don't avoid new things just because you think the users are too stupid to cope.
The highlight/MMB method is convenient, but it is also confusing to have two separate copy/paste mechanisms (MMB and Ctrl-C). On my Solaris box, Ctrl-C would only work in half the applications. OTOH, the MMB method would always work, but it would copy only ASCII, not styled text (or, heaven forbid, images). And each method had its own clipboard.
I don't agree entirely with your "progress dies" comment. Every major release of the Mac OS I've seen (i.e. all of them since 6.0.7) had new additions to the UI. There was real progress there; some inspired by the user community, with Apple sometimes incorporating additions that had been separate programs, while other new additions were from Apple's own UI research.
Now, you're right in that this progress was just a selection from what was possible. But to say that there's no progress at all is going too far.
Are there UI designers and other human factors specialists that contribute to Linux?
The problem is that the unified interface approach is exclusive.
The problem with choice in interfaces is that consistency goes out the window. Now, I'm not familiar with current Linux UIs, but when I last used a Solaris system, about half the apps used Motif and the others OpenLook. Result: I had a mix of applications with wildly different UIs, each with its own set of conventions, increasing the time I spent finding the functions I needed, rather than being productive.
The current proliferation of UIs will make this even worse, I expect. We'll be back in the bad old days of MS-DOS, when every application had its own UI.
And it's not just the UI that suffers. Simple interapplication interaction like copy and paste was horribly limited.
I can see the utility of making a UI customizable (although that, too, has its drawbacks, e.g. for system administrators and help desk people). But having entirely separate UIs isn't such a good thing IMO.
I didn't. The article said "Linux on the desktop is toast", and various Linuxites responded with "that's bullshit, if you just RTFM Linux is just as easy to use as Windows".
I've argued that that's incorrect, and I've shown why.
Several posters have argued that to succeed on the desktop, Linux needs a better UI. I agree with them. You haven't provided any arguments against that.
Linux and the various BSDs may have been around long before..., but they were always accessible/usable only to a small subset of the computer-using population. It's good to see steps being taken in the right direction.
So you'll also assume that grandma will do Windows-configuration by hacking the registry?
No. I was just reacting to a post that seemed to imply editing configuration files was easier than clicking in a control panel (the "I know my way around/etc " comment).
If SuSE has a control panel, well, more power to them. That's the way Linux should be evolving.
You're only a newbie once, or on a slow day, maybe twice. Thereafter you're a serious user,
Right. How often do you change, say, network settings? On my system, I see the network settings maybe once a year. Only someone with total recall would be able to change a text file without any syntax errors a year after he's consulted teh manual.
Actually, no, the answer is to learn to *read*.
Actually, no, the answer (for the majority of users) is to provide a system that allows them to change settings with minimal knowledge of the internals. There is no way you can convince me that checking a check box marked "use DHCP" is more difficult than entering a line that says "DHCP=1" (or whatever) in/etc/asdjhf.cfg.
Let's compare the knowledge required for both approaches.
Control panel
To change network settings, let's look in the list of control panels. I see there is a control panel called "network", maybe that does what I need.
My network administrator told me to use DHCP, so let's see if there is something named "DHCP" in this panel.
There's a rectangular box next to it, apparently it needs to be checked for the option to switch ON.
Text file
To change the network sttings, I need to open the file/etc/asdjhf.cfg in a text editor.
To add DHCP (or was it DCHP?-Hello, syntax errors) I need to add a line that says "DHCP=1".
Oh, and that line must not precede the network card ID (or whatever), otherwise, it's my first NIC and not my second that gets DHCP'd.
Oops, no spaces between P and =
It's obvious that editing a text file requires much more knowledge than using a control panel.
This boils down to educating the user-base, and in particular, those who don't want to be educated... and it makes me sick.
If you can take two roads to your destination, one takes an hour and the other takes three days, and the end result is the same, which do you take?
Users don't want to be educated because there are alternatives that don't require education. Those alternatives are cheaper. Many users don't want to know about their computer's internals, they just want to use the damn things. Whining about that won't make Linux magically appear on corporate desktops. Improving the UI to make it more accessible to non-geeks will.
