> I made sure the site's still usable, even if it doesn't look quite right.
That's how I handle IE6 as well. I go ahead and use transparent png images, knowing full well that IE6 is going to put gray rectangles behind them. Users who don't like gray rectangles can jolly well upgrade. But if there's a problem that makes the site unusable or illegible in IE6, I definitely want to know about that and fix it. I think this is a reasonable level of support for a nine-year-old browser.
But I *don't* think it would be reasonable to impose on your design the kinds of limitations you would have to impose to get the site to look the same in IE6 as it does in a modern browser. (This is what some people mean when they say "support IE6".) That would mean you couldn't use any feature less than ten years old, and your site would feel very dated. I'm opposed to that.
> A lot of companies have web apps and intranet pages that do not render > correctly in anything other than IE6. Who pays to update all of that?
This is easy. You put a link on the desktop labeled "Corporate Intranet" (or whatever you call the thing, which points to IE, which has your intranet thingy set as its start page. You put another link on the desktop that says "Browse the Internet" and points to a modern web browser (something less than five years old), which has something else for its start page (perhaps the public website, or Google, whatever).
For bonus points, you set the security settings in IE6 so it can *only* be used on the intranet, because using it on the public internet is an unnecessary risk.
> IE6 is mandatory as it is still in use by more than 5% of our audience.
Some level of support for it might be mandatory, but I don't think it's necessary for absolutely every aspect of the site to be every bit as good in IE6 as it is in modern browsers. Let people have *some* incentive to finally upgrade.
Personally, I have IE6 in my third-tier of support, which means the site should actually be usable, but it's okay if it looks a little wrong. For instance, I go ahead and use transparent png images, knowing full well that IE6 will put an ugly background behind them. If you don't like the ugly gray rectangles, you could always upgrade to a browser that's less than eight years old.
There's support and then there's support and then there's support.
As a web developer, I have four tiers of support (if you count non-support as the bottom tier):
The first tier is my primary target, the browser I use myself and recommend to people who ask me what browser is good. I don't implement anything that doesn't work right on this browser. The standard here is, on this browser the site should be exactly the way I want it. Currently this is Firefox 3.x.
The second tier is mostly inhabited by other browsers that I consider to be reasonably good and have convenient access to for testing. For example, the latest Opera falls into this category.
Any significant changes I make get tested in the tier-two browsers to make sure they're okay, and if they goof anything up, including minor layout glitches (within reason; I'm not going to worry about a border width being off by a pixel or something), I look into it and see if there's anything I can do to make it better. However, sometimes I go ahead and implement new things (like, say, @font-face) before all of these browsers support them, as long as they degrade gracefully. The standard here is, the site should look like it was intended to render the way it does render. No obvious glitches.
The latest version of the single most popular browser (currently, IE8) generally falls into tier two whether it's any good or not, and whether it's convenient for me to test with it or not. I'd prefer to put IE in tier three, because it doesn't run on my OS, but it's too popular for that, so I use rdesktop to test with it. (However, when all development on IE6 had totally stopped for several years, I did eventually drop it to tier three, even though IE7 wasn't out yet. Because IE6 was ancient and terrible, and I really needed alpha-channel transparency to do what I wanted, and every other major browser had supported it since the nineties, and I was tired of waiting.)
Links, since you mention it, also basically falls into the second tier, for the most part, but there are some things I just don't try to make it do, because of its inherent limitations. For example, I don't try to make photo galleries look good in Links, because that would just be silly. You get alt text, of course, but it might just say "[photo]" for every single image in the gallery, because why are you using Links to look at a photo gallery in the first place? No, I'm not going to type a long description of every single photo. If you simply *must* use something terminal-based to look at image galleries, you can always use wget to slurp them all down and compile some image preview software against aalib or something. Whatever floats your boat, but don't ask me to put in a lot of extra effort for it. I *do* check links for things like the ability to display the whole width of the page (i.e., stuff's not getting lost off the right edge) at eighty columns.
The third tier is inhabited by browsers that I only test in occasionally, usually because I don't have convenient access to them. Somewhat out-of-date (but not extremely ancient) versions of the major browsers fall here (e.g., Firefox 1.x), and also niche browsers like Konqueror, and also browsers that aren't available for my preferred OS (e.g., Safari). I try to at least get someone (or a service like browsershots) to send me screenshots occasionally, and if somebody reports a problem I'll see what I can do. My standard for these browsers is somewhat lower than for tier two. If stuff doesn't look quite right (e.g., the ugly background IE6 puts behind transparent png images), I don't really care. Upgrade or put up with it, your choice. However, if there's a problem that significantly interferes with the ability of the user to actually read the site, I definitely want to know about that and fix it.
The fourth tier is stuff I refuse to support at all. Extremely ancient browsers from the days of yore (IE4, old AOL browsers from before they acquired Netscape, etc), bizarrely obscure browsers ("hey, I'm
> There isn't a good reason to use AAs for a device > which goes through one charge in a matter of days
Depends where you're shipping the things.
For someone living in the developed world, you're probably right.
