> I used to change my user agent to say it was running IE on CP/M-86.....
I like to play fun games with my ua string, too. One of my favourite tricks is to claim to be running my browser on an X11 GUI on PC-DOS 3.3, but claiming MSIE on X11 is fun too (especially, MSIE on an X11 GUI on a Microsoft OS). Other user-agent jokes I've seen include the following: * Claim to be running a significantly future version, (e.g., claim MSIE 11.5
or Mozilla/7.0 or use a future Gecko build date, et cetera) * Claim to be both MSIE and Gecko in the same user-agent string * List Emacs as the operating system * List Klingon, Quenya, or Sanskrit as the localization language * Claim an utterly impossible browser/OS/hardware combo, like iCab on
OpenVMS on SPARC, or, even better, claim a combination that's not only
impossible but also ancient, like NCSA Mosaic on ITS on a PDP8. * Claim a virtual machine architecture (e.g., the z-machine, glulx,
parrot, jvm,... anything that's never been implemented in hardware)
as your hardware architecture. * Make wrong and incompitible version claims (e.g., start with Mozilla/2.0
and then give a 2003 Gecko build date or claim to be MSIE 6.0) * Claim to be running on Hurd, BeOS 6, or some other vaporware. * "NoBrowserNeeded (My TCP/IP stack is connected directly to my brain.)"
You didn't tell us -- are you protecting against vandalism (some clown messing up the settings, deleting stuff, whatever) or against information theft? The solution will be completely different.
To protect against vandalism, nothing beats nightly offsite backups, nothing.
To protect against information theft, how about storing the informationg in question on an external device that you keep on your person? Then when they go to steal it, it's not there. Hard to beat that.
> That's why I'm specifically interested in actual experiences with the company.
You're unlikely to get a lot of that sort of response, for a couple of reasons.
1. As uninformed and illogical as the slashdot crowd can be, I don't think
most of us are gullible *enough* to fall for something like *that*. 2. Those who are probably don't want to publically admit it; even if they
did post, it would almost surely be AC. Better browse at -1 if you
want to have any chance at all of seeing them.
Let me get this straight -- this outfit wants to charge you money up front for the valuable opportunity to do business with them and possibly get a job, and you're sufficiently unsure whether this is a good idea that you have to ask on slashdot?
Dude, nobody legitimate has to charge you for the chance to work. A temp agency might charge your employer more than they pay you and so skip off the top, but they don't get a *dime* until you get a paycheck. Similarly, a more traditional employer might place expectations on you (dresscode or whatnot) that might result in your spending money before your first day on the job, but you don't pay *them* anything.
Work-from-home is notoriously fertile ground for scammers of all kinds, but if they're wanting to charge you up front for a lifetime membership for the opportunity to work, that should set off loud alarms and flashing red lights in your brain even if the job in question were more traditional (e.g. factory).
Interesting. But you don't have to know that to figure out that it's on most of the world's power buttons and guess that it must be the symbol for power or perhaps "on".
> I have to defend the old scissors icon for "cut". It's always made perfect > sense to me. But, then again, I grew up actually doing "real" cut-&-pasting > with some scissors, glue and a photocopier.
I understood it, because my dad used to do this kind of copy-and-paste work when I was a kid. But my point was, it's been ten years since anybody's done copy and paste that way, but it still makes a good icon, and it will still make a good icon fifty years from now when there's nobody left alive who remembers doing it that way, because it's simple, visually distinctive, and standardized across many applications. The same is true of the floppy disk icon for save. Keep it.
Quick, off the top of your head, what does a red octagon with a white outline represent? How about a button on a GUI that looks like a pair of scissors? What about a red circle with a red line across it from the lower left to the upper right? A button on the corner of a screen window that has an X in it? Do *any* of these things actually look like the object or process that they represent? Does it matter?
A good icon is simple, visually distinctive, easy to recognize instantly, consistent across many interfaces. The floppy disk icon for save is all of these things, and it's also familiar to almost every experienced computer user. It could be simplified a little (removing some superfluous details, like the label and the little readonly-lock thingydo), but the basic visual is already quite simple and distinctive. Nobody's going to mistake it for (say) the paste button. Sure, it's an anachronism, but the standard icons for cutting and pasting are scissors and paste, respectively, and nobody's used *that* method of cutting and pasting since word processing came into vogue. So what? The icons are visually distinctive enough (well, the scissors are; they should probably have used a roll of transparent tape for paste, but it's too late to change that now) and their meaning is well established.
Have you looked at the icon on a power button lately? (No, not your old 8-bit micro with the toggle rocker with 0 for off and 1 for on; something that was manufactured this century.) On virtually every device it's the same. Why exactly that specific symbol means "power" is quite beyond me (why not a lightning bolt or something?), but everybody knows it's the power button because it's the power button on everything -- computers, monitors, UPS units, even a growing number of kitchen appliances. This is a Good Thing(TM).
So, take that picture of a floppy, simplify it into a basic icon, and use it to represent the concept of saving from now on. It doesn't matter if half the people clicking on it have never seen an actual factual floppy diskette and don't know the history behind the symbol; they won't have to look at very many applications before they learn it's the universal symbol for "save changes".
> You should forward all of your spam to uce@ftc.gov.
I used to do that, but a couple of years ago it got to the point where it was using too much of my bandwidth, and I had to stop. (My spam volume has gone up several hundred percent since then...)
> I wonder if I could make money with uselesscrapyouthinkyouwant.com...
Calling the site that is poor marketing. You'll sell much more useless crap if you market it properly. Seriously, with the right advertising campaign, I'm convinced you can get people to buy used dental floss or anything, but probably not many will buy it if you call it "useless crap".