Fortunately, some Linux distributions seem to take steps in this area, providing GUI front ends to configuration.
The "control panel" system has one huge advantage over/etc. (note: I'm assuming configuration is done by editing text files, without a GUI front end to the settings)
In a control panel, every available option is visible. When I want to add a network connection, I get a list of available protocols. When I enter a parameter, my input is checked for validity.
Contrast this with text files. Yes, you do have access to all the options, and yes, for an experienced user it can be faster than having to go through a wizard. But one typo can result in a nonfunctional system. The only way to make sure your input is correct is by manually checking everything, consulting with the manpage on every step.
With the "control panel" system, it's a lot easier to get comfortable with changing the configuration. I don't want to have to remember which options are valid in/etc/whatever. In a control panel system, I only need to remember the general location of an option, as opposed to its exact syntax.
Someone who's used neither before but has a working knowledge of computer technology in general, can be up and running much faster with a control panel system than with text files.
Now, this is not to say the control panel system is perfect. Wizards can be a nuisance for experienced users, some options are hidden in non-intuitive places, etc.
But if the Linux developer community has no interest in making life easier for the mythical `luser', then their presence on the desktop deserves to remain marginal.
The problem is most poeople who complain about smap in fact Opt-in to this spam when the sign up for wonderful free services like Yahoo mail and the what not.
I doubt that. Services like Yahoo will use your mail address themselves (by default), but I haven't seen many that will actually sell you address on.
Most of the spam I receive is because I used to post to Usenet groups with my real e-mail address. I suspect my address made it from there into one or more of those "80 million mail addresses for $10" scams. And why spend serious money buying addresses from Yahoo if you can have them at $0,000000125 a pop from J. Random Asshole?
Dutch news site Nu.nl has an articlestating that Xenu's ISP is under fire now. Apparently, xtdnet's upstream provider, Priority Telecom has disconnected xtdnet over Xenu. Xtdnet has an article with more detail.
Um, resource and data forks have nothing to do with the use of type/creator information. OK, the metadata may be stored in the resource fork (I'm not sure about this), but type/creator information can be stored in a non-forked structure as well.
But you're right about the type/creator system being vastly superior to file name extensions.
No, they're not. Icons (with VERY few exceptions) are only obvious after they've been explained. If the icons are good, and there aren't too many of them, you need to explain them only once. Try to replace everything with icons (which you're going to have to do if you want no localization problems), you end up with way too many icons for anyone to remember. Do you know (without looking at the tooltip) what every single icon in say, Word is for? Didn't think so.
Also, icons don't completely solve the localization problem. Images (especially the abstract drawings used in icons) can have different meanings in different cultures.
My IE starts instantaneously, whereas older Netscape loaded and loaded some modules or other,
Well, Microsoft is cheating here by loading most of IE at startup to use it as the file browser (you may say it's a valid technique, but this particular technique can be used only by Microsoft on Windows, so it does muddy the comparison).
Is this (e.g. Office) on the same computer you used those older versions on? That makes the comparison even more difficult. New versions of applications often seem (to me) faster only because the new version is used on massively faster hardware than the older version.
Apple Relizes that this is a share of the market place that it is missing - and I believe Apple is making it there - slowly.
I think you're right. Rumors on "Thing 2" (a rackmountable server) have been floating around for a while now.
If I were Apple, I'd combine a 1U rack unit with a variation of Microsoft's Mira: a wireless display that can connect to any Mac in the vicinity. Add USB to the display (for keyboard and mouse) and you've got the coolest KVM switch ever.
If you're complaining that new features are useless to you, and slow you down, don't install it.
No, I was saying that the new features do not justify a ~10x speed decrease. Now, this is just one example, and an extreme one at that (the 'new features' included radical changes like multithreading), but...
I'd also like to point out that just because program X (in this case, Eudora) became slower, that does not mean that is true for all programs.
Of course. But *in my experience*, very few programs get any faster with a new release. In fact, the Mac OS X updates (from 10.0.0 to 10.1.3) are the only counterexamples I can come up with right now.
