P(30,1), and in
practice (since some combinations will invariably be insensible)
probably not more than P(30,3) (the number of permutations of 30
strokes taken 3 at a time), which comes to fewer than 25000 words
and is probably generous at that.
As an English speaker, that sounds like an incredibly minute
vocabulary to me. Perhaps this is due mainly to the expansive
vocabulary of English, generally considered to exceed that of
any other language, but still... 25000 is SO few, even the
_smallest_ English lexicons, intended for gradeschool children,
boast _substantially_ more entries than that.
If by "word" you really meant "syllable", then what you say makes
more sense. Omniglot says there are about 1700 syllables possible
in Mandarin (presumably this is since it has fewer consonants than
English). This does not account for homophones (some are
distinguishable to native speakers, Mandarin being a tonal
language, and more importantly for our purposes they are written
distinctly), but the 25000 figure has ample room in it for that.
Still, some words can be numerous syllables long. One example
given at omniglot is schizophrenia, at five syllables. 5x3=15;
if a couple of syllables don't use all three strokes that's maybe
12 strokes. (Yes, I'm guesstimating.) Sure, it's a moderately
complicated word and is 13 letters in English too, but that's
pretty similar; as near as I can tell that makes the two languages
essentially comparable; claims that Mandarin is significantly
_faster_ to type would seem to be unwarranted.
What you fail to point out is that besides the traditional
ideographic system (Zhongwen), Chinese is also written with
assorted other systems, including at least one alphabetic system
(which still looks very like chicken scratchings to the Western
eye, but nevermind _that_).
It is trivial to demonstrate that alphabetic writing systems
have significant advantages from a programming standpoint.
However, it's also a well-regarded maxim in the (modern)
computer industry that it's easier to make the programmers do
hard things than it is to make the users do hard things. Which
raises the question: what is the big advantage (to humans) of
ideographic writing that justifies the extra programming effort?
Other than "we've always done it that way", I mean.
As I understand it (and my understanding is limited here, so
if there's an actual philologist or Chinese scholar about,
please chime in), the primary difficulty with using Chinese
alphabetic writing systems is that the pronunciation differs
significantly from region to region (way moreso than with the
various regional accents in the US), and using an alphabetic
system requires a (relatively) standardised pronunciation, if
people from various areas need to read it. Ideographic systems
have more latitude in that regard, since the way a word is
written is not tied nearly so directly to the exact details of
how it is pronounced.
That (as I understand it) is why it's worth the effort for someone
to do the (substantial) extra work to create full software support
for the various non-alphabetic Asian writing systems, of which
Chinese (Zhongwen) is probably foremost. (The primary Japanese
writing system, as I understand it, is a syllabary, which should
be easier to support than ideographs, at least in theory. Hangul
("Korean"), it seems, uses an alphabet.)
> Reading Ideograms IS FASTER THAN READING ALPHABET BASED SCRIPTS.
This is at least partly because most alphabetic writing systems
pad most words out with fluff that makes _learning_ to read
them easier. Vowells as distinct characters (as opposed to
diacritical marks as in Hebrew), spaces, punctuation, capitals.
These all help a society increase its literacy rate (well,
capitals are disputable), because they ease learning, but they
are unnecessary for a person skilled in the language (and
skilled in the writing system and practiced at reading without
spaces and distinct vowells) and make the writing take up more
space, making reading less efficient.
Anyway, speed of reading is less important to a computer
interface than speed of data entry. The latter is always
a good deal slower, and thus is the bottleneck. Yes, I
know the writing system was not designed for computers,
but we were talking (originally, at any rate) about why
Chinese is harder for computers to support properly.
> The key to Descent was the fact that you could simultaneously
> move in three directions with control.