There are different ways to spin ("market") the concept of useless crap. Calling it "Products You Never Knew Existed" or "Stuff You Didn't Know You Needed Until You Saw It" would be a little better, but these slogans are far too long to be really effective. "Awesome Stuff" is better, but it loses some of its punch due to being somewhat hackneyed. Probably what you want is to name your company something abstract and then advertise individual products or lines of products.
Re:Still doesn't compare to OS X
on
GTK 2.4.0 Released
·
· Score: 2, Insightful
> 3) You need to access everything from the keyboard - with no third party > extensions, which widget set do you pick, OS X's or GTK+?
Actually, for that, the Win32 widget set rather rocks. Unfortunately, it's not very portable.
Re:Separate windows are fine
on
GTK 2.4.0 Released
·
· Score: 2, Interesting
> The real question is: why are the standard window managers for GNOME and > KDE (metacity and kwin) not implementing something like this
Not sure about kwin, but the official metacity project slogan is "no features" (apparently). A lot of Gnome users swap it out for a different window manager such as sawfish or Enlightenment. Fortunately, the architecture of Gnome makes this possible (though the wm in question has to support certain Gnome things to get everything working properly (e.g., the panel task list, having certain panels be avoided by maximize, and so on), which does limit exactly which window managers you can choose; you can _theoretically_ choose any wm, but if you choose one that doesn't support Gnome stuff won't work; sawfish and Enlightenment are the major alternatives that handle Gnome stuff right, that I know about; there may be one or two others).
usenet is not a print source, so I don't think it's admissible. Mailing list traffic certainly wouldn't be admissible, as letters aren't and so email is essentially the same sort of thing, and definitely very informal. The thinking here is that when a word starts to pass from the realm of slang into the realm of mainstream vocabulary, it starts being used in print.
> I don't understand: why does a usage have to be on paper to count for this > project?
When a word passes out of the realm of slang and into the realm of general mainstream vocabulary, it starts showing up in print sources. I strongly suspect that "cloaking device" will be located in print sources dating back to the sixties, because Star Trek was sufficienly popular that people would have picked up the term there and used it elsewhere. But yes, the OED is put together with the sort of thoroughness you expect from Oxford, and so of course they have to see actual evidence of the term being used in print. In the case of "cloaking device", I'm pretty sure said evidence will turn up.
I always thought they preferred to see examples of the word being used in print sources; I didn't think they generally went by private letters or whatnot, and I'm thinking usenet would be in the latter category. (Informal sources often show slang terms that have not yet passed into general usage; when you start seeing the word in books and journal articles, it starts being mainstream.)
Morph as a _root_ is very old (we're talking at least a couple of millenia here), but I don't remember its being used as a verb in the English language in its current sense much before the early-to-mid nineties.
Can you find the specific articles in Scientific American that you're thinking about? If they use "morph" as a verb in its current sense, they might count as an earlier use of the term.
The etymology of "morph" is of course straightforward, and we've had other words in the English language that use the root for a while (e.g. morphology, metamorphosis,...), but the question is about "morph" as a word with its current meaning.
Why keep C on life support? C was designed for computing in an era when every byte mattered and you would store some numbers in 16 bits and others in 32, depending on how large they would probably get, an era when you would manually allocate each piece of memory and free it when you were done because you didn't want to have any allocated that you didn't need, not even until the end of a block of code, an era when CPU time was so precious that garbage collection was considered expensive and wasteful, an era when, in short, the computer's resources were so valuable that you would spend programmer time willy nilly to conserve them. It's an anachronism, a language that *needs* to fall into disuse (except for a few inherently low-level tasks like writing bootloaders and kernels), and the sooner the better. Portability is not the language's only (or even its most important) weakness! Do you really want to continue to have to comb your application source for every single malloc and hunt down the corresponding free, to doublecheck every buffer to make sure you've checked its bounds every time you read or copy anything into it, and to write pages and pages of code where a paragraph would get the job done in a more modern language? Do you really want to have to continue to construct simple structures like lists *by hand* using pointers, when in other languages they're a fundamental data type? (Yes, there are libraries now to handle some of these things, sort-of, but the language just isn't geared that way; it's geared toward low-level bit-fiddling and going through the programmer's time like water in order to conserve the computer's resources.)
I know you have fond memories of the language, and surely it was good in its time, but its time has come -- and gone. Sooner or later you've got to let it go. Say a nice eulogy and let's get on with our lives now.
> This is a nominally 17" Proview CRT that I bought for $200 about three > and a half years ago.
If you spent that amount on a monitor today, you'd expect better.
> Then what happens when I want to add an app to the computer?
Get the old version that was current when the hardware was new?
I know, everybody says, "Get the latest version", but that's only true if you've got relatively recent hardware; if your hardware is old, you're usually better served with old software too.
> What happens when I receive a document in a format created by a newer > version of a particular app that the version installed on my computer > can't read?
The same thing that happens when someone sends you a document in the obscure proprietary format of a random app they happen to have that you don't have: you ask them to please send you the information in a standard format. If you try to have all the versions of all the applications anyone might ever use to create a document that they might send you, you'll be spending more money on software than you spend on hardware. A lot more. You'll go out and buy the latest MS Office every year ($650), and then some clown will send you a FoxPro spreadsheet, so you'll buy the Corel Word Perfect Suite ($299), and then somebody will send you a MS Publisher document, so you'll shell out for the latest version of that, and then some machead will send you something they did in Pagemaker or Illustrator or InDesign... if you can afford all that, you can *certainly* afford some extra RAM. You can get 512MB of SDRAM for less than the average cost of *one* of these software upgrades.