Just because Moore's Law has been more or less correct for the entire history of PCs up to now, doesn't mean it should apply to software as well. With all the improvement in hardware speed we've seen over the years, have things like starting a program (or even starting the computer itself) become any faster? No. Sure, we can now do things that weren't feasible ten years ago (DVD playback). But in the meantime, processes that should take no time at all by now are still no faster than back then.
Nice theory. In practice, when I upgrade an application, the new version almost always turns out to be slower than the old one: example. So much for modularity.
I'm having a hard time figuring out what his point is. He's saying "we need a policy for archiving e-mail" and then he talks about Enron, where any policy regarding e-mail would have resulted in evidence being destroyed. Is he saying we need to start pre-emptively destroying email in case there's something incriminating in it?
"Digging up the dirt" isn't a new problem. Back when everything was done on paper, you could make copies and stash them somewhere, so shredding the original was never enough to ensure the document didn't exist anymore.
And as for saying "e-mail will play a role in many other unfolding corporate stories", well, duh!
While he does make some good points, I do have the impression he's being too optimistic.
But, of course, that money isn't lost to the economy. What happens to the money that companies save by using GPL software?
This (and the argument that follows it) seems to imply that most, if not all, of the money saved on software ends up paying for GPl sofware. But it doesn't. Even if 50% ends up financing GPL in some indirect way, that means your 'software budget' has been cut in half.
How, then, will it fund R&D when software sales aren't its primary money-maker? They will treat software as enabling technology and will continue to fund it because their service business won't work without technology.
Even if Microsoft manages to make as much money on services as they do now on selling software, they'll have to increase their workforce to provide the services as well as create the SW that runs it. Which means it'll be a lot less profitable. For Microsoft, this may not be a problem, but lots of smaller software houses will be up shit creek.
Somebody paid for all of that free software: Individual developers paid with their free time,...
For now, that will work because the amount of open-source SW is limited. But soon, we'll run out of programmers who are willing to donate free time. So I don't see the... larger development staff than any one company could support... happening in general. Only for the few Good Causes (in programmer community opinion) will an ample (free) workforce be available.
All in all, while GPL may be the wave of the future, I'm not sure it's the Promised Land. It might actually end up decreasing the amount of SW available.
Of course Internet Applications are going to be big with consumers. I don't see why they wouldn't. They don't require installation, they don't crash, they don't take space on the harddrive and they're easier to use.
Let's see:
They require an internet connection. Given the number of times I can't log on to my ISP, I don't want to be reliant on that.
They reduce the number of applications available. How many tools do you use now that were developed as shareware by a single person, who has no hope in hell of being able to provide this tool as an Internet application?
Where does this leave my data? Sensitive or not, I want complete control over whatever data I create, not find everything deleted because I didn't access my account in three months.
They reduce user choice. "This application is written in ActiveX and runs only on Windows with IE".
Many internet apps run inside a browser. For nontrivial apps, this means the UI sucks: no menu bar, for one. No windows. No palettes.
The UI also sucks because there'll be no Human Interface Guidelines to follow. Everyone will try and invent the wheel again. Welcome to the bad old days of MS-DOS!
That internet connection is SLOW. When I create a new document in a local application, it's there in the blink of an eye. With an internet app, go wait for the page to load. The entire UI gets squeezed down that narrow pipe every time you do something.
Integration between apps sucks. Here I am, writing a/. comment. I'm handcoding HTML, for goodness' sake!
IOW, Internet applications may become big, but I fear demand is more driven by IT departments (who but into the "no installation hassle" advantage) than by consumers.
I've been working for a company that created a complex application for storage and manipulation of images. They had a Windows version and a Web-based version. The Web-based version was less functional, looked like shit and was bloody annoying because of the download times.
Shell has used this technology for a while now (at least a year) with its "Easypay" system. I haven't seen any fraud stories in the media yet.
It uses a PIN number, so an RF intercept alone isn't going to be enough to duplicate a key.
Also, its range is only about 10cm.
Battery life is not an issue. The cards seem to be passive, and are powered by an EM field that's generated by te card reader.
How can someone possibly think that a 7XX MHz G3 running with slow SDRAM, and a slow IDE harddrive is soo much faster than a 14XX MHz Athlon running DDRAM, and a faster IDE harddrive or a faster scuzzy drive.