But AA is the single most standard, widely-available battery form factor in existence by a WIDE margin. They're more common than the next four most common form factors (C, 9-volt, AAA, and D) combined. You can get AA batteries absolutely EVERYWHERE, including places where you can only get Coca-Cola sporadically. The things are easier to find than T-shirts.
For a computer that you're going to replace in three years, this is unimportant. The built-in lithium-ion battery will last the life of the device. So for most of the slashdot crowd, yeah, the appeal is just not there.
But for a device that you want to be able to keep using well past the five-year mark, it can be VERY important. Four years after you buy the thing, the particular model of lithium-ion battery it came with is often no longer available even in the US.
Now, I'm not entirely sure the EduBook really qualifies as such a device. I mean yeah, they say they're shipping them to Nigeria, but how MANY of them are they really shipping there, honestly, and how many of those that they do ship to Nigeria are going anywhere beyond Lagos or Port Harcourt? And how useful would the devices even be in the middle of nowhere? Internet access can be pretty hard to find in the rural third world, and when you can find it it ain't free. How useful are these things without the internet? So maybe they're fooling themselves on that count.
But if it were the kind of device that would be distributed to the rural third world, running on AA batteries would be a major plus. It's a pretty big deal for cheap AM/FM receivers, for instance.
> they claim it's not really a bug because there is no > standard that say that a utility window should be on top.
Ah. That makes sense.
> I really don't understand how there can be a X window > manager standard which say that a window can be an > "utility window", without describing what kind of > behaviour such a window should have.
I take it you are (or recently were) a Windows or Mac user?
The reason such behavior would not be specified in a standard, is because not everyone wants the same behavior. The ability to use a different window manager and/or configure your window manager differently in order to get different behavior is part of the deliberately designed-in flexibility of the X windowing system, which allows different users to have things set up differently to suit the way they like to do things.
And yes, how various kinds of windows are handled is configurable on many window managers. Sawfish, for instance, gives you options to control, among other things, whether minimizing or unminimizing a window should minimize or unminimize its transients, whether selected windows are automatically unshaded, whether maximizing is a toggle or simply an automatic resize operation that leaves the window still resizeable, whether transient windows are stacked above their parents and/or automatically raised with them, whether window positions and dimensions are shown when moving and resizing and whether dialog windows are centered on their parents or placed independently (and how). Some window managers are much *more* configurable than sawfish, and even ones that are less configurable overall often have different options and features so that they appeal to certain people who like to do things that way.
If you haven't experimented with at least half a dozen different window managers, you can't fully appreciate the flexibility you have with X, to set things up to work the way you want them. I suggest trying out, at minimum, the following: sawfish, Xfce, Enlightenment, fluxbox, WindowMaker, wmii, and twm. You will probably find that most of them don't really do things quite the way you like. But seeing the wide range of options opens your eyes to the fact that you *aren't* stuck with the One And Only Way of doing things. You've got choices. Meaningful choices that have a real impact on how you use your computer.
Or you can get a Mac, and your choices basically boil down to "Do I want the window borders to be blue, silver, or green?" MS Windows is only a little more flexible.
Ah, that. I didn't think of it, because it's not specific to open/save dialogs. It's a more general problem with.lnk file support in cross-platform and open-source software.
To my knowledge, Emacs is the only piece of open-source code that knows how to parse Windows.lnk files more-or-less correctly. (In fairness to the people doing the porting, they don't *have* to parse link files on *nix systems, because symlink semantics are built into the filesystem and/or the API. Still, the.lnk format was introduced in 1996, so you'd think someone would figure out how to support it by now.)
> 2) (On OS X)
I haven't used OS X enough to comment on that.
The last time I had access to an OS X system, getting GTK software to run on it could ONLY be done via the X server, and getting rootless mode to work was a new and esoteric trick most people hadn't tried out yet. (Yeah, it's been a while. The Mac in question was one of those ridiculous-looking malformed-lamp iMacs, which had just come out recently. Adobe software at the time all still required Classic, including Acrobat Reader.)
So I can really only comment on the PC side of things.
> 3) no voice control, no tablet stylus support, no > touchscreen support (that's universal among all GTK+
Ah. I've never had the hardware for any of that stuff, so naturally I was blissfully unaware of any problems with it.
> 4) No open to move, delete, create shortcut, map > network drive or any of those other options normal > Open/Save dialogs have
Not can sentence fully this understand I.
Are you saying you use open and save dialog boxes as a file manager? So, like, if you want to copy some files, you pull up a word processor or something, hit Save As, find the location where the files are that you want to copy, and then, umm, somehow use the dialog box to copy them?
I was not aware that was even possible. Sounds cumbersome.
You do know about Windows Explorer (alias My Computer), right?
Call me weird, but personally I use file open and save dialog boxes for opening and saving.
> 5) It's been awhile, but I seem to recall it was incapable > of accessing FTP servers mapped as drives in Windows,
If it's mapped as a drive with a letter (the closest analog, on Windows, to mounting it), the software shouldn't have to do anything special to access it. It should work just like accessing any other part of the filesystem. The app shouldn't even have to *know* that it's actually an ftp server (or nfs, or cifs, or whatever). If it's mounted (i.e., has a drive letter), it should Just Work. If it doesn't, that's a bug in the operating system, and a fairly major one.
