No, the music that really matters and defines our culture is the music people listen to year after year after year. Such music seldom hits the "top 40" charts during any given year, but over the long haul it has a larger impact than the ostensibly more popular flash-in-the-pan stuff. Several of the serious biggies from the baroque era, while they don't ever "hit the charts", have been listened to, over the years, more times than all the song on the charts (at any given time) ever will be _combined_. They aren't what you think about when you think, "what's my favourite song" (well, not most of you), but you hear them hundreds upon hundreds of times over the course of your life anyway.
The only song from the modern era that I can think of that falls into this category is Happy Birthday, but it's an excellent example. You don't think of it, but just try to estimate how many times you have heard it.
In software, I think the equivalent might be something like winsock, something the casual user doesn't even think about using, but uses constantly. (Yeah, the example is from Windozeland, but the unices have their ubiquitous but mostly-unthought-about software too, such as mingetty.)
Any kid's electronics kit (the kind with components mounted permanently on a board, with the little springs that you stick the wires into to make the connections, and an instruction book with a few predetermined circuits that do different things, like make LEDs blink or sound the buzzer) comes with a diagram for one of those dual-circuit tripwire type circuits. (I forget the technical term for them.) I believe a single transistor is the soul of the beast. Anyway, there's one circuit that carries some current all the time, but if you break it, the other circuit fires and sounds the buzzer or whatever. Some primitive burglar alarms use this principle.
The reason I mention this is, it's easier to make a switch that will _open_ easily, than to make one that can be _closed_ easily. All you need is to ballance a wire precariously across two contacts, and just about any accidental touch will knock it off at least one of them -- instant open circuit.
_Somewhere_ I think I have a circuit diagram for one of these things... it's probably in my closet, but you can probably find instructions for it on the web faster than I can dig it out of there.
There are exceptions, like UPS units, but most non-software products state on the warrantee card that the liability of the manufacturer is limited to repair or replacement of the product or refund of the purchase price. Dammages to other items is normally not covered. If your stereo's tape deck starts eating cassette tapes while it's still under warrantee (usually not more than 3 years tops, sometimes 1), they'll fix or replace the tape deck, but if you lost a unique tape to the incident, you are unlikely to collect dammages without hiring expensive lawyers. (Sufficiently expensive lawyers can collect dammages for almost anything, warrantee or no warrantee, but that is another topic for another day.)
> I'm curious, given than one can run X11 apps on OS X with > Fink if one *really* needs to,
You can, and I can (and have), but most users cannot. fink is great, but most users don't want to deal with anything like that. They want to click on a "download now" link, bop okay on the "open with" dialog, and have the thing be automatically decompressed and the drive image mounted so that a window pops up containing a folder and text that says, "drag this folder to your hard drive to install". Inside the folder, there's an icon you click that starts the app. Like the way Mozilla installs. It would be a great boon if X11 apps could be compiled against a compat library and distributed this way for OS X, and come out looking reasonably native (a la OroborOSX).
> what X11 apps are worth porting to OS X / Aqua?
Gimp for darn sure (yes, it _can_ be got running on OS X, but it's WAY too hard), OpenOffice, several dozen mildly amusing simple games (Iagno, for example), FreeCiv, giram, xfrotz, xtads, graphics support for Emacs, possibly Evolution (not sure; haven't tried it; I use Gnus), gnome-terminal. These are just the ones I can think of off the top of my head. I suspect most X11 users out there could name at least one app worth porting over, and (except maybe for Gimp) I don't think everyone's going to say the same app.
> I mean, there are plenty of Unix apps critical to have, > python, perl, apache, postgresql, etc,
Those require less work to port though, typically. For one thing, Perl and Apache are so cross-platform in their design that porting them is like cake. But what about Perl/Tk and Python/Tkinter?
> but which X11 apps need to be ported? > Thinking... There is the Gimp, Sodipodi, Dia.
See, you named two I've never used. I think a lot of people could name ones we didn't think about.
> Mozilla is already ported.
Mozilla is just very cross-platform in its design. Which is a good thing, and most newer apps are getting better about this, but there are still a ton of legacy apps, and compatibility libraries are therefore a good thing.
> Rather than working on an X11 API for Quartz, how about > incorporating a Display Postscript API and AppKit > framework in Gnome or KDE?
Actually, I'd like to see the following, in no particular order:
* A Qt-compatible (TrollTech, not QuickTime) library
implemented as a wrapper around GTK, so that apps
intended for KDE can integrate more smoothly into
a Gnome desktop.
* A GTK-compatible library implemented as a wrapper
around Qt, so that apps intended for Gnome can
integrate more smoothly into a KDE desktop.
* X11 and GTK-compatible and Qt-compatible libraries
written as wrappers around the Win32 native stuff,
so that apps intended for X11 and Qt, given a POSIX
layer such as cygwin, can integrate smoothly and
appear native on Win32 systems.
* A Cocoa-compatible API* implementation for X11, to
enable apps written for Aqua to be ported by their
developers to X11. This would be most useful if the
ported apps were not required to be under any specific
license, since that would allow the greatest number
of apps to be ported.
* A complete implementation of the Win32 API* as a
wrapper around POSIX/X11 and either GTK or Qt, so
that reasonbly well-written Windows apps can be
easily compiled for the OSS desktop. This leaves
issues like hardcoded paths for the author of the
app to deal with, but it would help. Apps that
don't rely on too many OS quirks should be possible
to port in this fashion without excessive trouble.
Again, to allow the greatest number of apps to be
ported, this would be most useful if the ported
apps were not required to be under any particular
license.
(There ARE app developers who want to do this, if it were only less work. David Harris is one very good example, and Pegasus Mail would be a huge boon to the X11 desktop, even though it is still closed-source. Evolution may boast of being "like Outlook", but Pegasus Mail (without being less functional in any important way except usenet (and end users don't grok usenet anyway)) makes Outlook look hopelessly complex and unintuitive by comparison.)
* Only the parts of the API that are documented for
developers to use would need to be implemented, IMO.
I think my favourite free-as-in-zero-dollars computer game
of all time is
Curses.
This game will keep you up late at night and get you up
at all hours of the morning. Eventually you'll find
yourself searching google groups (rec.games.int-fiction)
for solutions to some of the trickier bits. (Like that
%$#! flashlight battery hidden right under your nose
where you can't get it without the [spoiler ommitted].)
You'll also need a z-machine to play it. For the classic
MacOS, the obvious choice is
MaxZip,
though there are of course numerous others.