At some point, you have to stop buying every piece of software on the market and ask people to send you the information in a standard format. Unless you have an infinite budget.
> Does anybody really need 1 GHz for a word processor or spreadsheet?
No. 300Mhz is more than plenty to handle that stuff comfortably, as long as you've got adequate RAM. That was my point.
> "The software I've already got" has known vulnerabilities whose only patch > is an upgrade to a new operating system that requires computers that cost > $1,000 per seat that I don't have.
Security is a separate issue from RAM usage. Let me let you in on a secret: the latest and most current version of Windows XP[1] has known vulnerabilities for which no patch is available. (Actually, it's not all that secret. I'm talking about stuff that is publically known, some of which is even in the MSKB.) Windows systems should always be behind a firewall; nothing new there.
[1] Here I'm including some of the apps that are bundled with it; technically
speaking it's possible to disable Outlook Express for example, but again,
security is a separate topic from RAM usage and is really off-topic for
this thread.
> Redoing drywall is actually pretty cheap, especially if it's just a > small patch.
If you have to rip out enough drywall to drill through a bunch of studs and run a cable, "just a small patch" is going to turn into Famous Last Words, and that's before you have to also patch the wallpaper...
> It also looks impressive to the ladies without being hard at all.
Running the cable does that well enough, especially if you wire the jack yourself. ("Yeah, this is a standard type-B wiring for RJ45. See, you keep each pair twisted until it's right up next to the jack. The orange pair goes here and here...")
> You're probably doing this, but many other sites (heh, usually IE-only > sites) sure as heck aren't. What about 320x480, 400x600, 640x1024?
If it works at 640x480 and 1600x1200, it'll work fine at 640x1024 also. The less-than-600 widths are another matter. I am aware that there are people who size windows to half the width of their screen, but I am also aware that they have the capability to size them differently without causing a disruption to their entire environment (only the browser is impacted, generally) if they desire. I make some attempt to see what my pages look like at 320x200, but at some point a multicolumn page just can't be crammed into that, and if the user has to scroll horizontally to read the other column I generally am willing to live with that if they've got their browser window sized to less than 600 pixels across. Now, if a block of text is going wider than the window so that they have to scroll horizontally as they read each line, then I would consider that to be something I should look into fixing, down to resolutions of about 320 across. If their browser is much narrower than that, it's a pathological case and there's not much that can be done.
There are people out there who want all web pages to be designed for their wristwatch systems that boast a black-and-white (not grayscale; 1-bit color depth, if you can call that color depth) display with two-digit dimensions in pixels, but I've got to draw the line somewhere. If they really need to browse on a display like that, they should use a browser that discards all the layout info from the page (probably all style info, actually) and just extracts the bare information itself and hyperlinks.
My philosophy on color depths is similar; I try to pick colors that will be *legible* down to 256-color mode, but I can't promise they page will look *good* at anything less than 24-bit color. (Anyway, you can always do what I (usually) do and discard the color information the page specifies in favor of your own color preferences.)
> Value priced computers do not come packed with a "remotely decent" monitor.
Many of them don't come with a monitor at all; you either buy that separately or use one you already have. Some (especially the mail-order ones) do come with a monitor included, but even some of those include a halfway decent one.
> I guess that by calling the optional feature of different resolutions > for different workspaces worthless in practice, I underestimated the > percentage of "remotely decent" monitors on the market
You're probably looking at the extreme low end, the sort of monitor that's bundled with computers that include a bundled monitor, the ones that have less than sixteen inches (diagonal) of viewable picture space, e.g., the cheapest of the cheap of "17-inch" monitors (the ones that are like 15.3" viewable or whatever) and, of course, the ones that don't even pretend to be 17". If you walk into a place like Circuit City, they'll have maybe three of these on display; mostly they'll have higher-end 16-viewable monitors, 19-inchers (18 viewable; these are the midrange), and LCDs. They might also have one large high-end monitor.
Now, it is true that the low-end ones sell a disproportionate number, because maybe 30-40% of the people buy the absolute rock-bottom cheapest thing they can get, but still there are quite a few people out there with a nicer model. Nineteen-inch monitors are no longer a novelty item; ecconomy of scale is such that they're *almost* as cheap as the 17" monitors. For example, a place[1] that sells an ultra-low-end 13.8" viewable monitor for $122 and a range of 16"-viewables for anywhere from $126 - $158 might sell 18-viewables for $163 - $220. This is the sweet price spot, because their 20-viewables might run $472 and up. Most of those 16" displays and probably all of the 18" ones can handle changing display resolutions without a hitch.
LCDs, of course, have serious trouble with resolutions other than the "native" one, but LCDs also have serious trouble with color accuracy and viewing angle; if they weren't so much smaller physically there'd be no reason at all to use them (other than in laptops, where every watt of power matters deeply for the battery life).
And yeah, the low-end monitors like the 13.8" viewable are probably not going to give you a lot of pleasure. But buying those only saves you in the general vicinity of ten bucks or less versus a decent 16" viewable. (In my example prices[1], it's a different of four whole bucks, woo.)
The five-year-old computer presumably has an existing operating system; it doesn't need a new one.
> The problem here is that few companies still publish applications for > Windows intended to run on the five-year-old computers that schools > still have. Many X11 apps, on the other hand, seem to demand much less > RAM. (No, just because top counts VRAM in its total doesn't make X11 a > memory hog.)
