Apple also do software updates for security fixes. So we can make the following correspondence:
Apple software update = Windows update
Apple point release (10.5.0 -> 10.5.1) = Windows service pack (XP -> XP SP1)
Apple major release (10.4.11 -> 10.5.0) = Windows major release (XP -> Vista)
So both companies do hot fixes that are downloaded and available gratis, do bundled fixes that are also available gratis, and major releases, which cost money.
Individual major releases from Microsoft cost more, but there have been fewer of them in recent years.
Apple has produced more point releases than Microsoft has service packs, but there are fewer new features in the individual point releases.
Apple uses a more consistent version-numbering scheme than Microsoft, which makes it easier to keep track of which version is which.:-)
Well as a counterexample, I just pulled my Sony Ericson K700 out of my pocket and it has a blank screen and no blinking lights. It looks as though it were switch off. If I press any of the keys, the display lights up and thus I can tell it is switched on.
I remember seeing phones that blinked to show they were still switched on but that was years ago. It was more important in the olden days when batteries would run out if you did not use your phone sparingly.
I think a solution to the iPhone problem would be for the setting that controls how often to check for email to be split in to two: one for the home network and one for when roaming. The latter would be set to 'manual' by default. Switching this on would be a good time for it to display a warning about data charges.
Trees both consume and produce CO2 for the first few years of life; it is only old trees that contain sequestered carbon. This is why planting new forests to replaced felled forests does not work.
Paper buried in landfill will break down eventually, releasing the CO2 one way or another.
We maintain an SQL script that creates the database or, when run in an existing database, upgrades any stored procedures that are out of date, alters tables, etc. This script (actually the smaller scripts it is assembled from) is checked in to Subversion like any other source code.
Our unit tests work at the C# level, not SQL (they test the objects implemented using the database, rather than the database itself). Most tests start by running the creation script to create a fresh database, do things to it, and then throw it away when the test is done. This way tests are isolated from each other and all the usual unit-testy requirements are met.
The down side is that tests that involve the database are a lot slower than unit tests that don't.
The installer (.msi file) uses the same script to install or upgrade the database on the production server. Testers work from their own copies of the database, completely isolated from whatever freakish mess we developers have perpetrated in our development databases.
Apple don't need to get people to switch to Mac OS X; they need to get them to buy Apple's computers.
Supporting Windows makes it easier for people to decide to try a Mac, because they don't have to worry about losing familiar applications like regedt32 and minesweeper. Apple hopes that they will then discover that they don't need Windows after all.
This makes much more sense talking about bicycle locks: if the thief is assumed to only want to steal one bike, you arrange that yours is not the easiest to steal. (The assumption is the theif is stealing your bike to ride it, not to sell it.)
Where there are a lot of places a thief might steal things from, then the criterion for a lock is that it cost more to break the lock than the cost of the valuables inside.
Most gadgets with pop-up input areas are horizontally rather than vertically oriented. Could Dasher be rotated 90% and work moving down the screen rather than accross? Or does that feel too much like a downhill slalom to be comfortable...?
Actually there has been some work on compatibility: Firefox can work with SVG linked to with embed tags (which is a good thing since object tags cannot be relied upon), and a certain amount of effort has been expended on other things that make it possible to embed SVG in a way that works with Adobe SVG Viewer and Firefox.
In some cases there are incompatibilities that their developers have argued over for some years now, but will not fix because they contradict web standards. For example, compressed SVG files (.svgz) do not work until you tweak your Apache settings to label them properly. I am willing to forgive this (assuming it works) because this is a case when cleaving closer to web standards fixes the problem.
The biggest disappointments for me are the total lack of declarative (SMIL) animation and the non-support of SVG fonts. But at least these are cleanly omitted features, not buggy half-baked implementations, which would be worse.
SVG works on Windows XP (using something called GDI+). I gather than on Linux, where it uses a different graphics library, it might not quite be there yet. So it might be that SVG support depends on your platform.
Slashdot charges 1c a page; at present this means users have to subscribe in advance. With a micropayments system that just worked, you could have a checkbox that hid adverts for 1c and all the hassle of aggregating the pennies would be done by the micropayments system: CmdrTaco and co would not have needed to write their own subscriptions system.