OS X can probably run
frotz
(though you may have to compile it). Anyway, whatever platform
you need it for, you should be able to find something here.
They've got z-machines for everything, including certain
brands of pocket toasters, or so it seems. (The z-machine
was originally developed for Zork.)
> Also as an aside, how long have homosexuals been unable to vote?
Exactly. Homosexuals are citizens like everyone else, except that they feel the need to WHINE a little more. If there's a group of people we should be revising the constitution to protect, by granting them citizenship, it's the young. Why can't children vote, for example? Aren't we suppressing them by legislating them without representation? (Okay, well, I'm not sure I believe that either, but it's something to think about anyway. Children are much further from being first-class citizens than homosexuals are, in any event.)
Although, politics being what they are, we're slightly more likely to grant rights to animals, plants, and software...
> I don't think that we need to go into a definition of "spam" > here. We all "know it when we see it".
The legislations in question all define it, and among other things their definition stipulates that it's commercial in nature. Personal mail doesn't qualify. Although, the forgery stuff at the end of the Ohio bill seems to apply to *all* electronic messages, but what legitimate user forges headers on personal mail? Hmm... What about usenet? Can a lawyer comment on the common usenet practice of address-mungeing with respect to section H of the Ohio bill? Do messages posted to usenet qualify as electronic messages under the bill, even though usenet is a broadcast rather than a person-to-person medium?
Great to see my home state finally doing _something_. It's too little, of course. For a consumer to prosecute, he's got to have followed the spammer's "remove me" instructions, which is absurd. So that part of the bill is useless, or nearly.
The ISP part is more interesting. If I read it right, they (the ISP) have to put their AUP on a public webpage (which they all already do, pretty much) _and_ make sure the SMTP server notifies the sending server of this during the SMTP transaction. (The bill doesn't say SMTP, of course, and actually is general enough to cover IM and such as well, I think.) Should be no problem; just tweaking your 220 response should do, if I read it right (though IANAL).
But, as my subject hints, the MOST interesting part of this bill by FAR is the last little thing, tagged on almost like an afterthought, point H. If I read _that_ right (IANAL), forging any mail headers in any way, and _specifically_ the From: header, is now _officially_ forgery, according to Ohio's definition of forgery. Now, IANAL, but that sounds more interesting to me than the whole rest of it together. Forgery -- isn't that a _criminal_ charge? Couldn't we potentially be talking _jailtime_ for that? _That_ might be a deterrent. Plus, they deserve it. And almost all spammers forge the From: header routinely, so they should just about all qualify, if anyone takes the trouble to prove who sent the $#@! stuff.
Of course, getting the spammers to come to Ohio so we can sue or prosecute them, that could be the hangup. Guess we'll just have to get other states and countries to enact similar measures. But this adds to the amount of extant precedent, which is surely a good thing.
Unfortunately, this and all other legal definitions of spam that I've found only cover _commercial_ spam. Still, commercial spam is far and away the majority of all spam, so controlling that is the most important thing. The noncom stuff, while annoying, doesn't have quite the same ability to inundate until the real mail gets lost in the shuffle.
> I don't call an uptime of three months+ on my system (taken > down due to power failure) without a largely noticeable > slow-down comparable to Windows.
Err, uptime isn't what they're talking about.
> Doesn't (didn't?) Microsoft even officially recommend > rebooting win 9x system daily?
I don't know what MS says about it, but my recommendation is 40 hours for Windows 95 or 20 hours for Windows 98, and then reboot. Divide those numbers in half for every instant messaging client or p2p filesharing thingy you have installed, and then divide them in half again if you don't know exactly what's running out of the various Run registry entries. I don't know Me well enough to assign a figure for it here.
But like I said, uptime isn't what they're talking about. Any performance degradation that occurs due to mere uptime can be solved just by rebooting (which, if you're used to using Windows, is not a big deal). The kind of cruft the article talks about takes more significant effort to fix.
IMO, they've got it completely backwards. Out of the box, any system is totally unusable, as far as I'm concerned. It takes DAYS to just install all the apps I use regularly and get the various settings and options and preferences just _roughly_ the way I want them. It takes _weeks_ to fine-tune things until I can get comfortable with the system. Then there are those obscure little apps and utilities that you _occasionally_ need and go months without realising you forgot to install them... it can take _years_ to get a system truly _right_.
Linux is a little better OOTB than Windows, because the distributions bundle more things, and this can save a couple of days worth of download time initially, but there are always still lost of little pieces missing. Every so often I discover something that's missing, something the distribution did not include, that I want. This becomes, over the years, gradually less frequent. Discovering that package x is badly obsolete doesn't become less frequent; that's more or less constant. But in a pinch you can get by with an old version; whereas, if you never previously installed (say) a TADS runtime, when you find that you need one, you can't proceed until you go hunt it down.
As far as having things decently up to date, I find that it mostly only matters for things you use with any frequency. I have the latest browser, the latest Emacs, and so on, but if Python is a bit out of date, I don't care until I go to install or upgrade something that requires a newer version. By the time I need to upgrade to a newer version of mkswap my hardware will probably be on its last legs.
> X11 has a very un-mac-like interface - or rather, it > has a very crappy interface
X11 in itself doesn't have a user interface at all. The user interfaces we typically build on top of it are un-mac-like (i.e., versatile, at the expense of simplicity and consistency), but it wouldn't have to be that way. Actually, I have to admit, the _default_ settings in Gnome and KDE suck. Fortunately, we're not stuck with the defaults. I like to keep a panel on the left side of the screen, with launchers for my favourite apps, set so that maximised apps avoid it, but keep the task list autohidden on the bottom, and a floating clock. YMMV. The interface in OS X has its merits too, although the bottom is not a very sensible default place for the dock; fortunately, it's not stuck there. The magnify feature of the dock is something I'd like to see Gnome clone off for the panel in the next major Gnome version.
The OS X interface would not be terribly difficult to mimic on X11, if the alpha channel support were put in. (Aren't there plans for that anyway?) Doing a window manager that looks like the one in Aqua and hangs dialogs off the parent would be quite possible now. You also need a file manager with Finder's different view options -- I believe the column interface is the only one with no analog in current X11-based file managers, though now Finder has spring-loaded folders back too. The Dock wouldn't be substantially harder than the Gnome panel -- less flexible in some ways. The hard part might be moving application windows from the application to the global menu panel, but that part of the Mac interface is one I'm not sure you really would want to keep; it confuses the bejeebers out of end users -- they can't figure out why they can't find a special menu (to shut down or whatever), when Finder isn't in foreground. Anyway, I'm sure Apple could have done it if they'd chosen to do so. Apple _could_ have chosen to build the same Aqua interface based on X11 instead of Quartz. They didn't, and there are assorted reasons why they didn't -- some reasonable (such as the desire to build on top of something that supported alpha channels from the ground up) and some less so (such as the desire to get mileage out of the NeXT investment). I'm sure there are considerations we don't know about, as well. Ultimately, this is a tradeoff.