A five-year-old computer, in addition to already having an OS, might also already have some apps installed, perhaps, don't you think? At least, it would have all of the ones that had been being used all along? Sure, it would be nice to upgrade, but it would be _nice_ to upgrade the hardware, too; sometimes on a budget you let upgrades wait, right?
I do *not* think it's reasonable to expect any software project, commercial or non-commercial, proprietary or open-source, to necessarily produce new applications for old hardware. Yes, some projects do do this, in some cases because the problem domain is such that it is possible (firewalls come to mind here), but in general I don't think it's realistic to expect all new software to run on arbitrarily old hardware. New wine, old wineskins. If you want to run all of the latest and greatest software, then you might have to shell out for new hardware. This doesn't make old hardware useless;
That hit the nail on the head. If you do the conduit right, with nice big junction boxes at *all* corners, conduit running to multiple locations in every room, and so on, then you can run whatever kind of cable you want at any time in the future very easily. Audio cable, video cable, fibre, Cat12b, you name it, you'll be able to run it. What kind of network cable will you (or whoever lives there) want in fifteen years? Fifty years? You have no clue, right now. But you know it'll be easy to run it; take a screwdriver, take the faceplates off, pull the cable through, and you know who Bob is.
One more thing:
> Run all your conduit to a central location (probably in the basement). > You'll want a nice (rack even?) open area that you can mount equipment > as well as patch panels, etc. Wire ties are your friend.
If it were me, I'd put nice boxes (kinda like a breaker box, but without the breakers) every fifty feet or so around the outside wall of the basement, with a nice fat conduit running straight up from each one to a junction box, accessible at the top by removing a faceplate (like a lightswitch cover). In any part of the basement that's going to be "finished", I'd also run horizontal conduit between these boxes. I'd put an electrical outlet near each of these boxes, so that a hub or switch can easily be put there.
Then, N years from now, when you want to run your new Terabit Ethernet cable or whatever from the sewing room to the kitchen, you pick one of the empty faceplates in the sewing room, run the cable from there to the nearest junction box and down to the box in the basement, where you put a hub. In the kitchen you do the same thing; then you run a cable around the horizontal conduit to connect your hubs, and you're done; with a hub at each of the basement boxes, you never need more than one horizontal cable of any given type, no matter how many things you run in the room above.
So, you want more faceplates than you need right now in every single room (yes, the bathroom; yes, the garage, too; I'd put one in each closet also), and a system of conduit connecting them all. You do NOT want to have to tear up your drywall later because there's not conduit going to such-and-such a location.
This isn't cheap; conduit costs more than cable. It costs less than redoing your drywall later, though. Run the conduit. You'll be glad you did.
> what if you don't have a windows computer to see how 'it looks' under IE?
Exchange screenshots with another webmaster who does use IE. I've got a couple of people I trade screenshots with regularly. They like this arrangement, because my screenshots show some edge cases that most people would miss. I always take a series of shots showing scalability from 640x480 up through at least 1280x1024, and I always show what the site looks like with and without page colors turned on (and my system colors are medium-contrast light-on-dark, which shows up stuff that gets missed if you use black-on-white). Also I tend to take screenshots with about three different rendering engines (always Gecko, plus usually Konqueror and one or more of Opera, W3, Links, Lynx). So my approach shows up a lot of edge cases that more typical setups (black-text, white-background, MSIE, 800x600, page colors enabled) won't see.
Yes, it's possible to design a web page that looks "right" under all of the above settings. (By "right", I mean it looks like it was designed for those settings.) Eye candy in the graphical browsers, without breaking the text browsers; client-side scripts that automate things if scripts are enabled, without breaking the site if scripts aren't enabled. Images that look good (no jaggy edges) against either a light _or_ dark background. (This is tricky if you have to support browsers with no proper alpha channel, but it can be done; the trick is to set the background color when you save your PNG images, so that non-alpha-channel browsers (*cough* MSIE) will antialias against that color -- set that to the same color as the surrounding background and you get to be Bob's nephew. Or use the MSIE PNG-alpha-channel hack, but not all versions of MSIE support that, and it still leaves old versions of Opera out in the cold.)
Green on black is _okay_ (unlike black on blinding white, which will make you snowblind if you set the contrast high enough to show detail properly), but there are better color schemes. Amber on black looks uglier for the first five minutes, but after your eyes get used to it you can stare at it for much longer periods of time with zero eyestrain.
Even better, I've found, is a tertiary color scheme. Set your system foreground color to #FFE6BC and your background to #294D4A. Set this system-wide. If you use Gnome, there's a theme available called themacs (GTK1) or eMaCs (GTK2). On Windows you can just go into the Display properties under the Appearances tab (clicking on Advanced if you're using WinXP) and set it up manually easily enough. It's possible in KDE/Qt also, though I forget the exact steps. Anyway, Also set your web browser to use these colors exclusively (ignoring the colors the web page author specifies). (You'll need to change your link colors also; blue on slate green isn't the best combination.)
Your first reaction will be, "boring", but use it for a week and then ask yourself when was the last time your eyes got worn out looking at your display; it'll be a week ago, right before you switched to this scheme.
Oh, and do make sure your refresh rate is set as high as it will go.
Slashdot Mirror
> I used to change my user agent to say it was running IE on CP/M-86.....
... anything that's never been implemented in hardware)
I like to play fun games with my ua string, too. One of my favourite tricks
is to claim to be running my browser on an X11 GUI on PC-DOS 3.3, but claiming
MSIE on X11 is fun too (especially, MSIE on an X11 GUI on a Microsoft OS).