They can because for a lot of tasks, processor speed is not the limiting factor. It's the user. And on a Mac, it often takes the user less time to accomplish a task due to the superior UI.
Apple's "there is no step three" ad campaign was much more relevant to real-world computer usage than the MHz-flaunting going on elsewhere.
Slashdot Mirror
don't avoid new things just because you think the users are too stupid to cope.
The highlight/MMB method is convenient, but it is also confusing to have two separate copy/paste mechanisms (MMB and Ctrl-C). On my Solaris box, Ctrl-C would only work in half the applications. OTOH, the MMB method would always work, but it would copy only ASCII, not styled text (or, heaven forbid, images). And each method had its own clipboard.
I don't agree entirely with your "progress dies" comment. Every major release of the Mac OS I've seen (i.e. all of them since 6.0.7) had new additions to the UI. There was real progress there; some inspired by the user community, with Apple sometimes incorporating additions that had been separate programs, while other new additions were from Apple's own UI research.
Now, you're right in that this progress was just a selection from what was possible. But to say that there's no progress at all is going too far.
Are there UI designers and other human factors specialists that contribute to Linux?
The problem is that the unified interface approach is exclusive.
The problem with choice in interfaces is that consistency goes out the window. Now, I'm not familiar with current Linux UIs, but when I last used a Solaris system, about half the apps used Motif and the others OpenLook. Result: I had a mix of applications with wildly different UIs, each with its own set of conventions, increasing the time I spent finding the functions I needed, rather than being productive.
The current proliferation of UIs will make this even worse, I expect. We'll be back in the bad old days of MS-DOS, when every application had its own UI.
And it's not just the UI that suffers. Simple interapplication interaction like copy and paste was horribly limited.
I can see the utility of making a UI customizable (although that, too, has its drawbacks, e.g. for system administrators and help desk people). But having entirely separate UIs isn't such a good thing IMO.
Whoever said we actually need you people?
I didn't. The article said "Linux on the desktop is toast", and various Linuxites responded with "that's bullshit, if you just RTFM Linux is just as easy to use as Windows".
I've argued that that's incorrect, and I've shown why.
Several posters have argued that to succeed on the desktop, Linux needs a better UI. I agree with them. You haven't provided any arguments against that.
Linux and the various BSDs may have been around long before..., but they were always accessible/usable only to a small subset of the computer-using population. It's good to see steps being taken in the right direction.
So you'll also assume that grandma will do Windows-configuration by hacking the registry?
No. I was just reacting to a post that seemed to imply editing configuration files was easier than clicking in a control panel (the "I know my way around /etc " comment).
If SuSE has a control panel, well, more power to them. That's the way Linux should be evolving.
You're only a newbie once, or on a slow day, maybe twice. Thereafter you're a serious user,
Right. How often do you change, say, network settings? On my system, I see the network settings maybe once a year. Only someone with total recall would be able to change a text file without any syntax errors a year after he's consulted teh manual.
Actually, no, the answer is to learn to *read*.
Actually, no, the answer (for the majority of users) is to provide a system that allows them to change settings with minimal knowledge of the internals. There is no way you can convince me that checking a check box marked "use DHCP" is more difficult than entering a line that says "DHCP=1" (or whatever) in /etc/asdjhf.cfg.
Let's compare the knowledge required for both approaches.
Control panel
Text file
It's obvious that editing a text file requires much more knowledge than using a control panel.
This boils down to educating the user-base, and in particular, those who don't want to be educated ... and it makes me sick.
If you can take two roads to your destination, one takes an hour and the other takes three days, and the end result is the same, which do you take?
Users don't want to be educated because there are alternatives that don't require education. Those alternatives are cheaper. Many users don't want to know about their computer's internals, they just want to use the damn things. Whining about that won't make Linux magically appear on corporate desktops. Improving the UI to make it more accessible to non-geeks will.
Fortunately, some Linux distributions seem to take steps in this area, providing GUI front ends to configuration.