What they need to do now is provide an X11 implementation as a wrapper around the native GUI toolkit (called CarbonX or something), so that X apps can be compiled against it and then run without hassle by users. Currently, it is a bigger pain to try to get an X application running on OS X than most users are willing to attempt. Which is bad, because there are quite a few more useful X11 apps freely available than there are useful MacOS apps freely available. Doing the X11 compatibility layer, as a wrapper around Aqua, would give them _most_ of the advantages they would have had from building on X11 instead of Quartz, without the disadvantages. While they're at it, they should do GTK and TrollTech compatibility libraries as wrappers around the native widget set, to give the X11 apps a more completely native look and feel. If they did that, they'd have the bulk of the OpenOffice port for free.
There are some X11 apps that wouldn't be able to run with the compat library, without a real X server -- window managers, things that draw on the root window (mainly xscreensaver -- a sore loss, but unavoidable), and probably the panels (Gnome and KDE) and docks (WindowMaker) and other such utilities. But most normal applications should be able to be easily ported this way. Besides OpenOffice, they'd get graphics support in Emacs (only important to geeks, but still...), and the majority of the other X11 applications out there.
> "We need to build a vibrant and healthy developer community. > That's the lesson Linux has taught us. Having people to help. > Knowing where to get questions answered,"
That quote is what struck me too. One of the big reasons we all hate MS so _badly_ (aside from their determination to remove choice from the marketplace) is that for years their fundamental philosophy has been to cater primarily to the end user at the expense of the developer, because developers will march barefoot and blind through the valley of the shadow of death to produce software that will run on the platform the users choose. And that was true, up to a point.
But Linux is teaching them that there are limits, that if the developer experience is bad enough, developers will leave the valley and settle on the mountainsides, making their own trails and their own settlements. Some of them may venture into the valley from time to time, to sell to the plodding users, but the worse the valley gets, the more developers move to the mountainsides and stay there.
What really scares Microsoft is that the mountainside settlements may attract more than casual attention of the users in the valley.
> The problem is that DiskDrake does not allow you to type > in specific values for start/end sectors when partitioning. > You have to use their slider bars. This means holding down > your mouse button for an hour or so to get to the right spot, > or just getting 'close enough' and wasting a couple hundred > meg of space. I went for the latter, and am not too happy > with it. So, do they allow you to fine-tune your partitions > yet?
I believe the new release includes an innovative DiskDrake alternative that handles this situation very nicely, allowing you to specify every detail about your partitions. I think it's called fdisk or something like that.
Spam is a double-edged sword. It has an impact in both directions. Whether the revenue-increasing impact outweighs the revenue-decreasing impact depends on a number of factors. Both impacts are directly proportional to the number of unique recipients, but other factors increase or reduce one impact more than the other. A small handul of examples first (and then I'll draw a conclusion afterward):
* The revenue-decreasing impact is proportional to the
legitimacy and fame of the advertised product, as well
as the legitimacy and fame of the seller. This is
because if your company and product are already selling
well, it only takes a very small percentage of negative
responses (people avoiding you because they associate
you with spam) to have a significant impact.
* The revenue-increasing impact is proportional to the
moron-appeal of the product. Because morons have less
discernment, and are less likely to realise that the
spam is bulkmail, less likely to realise that they did
not in fact subscribe as you claim, less likely to be
skeptical of the legitimacy of a business that has to
resort to spam to make sales, and so on. Similar
reasoning applies to the moron-appeal of the spiel.
* The revenue-decreasing impact is proportional to the
number of duplicate recipients (people who get N copies
of it within the span of short-term memory).
* The revenue-increasing impact is inversely proportional
to the difficulty of responding and making a purchase.
* The revenue-decreasing impact is inversely proportional
to the difficulty of tracking down the identity of the
seller, which often correlates with the difficulty of
responding and making a purchase. That is, the things
that will enable intelligent consumers to avoid you are
(some of) the same things that allow morons to make the
purchase.
There are others.
The conclusion I draw is that we'll continue to see spam advertising fly-by-night companies and dubious products, but as the advertising industry begins to understand how spam works, and why it works (which will be a gradual process over probably several decades) we should see a nearly entire dearth of spam advertising legitimate products and companies. This will decrease the desirability (if that be possible) of making a purchase from a spammer, while simultaneously increasing the effectiveness of spam as a whole (since it will only be sent in cases where it would be effective). Although the increasing total amount of spam sent may hold the effectiveness of spam as a whole in check or even destroy it (by saturation).
Just within the last couple of months, I have started to see Transmeta-driven systems in catalogs (such as MicroWarehouse) for the first time. Or at least, *I* never noticed them before, and now, there they are, all over the place. If I had to venture a guess, I'd say Transmeta has finished development of its first product line and wants to see some profit before pushing hard toward the next product line.
Or something along those lines.
The other possible interpretation is that the Transmeta-based systems I was seeing in the catalogs weren't selling...
This raises an interesting point (though one that goes a bit off topic for the Apple update): What happens when some math grad student discovers a generalised way to determine a private key given the corresponding public key? Just something to think about.
Because it's a cool game that we all spend entirely too many hours playing, and if there's a new improved version we may as well spend those hours using that instead of the old version.
I block it on my LAN. (Rather than blocking ports, I just make sure the domain resolves to an IP address that doesn't have a web server, but the result is the same.)