Other user-agent jokes I've seen include the following:
* Claim to be running a significantly future version, (e.g., claim MSIE 11.5
or Mozilla/7.0 or use a future Gecko build date, et cetera)
* Claim to be both MSIE and Gecko in the same user-agent string
* List Emacs as the operating system
* List Klingon, Quenya, or Sanskrit as the localization language
* Claim an utterly impossible browser/OS/hardware combo, like iCab on
OpenVMS on SPARC, or, even better, claim a combination that's not only
impossible but also ancient, like NCSA Mosaic on ITS on a PDP8.
* Claim a virtual machine architecture (e.g., the z-machine, glulx,
parrot, jvm,
as your hardware architecture.
* Make wrong and incompitible version claims (e.g., start with Mozilla/2.0
and then give a 2003 Gecko build date or claim to be MSIE 6.0)
* Claim to be running on Hurd, BeOS 6, or some other vaporware.
* "NoBrowserNeeded (My TCP/IP stack is connected directly to my brain.)"
You didn't tell us -- are you protecting against vandalism (some clown messing
up the settings, deleting stuff, whatever) or against information theft? The
solution will be completely different.
To protect against vandalism, nothing beats nightly offsite backups, nothing.
To protect against information theft, how about storing the informationg in
question on an external device that you keep on your person? Then when they
go to steal it, it's not there. Hard to beat that.
> Yes, that's my first reaction too...
There's a reason.
> That's why I'm specifically interested in actual experiences with the company.
You're unlikely to get a lot of that sort of response, for a couple of reasons.
1. As uninformed and illogical as the slashdot crowd can be, I don't think
most of us are gullible *enough* to fall for something like *that*.
2. Those who are probably don't want to publically admit it; even if they
did post, it would almost surely be AC. Better browse at -1 if you
want to have any chance at all of seeing them.
Let me get this straight -- this outfit wants to charge you money up front for
the valuable opportunity to do business with them and possibly get a job, and
you're sufficiently unsure whether this is a good idea that you have to ask on
slashdot?
Dude, nobody legitimate has to charge you for the chance to work. A temp
agency might charge your employer more than they pay you and so skip off the
top, but they don't get a *dime* until you get a paycheck. Similarly, a more
traditional employer might place expectations on you (dresscode or whatnot)
that might result in your spending money before your first day on the job,
but you don't pay *them* anything.
Work-from-home is notoriously fertile ground for scammers of all kinds, but
if they're wanting to charge you up front for a lifetime membership for the
opportunity to work, that should set off loud alarms and flashing red lights
in your brain even if the job in question were more traditional (e.g. factory).
Interesting. But you don't have to know that to figure out that it's on most
of the world's power buttons and guess that it must be the symbol for power or
perhaps "on".
> I have to defend the old scissors icon for "cut". It's always made perfect
> sense to me. But, then again, I grew up actually doing "real" cut-&-pasting
> with some scissors, glue and a photocopier.
I understood it, because my dad used to do this kind of copy-and-paste work
when I was a kid. But my point was, it's been ten years since anybody's done
copy and paste that way, but it still makes a good icon, and it will still
make a good icon fifty years from now when there's nobody left alive who
remembers doing it that way, because it's simple, visually distinctive, and
standardized across many applications. The same is true of the floppy disk
icon for save. Keep it.
I would have thought John Cage would sue them blind for selling a track of
nothing but silence and not paying him royalties.
Quick, off the top of your head, what does a red octagon with a white outline
represent? How about a button on a GUI that looks like a pair of scissors?
What about a red circle with a red line across it from the lower left to the
upper right? A button on the corner of a screen window that has an X in it?
Do *any* of these things actually look like the object or process that they
represent? Does it matter?
A good icon is simple, visually distinctive, easy to recognize instantly,
consistent across many interfaces. The floppy disk icon for save is all of
these things, and it's also familiar to almost every experienced computer user.
It could be simplified a little (removing some superfluous details, like the
label and the little readonly-lock thingydo), but the basic visual is already
quite simple and distinctive. Nobody's going to mistake it for (say) the paste
button. Sure, it's an anachronism, but the standard icons for cutting and
pasting are scissors and paste, respectively, and nobody's used *that* method
of cutting and pasting since word processing came into vogue. So what? The
icons are visually distinctive enough (well, the scissors are; they should
probably have used a roll of transparent tape for paste, but it's too late to
change that now) and their meaning is well established.
Have you looked at the icon on a power button lately? (No, not your old 8-bit
micro with the toggle rocker with 0 for off and 1 for on; something that was
manufactured this century.) On virtually every device it's the same. Why
exactly that specific symbol means "power" is quite beyond me (why not a
lightning bolt or something?), but everybody knows it's the power button
because it's the power button on everything -- computers, monitors, UPS units,
even a growing number of kitchen appliances. This is a Good Thing(TM).
So, take that picture of a floppy, simplify it into a basic icon, and use
it to represent the concept of saving from now on. It doesn't matter if
half the people clicking on it have never seen an actual factual floppy
diskette and don't know the history behind the symbol; they won't have to
look at very many applications before they learn it's the universal symbol
for "save changes".
> You should forward all of your spam to uce@ftc.gov.
I used to do that, but a couple of years ago it got to the point where it was
using too much of my bandwidth, and I had to stop. (My spam volume has gone
up several hundred percent since then...)
> I wonder if I could make money with uselesscrapyouthinkyouwant.com...
Calling the site that is poor marketing. You'll sell much more useless crap
if you market it properly. Seriously, with the right advertising campaign,
I'm convinced you can get people to buy used dental floss or anything, but
probably not many will buy it if you call it "useless crap".