Exactly. Descent was truly 6DOF. The premise was that you
were either in an asteroid with very little gravity, or your
ship's computers took care of automatically detecting and
compensating for gravity, so all you had to do was fly. The
Pyro GX was *way* easier to fly than any flight sim I've ever
seen, and it had *amazing* maneuverability. This made for
good gameplay: once you remapped your controls (the defaults
were horrible) and learned them, you could forget entirely
about the flight controls; your Pyro GX became like an
extension of your body, and you could concentrate just on
rooting out and blasting the droids without getting shot to
pieces, finding your way around the mines, and accomplishing
your objectives (mainly, blowing up the place and getting
out alive).
> Mozilla's been around for what, about 4 years?
Longer than that. Since the mid 1990s at least. I don't know
when the red trex logo was first introduced, though; that may
have been more recent.
> "What you will see from Sun is a lot more attention paid to
> Linux on the desktop, because there is a lot more growth there
> than anyone is willing to suggest," said Jonathan Schwartz,
> executive vice president for Sun's software group.
This is in direct contrast to IBM's approach, and IMO it makes
sense for Sun, because it hurts a certain competitor with a very
large market share more than it hurts Sun. Think about it: Sun
doesn't want to commoditise the server market if they have any
brains, because that's where they make their money. But they
*do* want to commoditise the desktop market, because that will
prevent anyone from leveraging control of the desktop market
(since no one entity can control a commoditised market) to push
Sun (along with other competitors) out of the server market.
This is Sun being smart. *And* it's something the Linux
community really needs badly: a major desktop OEM.
Now, granted, this is highly speculative, since the product
they're unveiling right now is a low-end server. But I would
very much like to see Sun (or any major OEM -- sorry, WallMart
doesn't count as a major OEM) unveil an affordable Linux-based
desktop system.
It's different for IBM, because they make a lot of money on
the consulting and support end of the business, so that if
the server becomes a commodity, it doesn't hurt them really.
Sun has a bottom line in the server market to worry about,
but they can better afford to commoditise the desktop, since
that's a natural complement of the server.
Am I making any sense?
> Why compare with Windows? The interesting thing about Windows
> is how long it takes to erase.
Not very long if you use fdisk.
Seriously, though, with Windows 9x the (very significant)
amount of pain it took to install was relevant, because it
needed to be periodically _re_installed. With NT, it really
doesn't matter so much, because almost everyone is going to
be working with an OEM install.
> Having reinstalled Windows many times, I know what makes
> it better for me: automatic hardware detection.
Automatic hardware detection if you don't mind having no
working ethernet card and 640x480x8bpp graphics at 60Hz
(i.e., much visual pain). The majority of systems I've
installed Windows on, I've had to visit at LEAST two
manufacturer websites to get working drivers (please,
don't talk to me about the drivers on the CD that the
manufacturer ships with the hardware, grumble), and the
video driver is almost always one of the two, meaning
you have to spend five or ten minutes trying to navigate
a corporate website (heaven help you if it's one of the
major OEMs that make you jump through six hoops just to
FIND the downloads section) using a painfully bad display.
I've never had that problem with Mandrake. Just tell
it what video mode to use, and it _does_ it, no flak
about needing a $#@! driver. Granted, I've only installed
Mandrake on about four systems (verses Windows on lots
more systems than that, sometimes numerous times on the
_same_ system), but the ethernet and video have in my
experience always been correctly detected automagically.
Now, granted, soundcards (especially onboard ones)
and software modems are more of a problem under Linux.
But to say that Windows does "automatic hardware
detection" is a pretty generous assessment. It TRIES
to do automatic hardware detection... _sometimes_ it
actually gets it right. On the whole, Windows is more
likely to detect your soundcard and modem, and Linux
is more likely to (correctly) detect your graphics card
and NIC. (Okay, so Windows always _detects_ them, if by
that you mean detects that they exist, but often it
reports that it has detected "unknown hardware" and asks
you where to find the driver... I don't consider that
form of detection to be successful.)
Still, for easy-to-install, no OS can beat the BeOS.