Why? First, I'll tell you what's *not* the reason. This has _nothing_ to do with DRM. I don't care about that. It's not about bandwidth either; I can run half a dozen simultaneous instances of wget, no problem. (Okay, that makes things take longer, but eventually it all gets done.) So, why then do I block KaZaA? Because of what it does to the TCP/IP over PPP that connects me to the internet. Something about the way it forms its traffic causes very significant issues. Let it run on one workstation for just a few minutes, and my link up to my ISP eats flaming death. I can't even ping the upstream gateway then. I have to take networking down and back up to get things going again. And they don't _stay_ going until I kill KaZaA. My conclusion is that KaZaA does something invalid with its packets. I don't know the details. Perhaps if KaZaA were open source we could find out. But I'm fairly certain my ISP isn't doing anything to KaZaA deliberately; I have a medium-sized regional ISP (bright.net), and while it _is_ dialup, I've been more than pleased with the service they provide. Their nntp server is reliable and gets a full feed of everything, even free.* The phone support is good. The bandwidth is limited by the phone lines in the area, not by caps. Pretty much the only things they don't provide are shell access and cgi -- for security reasons. They know I'm using IP Masq to share my connection, and they don't care (though they don't provide support for that). This is a good ISP. But something about their configuration goes haywire when KaZaA runs over it. Maybe it's the type of system they use for their dialup servers; maybe it's even a bug in some vendor's IP stack, I don't know. It doesn't really matter; regular traffic doesn't trigger it, and KaZaA does. So if I want to share files, I do it some other way. I can run proftpd on my IP Masq gateway if I want, post binaries to usenet, whatever... this is not about bandwidth consumption or DRM. It's about what KaZaA's traffic does to a network. Not because of the amount of traffic, but because of _how_ it's done. (Again, I don't know the details; wish I did.)
Now, I blocked it months ago, and it's possible that it doesn't have these problems anymore, but I didn't feel the need to expose my network to the issue again. Especially after I did a web search and found out about the known ties with spyware and who knows what. There are other ways to share files that don't have these problems.
Quoting from the article: >... the major record labels [working through the courts]... have > now made a serious move that, if successful, will... apply... > hefty fees to broadcasters,... retroactive to 1998.
Retroactive to 1998? Yeeeesh. If that's true, it would represent a serious abuse of power, or I'm missing something. Lawmakers can't even _think_ about levying fees retroactive to 1998 (Article I Section 9). But now the courts _can_? The courts are supposed to interpret the law, not go off on their own doing things that *can't* be made into law because the constitution won't allow it. Or is there some twisted interpretation by which some extant law can be construed to indicate that these fees should have been paid all along? Can someone explain this, before I lose my last shreds of faith in our legal system?
> 16.1"? Why? Whats the point? Maybe so that you can _use_ it. Without going crosseyed.
> Aren't laptops supposed to be/portable/? That's the traditional line, but frankly, any system with less than 16" viewable display is not usable for regular work. You could get by with it for a few minutes in a serious pinch, but doing day-to-day work with it would be *painful*.
Okay, so you buy this thing to use for a small handful of minutes at a time, but to actually sit down and do anything you've got to get back to your desktop system. Granted, with ethernet being cheap as it is now you can mount the laptop's filesystem easily enough, but now you have to maintain two systems. Every software you use, you install twice, unless you just can't use it on the laptop. (Gimp, for example, would be utterly worthless on a system with less than 16" of display.) Everything you upgrade or configure, you upgrade or configure twice. Ick.
The other thing is, a large-side-of-typical 15" diagonal laptop measures perhaps about 9"x12" or thereabouts. 12" is really too small to prop on my lap -- the back end would keep falling between my knees. So you need a table to use it. IMO that rather defeats the purpose of portable. And it certainly takes backseat, in terms of portability, to handheld units, but is barely more useful -- the keyboards on those things are too small for touch-typing, and the screens are too small for running most applications. This is why up until now I haven't spent any money on a laptop. A desktop my not be portable, but at least it's useable. Give me something I can carry around -- all one piece, folds up nice and neat like a (large) book, et cetera, but give me something I can *use*, so I can use it as my regular system. 12"x16" seems quite reasonable to me, and would allow well over 19" of display -- a quite reasonable size even for a desktop, currently -- as well as a normal sized and so fully usable keyboard plus trackball. Then I can take my desktop system and turn it into some kind of networking appliance for my home LAN and use the laptop as my regular workstation, and still be able to take it with me when I travel.
Build it "one inch thin" when closed, make it out of whatever alloy sounds impressive to the marketroids, and call it "the desktop you can carry onto the plane", or "the laptop you can comfortably use at your desk", or both.
Now, PDAs that can be held in one hand make sense, maybe. But nobody will mistake those as a substitute for a computer; they are a supplement, more of a peripheral almost, and no workstation in any case. Laptops up to this point have been largely a horrid compromise between portability and usefulness. It doesn't have to be that way.
> The last thing I want is to carry around a 16.1" diagonal behemoth
Oh, come on. They're making them thin and light these days, if you haven't noticed. A 17" diag laptop today probably weighs less than a 13.5" viewable screen laptop from the mid nineties. But the portability comes mostly from its being all one piece, and also from the thinness and lack of weight and battery power. A large textbook can be half again as heavy as the heaviest modern laptop, and students routinely carry 2-3 of them under one arm while dashing from class to class.
A laptop that I can't use is of no use to me at all. If I buy a laptop, I'm going for the largest screen size I can reasonably afford (within reason -- over 20" would be overkill, I suppose), provided the other hardware is acceptable. 16" viewable is the absolute minimum I would consider; I would prefer a couple more inches than that, if the price increase weren't too horrific. Of course, 3D goggles with good resolution have the potential to make this a moot point... but they're just not _ready_ for mainstream adoption yet, it seems. Especially for a portable system, where they'd need to support translucency (i.e., let the real world show through, and let the user configure how _well_ the real world shows through; on a bus you'd turn the translucency down and do work; walking, you'd turn the translucency up over 50%, but you could still read/.). But now I'm dreaming.
Perhaps they are thinking of the new eMac. That's not a flat _panel_ per se (as in LCD), but it's 17", and it's flat (as in, flat surface, no curvature of the screen), and while they've changed the inane prefix letter, it _looks_ just like one of the old iMacs.
Although I'm fairly sure a flat-panel-on-a-stick iMac with a larger panel would sell. 15" is so _small_. Almost all the new systems are touting at least 16" viewable these days.
Slashdot Mirror
No, the music that really matters and defines our culture
is the music people listen to year after year after year.
Such music seldom hits the "top 40" charts during any given
year, but over the long haul it has a larger impact than the
ostensibly more popular flash-in-the-pan stuff. Several of
the serious biggies from the baroque era, while they don't
ever "hit the charts", have been listened to, over the
years, more times than all the song on the charts (at any
given time) ever will be _combined_. They aren't what you
think about when you think, "what's my favourite song" (well,
not most of you), but you hear them hundreds upon hundreds
of times over the course of your life anyway.
The only song from the modern era that I can think of that
falls into this category is Happy Birthday, but it's an
excellent example. You don't think of it, but just try to
estimate how many times you have heard it.