There are different ways to spin ("market") the concept of useless crap.
Calling it "Products You Never Knew Existed" or "Stuff You Didn't Know You
Needed Until You Saw It" would be a little better, but these slogans are far
too long to be really effective. "Awesome Stuff" is better, but it loses
some of its punch due to being somewhat hackneyed. Probably what you want is
to name your company something abstract and then advertise individual products
or lines of products.
> 3) You need to access everything from the keyboard - with no third party
> extensions, which widget set do you pick, OS X's or GTK+?
Actually, for that, the Win32 widget set rather rocks. Unfortunately, it's
not very portable.
> The real question is: why are the standard window managers for GNOME and
> KDE (metacity and kwin) not implementing something like this
Not sure about kwin, but the official metacity project slogan is "no features"
(apparently). A lot of Gnome users swap it out for a different window manager
such as sawfish or Enlightenment. Fortunately, the architecture of Gnome makes
this possible (though the wm in question has to support certain Gnome things
to get everything working properly (e.g., the panel task list, having certain
panels be avoided by maximize, and so on), which does limit exactly which
window managers you can choose; you can _theoretically_ choose any wm, but
if you choose one that doesn't support Gnome stuff won't work; sawfish and
Enlightenment are the major alternatives that handle Gnome stuff right, that
I know about; there may be one or two others).
Funny you should mention Greek; "morph" also comes from Greek. (Not sure about
avatar; that one might come from Latin; it doesn't sound Greek to me.)
usenet is not a print source, so I don't think it's admissible. Mailing list
traffic certainly wouldn't be admissible, as letters aren't and so email is
essentially the same sort of thing, and definitely very informal. The thinking
here is that when a word starts to pass from the realm of slang into the realm
of mainstream vocabulary, it starts being used in print.
> I don't understand: why does a usage have to be on paper to count for this
> project?
When a word passes out of the realm of slang and into the realm of general
mainstream vocabulary, it starts showing up in print sources. I strongly
suspect that "cloaking device" will be located in print sources dating back
to the sixties, because Star Trek was sufficienly popular that people would
have picked up the term there and used it elsewhere. But yes, the OED is
put together with the sort of thoroughness you expect from Oxford, and so of
course they have to see actual evidence of the term being used in print. In
the case of "cloaking device", I'm pretty sure said evidence will turn up.
I always thought they preferred to see examples of the word being used in print
sources; I didn't think they generally went by private letters or whatnot, and
I'm thinking usenet would be in the latter category. (Informal sources often
show slang terms that have not yet passed into general usage; when you start
seeing the word in books and journal articles, it starts being mainstream.)
Morph as a _root_ is very old (we're talking at least a couple of millenia
...), but the question is about "morph" as a word with its
here), but I don't remember its being used as a verb in the English language
in its current sense much before the early-to-mid nineties.
Can you find the specific articles in Scientific American that you're thinking
about? If they use "morph" as a verb in its current sense, they might count
as an earlier use of the term.
The etymology of "morph" is of course straightforward, and we've had other
words in the English language that use the root for a while (e.g. morphology,
metamorphosis,
current meaning.
Why keep C on life support? C was designed for computing in an era when every
byte mattered and you would store some numbers in 16 bits and others in 32,
depending on how large they would probably get, an era when you would manually
allocate each piece of memory and free it when you were done because you didn't
want to have any allocated that you didn't need, not even until the end of a
block of code, an era when CPU time was so precious that garbage collection
was considered expensive and wasteful, an era when, in short, the computer's
resources were so valuable that you would spend programmer time willy nilly to
conserve them. It's an anachronism, a language that *needs* to fall into
disuse (except for a few inherently low-level tasks like writing bootloaders
and kernels), and the sooner the better. Portability is not the language's
only (or even its most important) weakness! Do you really want to continue
to have to comb your application source for every single malloc and hunt down
the corresponding free, to doublecheck every buffer to make sure you've
checked its bounds every time you read or copy anything into it, and to write
pages and pages of code where a paragraph would get the job done in a more
modern language? Do you really want to have to continue to construct simple
structures like lists *by hand* using pointers, when in other languages
they're a fundamental data type? (Yes, there are libraries now to handle
some of these things, sort-of, but the language just isn't geared that way;
it's geared toward low-level bit-fiddling and going through the programmer's
time like water in order to conserve the computer's resources.)
I know you have fond memories of the language, and surely it was good in its
time, but its time has come -- and gone. Sooner or later you've got to let
it go. Say a nice eulogy and let's get on with our lives now.
> This is a nominally 17" Proview CRT that I bought for $200 about three
> and a half years ago.
If you spent that amount on a monitor today, you'd expect better.
> Then what happens when I want to add an app to the computer?
Get the old version that was current when the hardware was new?
I know, everybody says, "Get the latest version", but that's only true if
you've got relatively recent hardware; if your hardware is old, you're usually
better served with old software too.
> What happens when I receive a document in a format created by a newer
> version of a particular app that the version installed on my computer
> can't read?
The same thing that happens when someone sends you a document in the obscure
proprietary format of a random app they happen to have that you don't have:
you ask them to please send you the information in a standard format. If you
try to have all the versions of all the applications anyone might ever use to
create a document that they might send you, you'll be spending more money on
software than you spend on hardware. A lot more. You'll go out and buy the
latest MS Office every year ($650), and then some clown will send you a
FoxPro spreadsheet, so you'll buy the Corel Word Perfect Suite ($299), and
then somebody will send you a MS Publisher document, so you'll shell out for
the latest version of that, and then some machead will send you something
they did in Pagemaker or Illustrator or InDesign... if you can afford all
that, you can *certainly* afford some extra RAM. You can get 512MB of SDRAM
for less than the average cost of *one* of these software upgrades.