For a long time the theory has been that on-line comics would do the same: charge a tiny amount per page and have no ads. (The closest approach I know of is bitpass.com, where you typically charge 25c for a story (several comics pages).) A WWW with micropayments might be a lot nicer for individuals promoting their own content, but less profitable for advertisers.
I assume that the apple marketing people, then, are miserable about the iPod.
Let me put it another way. Given two products, product A which has few features, and product B which has all the features of product A and some more features, then the average marketing unit will find it more of a challenge writing copy about product A, and that said marketing unit may therefore prefer the more complex product from this limited point of view of saving them from having to think so hard. This does not mean that, given the budget to purchase a bigger-brained marketing unit, you cannot write copy that makes A look good and sells well and all the rest. Apple demands harder work from their marketing department than the average box-shifting computer company, and they reap the benefits.
I meant faster in general. Mostly you want to click on things; popping up context menus is the exception, not the rule.
I say 'should' because you have to design the user interface accordingly for this to work. If you can change something with drag and drop and direct manipulation, then that can be faster than popping up a menu or context menu. If you design the GUI around direct manipulation rather than reading menus, then making primary-button use more efficient at the expense of secondary-button use is a win overall.
But it would appear that Apple have decided that this is not the approach they want to take from now on, and that designing user interfaces around invisible context menus is more cool because everyone else does it, and therefore a multi-button mouse is going to be required instead.
If you're going to have more than one button, why not three? I remember SunWindows, where they actually had a system for what sort of thing each mouse button was used for -- following the lead of the old Smalltalk systems, where I get the impression buttons were colour-coded (I've never seen one). I also saw a system whose 'puck' had four buttons arranged in a diamond pattern. Crazy.
Compared to SunWindows, X10 and X11 had more of a free-for-all, with mouse buttons assigned functions almost at random. I imagine the Gnome and KDE guidelines have attempted to reign all this variation in a tad.
Marketing people always prefer the complex to the simple -- it gives them more features to boast about.
The advantages of the single-button mosuse for the experienced user should be that it is faster (since manipulating it needs less precise movements) and your index finger does not wear out so quickly because you can split the load between your first two fingers.
But you would hardly expect them to say that on the page describing their new multi-button wonder mouse. In order to sell the new model, you always have to rubbish the old model, even if last week you were describing it as the acme of human achievement...
Yes, the company that provided the SVG Viewer plug-in is now the company that wants SVG to fail...
On the other hand, it's not the fatal blow it might have been a couple of years ago; other companies are producing SVG renderers, and Mozilla Firefox 1.1 and a future Opera verison will have SVG support built-in. The intermingling of HTML and SVG code might allow for nifty effects that make Flash look old hat, at that.
'Microsoft, with Windows, has to support every reasonable configuration of x86 hardware there is[...]'
Except are there not great swathes of people whose old PCs are not compatible with Windows XP because of some BIOS limitation or other? Windows *used* to run on any PC, but the last few versions have been less than 100% compatible.
> Ok. I'm not a beginner. Can I please have a non-beginner apple mouse?
Yes, you can; multi-button mouses and scroll wheels are supported by Mac OS X, so you can have any mouse you like, so long as it's USB.
The Mac Mini (which I think is where this discussion started) is sold without a mouse, so you will be buying one or reusing an old mouse in any event.
You could turn this question around and ask whether other operating systems should have an option to work with a one-button mouse for the sake of people who can't use a multi-button mouse for whatever reason. (People with a persistent tremor will find a conventional mouse difficult to control, for example.) I remember having all sorts of trouble using a PC 2-button mouse to control applications designed for Sun's 3-button devices. Nowadays we would view a program that REQUIRES a middle button as badly designed.
It's true there are some usability features old-timers miss from Mac OS 9 -- but don't forget how many complaints there were when GNOME took on board some of these selfsame features (such as 'spacial' file-browsing windows). The Finder may make some people cross but it works well enough for my purposes.