> or any network drive that didn't have a drive letter (could > be wrong on this one)
If you think reinstalling from an XP SP1 disk is a pain, try using a pre-SP1 XP disk. The version of IE it comes with isn't good enough for Windows Updates, so you have to download and install an IE6 update first (which, naturally, requires a reboot), before you can even get started.
>...which shows what is wrong with Microsoft's kernel
It's not the kernel. It's the filesystem.
Most filesystems, and in particular all the ones that are popular in the Unix world, have an abstraction/redirection layer sitting between a file's directory entry and the actual file contents. Unix people call them "inodes". The details vary somewhat depending on exactly what filesystem you're using, but in general the directory entry points to the inode, and the inode points to the actual file contents wherever they're stored. Because of this, a file can be changed or even replaced in situ, even while another process has the file open and is using it. The inode for the old file remains until the process that was using it lets go, but the directory entry is updated to point to the new inode.
FAT and NTFS don't have inodes, so it's not safe to alter a file while another process has it open. So you have to stop every process that's using the file, before you can do that. The easiest way to do that (and in some cases the only way, e.g., if the file is a shared library that lots of programs use) is to reboot.
> > updates that require a reboot so they can install > > another update that then requires another reboot. > Ah, see now you're confusing Microsoft with Adobe.
No, Microsoft does it too. The only way you don't notice is if you keep up-to-date on your updates all the time and never ever have to do a reinstall. Good luck with that second part, seeing as how this is Windows we're talking about.
Just for grins sometime, reinstall Windows and *count* how many reboots are required before you are fully up to date. Don't stop counting Windows Update, run right after a reboot, tells you there are no more updates to install.
The worst I've ever seen was a Windows XP system that didn't have any of the service packs included on the install CD. First, just to get to a desktop, you have to reboot at least twice (not counting initially booting to the install CD to get started). Then Windows Updates won't do squat until you install an IE6 update, because the version that comes with isn't good enough for Windows Updates. The IE6 update requires a reboot, naturally. Then you can start getting Windows Updates...
Oh, and then you've got allergies, in which case your immune system isn't directly attacking you per se, but it's overactively attacking something it doesn't really need to, and you're sort of getting caught in the crossfire.
> Actally having too strong an immune system IS bad; that's > what arthritis is, your body's immune system attacking you.
Arthritis, lupus, Crohns, Grave's, Psoriasis,... all kinds of fun stuff.
> Too bad it will be "a few decades", I'll be dead by then.
Heh. When speaking of the future development of a certain technology (medical or otherwise), "a few decades" typically either means "less than twenty years" or "never". In this case, let's just say I won't be betting any money on the former timeframe.
> That we can't actually see a majority of diseases under a microscope
No, that's wrong. Well, it could be technically right (depending on how you define "disease"), but not in the way you were probably thinking.
If the human immune system is making antibodies to fight the disease, then there's a pathogen involved, which we most certainly can see under a microscope. (These days, microscopes can image amazingly small objects, much smaller than any protein. Not in color, of course, because we're not talking about light-refracting microscopes here. But that's really beside the point. Anyway, I'm not sure what "color" would even mean at the molecular level.)
The larger issue is that we understand the human immune system about as well as we understand the human brain. Talking about re-engineering it and making "Immune System 2.0" is pretty arrogant and unrealistic, IMO. Giving the immune system a sample of what it's supposed to fight and letting it learn to fight that (vaccination, in other words) is one thing. That's like letting the search dog sniff a scrap of the missing person's clothing so they know who they're supposed to find. Doing the actual searching is still up to the dog. Trying to improve on the actual disease-fighting process would be something else quite again. That's like trying to design a better dog. Human knowledge of how living systems work is just plain not up to that yet.
I suspect it's more likely a sign that they know full well nobody *needs* their "FAST" search product on a Unix system, because it's redundant with standard basic system utilities that every Unix geek has taken for granted since the seventies. They would never say it that way, of course, because that way makes Windows sound inferior.
> But as you found out, this is not what really > happens, so either its a bug in you window manager
It could be a window manager bug, or it could also be a deliberate design choice. But either way the difference you experienced is the window manager's doing. The setting in question does work as described with most window managers.
> Considering how wrong GTK gets Open/Save dialogs *alone* in Windows...
I hadn't noticed.
The main programs I've seen with GTK widgets on Windows are Gimp, Inkscape, Emacs, and Freeciv. Maybe they don't suffer from the problem? Emacs obviously wouldn't because it uses the minibuffer in lieu of Open/Save dialogs. Gimp does some of its own custom stuff in the save dialog (notably, instead of automatically filling in the extension when you choose a filetype, it automatically fills in the filetype based on the extension), so maybe it doesn't use the regular GTK save dialog either? Oh, and civserver takes savegame filenames on the command line and generates save files automatically based on server options. Not sure why I haven't noticed anything wrong in Inkscape.
