Most people don't want to make choices. They want a system that works easily and does the stuff they want. A Gentoo install requires knowledge of computers and the use of a command line. That's not good enough.
Gentoo isn't something I'd recommend to novice users. Ubuntu maybe, but it still has some kinks that need worked out (but it is a step in the right direction).
I assume you are clicking on the link from an Office application and not just Windows?
Tried Outlook and Word. I have to try this one at work since I use OO.o at home, though.
If you click on the standard Windows Update link that shows up in the 'start' menu (generally under Programs) on any newly installed Windows, you get the Windows update webpage displayed in whatever browser is chosen as your default. The same bad behavior re-appears after successfully updating to Microsoft Update (the link Windows/Microsoft places next to the Windows Update link for Microsoft Update takes you to your default browser). The same bad behavior happens in Office applications when you choose updates from the right-most menu. It sends the link to whatever your default browser is. This, if your default browser is chosen to be non-MS, passes the link to a non-MS app.
I have never encountered this — I've never known of any place to get to updates other than through "Help and Support", though (I still can't see one now). As I said, though, these things should probably be done through a non-browser system anyway, which is what automatic updates (enabled by default since SP2) does.
No. There is a definite delay between the time you click a link and the non-MS browser even begins to open.
Not on my system. Perhaps Microsoft have singled you out as an enemy;).
If Microsoft doesn't think it is a good idea to use non-IE browsers due to security (but somehow, as this current Slashdot article attests, they are okay with their DRM running on non-MS browsers now) and Microsoft knows the enduser can choose to have a non-IE browser be default (due to court order), why does Microsoft permit the bad behavior of letting these links be passed to non-MS apps?
You have completely misunderstood this, I think. It's not MS which won't allow the controls to run on Firefox, it's Firefox which won't allow the controls to run on its system. It uses ActiveX, I believe, which isn't supported by Firefox. Like I said, you can install a plugin to use ActiveX on Mozilla, but I really wouldn't recommend it.
For the record the only place that I can find to open Windows Update uses the Windows renderer (which doesn't appear to require IE to be installed) and doesn't pass the link to a non-MS app.
I'm led to believe that it's more the renderer than anything else. I could quite easily be misinformed, however. Some components of IE, at least, are cached, however.
I always assumed that IE opened faster from OE because the general browser framework was so integrated into the operating system, like being able to access your favorites from the start menu or Windows Explorer. Wasn't that the big controversy? That IE went from being a 6 meg add-on to Windows to being the heart and lungs (or at least the rib-cage) of the operating system while Netscape remained a 6 meg add-on? On the same note, however, the start button/windows explorer still only open IE favorites and don't (to my knowledge at least) have a way of switching to the Firefox Bookmarks
That's basically what I meant when I said it was memory-resident. The renderer is used for lots of things in Windows so parts of it are cached, I believe. Firefox is a huge app, on top of this, though, while IE is quite a small app (at least partly attributable to it being so bloody old, though).
There isn't a version of Windows Media Player for Linux. The plugin uses the components of the full player (as with most media plugins which aren't restricted to the browser). Basically, since MS haven't made a Windows Media port for Linux, embedding it in a browser is completely impossible. One wonders how hard it would be for them to make one, though (it can't be that far from the OSX version, other than — as you mention — the drawing to screen).
The slow link response seems to be because IE is generally memory-resident (and smaller) and takes less time to open than Firefox in general (certainly I notice no difference in times between opening a link from Outlook and opening from XChat or whatever), so I'm not convinced this is really an issue of their bias.
The updates use controls that wouldn't go on Firefox because they'd be deemed unsafe (I believe there's a plugin to make them work, I wouldn't recommend it though). The best way for Microsoft to get around this would be to stop relying on a browser at all for Windows Updates, which is basically what they've done by "forcing" Automatic Updates (which isn't browser-based) on in XP SP2.
I don't really see a problem, though. It seems reasonable to make an exception with open formats for those who need aid. We let seeing-eye dogs in where pets aren't allowed.
The point is less allowing them to use other formats than it is using a format that they cannot, I believe. It's not a good solution if, every time they want to read an in-house document, they have to use some random conversion tool, for example.
Agreed. The difference here is that Linux systems are almost always customised to some degree, and many optimised versions exist, whereas OSX systems are almost never customised, meaning tailoring them to this hardware is potentially more difficult.
That tells me that this project is doomed right now.
Or, much more likely, they want to alter it to run well on the extremely unique hardware that they were developing for the project. As far as I can tell, they wanted the designers to be able to tinker, rather than the recipients themselves.
Yeah, I didn't link that because it only has the download for the old version (since the new one is BSD licenced and there's some weirdness about putting those on the Drupal homepage).
Clearly and accurately documenting what the program is supposed to do is a huge first step.
Agreed, but I find that a lot of the time, particularly when working on projects where the requirements can change, unit testing provides a kind of documentation in itself. Even if the programmer has been too lazy to update the documentation for a method, the test itself must have been updated, so the functionality — provided that the tests have sufficient coverage — is described in the expectations of these tests.