In software, I think the equivalent might be something
like winsock, something the casual user doesn't even think
about using, but uses constantly. (Yeah, the example is
from Windozeland, but the unices have their ubiquitous
but mostly-unthought-about software too, such as mingetty.)
Any kid's electronics kit (the kind with components mounted
permanently on a board, with the little springs that you stick
the wires into to make the connections, and an instruction book
with a few predetermined circuits that do different things,
like make LEDs blink or sound the buzzer) comes with a diagram
for one of those dual-circuit tripwire type circuits. (I
forget the technical term for them.) I believe a single
transistor is the soul of the beast. Anyway, there's one
circuit that carries some current all the time, but if you
break it, the other circuit fires and sounds the buzzer or
whatever. Some primitive burglar alarms use this principle.
The reason I mention this is, it's easier to make a switch
that will _open_ easily, than to make one that can be
_closed_ easily. All you need is to ballance a wire
precariously across two contacts, and just about any
accidental touch will knock it off at least one of them --
instant open circuit.
_Somewhere_ I think I have a circuit diagram for one
of these things... it's probably in my closet, but you
can probably find instructions for it on the web faster
than I can dig it out of there.
There are exceptions, like UPS units, but most non-software
products state on the warrantee card that the liability of
the manufacturer is limited to repair or replacement of the
product or refund of the purchase price. Dammages to other
items is normally not covered. If your stereo's tape deck
starts eating cassette tapes while it's still under warrantee
(usually not more than 3 years tops, sometimes 1), they'll
fix or replace the tape deck, but if you lost a unique tape
to the incident, you are unlikely to collect dammages without
hiring expensive lawyers. (Sufficiently expensive lawyers
can collect dammages for almost anything, warrantee or no
warrantee, but that is another topic for another day.)
> I'm curious, given than one can run X11 apps on OS X with
> Fink if one *really* needs to,
You can, and I can (and have), but most users cannot. fink
is great, but most users don't want to deal with anything
like that. They want to click on a "download now" link,
bop okay on the "open with" dialog, and have the thing be
automatically decompressed and the drive image mounted so
that a window pops up containing a folder and text that
says, "drag this folder to your hard drive to install".
Inside the folder, there's an icon you click that starts
the app. Like the way Mozilla installs. It would be a
great boon if X11 apps could be compiled against a compat
library and distributed this way for OS X, and come out
looking reasonably native (a la OroborOSX).
> what X11 apps are worth porting to OS X / Aqua?
Gimp for darn sure (yes, it _can_ be got running on OS X,
but it's WAY too hard), OpenOffice, several dozen mildly
amusing simple games (Iagno, for example), FreeCiv, giram,
xfrotz, xtads, graphics support for Emacs, possibly
Evolution (not sure; haven't tried it; I use Gnus),
gnome-terminal. These are just the ones I can think of
off the top of my head. I suspect most X11 users out there
could name at least one app worth porting over, and (except
maybe for Gimp) I don't think everyone's going to say the
same app.
> I mean, there are plenty of Unix apps critical to have,
> python, perl, apache, postgresql, etc,
Those require less work to port though, typically. For
one thing, Perl and Apache are so cross-platform in their
design that porting them is like cake. But what about
Perl/Tk and Python/Tkinter?
> but which X11 apps need to be ported?
> Thinking... There is the Gimp, Sodipodi, Dia.
See, you named two I've never used. I think a lot
of people could name ones we didn't think about.
> Mozilla is already ported.
Mozilla is just very cross-platform in its design. Which
is a good thing, and most newer apps are getting better
about this, but there are still a ton of legacy apps, and
compatibility libraries are therefore a good thing.
> Rather than working on an X11 API for Quartz, how about
> incorporating a Display Postscript API and AppKit
> framework in Gnome or KDE?
Actually, I'd like to see the following, in no particular
order:
* A Qt-compatible (TrollTech, not QuickTime) library
implemented as a wrapper around GTK, so that apps
intended for KDE can integrate more smoothly into
a Gnome desktop.
* A GTK-compatible library implemented as a wrapper
around Qt, so that apps intended for Gnome can
integrate more smoothly into a KDE desktop.
* X11 and GTK-compatible and Qt-compatible libraries
written as wrappers around the Win32 native stuff,
so that apps intended for X11 and Qt, given a POSIX
layer such as cygwin, can integrate smoothly and
appear native on Win32 systems.
* A Cocoa-compatible API* implementation for X11, to
enable apps written for Aqua to be ported by their
developers to X11. This would be most useful if the
ported apps were not required to be under any specific
license, since that would allow the greatest number
of apps to be ported.
* A complete implementation of the Win32 API* as a
wrapper around POSIX/X11 and either GTK or Qt, so
that reasonbly well-written Windows apps can be
easily compiled for the OSS desktop. This leaves
issues like hardcoded paths for the author of the
app to deal with, but it would help. Apps that
don't rely on too many OS quirks should be possible
to port in this fashion without excessive trouble.
Again, to allow the greatest number of apps to be
ported, this would be most useful if the ported
apps were not required to be under any particular
license.
(There ARE app developers who want to do this, if it were only
less work. David Harris is one very good example, and Pegasus
Mail would be a huge boon to the X11 desktop, even though it is
still closed-source. Evolution may boast of being "like Outlook",
but Pegasus Mail (without being less functional in any important
way except usenet (and end users don't grok usenet anyway)) makes
Outlook look hopelessly complex and unintuitive by comparison.)
* Only the parts of the API that are documented for
developers to use would need to be implemented, IMO.
I think my favourite free-as-in-zero-dollars computer game of all time is Curses. This game will keep you up late at night and get you up at all hours of the morning. Eventually you'll find yourself searching google groups (rec.games.int-fiction) for solutions to some of the trickier bits. (Like that %$#! flashlight battery hidden right under your nose where you can't get it without the [spoiler ommitted].)
You'll also need a z-machine to play it. For the classic MacOS, the obvious choice is MaxZip, though there are of course numerous others.
OS X can probably run frotz (though you may have to compile it). Anyway, whatever platform you need it for, you should be able to find something here. They've got z-machines for everything, including certain brands of pocket toasters, or so it seems. (The z-machine was originally developed for Zork.)
> Also as an aside, how long have homosexuals been unable to vote?
Exactly. Homosexuals are citizens like everyone else, except
that they feel the need to WHINE a little more. If there's a
group of people we should be revising the constitution to
protect, by granting them citizenship, it's the young. Why
can't children vote, for example? Aren't we suppressing them
by legislating them without representation? (Okay, well,
I'm not sure I believe that either, but it's something to
think about anyway. Children are much further from being
first-class citizens than homosexuals are, in any event.)