Out of curiousity, what, exactly, does GTK get wrong about file open/save dialogs on Windows?
> Linux might be a different story.
Linux is what I use on my own workstation, both at home and at work, but I'm significantly familiar with Windows systems from working briefly with other people's computers, when I'm away from my desk for whatever reason, or when somebody needs help with something. I guess I probably use Windows about ten or twelve hours a week, all told.
I did copy and paste, of course. (Well, technically I selected the text, then pasted the current selection with a middle-click, without involving the copy clipboard; but that distinction is probably lost on kids these days, I suppose.)
I substituted the abbreviation because spelling it out roughly doubled the length of what I was quoting. Using the shorter form made it more likely to stay on one line, among other benefits.
As for quote tags, I don't use them because I don't like them. As far as I'm concerned, actual quote marks along the left edge are much better, a more readily apparent visual indication of quoting. So that's what I use. Slashdot even has a post setting (Extrans) that makes this easy, and a preference you can turn on that makes this the default for all your posts. Slashdot has such facilities because they have deliberately chosen to put up with, and even deliberately accommodate, geezers who remember usenet fondly. If you don't like this, you can always go find a site that doesn't offer such options. MyBook or FaceSpace or something along those lines, perhaps. Personally, I'd rather be here, hanging out with the old geezer computer nerds who actually remember usenet, but YMMV.
> a lot of people (ie older parents/grandparents) > buy a Mac because it's "easier" and are more > inclined to be on a 5-10 year cycle.
In the Windows world, a lot of people are on an 8-12 year upgrade cycle.
I wouldn't be terribly surprised if more people are still using Windows 98 than the total number of Mac users, all versions combined (actual general-purpose computers, I mean, not handheld music players and such). Obviously there's no reliable way to get actual numbers for how many Windows 98 systems are still in use, but I bet it's a lot higher than most computer geeks realize. People's tendency to upgrade promptly increases (more or less) geometrically with their level of computer knowledge, and the people whom an IT professional knows personally tend to be significantly above average.
> I maintain a 15k line Javascript application > that's compatible with IE 5.5.
Ouch.
I'm pretty sure I'd rather support a punchcard-driven mainframe than IE 5.5.
If your IE-5.5-supporting Javascript code is by some miracle of herculean effort *also* compatible with modern browsers in standards-based rendering modes, you ought to win some kind of prize. (Do they give out prizes for masochism? What do they do, pin a medal on your bare chest?)
> GTK+ looks like an total alien on every OS except Linux.
IMO, GTK is significantly more consistent with "native" (Win32) apps on Windows than it is with non-GTK apps on Linux (e.g., KDE stuff, or the occasional Tk or Xaw application).
Personally, I'm more interested in functionality and flexible customizeability than absolute visual consistency, so having GTK and Qt apps (say) on the same system doesn't really bother me, as long as they're all good solid well-designed and properly configurable applications that meet my needs.
(Xaw does bother me, but that's because it's terrible.)
> Newer Firefox will probably still work on Linux on G4,
That isn't necessarily going to make Mac users happy. Typically, they didn't buy the G4 to run Linux. If they wanted to run Linux, they could have gotten a better-performing Celeron system for less money.
> or in X11 on OSX 10.4.
If so, they should be specifically saying so. "To run our software on your OS, you will need to install an optional library, which your OS vendor offers free of charge" is not entirely the same thing as "our software will no longer run on your OS".
> Now, if this had been posted to a WoW forum...;-)
Meh. Really the modern equivalent of the old slashdot effect these days is when the Google doodle returns your site as the first result. Hopefully your hosting provider doesn't bill by the megabyte...
> if my life is on the line, I'm trusting > nothing less than a hot loaded 9mm JHP.
Why so small? I prefer to cart around an M109. For delivering a lot of kinetic energy to my would-be attacker's chest, what could be better than 155mm rocket-assisted rounds?
Mozilla, as an organization, still hasn't figured out that whether the server I interact with today has the same private key as the a server I interacted with previously is more important than whether the people running the server paid money to Verisign (or whomever) for a cert signed by somebody on The List.
Mozilla, as an organization, still hasn't figured out that whether the DNS entry that resolves the FQDN to the server's IP address follows a continuously signed path up from the root is more important than whether the people running the server paid money to Verisign (or whomever) for a cert signed by somebody on The List.
Mozilla, as an organization, seems completely unaware of the inherent problems with HTTPS. They seem to think EV (which basically boils down to getting site operators to pay even more for their certs) is a meaningful solution. It is clear to me that they, as an organization, don't understand security at all.
Perhaps there are some people within the organization who do understand security. If so, they are not the people driving policy.
For what it's worth, I'm not aware of any other browser that does any better. But that doesn't mean it's all good. It's not all good. As things stand right now, people should think of https as inherently completely insecure. Because it is.
If you need to transfer information securely, don't use https.
Slashdot Mirror
> I made sure the site's still usable, even if it doesn't look quite right.