BeOS has some problems (like, the company that created
it became insolvent, for example), but it had the ease
of installation thing *nailed*. Way ahead of its time
on that front, BeOS was. Shame about what happened to
the company.
> You might not get paid much flipping burgers but at least you
> won't be asked to work a 7 day week and you can actually take
> a lunch break or even, gasp, a holiday!
Hate to burst your bubble, but McDs won't _ask_ you to work ten
hour days eight days a week, they'll just _schedule_ you for it
if that's what they feel like, and the _only_ thing you can do
about it is quit. The worst they ever scheduled me was from
midnight to 1pm Sunday-Monday and Monday-Tuesday (so I could
do overnight cleaning and then work grill area (kitchen) for
breakfast and lunch) then 5am-1pm (breakfast and lunch) Wed.
through Saturday. All the 1pm times turned into 1:30 or 2,
of course. I had to do that three weeks running when they
were getting ready for corporate inspections. There were
some other people working shifts like 9am-1am around the
same time. None of us had any choice, other than quitting.
As for taking a holiday, the scheduling manager literally
started taking the request book home and "forgetting" to bring
it in (for weeks on end), because too many people were requesting
time off. Turn in a request on a piece of ordinary paper? Nope,
they'd throw that away, because of the disorganisation it would
create, lots of pieces of paper everyplace -- couldn't have that.
You want time off, it's got to be in the request book... which
isn't here today. What I can't figure out in retrospect is why
I put up with that for as long as I did. If I'd had any sense,
the first time I couldn't put in a request for time off, I'd
have put in notice instead.
The one thing is, there's always another burger joint you
can go to and put in an application, and most of them are
pretty much always hiring. But I for one am quite glad to
have the job I have now. (I have the privilege of being
"the computer guy" at the local public library. I work in
an air-conditioned building, and we're CLOSED on Sundays.
It's much nicer.)
The thing is, my job at McDs wasn't one of the really bad
ones. If I have to, I'll work fast food again (though I
don't relish the thought), but I'll NEVER work fast food
_management_. I've seen the manager's job, and I do NOT
want it.
It seems a lot of people are under the impression that the library's
computer remembers everything you ever checked out. While it is
theoretically possible that there may be libraries whose automation
systems do that, it is certainly not usual. I work at a library,
and our vendor (Gaylord) produces two of the major catalog systems
on the market, Galaxy (which we use) and Polaris (which is newer
and less, erhm, mature). Neither provides even the _option_ of
storing this information.
The library _does_ of course know what books you _currently_
have checked out. That's sort of necessary for them to be able
to hold you accountable if you fail to return the item. They may
even be able to check a book that recently came back and see who
just had it out, but that information is not stored forever, either.
(On our system, it's stored either for three days, or until someone
else checks the item out and returns it, whichever is sooner.
There is no way to look it up on a per-person basis, not even
with the report-generation facilities.)
So, if you are worried that having checked out a book on bomb
making a couple of years back for a report will make you a
suspect when the next terrorist attack rolls around, set your
mind at ease.
Furthermore, it is in many states (including Ohio) illegal for
a public library to disclose to anyone outside the library
your personal information (such as what you have out or what
your phone number is) except in certain special situations,
such as at the request of a parent of a minor patron, or a
court order.
So, to summarise the risk, the feds could, with a court order,
find out what you _currently_ have out, and your address and
such. Actually, I'd be more concerned about J. Random Criminal
(or someone who decides to hold a grudge for some reason) walking
up to an unattended circ terminal while the librarian on duty
is off helping a patron in the stacks (this happens quite a bit
at smaller libraries) and quickly looking up your address, or
charging you fines, or whatever. Very little computer knowledge
would be required to do this, because library computer systems
are designed for librarians, many of whom are not geeks.