At some point, you have to stop buying every piece of software on the market
and ask people to send you the information in a standard format. Unless you
have an infinite budget.
> Does anybody really need 1 GHz for a word processor or spreadsheet?
No. 300Mhz is more than plenty to handle that stuff comfortably, as long
as you've got adequate RAM. That was my point.
> "The software I've already got" has known vulnerabilities whose only patch
> is an upgrade to a new operating system that requires computers that cost
> $1,000 per seat that I don't have.
Security is a separate issue from RAM usage. Let me let you in on a secret:
the latest and most current version of Windows XP[1] has known vulnerabilities
for which no patch is available. (Actually, it's not all that secret. I'm
talking about stuff that is publically known, some of which is even in the
MSKB.) Windows systems should always be behind a firewall; nothing new there.
[1] Here I'm including some of the apps that are bundled with it; technically
speaking it's possible to disable Outlook Express for example, but again,
security is a separate topic from RAM usage and is really off-topic for
this thread.
> Redoing drywall is actually pretty cheap, especially if it's just a
> small patch.
If you have to rip out enough drywall to drill through a bunch of studs and
run a cable, "just a small patch" is going to turn into Famous Last Words,
and that's before you have to also patch the wallpaper...
> It also looks impressive to the ladies without being hard at all.
Running the cable does that well enough, especially if you wire the jack
yourself. ("Yeah, this is a standard type-B wiring for RJ45. See, you
keep each pair twisted until it's right up next to the jack. The orange
pair goes here and here...")
> You're probably doing this, but many other sites (heh, usually IE-only
> sites) sure as heck aren't. What about 320x480, 400x600, 640x1024?
If it works at 640x480 and 1600x1200, it'll work fine at 640x1024 also. The
less-than-600 widths are another matter. I am aware that there are people
who size windows to half the width of their screen, but I am also aware that
they have the capability to size them differently without causing a disruption
to their entire environment (only the browser is impacted, generally) if they
desire. I make some attempt to see what my pages look like at 320x200, but
at some point a multicolumn page just can't be crammed into that, and if the
user has to scroll horizontally to read the other column I generally am
willing to live with that if they've got their browser window sized to less
than 600 pixels across. Now, if a block of text is going wider than the
window so that they have to scroll horizontally as they read each line, then
I would consider that to be something I should look into fixing, down to
resolutions of about 320 across. If their browser is much narrower than
that, it's a pathological case and there's not much that can be done.
There are people out there who want all web pages to be designed for their
wristwatch systems that boast a black-and-white (not grayscale; 1-bit color
depth, if you can call that color depth) display with two-digit dimensions
in pixels, but I've got to draw the line somewhere. If they really need
to browse on a display like that, they should use a browser that discards
all the layout info from the page (probably all style info, actually) and
just extracts the bare information itself and hyperlinks.
My philosophy on color depths is similar; I try to pick colors that will be
*legible* down to 256-color mode, but I can't promise they page will look
*good* at anything less than 24-bit color. (Anyway, you can always do what
I (usually) do and discard the color information the page specifies in
favor of your own color preferences.)
> Value priced computers do not come packed with a "remotely decent" monitor.
Many of them don't come with a monitor at all; you either buy that separately
or use one you already have. Some (especially the mail-order ones) do come
with a monitor included, but even some of those include a halfway decent one.
> I guess that by calling the optional feature of different resolutions
> for different workspaces worthless in practice, I underestimated the
> percentage of "remotely decent" monitors on the market
You're probably looking at the extreme low end, the sort of monitor that's
bundled with computers that include a bundled monitor, the ones that have less
than sixteen inches (diagonal) of viewable picture space, e.g., the cheapest
of the cheap of "17-inch" monitors (the ones that are like 15.3" viewable or
whatever) and, of course, the ones that don't even pretend to be 17". If you
walk into a place like Circuit City, they'll have maybe three of these on
display; mostly they'll have higher-end 16-viewable monitors, 19-inchers (18
viewable; these are the midrange), and LCDs. They might also have one large
high-end monitor.
Now, it is true that the low-end ones sell a disproportionate number, because
maybe 30-40% of the people buy the absolute rock-bottom cheapest thing they
can get, but still there are quite a few people out there with a nicer model.
Nineteen-inch monitors are no longer a novelty item; ecconomy of scale is
such that they're *almost* as cheap as the 17" monitors. For example, a
place[1] that sells an ultra-low-end 13.8" viewable monitor for $122 and a
range of 16"-viewables for anywhere from $126 - $158 might sell 18-viewables
for $163 - $220. This is the sweet price spot, because their 20-viewables
might run $472 and up. Most of those 16" displays and probably all of the
18" ones can handle changing display resolutions without a hitch.
LCDs, of course, have serious trouble with resolutions other than the "native"
one, but LCDs also have serious trouble with color accuracy and viewing angle;
if they weren't so much smaller physically there'd be no reason at all to use
them (other than in laptops, where every watt of power matters deeply for the
battery life).
And yeah, the low-end monitors like the 13.8" viewable are probably not
going to give you a lot of pleasure. But buying those only saves you in
the general vicinity of ten bucks or less versus a decent 16" viewable.
(In my example prices[1], it's a different of four whole bucks, woo.)
The five-year-old computer presumably has an existing operating system; it doesn't need a new one.