The "control panel" system has one huge advantage over /etc. (note: I'm assuming configuration is done by editing text files, without a GUI front end to the settings)
In a control panel, every available option is visible. When I want to add a network connection, I get a list of available protocols. When I enter a parameter, my input is checked for validity.
Contrast this with text files. Yes, you do have access to all the options, and yes, for an experienced user it can be faster than having to go through a wizard. But one typo can result in a nonfunctional system. The only way to make sure your input is correct is by manually checking everything, consulting with the manpage on every step.
With the "control panel" system, it's a lot easier to get comfortable with changing the configuration. I don't want to have to remember which options are valid in /etc/whatever. In a control panel system, I only need to remember the general location of an option, as opposed to its exact syntax.
Someone who's used neither before but has a working knowledge of computer technology in general, can be up and running much faster with a control panel system than with text files.
Now, this is not to say the control panel system is perfect. Wizards can be a nuisance for experienced users, some options are hidden in non-intuitive places, etc.
But if the Linux developer community has no interest in making life easier for the mythical `luser', then their presence on the desktop deserves to remain marginal.
Freedom of speech != freedom to advertise. It's a political, not a commercial freedom.
The problem is most poeople who complain about smap in fact Opt-in to this spam when the sign up for wonderful free services like Yahoo mail and the what not.
I doubt that. Services like Yahoo will use your mail address themselves (by default), but I haven't seen many that will actually sell you address on.
Most of the spam I receive is because I used to post to Usenet groups with my real e-mail address. I suspect my address made it from there into one or more of those "80 million mail addresses for $10" scams. And why spend serious money buying addresses from Yahoo if you can have them at $0,000000125 a pop from J. Random Asshole?
Dutch news site Nu.nl has an articlestating that Xenu's ISP is under fire now. Apparently, xtdnet's upstream provider, Priority Telecom has disconnected xtdnet over Xenu. Xtdnet has an article with more detail.
Xenu is still up, though.
Um, resource and data forks have nothing to do with the use of type/creator information. OK, the metadata may be stored in the resource fork (I'm not sure about this), but type/creator information can be stored in a non-forked structure as well.
But you're right about the type/creator system being vastly superior to file name extensions.
Use pictograms. They are much more intuitive.
No, they're not. Icons (with VERY few exceptions) are only obvious after they've been explained. If the icons are good, and there aren't too many of them, you need to explain them only once. Try to replace everything with icons (which you're going to have to do if you want no localization problems), you end up with way too many icons for anyone to remember. Do you know (without looking at the tooltip) what every single icon in say, Word is for? Didn't think so.
Also, icons don't completely solve the localization problem. Images (especially the abstract drawings used in icons) can have different meanings in different cultures.
My IE starts instantaneously, whereas older Netscape loaded and loaded some modules or other,
Well, Microsoft is cheating here by loading most of IE at startup to use it as the file browser (you may say it's a valid technique, but this particular technique can be used only by Microsoft on Windows, so it does muddy the comparison).
Is this (e.g. Office) on the same computer you used those older versions on? That makes the comparison even more difficult. New versions of applications often seem (to me) faster only because the new version is used on massively faster hardware than the older version.
Apple Relizes that this is a share of the market place that it is missing - and I believe Apple is making it there - slowly.
I think you're right. Rumors on "Thing 2" (a rackmountable server) have been floating around for a while now.
If I were Apple, I'd combine a 1U rack unit with a variation of Microsoft's Mira: a wireless display that can connect to any Mac in the vicinity. Add USB to the display (for keyboard and mouse) and you've got the coolest KVM switch ever.
Oh, and there's a petition underway, too.
If you're complaining that new features are useless to you, and slow you down, don't install it.
No, I was saying that the new features do not justify a ~10x speed decrease. Now, this is just one example, and an extreme one at that (the 'new features' included radical changes like multithreading), but...
I'd also like to point out that just because program X (in this case, Eudora) became slower, that does not mean that is true for all programs.
Of course. But *in my experience*, very few programs get any faster with a new release. In fact, the Mac OS X updates (from 10.0.0 to 10.1.3) are the only counterexamples I can come up with right now.