That's how I handle IE6 as well. I go ahead and use transparent png images, knowing full well that IE6 is going to put gray rectangles behind them. Users who don't like gray rectangles can jolly well upgrade. But if there's a problem that makes the site unusable or illegible in IE6, I definitely want to know about that and fix it. I think this is a reasonable level of support for a nine-year-old browser.
But I *don't* think it would be reasonable to impose on your design the kinds of limitations you would have to impose to get the site to look the same in IE6 as it does in a modern browser. (This is what some people mean when they say "support IE6".) That would mean you couldn't use any feature less than ten years old, and your site would feel very dated. I'm opposed to that.
> A lot of companies have web apps and intranet pages that do not render
> correctly in anything other than IE6. Who pays to update all of that?
This is easy. You put a link on the desktop labeled "Corporate Intranet" (or whatever you call the thing, which points to IE, which has your intranet thingy set as its start page. You put another link on the desktop that says "Browse the Internet" and points to a modern web browser (something less than five years old), which has something else for its start page (perhaps the public website, or Google, whatever).
For bonus points, you set the security settings in IE6 so it can *only* be used on the intranet, because using it on the public internet is an unnecessary risk.
> IE6 is mandatory as it is still in use by more than 5% of our audience.
Some level of support for it might be mandatory, but I don't think it's necessary for absolutely every aspect of the site to be every bit as good in IE6 as it is in modern browsers. Let people have *some* incentive to finally upgrade.
Personally, I have IE6 in my third-tier of support, which means the site should actually be usable, but it's okay if it looks a little wrong. For instance, I go ahead and use transparent png images, knowing full well that IE6 will put an ugly background behind them. If you don't like the ugly gray rectangles, you could always upgrade to a browser that's less than eight years old.
There's support and then there's support and then there's support.
As a web developer, I have four tiers of support (if you count non-support as the bottom tier):
The first tier is my primary target, the browser I use myself and recommend to people who ask me what browser is good. I don't implement anything that doesn't work right on this browser. The standard here is, on this browser the site should be exactly the way I want it. Currently this is Firefox 3.x.
The second tier is mostly inhabited by other browsers that I consider to be reasonably good and have convenient access to for testing. For example, the latest Opera falls into this category.
Any significant changes I make get tested in the tier-two browsers to make sure they're okay, and if they goof anything up, including minor layout glitches (within reason; I'm not going to worry about a border width being off by a pixel or something), I look into it and see if there's anything I can do to make it better. However, sometimes I go ahead and implement new things (like, say, @font-face) before all of these browsers support them, as long as they degrade gracefully. The standard here is, the site should look like it was intended to render the way it does render. No obvious glitches.
The latest version of the single most popular browser (currently, IE8) generally falls into tier two whether it's any good or not, and whether it's convenient for me to test with it or not. I'd prefer to put IE in tier three, because it doesn't run on my OS, but it's too popular for that, so I use rdesktop to test with it. (However, when all development on IE6 had totally stopped for several years, I did eventually drop it to tier three, even though IE7 wasn't out yet. Because IE6 was ancient and terrible, and I really needed alpha-channel transparency to do what I wanted, and every other major browser had supported it since the nineties, and I was tired of waiting.)
Links, since you mention it, also basically falls into the second tier, for the most part, but there are some things I just don't try to make it do, because of its inherent limitations. For example, I don't try to make photo galleries look good in Links, because that would just be silly. You get alt text, of course, but it might just say "[photo]" for every single image in the gallery, because why are you using Links to look at a photo gallery in the first place? No, I'm not going to type a long description of every single photo. If you simply *must* use something terminal-based to look at image galleries, you can always use wget to slurp them all down and compile some image preview software against aalib or something. Whatever floats your boat, but don't ask me to put in a lot of extra effort for it. I *do* check links for things like the ability to display the whole width of the page (i.e., stuff's not getting lost off the right edge) at eighty columns.
The third tier is inhabited by browsers that I only test in occasionally, usually because I don't have convenient access to them. Somewhat out-of-date (but not extremely ancient) versions of the major browsers fall here (e.g., Firefox 1.x), and also niche browsers like Konqueror, and also browsers that aren't available for my preferred OS (e.g., Safari). I try to at least get someone (or a service like browsershots) to send me screenshots occasionally, and if somebody reports a problem I'll see what I can do. My standard for these browsers is somewhat lower than for tier two. If stuff doesn't look quite right (e.g., the ugly background IE6 puts behind transparent png images), I don't really care. Upgrade or put up with it, your choice. However, if there's a problem that significantly interferes with the ability of the user to actually read the site, I definitely want to know about that and fix it.
The fourth tier is stuff I refuse to support at all. Extremely ancient browsers from the days of yore (IE4, old AOL browsers from before they acquired Netscape, etc), bizarrely obscure browsers ("hey, I'm
> There isn't a good reason to use AAs for a device
> which goes through one charge in a matter of days
Depends where you're shipping the things.
For someone living in the developed world, you're probably right.
But AA is the single most standard, widely-available battery form factor in existence by a WIDE margin. They're more common than the next four most common form factors (C, 9-volt, AAA, and D) combined. You can get AA batteries absolutely EVERYWHERE, including places where you can only get Coca-Cola sporadically. The things are easier to find than T-shirts.