Although, politics being what they are, we're slightly more
likely to grant rights to animals, plants, and software...
> I don't think that we need to go into a definition of "spam"
> here. We all "know it when we see it".
The legislations in question all define it, and among other
things their definition stipulates that it's commercial in
nature. Personal mail doesn't qualify. Although, the
forgery stuff at the end of the Ohio bill seems to apply
to *all* electronic messages, but what legitimate user
forges headers on personal mail? Hmm... What about
usenet? Can a lawyer comment on the common usenet practice
of address-mungeing with respect to section H of the Ohio
bill? Do messages posted to usenet qualify as electronic
messages under the bill, even though usenet is a broadcast
rather than a person-to-person medium?
Great to see my home state finally doing _something_. It's too
little, of course. For a consumer to prosecute, he's got to
have followed the spammer's "remove me" instructions, which is
absurd. So that part of the bill is useless, or nearly.
The ISP part is more interesting. If I read it right, they
(the ISP) have to put their AUP on a public webpage (which they
all already do, pretty much) _and_ make sure the SMTP server
notifies the sending server of this during the SMTP transaction.
(The bill doesn't say SMTP, of course, and actually is general
enough to cover IM and such as well, I think.) Should be no
problem; just tweaking your 220 response should do, if I read
it right (though IANAL).
But, as my subject hints, the MOST interesting part of this
bill by FAR is the last little thing, tagged on almost like
an afterthought, point H. If I read _that_ right (IANAL),
forging any mail headers in any way, and _specifically_ the
From: header, is now _officially_ forgery, according to
Ohio's definition of forgery. Now, IANAL, but that sounds
more interesting to me than the whole rest of it together.
Forgery -- isn't that a _criminal_ charge? Couldn't we
potentially be talking _jailtime_ for that? _That_ might
be a deterrent. Plus, they deserve it. And almost all
spammers forge the From: header routinely, so they should
just about all qualify, if anyone takes the trouble to
prove who sent the $#@! stuff.
Of course, getting the spammers to come to Ohio so we
can sue or prosecute them, that could be the hangup.
Guess we'll just have to get other states and countries
to enact similar measures. But this adds to the amount
of extant precedent, which is surely a good thing.
Unfortunately, this and all other legal definitions
of spam that I've found only cover _commercial_ spam.
Still, commercial spam is far and away the majority
of all spam, so controlling that is the most important
thing. The noncom stuff, while annoying, doesn't have
quite the same ability to inundate until the real mail
gets lost in the shuffle.
Oh. I always thought Rhode Island was a city in Massachussetts.
Transmeta. There are Crusoe-based laptops on the
market now. See, for example, www.emperorlinux.com
Is 64-bit really enough? I think we should push for the immediate
development of x86-1024, so that the home user can afford a REAL
computer...
> I don't call an uptime of three months+ on my system (taken
> down due to power failure) without a largely noticeable
> slow-down comparable to Windows.
Err, uptime isn't what they're talking about.
> Doesn't (didn't?) Microsoft even officially recommend
> rebooting win 9x system daily?
I don't know what MS says about it, but my recommendation
is 40 hours for Windows 95 or 20 hours for Windows 98,
and then reboot. Divide those numbers in half for every
instant messaging client or p2p filesharing thingy you
have installed, and then divide them in half again if
you don't know exactly what's running out of the various
Run registry entries. I don't know Me well enough to
assign a figure for it here.
But like I said, uptime isn't what they're talking about.
Any performance degradation that occurs due to mere uptime
can be solved just by rebooting (which, if you're used to
using Windows, is not a big deal). The kind of cruft the
article talks about takes more significant effort to fix.
IMO, they've got it completely backwards. Out of the box,
any system is totally unusable, as far as I'm concerned.
It takes DAYS to just install all the apps I use regularly
and get the various settings and options and preferences
just _roughly_ the way I want them. It takes _weeks_ to
fine-tune things until I can get comfortable with the
system. Then there are those obscure little apps and
utilities that you _occasionally_ need and go months
without realising you forgot to install them... it can
take _years_ to get a system truly _right_.
Linux is a little better OOTB than Windows, because the
distributions bundle more things, and this can save a
couple of days worth of download time initially, but
there are always still lost of little pieces missing.
Every so often I discover something that's missing,
something the distribution did not include, that I want.
This becomes, over the years, gradually less frequent.
Discovering that package x is badly obsolete doesn't
become less frequent; that's more or less constant. But
in a pinch you can get by with an old version; whereas,
if you never previously installed (say) a TADS runtime,
when you find that you need one, you can't proceed until
you go hunt it down.
As far as having things decently up to date, I find that
it mostly only matters for things you use with any frequency.
I have the latest browser, the latest Emacs, and so on, but
if Python is a bit out of date, I don't care until I go to
install or upgrade something that requires a newer version.
By the time I need to upgrade to a newer version of mkswap
my hardware will probably be on its last legs.
> X11 has a very un-mac-like interface - or rather, it
> has a very crappy interface
X11 in itself doesn't have a user interface at all. The
user interfaces we typically build on top of it are
un-mac-like (i.e., versatile, at the expense of simplicity
and consistency), but it wouldn't have to be that way.
Actually, I have to admit, the _default_ settings in
Gnome and KDE suck. Fortunately, we're not stuck with
the defaults. I like to keep a panel on the left side
of the screen, with launchers for my favourite apps, set
so that maximised apps avoid it, but keep the task list
autohidden on the bottom, and a floating clock. YMMV.
The interface in OS X has its merits too, although the
bottom is not a very sensible default place for the
dock; fortunately, it's not stuck there. The magnify
feature of the dock is something I'd like to see Gnome
clone off for the panel in the next major Gnome version.
The OS X interface would not be terribly difficult to
mimic on X11, if the alpha channel support were put in.
(Aren't there plans for that anyway?) Doing a window
manager that looks like the one in Aqua and hangs
dialogs off the parent would be quite possible now. You
also need a file manager with Finder's different view
options -- I believe the column interface is the only
one with no analog in current X11-based file managers,
though now Finder has spring-loaded folders back too.