I tried to change the default shell but never quite worked out the trick of it; my mistake was to assume it would be a user setting via the control panel rather than via Netinfo:-)
To be honest tcsh is not bad enough that I ever made a serious effort to change it. I was but speaking in jest.:-)
I don't think your analgy really applies; stone-age humans stopped using stone axes when they had invented something better (actually some time after inventing alternatives, because the alternatives started out too expensive). If we had run out of stones and had not invented bronze, we would have had to retreat to still-more-primitive technologies, and large portions of the population would have starved.
As things appear now, we do not have the luxury of switching to something cleaner, cheaper and easier to ship around the planet than oil.
Just to be pedantic... while old USSR rockets looked like they had CCCP on the side, it was really U+0421 U+0421 U+0421 U+0420 (pronounced Es Es Es Er, apparently). (I tried using the Cyrillic characters, but Slashdot is stripping them out.) The Cyrillic letter Es is no more a C than an O is a 0.
So with a Russian keyboard producing the correct characters, you should be able to enter the aforementioned Cyrillic characters in a IDN. If you wanted to allow people to also write cccp.su (using the homoglyphs) that's OK too. But a mixture like U+0440 U+0440 U+0440 U+0061 should at least generate a warning message.
I don't claim to be a Unicode expert, but I don't think that normalization form KC will convert U+0430 to 'a', because the Latin small letter A is not its compatability decomposition.
On the other hand, it would prevent spoofing of Greek names using mu and capital omega, or capital A with ring above, or ff ligatures, since there are characters that have them as compatibility decompositions.
For some people, what you consider international characters are national characters.
The test browsers should apply is whether a word in the URL contains letters from more than one script (e.g., using a Cyrillic letter in a word otherwise in the latin script is suspicious, whereas a word consisting entirely of Cyrillic is OK).
Might be problems if (say) Greek or Japanese companies insist on having trade names mixing native and latin leters.
In my opinion, the IRI specification should have forbidden mixing letters from different scripts: the phishing IRIs would then be forbidden and browsers would be entitled to refuse to follow those links.
Slashdot Mirror
My sister uses a wheelchair and she drives. Her car is modified to use hand controls instead of pedals.
Apple also do software updates for security fixes. So we can make the following correspondence:
:-)
Apple software update = Windows update
Apple point release (10.5.0 -> 10.5.1) = Windows service pack (XP -> XP SP1)
Apple major release (10.4.11 -> 10.5.0) = Windows major release (XP -> Vista)
So both companies do hot fixes that are downloaded and available gratis, do bundled fixes that are also available gratis, and major releases, which cost money.
Individual major releases from Microsoft cost more, but there have been fewer of them in recent years.
Apple has produced more point releases than Microsoft has service packs, but there are fewer new features in the individual point releases.
Apple uses a more consistent version-numbering scheme than Microsoft, which makes it easier to keep track of which version is which.
Well as a counterexample, I just pulled my Sony Ericson K700 out of my pocket and it has a blank screen and no blinking lights. It looks as though it were switch off. If I press any of the keys, the display lights up and thus I can tell it is switched on.
I remember seeing phones that blinked to show they were still switched on but that was years ago. It was more important in the olden days when batteries would run out if you did not use your phone sparingly.
I think a solution to the iPhone problem would be for the setting that controls how often to check for email to be split in to two: one for the home network and one for when roaming. The latter would be set to 'manual' by default. Switching this on would be a good time for it to display a warning about data charges.
Trees both consume and produce CO2 for the first few years of life; it is only old trees that contain sequestered carbon. This is why planting new forests to replaced felled forests does not work.
Paper buried in landfill will break down eventually, releasing the CO2 one way or another.
We maintain an SQL script that creates the database or, when run in an existing database, upgrades any stored procedures that are out of date, alters tables, etc. This script (actually the smaller scripts it is assembled from) is checked in to Subversion like any other source code.
Our unit tests work at the C# level, not SQL (they test the objects implemented using the database, rather than the database itself). Most tests start by running the creation script to create a fresh database, do things to it, and then throw it away when the test is done. This way tests are isolated from each other and all the usual unit-testy requirements are met.
The down side is that tests that involve the database are a lot slower than unit tests that don't.
The installer (.msi file) uses the same script to install or upgrade the database on the production server. Testers work from their own copies of the database, completely isolated from whatever freakish mess we developers have perpetrated in our development databases.