For a computer that you're going to replace in three years, this is unimportant. The built-in lithium-ion battery will last the life of the device. So for most of the slashdot crowd, yeah, the appeal is just not there.
But for a device that you want to be able to keep using well past the five-year mark, it can be VERY important. Four years after you buy the thing, the particular model of lithium-ion battery it came with is often no longer available even in the US.
Now, I'm not entirely sure the EduBook really qualifies as such a device. I mean yeah, they say they're shipping them to Nigeria, but how MANY of them are they really shipping there, honestly, and how many of those that they do ship to Nigeria are going anywhere beyond Lagos or Port Harcourt? And how useful would the devices even be in the middle of nowhere? Internet access can be pretty hard to find in the rural third world, and when you can find it it ain't free. How useful are these things without the internet? So maybe they're fooling themselves on that count.
But if it were the kind of device that would be distributed to the rural third world, running on AA batteries would be a major plus. It's a pretty big deal for cheap AM/FM receivers, for instance.
> they claim it's not really a bug because there is no
> standard that say that a utility window should be on top.
Ah. That makes sense.
> I really don't understand how there can be a X window
> manager standard which say that a window can be an
> "utility window", without describing what kind of
> behaviour such a window should have.
I take it you are (or recently were) a Windows or Mac user?
The reason such behavior would not be specified in a standard, is because not everyone wants the same behavior. The ability to use a different window manager and/or configure your window manager differently in order to get different behavior is part of the deliberately designed-in flexibility of the X windowing system, which allows different users to have things set up differently to suit the way they like to do things.
And yes, how various kinds of windows are handled is configurable on many window managers. Sawfish, for instance, gives you options to control, among other things, whether minimizing or unminimizing a window should minimize or unminimize its transients, whether selected windows are automatically unshaded, whether maximizing is a toggle or simply an automatic resize operation that leaves the window still resizeable, whether transient windows are stacked above their parents and/or automatically raised with them, whether window positions and dimensions are shown when moving and resizing and whether dialog windows are centered on their parents or placed independently (and how). Some window managers are much *more* configurable than sawfish, and even ones that are less configurable overall often have different options and features so that they appeal to certain people who like to do things that way.
If you haven't experimented with at least half a dozen different window managers, you can't fully appreciate the flexibility you have with X, to set things up to work the way you want them. I suggest trying out, at minimum, the following: sawfish, Xfce, Enlightenment, fluxbox, WindowMaker, wmii, and twm. You will probably find that most of them don't really do things quite the way you like. But seeing the wide range of options opens your eyes to the fact that you *aren't* stuck with the One And Only Way of doing things. You've got choices. Meaningful choices that have a real impact on how you use your computer.
Or you can get a Mac, and your choices basically boil down to "Do I want the window borders to be blue, silver, or green?" MS Windows is only a little more flexible.
> 1) Shortcuts to folders were treated as files
.lnk file support in cross-platform and open-source software.
.lnk files more-or-less correctly. (In fairness to the people doing the porting, they don't *have* to parse link files on *nix systems, because symlink semantics are built into the filesystem and/or the API. Still, the .lnk format was introduced in 1996, so you'd think someone would figure out how to support it by now.)
Ah, that. I didn't think of it, because it's not specific to open/save dialogs. It's a more general problem with
To my knowledge, Emacs is the only piece of open-source code that knows how to parse Windows
> 2) (On OS X)
I haven't used OS X enough to comment on that.
The last time I had access to an OS X system, getting GTK software to run on it could ONLY be done via the X server, and getting rootless mode to work was a new and esoteric trick most people hadn't tried out yet. (Yeah, it's been a while. The Mac in question was one of those ridiculous-looking malformed-lamp iMacs, which had just come out recently. Adobe software at the time all still required Classic, including Acrobat Reader.)
So I can really only comment on the PC side of things.
> 3) no voice control, no tablet stylus support, no
> touchscreen support (that's universal among all GTK+
Ah. I've never had the hardware for any of that stuff, so naturally I was blissfully unaware of any problems with it.
> 4) No open to move, delete, create shortcut, map
> network drive or any of those other options normal
> Open/Save dialogs have
Not can sentence fully this understand I.
Are you saying you use open and save dialog boxes as a file manager? So, like, if you want to copy some files, you pull up a word processor or something, hit Save As, find the location where the files are that you want to copy, and then, umm, somehow use the dialog box to copy them?
I was not aware that was even possible. Sounds cumbersome.
You do know about Windows Explorer (alias My Computer), right?
Call me weird, but personally I use file open and save dialog boxes for opening and saving.
> 5) It's been awhile, but I seem to recall it was incapable
> of accessing FTP servers mapped as drives in Windows,
If it's mapped as a drive with a letter (the closest analog, on Windows, to mounting it), the software shouldn't have to do anything special to access it. It should work just like accessing any other part of the filesystem. The app shouldn't even have to *know* that it's actually an ftp server (or nfs, or cifs, or whatever). If it's mounted (i.e., has a drive letter), it should Just Work. If it doesn't, that's a bug in the operating system, and a fairly major one.