Just because Moore's Law has been more or less correct for the entire history of PCs up to now, doesn't mean it should apply to software as well. With all the improvement in hardware speed we've seen over the years, have things like starting a program (or even starting the computer itself) become any faster? No. Sure, we can now do things that weren't feasible ten years ago (DVD playback). But in the meantime, processes that should take no time at all by now are still no faster than back then.
Nice theory. In practice, when I upgrade an application, the new version almost always turns out to be slower than the old one: example. So much for modularity.
Now if either person sends a message with this same noise through
But the whole point of this method is that that's not going to happen. There's a limitless supply of noise, so no need to reuse it.
If you press a 'button' that's in the shadow of one of the other fingers, the keystroke would not be seen.
Rumors on "Thing 2" (a rackmountable server from Apple) have been floating around for some time now. And there's a petition underway.
And 2U is not enough (or too much...). With blade servers being the new wave, even a 1U unit may be seen as taking up too much space.
I'm having a hard time figuring out what his point is. He's saying "we need a policy for archiving e-mail" and then he talks about Enron, where any policy regarding e-mail would have resulted in evidence being destroyed. Is he saying we need to start pre-emptively destroying email in case there's something incriminating in it?
"Digging up the dirt" isn't a new problem. Back when everything was done on paper, you could make copies and stash them somewhere, so shredding the original was never enough to ensure the document didn't exist anymore.
And as for saying "e-mail will play a role in many other unfolding corporate stories", well, duh!
It looks like they use red-shift data
But using red-shift isn't creative. It's one of the standard ways of estimating astronomical distances.
Really creative, yes. Pity we don't get to hear about it. Or am I overlooking something?
While he does make some good points, I do have the impression he's being too optimistic.
But, of course, that money isn't lost to the economy. What happens to the money that companies save by using GPL software?
This (and the argument that follows it) seems to imply that most, if not all, of the money saved on software ends up paying for GPl sofware. But it doesn't. Even if 50% ends up financing GPL in some indirect way, that means your 'software budget' has been cut in half.
How, then, will it fund R&D when software sales aren't its primary money-maker? They will treat software as enabling technology and will continue to fund it because their service business won't work without technology.
Even if Microsoft manages to make as much money on services as they do now on selling software, they'll have to increase their workforce to provide the services as well as create the SW that runs it. Which means it'll be a lot less profitable. For Microsoft, this may not be a problem, but lots of smaller software houses will be up shit creek.
Somebody paid for all of that free software: Individual developers paid with their free time, ...
For now, that will work because the amount of open-source SW is limited. But soon, we'll run out of programmers who are willing to donate free time. So I don't see the ... larger development staff than any one company could support... happening in general. Only for the few Good Causes (in programmer community opinion) will an ample (free) workforce be available.
All in all, while GPL may be the wave of the future, I'm not sure it's the Promised Land. It might actually end up decreasing the amount of SW available.
Of course Internet Applications are going to be big with consumers. I don't see why they wouldn't. They don't require installation, they don't crash, they don't take space on the harddrive and they're easier to use.
Let's see:
IOW, Internet applications may become big, but I fear demand is more driven by IT departments (who but into the "no installation hassle" advantage) than by consumers.
I've been working for a company that created a complex application for storage and manipulation of images. They had a Windows version and a Web-based version. The Web-based version was less functional, looked like shit and was bloody annoying because of the download times.
Shell has used this technology for a while now (at least a year) with its "Easypay" system. I haven't seen any fraud stories in the media yet. It uses a PIN number, so an RF intercept alone isn't going to be enough to duplicate a key. Also, its range is only about 10cm.
Battery life is not an issue. The cards seem to be passive, and are powered by an EM field that's generated by te card reader.
How can someone possibly think that a 7XX MHz G3 running with slow SDRAM, and a slow IDE harddrive is soo much faster than a 14XX MHz Athlon running DDRAM, and a faster IDE harddrive or a faster scuzzy drive.
They can because for a lot of tasks, processor speed is not the limiting factor. It's the user. And on a Mac, it often takes the user less time to accomplish a task due to the superior UI.
Apple's "there is no step three" ad campaign was much more relevant to real-world computer usage than the MHz-flaunting going on elsewhere.