Perhaps the most interesting insight I have to offer here is
that librarians tend WAY further toward the privacy-nut view
on this issue than the typical citizen. A significant number
of patrons would prefer (some of them strongly, to the point
of being quite annoyed at our refusal) that we retain a
complete list of every item they have ever checked out, in
order to be able to inform them whether they've already read
a given book, which books we have by a given author that they
have not read, and so on. Our suggestions that they retain
such a list themselves fall on deaf ears. They don't want
to be troubled with that. They want the convenience. (I
personally am appalled that anyone could take the trouble
to read an entire book and then not remember the plot (or
the major points, or whatever), to say nothing of not even
remembering whether they've read it or not, but apparently I
am nearly alone in this view. Anytime I state it, people look
at like I've just announced I'm from Mars.) If there are
libraries that do retain such information, I'm quite sure
it's because they caved in to patron demand.
> It has nothing to do with addiction to lifestyle, it has to
> do with that technology being necessary for most people's
> LIVES, since so few people have any alternate means of
> transportation.
Malarke. The previous poster was being stupid, yes, but
your argument is just as lame. Lots of people get by
without a car all the time, with no discernible ill
effects.
> The majority of the population would have a hard time
> getting to and from a grocery store, typically located
> several miles from their home.
This is nonsense. Nearly half of the population of
North America live in communities not more than a couple
of miles from one end of town to the other, usually with
not one but _several_ groceries. Most of the _rest_ of
the population live in cities, where things are even
closer. Only the most extreme rural populations would be
unable, on pain of starvation, to walk to the grocery,
and most of those are near (or on) a farm. Quite aside
from that, location of domicile is part of lifestyle,
and if we're altering our transportation habits we would
presumably alter that as well in many cases, to say
nothing of most people needing to change jobs...
It's not life and death; it _is_ lifestyle. That said,
it's preposterous to suggest that it might ever be in
any way appropriate to ask every person in the world to
alter his lifestyle so you can conduct an experiment in
climatology.
The distinction between that and which is irrelevant, however.
(The relevant distinction is between sentient beings (such as
"someone") and insentient items (such as "something").)
Granted, we all knew what was meant. Nevertheless, get your
criticism right, or don't criticise.
The best way to have fun in a computer lab is with fractal
sound. I find a simple one-dimensional version of the
ever-popular "plasma clouds" fractal, using the range of
frequencies a PC speaker (yes, the old pre-soundcard kind)
emits for the values at each point, works great. Get about
three PCs in the lab doing this at once, and there WILL
be excitement if there's anybody else in the lab.
That's pretty extreme. You don't need to actually _kill_ anyone.
Just shocking their systems will do. My family tends to stay out
of food I cook, because they know I have a tendency to include
ingredients they don't like, such as beets, greens, or peppers.
I know a guy who in college had a roomie who was getting into the
food in the kitchen without even saying anything. This annoyed my
friend, because he'd go to make a recipe and be missing something.
So they had a little talk... then (and this was NOT planned),
one day, while my friend was doing homework, they guy picked up
the package of dried habaneros that my friend happened to have
around and asked, "can I have some". "Yeah, but be careful,
don't use too much". By which, they guy thought he meant too
many and figured he needed them for a recipe -- only that wasn't
what he meant, as you know if you've ever cooked with habanero.
A moment later there was a horrible noise and the faucet
turned on...
Now, he didn't do that to the guy on purpose, but you _could_
do something like that, and it would keep them out of your
food without killing them per se.
The good PR, I mean. MS has been a little short on good PR lately,
so this is a good thing for them. They could stand another couple
of PR boosts, but this is at least something. Really, to get their
PR up to a decent level, they could stand to do something that
seems truly magnanimous, like announce an across-the-board discount
(say, ten percent) on all of their products sold during the month
of September, or release the source code for Notepad and offer a
prize (say, a free copy of a compiler or something) for the best
enhancement, or announce that IE7 will ship with Sun's JRE, or
release a patch for a bug that hasn't been exploited yet, or
something. Really, Microsoft needs a new PR manager badly. But
this is _something_.