> or any network drive that didn't have a drive letter (could
> be wrong on this one)
I'll have to try that out at some point and see.
> 6) And of course BUTT FUGLY
This comment is too vague to be meaningful.
If you think reinstalling from an XP SP1 disk is a pain, try using a pre-SP1 XP disk. The version of IE it comes with isn't good enough for Windows Updates, so you have to download and install an IE6 update first (which, naturally, requires a reboot), before you can even get started.
> ...which shows what is wrong with Microsoft's kernel
It's not the kernel. It's the filesystem.
Most filesystems, and in particular all the ones that are popular in the Unix world, have an abstraction/redirection layer sitting between a file's directory entry and the actual file contents. Unix people call them "inodes". The details vary somewhat depending on exactly what filesystem you're using, but in general the directory entry points to the inode, and the inode points to the actual file contents wherever they're stored. Because of this, a file can be changed or even replaced in situ, even while another process has the file open and is using it. The inode for the old file remains until the process that was using it lets go, but the directory entry is updated to point to the new inode.
FAT and NTFS don't have inodes, so it's not safe to alter a file while another process has it open. So you have to stop every process that's using the file, before you can do that. The easiest way to do that (and in some cases the only way, e.g., if the file is a shared library that lots of programs use) is to reboot.
> > updates that require a reboot so they can install
> > another update that then requires another reboot.
> Ah, see now you're confusing Microsoft with Adobe.
No, Microsoft does it too. The only way you don't notice is if you keep up-to-date on your updates all the time and never ever have to do a reinstall. Good luck with that second part, seeing as how this is Windows we're talking about.
Just for grins sometime, reinstall Windows and *count* how many reboots are required before you are fully up to date. Don't stop counting Windows Update, run right after a reboot, tells you there are no more updates to install.
The worst I've ever seen was a Windows XP system that didn't have any of the service packs included on the install CD. First, just to get to a desktop, you have to reboot at least twice (not counting initially booting to the install CD to get started). Then Windows Updates won't do squat until you install an IE6 update, because the version that comes with isn't good enough for Windows Updates. The IE6 update requires a reboot, naturally. Then you can start getting Windows Updates...
Oh, and then you've got allergies, in which case your immune system isn't directly attacking you per se, but it's overactively attacking something it doesn't really need to, and you're sort of getting caught in the crossfire.
> Actally having too strong an immune system IS bad; that's
... all kinds of fun stuff.
> what arthritis is, your body's immune system attacking you.
Arthritis, lupus, Crohns, Grave's, Psoriasis,
> Too bad it will be "a few decades", I'll be dead by then.
Heh. When speaking of the future development of a certain technology (medical or otherwise), "a few decades" typically either means "less than twenty years" or "never". In this case, let's just say I won't be betting any money on the former timeframe.
> That we can't actually see a majority of diseases under a microscope
No, that's wrong. Well, it could be technically right (depending on how you define "disease"), but not in the way you were probably thinking.
If the human immune system is making antibodies to fight the disease, then there's a pathogen involved, which we most certainly can see under a microscope. (These days, microscopes can image amazingly small objects, much smaller than any protein. Not in color, of course, because we're not talking about light-refracting microscopes here. But that's really beside the point. Anyway, I'm not sure what "color" would even mean at the molecular level.)
The larger issue is that we understand the human immune system about as well as we understand the human brain. Talking about re-engineering it and making "Immune System 2.0" is pretty arrogant and unrealistic, IMO. Giving the immune system a sample of what it's supposed to fight and letting it learn to fight that (vaccination, in other words) is one thing. That's like letting the search dog sniff a scrap of the missing person's clothing so they know who they're supposed to find. Doing the actual searching is still up to the dog. Trying to improve on the actual disease-fighting process would be something else quite again. That's like trying to design a better dog. Human knowledge of how living systems work is just plain not up to that yet.
I suspect it's more likely a sign that they know full well nobody *needs* their "FAST" search product on a Unix system, because it's redundant with standard basic system utilities that every Unix geek has taken for granted since the seventies. They would never say it that way, of course, because that way makes Windows sound inferior.
> But as you found out, this is not what really
> happens, so either its a bug in you window manager
It could be a window manager bug, or it could also be a deliberate design choice. But either way the difference you experienced is the window manager's doing. The setting in question does work as described with most window managers.
> Considering how wrong GTK gets Open/Save dialogs *alone* in Windows...
I hadn't noticed.
The main programs I've seen with GTK widgets on Windows are Gimp, Inkscape, Emacs, and Freeciv. Maybe they don't suffer from the problem? Emacs obviously wouldn't because it uses the minibuffer in lieu of Open/Save dialogs. Gimp does some of its own custom stuff in the save dialog (notably, instead of automatically filling in the extension when you choose a filetype, it automatically fills in the filetype based on the extension), so maybe it doesn't use the regular GTK save dialog either? Oh, and civserver takes savegame filenames on the command line and generates save files automatically based on server options. Not sure why I haven't noticed anything wrong in Inkscape.