The Dock wouldn't be substantially harder than the
Gnome panel -- less flexible in some ways. The hard
part might be moving application windows from the
application to the global menu panel, but that part
of the Mac interface is one I'm not sure you really
would want to keep; it confuses the bejeebers out of
end users -- they can't figure out why they can't
find a special menu (to shut down or whatever), when
Finder isn't in foreground. Anyway, I'm sure Apple
could have done it if they'd chosen to do so. Apple
_could_ have chosen to build the same Aqua interface
based on X11 instead of Quartz. They didn't, and there
are assorted reasons why they didn't -- some reasonable
(such as the desire to build on top of something that
supported alpha channels from the ground up) and some
less so (such as the desire to get mileage out of the
NeXT investment). I'm sure there are considerations we
don't know about, as well. Ultimately, this is a tradeoff.
What they need to do now is provide an X11 implementation
as a wrapper around the native GUI toolkit (called CarbonX
or something), so that X apps can be compiled against it
and then run without hassle by users. Currently, it is a
bigger pain to try to get an X application running on OS X
than most users are willing to attempt. Which is bad,
because there are quite a few more useful X11 apps freely
available than there are useful MacOS apps freely available.
Doing the X11 compatibility layer, as a wrapper around Aqua,
would give them _most_ of the advantages they would have
had from building on X11 instead of Quartz, without the
disadvantages. While they're at it, they should do GTK
and TrollTech compatibility libraries as wrappers around
the native widget set, to give the X11 apps a more
completely native look and feel. If they did that,
they'd have the bulk of the OpenOffice port for free.
There are some X11 apps that wouldn't be able to run
with the compat library, without a real X server -- window
managers, things that draw on the root window (mainly
xscreensaver -- a sore loss, but unavoidable), and probably
the panels (Gnome and KDE) and docks (WindowMaker) and
other such utilities. But most normal applications
should be able to be easily ported this way. Besides
OpenOffice, they'd get graphics support in Emacs (only
important to geeks, but still...), and the majority of
the other X11 applications out there.
> "We need to build a vibrant and healthy developer community.
> That's the lesson Linux has taught us. Having people to help.
> Knowing where to get questions answered,"
That quote is what struck me too. One of the big reasons we
all hate MS so _badly_ (aside from their determination to
remove choice from the marketplace) is that for years their
fundamental philosophy has been to cater primarily to the
end user at the expense of the developer, because developers
will march barefoot and blind through the valley of the
shadow of death to produce software that will run on the
platform the users choose. And that was true, up to a point.
But Linux is teaching them that there are limits, that
if the developer experience is bad enough, developers will
leave the valley and settle on the mountainsides, making
their own trails and their own settlements. Some of them
may venture into the valley from time to time, to sell
to the plodding users, but the worse the valley gets, the
more developers move to the mountainsides and stay there.
What really scares Microsoft is that the mountainside
settlements may attract more than casual attention of
the users in the valley.
> The problem is that DiskDrake does not allow you to type
> in specific values for start/end sectors when partitioning.
> You have to use their slider bars. This means holding down
> your mouse button for an hour or so to get to the right spot,
> or just getting 'close enough' and wasting a couple hundred
> meg of space. I went for the latter, and am not too happy
> with it. So, do they allow you to fine-tune your partitions
> yet?
I believe the new release includes an innovative DiskDrake
alternative that handles this situation very nicely, allowing
you to specify every detail about your partitions. I think
it's called fdisk or something like that.
Spam is a double-edged sword. It has an impact in both
directions. Whether the revenue-increasing impact outweighs
the revenue-decreasing impact depends on a number of factors.
Both impacts are directly proportional to the number of unique
recipients, but other factors increase or reduce one impact
more than the other. A small handul of examples first (and
then I'll draw a conclusion afterward):
* The revenue-decreasing impact is proportional to the
legitimacy and fame of the advertised product, as well
as the legitimacy and fame of the seller. This is
because if your company and product are already selling
well, it only takes a very small percentage of negative
responses (people avoiding you because they associate
you with spam) to have a significant impact.
* The revenue-increasing impact is proportional to the
moron-appeal of the product. Because morons have less
discernment, and are less likely to realise that the
spam is bulkmail, less likely to realise that they did
not in fact subscribe as you claim, less likely to be
skeptical of the legitimacy of a business that has to
resort to spam to make sales, and so on. Similar
reasoning applies to the moron-appeal of the spiel.
* The revenue-decreasing impact is proportional to the
number of duplicate recipients (people who get N copies
of it within the span of short-term memory).
* The revenue-increasing impact is inversely proportional
to the difficulty of responding and making a purchase.
* The revenue-decreasing impact is inversely proportional
to the difficulty of tracking down the identity of the
seller, which often correlates with the difficulty of
responding and making a purchase. That is, the things
that will enable intelligent consumers to avoid you are
(some of) the same things that allow morons to make the
purchase.
There are others.
The conclusion I draw is that we'll continue to see spam
advertising fly-by-night companies and dubious products, but
as the advertising industry begins to understand how spam
works, and why it works (which will be a gradual process
over probably several decades) we should see a nearly entire
dearth of spam advertising legitimate products and companies.
This will decrease the desirability (if that be possible) of
making a purchase from a spammer, while simultaneously
increasing the effectiveness of spam as a whole (since it
will only be sent in cases where it would be effective).
Although the increasing total amount of spam sent may hold
the effectiveness of spam as a whole in check or even
destroy it (by saturation).
Just within the last couple of months, I have started to see
Transmeta-driven systems in catalogs (such as MicroWarehouse)
for the first time. Or at least, *I* never noticed them
before, and now, there they are, all over the place. If I
had to venture a guess, I'd say Transmeta has finished
development of its first product line and wants to see some
profit before pushing hard toward the next product line.
Or something along those lines.
The other possible interpretation is that the Transmeta-based
systems I was seeing in the catalogs weren't selling...
> Ever heared about public key cryptography?
This raises an interesting point (though one that goes a
bit off topic for the Apple update): What happens when
some math grad student discovers a generalised way to
determine a private key given the corresponding public
key? Just something to think about.
> why is this important?
Because it's a cool game that we all spend entirely too many
hours playing, and if there's a new improved version we may
as well spend those hours using that instead of the old version.
How was that not obvious?
I block it on my LAN. (Rather than blocking ports, I just
make sure the domain resolves to an IP address that doesn't
have a web server, but the result is the same.)
Why? First, I'll tell you what's *not* the reason. This
has _nothing_ to do with DRM. I don't care about that.
It's not about bandwidth either; I can run half a dozen
simultaneous instances of wget, no problem. (Okay, that
makes things take longer, but eventually it all gets done.)
So, why then do I block KaZaA? Because of what it does
to the TCP/IP over PPP that connects me to the internet.
Something about the way it forms its traffic causes very
significant issues. Let it run on one workstation for just
a few minutes, and my link up to my ISP eats flaming death.