If it helps with your numbers, my company uses it (along with NMock — whose site appears to be down, ironically because of a problem with Java — for isolation) for all.NET projects.
I love how they've used attributes in the implementation, makes things very easy. The Java (1.)5.0 implementation of JUnit uses this approach too, I believe.
Slashdot Mirror
Most people don't want to make choices. They want a system that works easily and does the stuff they want. A Gentoo install requires knowledge of computers and the use of a command line. That's not good enough.
Gentoo isn't something I'd recommend to novice users. Ubuntu maybe, but it still has some kinks that need worked out (but it is a step in the right direction).
Usability?
It's all fun and games until your "iPod sync" Perl script becomes self-aware and threatens life as we know it.
Someone should really create HQ9++, with some more library operations. DeCSS, web browser, stuff like that.
The principle is clearly that if you can do absolutely everything in 20 indecipherable characters, your code will never need to be maintained.
Diagonalisin' makes me feel good!
Computational complexity scares the living daylights out of everyone.
Well done for admitting you were wrong on Slashdot!
You appear to have restated the point of the post you were replying to?
Tried Outlook and Word. I have to try this one at work since I use OO.o at home, though.
I have never encountered this — I've never known of any place to get to updates other than through "Help and Support", though (I still can't see one now). As I said, though, these things should probably be done through a non-browser system anyway, which is what automatic updates (enabled by default since SP2) does.
Interesting. Doesn't really affect the time that it takes for the window to appear, though, but that is interesting.
Not on my system. Perhaps Microsoft have singled you out as an enemy ;).
You have completely misunderstood this, I think. It's not MS which won't allow the controls to run on Firefox, it's Firefox which won't allow the controls to run on its system. It uses ActiveX, I believe, which isn't supported by Firefox. Like I said, you can install a plugin to use ActiveX on Mozilla, but I really wouldn't recommend it.
For the record the only place that I can find to open Windows Update uses the Windows renderer (which doesn't appear to require IE to be installed) and doesn't pass the link to a non-MS app.
I'm led to believe that it's more the renderer than anything else. I could quite easily be misinformed, however. Some components of IE, at least, are cached, however.
That's basically what I meant when I said it was memory-resident. The renderer is used for lots of things in Windows so parts of it are cached, I believe. Firefox is a huge app, on top of this, though, while IE is quite a small app (at least partly attributable to it being so bloody old, though).
There isn't a version of Windows Media Player for Linux. The plugin uses the components of the full player (as with most media plugins which aren't restricted to the browser). Basically, since MS haven't made a Windows Media port for Linux, embedding it in a browser is completely impossible. One wonders how hard it would be for them to make one, though (it can't be that far from the OSX version, other than — as you mention — the drawing to screen).
The slow link response seems to be because IE is generally memory-resident (and smaller) and takes less time to open than Firefox in general (certainly I notice no difference in times between opening a link from Outlook and opening from XChat or whatever), so I'm not convinced this is really an issue of their bias.
The updates use controls that wouldn't go on Firefox because they'd be deemed unsafe (I believe there's a plugin to make them work, I wouldn't recommend it though). The best way for Microsoft to get around this would be to stop relying on a browser at all for Windows Updates, which is basically what they've done by "forcing" Automatic Updates (which isn't browser-based) on in XP SP2.
The point is less allowing them to use other formats than it is using a format that they cannot, I believe. It's not a good solution if, every time they want to read an in-house document, they have to use some random conversion tool, for example.
If there's no accessible software for the storage format, then it has everything to do with accessibility.
I don't see why this has turned into a Windows vs. Linux thing though. If it was just that they could run OO.o on Windows.
Agreed. The difference here is that Linux systems are almost always customised to some degree, and many optimised versions exist, whereas OSX systems are almost never customised, meaning tailoring them to this hardware is potentially more difficult.
On the other hand, you'd probably have gotten equivalents to all of the packages listed bundled with it, and then some.
Or, much more likely, they want to alter it to run well on the extremely unique hardware that they were developing for the project. As far as I can tell, they wanted the designers to be able to tinker, rather than the recipients themselves.
Yeah, I didn't link that because it only has the download for the old version (since the new one is BSD licenced and there's some weirdness about putting those on the Drupal homepage).
Try using this for Drupal, if the problem comes up again, I've been using it for a while and it's excellent.
Agreed, but I find that a lot of the time, particularly when working on projects where the requirements can change, unit testing provides a kind of documentation in itself. Even if the programmer has been too lazy to update the documentation for a method, the test itself must have been updated, so the functionality — provided that the tests have sufficient coverage — is described in the expectations of these tests.
If it helps with your numbers, my company uses it (along with NMock — whose site appears to be down, ironically because of a problem with Java — for isolation) for all .NET projects.
I love how they've used attributes in the implementation, makes things very easy. The Java (1.)5.0 implementation of JUnit uses this approach too, I believe.