Apple don't need to get people to switch to Mac OS X; they need to get them to buy Apple's computers.
0 52.html for a discussion
Supporting Windows makes it easier for people to decide to try a Mac, because they don't have to worry about losing familiar applications like regedt32 and minesweeper. Apple hopes that they will then discover that they don't need Windows after all.
See http://www.joelonsoftware.com/articles/fog0000000
This makes much more sense talking about bicycle locks: if the thief is assumed to only want to steal one bike, you arrange that yours is not the easiest to steal. (The assumption is the theif is stealing your bike to ride it, not to sell it.)
Where there are a lot of places a thief might steal things from, then the criterion for a lock is that it cost more to break the lock than the cost of the valuables inside.
Most gadgets with pop-up input areas are horizontally rather than vertically oriented. Could Dasher be rotated 90% and work moving down the screen rather than accross? Or does that feel too much like a downhill slalom to be comfortable...?
Actually there has been some work on compatibility: Firefox can work with SVG linked to with embed tags (which is a good thing since object tags cannot be relied upon), and a certain amount of effort has been expended on other things that make it possible to embed SVG in a way that works with Adobe SVG Viewer and Firefox.
In some cases there are incompatibilities that their developers have argued over for some years now, but will not fix because they contradict web standards. For example, compressed SVG files (.svgz) do not work until you tweak your Apache settings to label them properly. I am willing to forgive this (assuming it works) because this is a case when cleaving closer to web standards fixes the problem.
The biggest disappointments for me are the total lack of declarative (SMIL) animation and the non-support of SVG fonts. But at least these are cleanly omitted features, not buggy half-baked implementations, which would be worse.
SVG works on Windows XP (using something called GDI+). I gather than on Linux, where it uses a different graphics library, it might not quite be there yet. So it might be that SVG support depends on your platform.
Slashdot charges 1c a page; at present this means users have to subscribe in advance. With a micropayments system that just worked, you could have a checkbox that hid adverts for 1c and all the hassle of aggregating the pennies would be done by the micropayments system: CmdrTaco and co would not have needed to write their own subscriptions system.
For a long time the theory has been that on-line comics would do the same: charge a tiny amount per page and have no ads. (The closest approach I know of is bitpass.com, where you typically charge 25c for a story (several comics pages).) A WWW with micropayments might be a lot nicer for individuals promoting their own content, but less profitable for advertisers.
Let me put it another way. Given two products, product A which has few features, and product B which has all the features of product A and some more features, then the average marketing unit will find it more of a challenge writing copy about product A, and that said marketing unit may therefore prefer the more complex product from this limited point of view of saving them from having to think so hard. This does not mean that, given the budget to purchase a bigger-brained marketing unit, you cannot write copy that makes A look good and sells well and all the rest. Apple demands harder work from their marketing department than the average box-shifting computer company, and they reap the benefits.
I meant faster in general. Mostly you want to click on things; popping up context menus is the exception, not the rule.
I say 'should' because you have to design the user interface accordingly for this to work. If you can change something with drag and drop and direct manipulation, then that can be faster than popping up a menu or context menu. If you design the GUI around direct manipulation rather than reading menus, then making primary-button use more efficient at the expense of secondary-button use is a win overall.
But it would appear that Apple have decided that this is not the approach they want to take from now on, and that designing user interfaces around invisible context menus is more cool because everyone else does it, and therefore a multi-button mouse is going to be required instead.
If you're going to have more than one button, why not three? I remember SunWindows, where they actually had a system for what sort of thing each mouse button was used for -- following the lead of the old Smalltalk systems, where I get the impression buttons were colour-coded (I've never seen one). I also saw a system whose 'puck' had four buttons arranged in a diamond pattern. Crazy.
Compared to SunWindows, X10 and X11 had more of a free-for-all, with mouse buttons assigned functions almost at random. I imagine the Gnome and KDE guidelines have attempted to reign all this variation in a tad.
Marketing people always prefer the complex to the simple -- it gives them more features to boast about.