> The problem here is that few companies still publish applications for
> Windows intended to run on the five-year-old computers that schools
> still have. Many X11 apps, on the other hand, seem to demand much less
> RAM. (No, just because top counts VRAM in its total doesn't make X11 a
> memory hog.)
A five-year-old computer, in addition to already having an OS, might also
already have some apps installed, perhaps, don't you think? At least, it
would have all of the ones that had been being used all along? Sure, it
would be nice to upgrade, but it would be _nice_ to upgrade the hardware,
too; sometimes on a budget you let upgrades wait, right?
I do *not* think it's reasonable to expect any software project, commercial
or non-commercial, proprietary or open-source, to necessarily produce new
applications for old hardware. Yes, some projects do do this, in some
cases because the problem domain is such that it is possible (firewalls
come to mind here), but in general I don't think it's realistic to expect all
new software to run on arbitrarily old hardware. New wine, old wineskins.
If you want to run all of the latest and greatest software, then you might
have to shell out for new hardware. This doesn't make old hardware useless;
That hit the nail on the head. If you do the conduit right, with nice big
junction boxes at *all* corners, conduit running to multiple locations in
every room, and so on, then you can run whatever kind of cable you want at
any time in the future very easily. Audio cable, video cable, fibre, Cat12b,
you name it, you'll be able to run it. What kind of network cable will you
(or whoever lives there) want in fifteen years? Fifty years? You have no
clue, right now. But you know it'll be easy to run it; take a screwdriver,
take the faceplates off, pull the cable through, and you know who Bob is.
One more thing:
> Run all your conduit to a central location (probably in the basement).
> You'll want a nice (rack even?) open area that you can mount equipment
> as well as patch panels, etc. Wire ties are your friend.
If it were me, I'd put nice boxes (kinda like a breaker box, but without the
breakers) every fifty feet or so around the outside wall of the basement, with
a nice fat conduit running straight up from each one to a junction box,
accessible at the top by removing a faceplate (like a lightswitch cover).
In any part of the basement that's going to be "finished", I'd also run
horizontal conduit between these boxes. I'd put an electrical outlet near
each of these boxes, so that a hub or switch can easily be put there.
Then, N years from now, when you want to run your new Terabit Ethernet cable
or whatever from the sewing room to the kitchen, you pick one of the empty
faceplates in the sewing room, run the cable from there to the nearest
junction box and down to the box in the basement, where you put a hub. In
the kitchen you do the same thing; then you run a cable around the horizontal
conduit to connect your hubs, and you're done; with a hub at each of the
basement boxes, you never need more than one horizontal cable of any given
type, no matter how many things you run in the room above.
So, you want more faceplates than you need right now in every single room
(yes, the bathroom; yes, the garage, too; I'd put one in each closet also),
and a system of conduit connecting them all. You do NOT want to have to
tear up your drywall later because there's not conduit going to such-and-such
a location.
This isn't cheap; conduit costs more than cable. It costs less than redoing
your drywall later, though. Run the conduit. You'll be glad you did.
> what if you don't have a windows computer to see how 'it looks' under IE?
Exchange screenshots with another webmaster who does use IE. I've got a couple
of people I trade screenshots with regularly. They like this arrangement,
because my screenshots show some edge cases that most people would miss.
I always take a series of shots showing scalability from 640x480 up through
at least 1280x1024, and I always show what the site looks like with and without
page colors turned on (and my system colors are medium-contrast light-on-dark,
which shows up stuff that gets missed if you use black-on-white). Also I tend
to take screenshots with about three different rendering engines (always Gecko,
plus usually Konqueror and one or more of Opera, W3, Links, Lynx). So my
approach shows up a lot of edge cases that more typical setups (black-text,
white-background, MSIE, 800x600, page colors enabled) won't see.
Yes, it's possible to design a web page that looks "right" under all of the
above settings. (By "right", I mean it looks like it was designed for those
settings.) Eye candy in the graphical browsers, without breaking the text
browsers; client-side scripts that automate things if scripts are enabled,
without breaking the site if scripts aren't enabled. Images that look good
(no jaggy edges) against either a light _or_ dark background. (This is
tricky if you have to support browsers with no proper alpha channel, but
it can be done; the trick is to set the background color when you save your
PNG images, so that non-alpha-channel browsers (*cough* MSIE) will antialias
against that color -- set that to the same color as the surrounding background
and you get to be Bob's nephew. Or use the MSIE PNG-alpha-channel hack, but
not all versions of MSIE support that, and it still leaves old versions of
Opera out in the cold.)
Green on black is _okay_ (unlike black on blinding white, which will make you
snowblind if you set the contrast high enough to show detail properly), but
there are better color schemes. Amber on black looks uglier for the first five
minutes, but after your eyes get used to it you can stare at it for much longer
periods of time with zero eyestrain.
Even better, I've found, is a tertiary color scheme. Set your system foreground
color to #FFE6BC and your background to #294D4A. Set this system-wide. If you
use Gnome, there's a theme available called themacs (GTK1) or eMaCs (GTK2).
On Windows you can just go into the Display properties under the Appearances
tab (clicking on Advanced if you're using WinXP) and set it up manually easily
enough. It's possible in KDE/Qt also, though I forget the exact steps. Anyway,
Also set your web browser to use these colors exclusively (ignoring the colors
the web page author specifies). (You'll need to change your link colors also;
blue on slate green isn't the best combination.)
Your first reaction will be, "boring", but use it for a week and then ask
yourself when was the last time your eyes got worn out looking at your
display; it'll be a week ago, right before you switched to this scheme.
Oh, and do make sure your refresh rate is set as high as it will go.