Out of curiousity, what, exactly, does GTK get wrong about file open/save dialogs on Windows?
> Linux might be a different story.
Linux is what I use on my own workstation, both at home and at work, but I'm significantly familiar with Windows systems from working briefly with other people's computers, when I'm away from my desk for whatever reason, or when somebody needs help with something. I guess I probably use Windows about ten or twelve hours a week, all told.
I did copy and paste, of course. (Well, technically I selected the text, then pasted the current selection with a middle-click, without involving the copy clipboard; but that distinction is probably lost on kids these days, I suppose.)
I substituted the abbreviation because spelling it out roughly doubled the length of what I was quoting. Using the shorter form made it more likely to stay on one line, among other benefits.
As for quote tags, I don't use them because I don't like them. As far as I'm concerned, actual quote marks along the left edge are much better, a more readily apparent visual indication of quoting. So that's what I use. Slashdot even has a post setting (Extrans) that makes this easy, and a preference you can turn on that makes this the default for all your posts. Slashdot has such facilities because they have deliberately chosen to put up with, and even deliberately accommodate, geezers who remember usenet fondly. If you don't like this, you can always go find a site that doesn't offer such options. MyBook or FaceSpace or something along those lines, perhaps. Personally, I'd rather be here, hanging out with the old geezer computer nerds who actually remember usenet, but YMMV.
> a lot of people (ie older parents/grandparents)
> buy a Mac because it's "easier" and are more
> inclined to be on a 5-10 year cycle.
In the Windows world, a lot of people are on an 8-12 year upgrade cycle.
I wouldn't be terribly surprised if more people are still using Windows 98 than the total number of Mac users, all versions combined (actual general-purpose computers, I mean, not handheld music players and such). Obviously there's no reliable way to get actual numbers for how many Windows 98 systems are still in use, but I bet it's a lot higher than most computer geeks realize. People's tendency to upgrade promptly increases (more or less) geometrically with their level of computer knowledge, and the people whom an IT professional knows personally tend to be significantly above average.
> I maintain a 15k line Javascript application
> that's compatible with IE 5.5.
Ouch.
I'm pretty sure I'd rather support a punchcard-driven mainframe than IE 5.5.
If your IE-5.5-supporting Javascript code is by some miracle of herculean effort *also* compatible with modern browsers in standards-based rendering modes, you ought to win some kind of prize. (Do they give out prizes for masochism? What do they do, pin a medal on your bare chest?)
> GTK+ looks like an total alien on every OS except Linux.
IMO, GTK is significantly more consistent with "native" (Win32) apps on Windows than it is with non-GTK apps on Linux (e.g., KDE stuff, or the occasional Tk or Xaw application).
Personally, I'm more interested in functionality and flexible customizeability than absolute visual consistency, so having GTK and Qt apps (say) on the same system doesn't really bother me, as long as they're all good solid well-designed and properly configurable applications that meet my needs.
(Xaw does bother me, but that's because it's terrible.)
> Newer Firefox will probably still work on Linux on G4,
That isn't necessarily going to make Mac users happy. Typically, they didn't buy the G4 to run Linux. If they wanted to run Linux, they could have gotten a better-performing Celeron system for less money.
> or in X11 on OSX 10.4.
If so, they should be specifically saying so. "To run our software on your OS, you will need to install an optional library, which your OS vendor offers free of charge" is not entirely the same thing as "our software will no longer run on your OS".
Obviously, Verizon is trying to provide better service for their customers.
> Now, if this had been posted to a WoW forum... ;-)
Meh. Really the modern equivalent of the old slashdot effect these days is when the Google doodle returns your site as the first result. Hopefully your hosting provider doesn't bill by the megabyte...
> if my life is on the line, I'm trusting
> nothing less than a hot loaded 9mm JHP.
Why so small? I prefer to cart around an M109. For delivering a lot of kinetic energy to my would-be attacker's chest, what could be better than 155mm rocket-assisted rounds?
> Because Mozilla is capable of doing it
No, they're not.
Mozilla, as an organization, still hasn't figured out that whether the server I interact with today has the same private key as the a server I interacted with previously is more important than whether the people running the server paid money to Verisign (or whomever) for a cert signed by somebody on The List.
Mozilla, as an organization, still hasn't figured out that whether the DNS entry that resolves the FQDN to the server's IP address follows a continuously signed path up from the root is more important than whether the people running the server paid money to Verisign (or whomever) for a cert signed by somebody on The List.
Mozilla, as an organization, seems completely unaware of the inherent problems with HTTPS. They seem to think EV (which basically boils down to getting site operators to pay even more for their certs) is a meaningful solution. It is clear to me that they, as an organization, don't understand security at all.
Perhaps there are some people within the organization who do understand security. If so, they are not the people driving policy.
For what it's worth, I'm not aware of any other browser that does any better. But that doesn't mean it's all good. It's not all good. As things stand right now, people should think of https as inherently completely insecure. Because it is.
If you need to transfer information securely, don't use https.