I can't even ping the upstream gateway then. I have to take
networking down and back up to get things going again. And
they don't _stay_ going until I kill KaZaA. My conclusion
is that KaZaA does something invalid with its packets. I
don't know the details. Perhaps if KaZaA were open source
we could find out. But I'm fairly certain my ISP isn't
doing anything to KaZaA deliberately; I have a medium-sized
regional ISP (bright.net), and while it _is_ dialup, I've
been more than pleased with the service they provide. Their
nntp server is reliable and gets a full feed of everything,
even free.* The phone support is good. The bandwidth is
limited by the phone lines in the area, not by caps. Pretty
much the only things they don't provide are shell access and
cgi -- for security reasons. They know I'm using IP Masq to
share my connection, and they don't care (though they don't
provide support for that). This is a good ISP. But
something about their configuration goes haywire when KaZaA
runs over it. Maybe it's the type of system they use for
their dialup servers; maybe it's even a bug in some vendor's
IP stack, I don't know. It doesn't really matter; regular
traffic doesn't trigger it, and KaZaA does. So if I want
to share files, I do it some other way. I can run proftpd
on my IP Masq gateway if I want, post binaries to usenet,
whatever... this is not about bandwidth consumption or DRM.
It's about what KaZaA's traffic does to a network. Not
because of the amount of traffic, but because of _how_
it's done. (Again, I don't know the details; wish I did.)
Now, I blocked it months ago, and it's possible that
it doesn't have these problems anymore, but I didn't
feel the need to expose my network to the issue again.
Especially after I did a web search and found out about
the known ties with spyware and who knows what. There
are other ways to share files that don't have these
problems.
Quoting from the article: ... the major record labels [working through the courts] ... have ... apply... ... retroactive to 1998.
>
> now made a serious move that, if successful, will
> hefty fees to broadcasters,
Retroactive to 1998? Yeeeesh. If that's true, it would
represent a serious abuse of power, or I'm missing something.
Lawmakers can't even _think_ about levying fees retroactive
to 1998 (Article I Section 9). But now the courts _can_?
The courts are supposed to interpret the law, not go off
on their own doing things that *can't* be made into law
because the constitution won't allow it. Or is there some
twisted interpretation by which some extant law can be
construed to indicate that these fees should have been paid
all along? Can someone explain this, before I lose my last
shreds of faith in our legal system?
> 16.1"? Why? Whats the point?
/portable/?
/.). But
Maybe so that you can _use_ it. Without going crosseyed.
> Aren't laptops supposed to be
That's the traditional line, but frankly, any system with less
than 16" viewable display is not usable for regular work. You
could get by with it for a few minutes in a serious pinch, but
doing day-to-day work with it would be *painful*.
Okay, so you buy this thing to use for a small handful of
minutes at a time, but to actually sit down and do anything
you've got to get back to your desktop system. Granted, with
ethernet being cheap as it is now you can mount the laptop's
filesystem easily enough, but now you have to maintain two
systems. Every software you use, you install twice, unless
you just can't use it on the laptop. (Gimp, for example,
would be utterly worthless on a system with less than 16" of
display.) Everything you upgrade or configure, you upgrade
or configure twice. Ick.
The other thing is, a large-side-of-typical 15" diagonal
laptop measures perhaps about 9"x12" or thereabouts. 12" is
really too small to prop on my lap -- the back end would keep
falling between my knees. So you need a table to use it.
IMO that rather defeats the purpose of portable. And it
certainly takes backseat, in terms of portability, to handheld
units, but is barely more useful -- the keyboards on those
things are too small for touch-typing, and the screens are too
small for running most applications. This is why up until now
I haven't spent any money on a laptop. A desktop my not be
portable, but at least it's useable.
Give me something I can carry around -- all one piece, folds up
nice and neat like a (large) book, et cetera, but give me
something I can *use*, so I can use it as my regular system.
12"x16" seems quite reasonable to me, and would allow well
over 19" of display -- a quite reasonable size even for a
desktop, currently -- as well as a normal sized and so fully
usable keyboard plus trackball. Then I can take my desktop
system and turn it into some kind of networking appliance for
my home LAN and use the laptop as my regular workstation, and
still be able to take it with me when I travel.
Build it "one inch thin" when closed, make it out of whatever
alloy sounds impressive to the marketroids, and call it "the
desktop you can carry onto the plane", or "the laptop you can
comfortably use at your desk", or both.
Now, PDAs that can be held in one hand make sense, maybe. But
nobody will mistake those as a substitute for a computer; they
are a supplement, more of a peripheral almost, and no workstation
in any case. Laptops up to this point have been largely a horrid
compromise between portability and usefulness. It doesn't have
to be that way.
> The last thing I want is to carry around a 16.1" diagonal behemoth
Oh, come on. They're making them thin and light these days,
if you haven't noticed. A 17" diag laptop today probably
weighs less than a 13.5" viewable screen laptop from the mid
nineties. But the portability comes mostly from its being all
one piece, and also from the thinness and lack of weight and
battery power. A large textbook can be half again as heavy as
the heaviest modern laptop, and students routinely carry 2-3 of
them under one arm while dashing from class to class.
A laptop that I can't use is of no use to me at all. If I buy
a laptop, I'm going for the largest screen size I can reasonably
afford (within reason -- over 20" would be overkill, I suppose),
provided the other hardware is acceptable. 16" viewable is the
absolute minimum I would consider; I would prefer a couple more
inches than that, if the price increase weren't too horrific.
Of course, 3D goggles with good resolution have the potential
to make this a moot point... but they're just not _ready_ for
mainstream adoption yet, it seems. Especially for a portable
system, where they'd need to support translucency (i.e., let
the real world show through, and let the user configure how
_well_ the real world shows through; on a bus you'd turn
the translucency down and do work; walking, you'd turn the
translucency up over 50%, but you could still read
now I'm dreaming.
Perhaps they are thinking of the new eMac. That's not
a flat _panel_ per se (as in LCD), but it's 17", and
it's flat (as in, flat surface, no curvature of the
screen), and while they've changed the inane prefix
letter, it _looks_ just like one of the old iMacs.
Although I'm fairly sure a flat-panel-on-a-stick iMac
with a larger panel would sell. 15" is so _small_.
Almost all the new systems are touting at least 16"
viewable these days.
> who is profiting from the development of perl, and would get
> money for various reasons if perl6 were finished?
Anyone who sells techie books. O'Reilly, for example.
How _much_ money they'd make from Perl 6 books I don't know.