The advantages of the single-button mosuse for the experienced user should be that it is faster (since manipulating it needs less precise movements) and your index finger does not wear out so quickly because you can split the load between your first two fingers.
But you would hardly expect them to say that on the page describing their new multi-button wonder mouse. In order to sell the new model, you always have to rubbish the old model, even if last week you were describing it as the acme of human achievement...
Yes, the company that provided the SVG Viewer plug-in is now the company that wants SVG to fail...
On the other hand, it's not the fatal blow it might have been a couple of years ago; other companies are producing SVG renderers, and Mozilla Firefox 1.1 and a future Opera verison will have SVG support built-in. The intermingling of HTML and SVG code might allow for nifty effects that make Flash look old hat, at that.
'Microsoft, with Windows, has to support every reasonable configuration of x86 hardware there is[...]'
Except are there not great swathes of people whose old PCs are not compatible with Windows XP because of some BIOS limitation or other? Windows *used* to run on any PC, but the last few versions have been less than 100% compatible.
> Ok. I'm not a beginner. Can I please have a non-beginner apple mouse?
Yes, you can; multi-button mouses and scroll wheels are supported by Mac OS X, so you can have any mouse you like, so long as it's USB.
The Mac Mini (which I think is where this discussion started) is sold without a mouse, so you will be buying one or reusing an old mouse in any event.
You could turn this question around and ask whether other operating systems should have an option to work with a one-button mouse for the sake of people who can't use a multi-button mouse for whatever reason. (People with a persistent tremor will find a conventional mouse difficult to control, for example.) I remember having all sorts of trouble using a PC 2-button mouse to control applications designed for Sun's 3-button devices. Nowadays we would view a program that REQUIRES a middle button as badly designed.
It's true there are some usability features old-timers miss from Mac OS 9 -- but don't forget how many complaints there were when GNOME took on board some of these selfsame features (such as 'spacial' file-browsing windows). The Finder may make some people cross but it works well enough for my purposes.
I tried to change the default shell but never quite worked out the trick of it; my mistake was to assume it would be a user setting via the control panel rather than via Netinfo :-)
:-)
To be honest tcsh is not bad enough that I ever made a serious effort to change it. I was but speaking in jest.
Finally a reason to upgrade to Panther!
I wish they'd mentioned this in the advertising.
I don't think your analgy really applies; stone-age humans stopped using stone axes when they had invented something better (actually some time after inventing alternatives, because the alternatives started out too expensive). If we had run out of stones and had not invented bronze, we would have had to retreat to still-more-primitive technologies, and large portions of the population would have starved.
As things appear now, we do not have the luxury of switching to something cleaner, cheaper and easier to ship around the planet than oil.
That's a good point. Come to think of it, I wonder what mischief one could make with the use of half-width and full-width ASCII characters...
Just to be pedantic... while old USSR rockets looked like they had CCCP on the side, it was really U+0421 U+0421 U+0421 U+0420 (pronounced Es Es Es Er, apparently). (I tried using the Cyrillic characters, but Slashdot is stripping them out.) The Cyrillic letter Es is no more a C than an O is a 0.
So with a Russian keyboard producing the correct characters, you should be able to enter the aforementioned Cyrillic characters in a IDN. If you wanted to allow people to also write cccp.su (using the homoglyphs) that's OK too. But a mixture like U+0440 U+0440 U+0440 U+0061 should at least generate a warning message.
I don't claim to be a Unicode expert, but I don't think that normalization form KC will convert U+0430 to 'a', because the Latin small letter A is not its compatability decomposition.
On the other hand, it would prevent spoofing of Greek names using mu and capital omega, or capital A with ring above, or ff ligatures, since there are characters that have them as compatibility decompositions.
For some people, what you consider international characters are national characters.
The test browsers should apply is whether a word in the URL contains letters from more than one script (e.g., using a Cyrillic letter in a word otherwise in the latin script is suspicious, whereas a word consisting entirely of Cyrillic is OK).
Might be problems if (say) Greek or Japanese companies insist on having trade names mixing native and latin leters.
In my opinion, the IRI specification should have forbidden mixing letters from different scripts: the phishing IRIs would then be forbidden and browsers would be entitled to refuse to follow those links.