This is somewhat akeen to the stupidest "feature" in Windows, when you are dragging a scrollbar and move the mouse too far away from the bar, it will jump back to it's original position. Fucking moronic.
To some of us, this is an exceptionally useful feature.
Didn't you know that Windows NT is (kind-of) the successor to OS/2?
In name only. They don't have a common codebase from which both have been derived. One need look no further than the complete lack of anything substantial in common in their designs to see that.
I see no reason why, if you design your API correctly and extensibly in the first place, with good modularisation, your OS shouldn't be compatible with code in 50 or 100 years time, let alone 5. Backwards compatibility is useful. Especially in computing where projects are rarely maintained beyond the second or third stable release. I don't quite see why moving forward should necessarily leave old applications broken.
The vast bulk of "compatibility problems" (indeed, the vast bulk of problems in general) with Windows software are because the applications are poorly written.
Just to get it out of the way, I 100% support abortion, even for simple birth control reasons, however:
You have a uterus? Oh. Then you have no real say.
Why should the father, who may well have equal emotional stakes in the child[ren] (and will probably have equal stakes in the cost of the child[ren], even if he _doesn't_ want it/them) have no input on the matter ?
A virus that cannot write to any executable files is dead on restart. Such a virus will not stick around for long.
It is trivial to get something to start on login (~/.bashrc, etc).
True enough. But then, there can be no security if the user doesn't know what s/he is doing.
A point I continually try and make, but no-one seems to grasp. This is the fundamental problem - users have to decide what to run and what not to run and the vast majority of them are quite happy to run _anything_.
Hopefully the distro. It's simple enough and effective enough that such a system should be provided by the more user friendly distributions.
Your idea sounds very similar to System Restore on Windows, only expanded to cover user files as well as system files.
Don't get me wrong, I got a good chuckle out of that but CEOs make a lot of money because they're worth a lot of money.
Given that the old boys network never lets anyone into a major CEO role until they're already "worth a lot of money", it's kind of hard to know if anyone can do the same job for less...
Because the virus will likely not be able to use the machine to spread, [...]
Why not ? Think about the most popular vectors for malicious code for the last ten years, then explain which of those a program running as a regular user couldn't do.
[...] or at the very least it won't be able to modify any executables, so a simple restart will fix it.
This hasn't been a particularly common method of malware propogation for a _very_ long time. It's not like OS-level binary executable files are shared between machines very often these days.
Not to mention all those wonderful "user friendly" frontends to sudo banging around these days - conveniently giving $RANDOM_BINARY an easy way to prompt the user to elevate its permissions.
Plus a simple cron job can be designed to back up user files to a non user-writeable directory.
With a typical Linux installation, a compromised process only has as much access as the user name it was run under. That is, a virus that remotely compromises a process will not have root access.
Now, assuming this is a typical end user PC (to make the Windows comparison valid), why does this matter ?
They do this shit all the time in negotations with large firms. If two people use one machine then you need to pay for Excel twice.
No, you don't. This blatant FUD.
Indeed, Microsoft specifically have a licensing scheme that allows employees to have a copy of something like Excel at home and use it, without having to purchase an additional copy.
Maybe this will get rid of licensing models that are 'per cpu'. I've never understood the logic in charging per CPU, anyone care to explain? One computer, one license. Or even better, no licenses.
Because the easiest way to differentiate between a "high end" (ie: large, critical computer-based business processes, more able and more willing to pay more $$$) and a "low end" (ie: small, business is probably not completely dependent on computers, not a lot of $$$ to throw around) business is the presence of multiprocessor hardware.
Basically it's just a crude way of identifying rich customers from poor customers and charging both as much as possible.
This is not strictly true. The product that became Windows NT was going to be the "high end" version of OS/2 - portable, multiuser, heavily multithreaded, SMP capable, microkernel[ish], etc, etc. This was the OS/2 Microsoft was working on 1988 - onwards.
The product that was (at the time, yet to be released) OS/2 2.x was going to be the "consumer" version of OS/2. Basically the OS/2 equivalent of Windows 9x. It wasn't portable, wasn't SMP, was single user, etc, etc. This was the OS/2 IBM were working on with Microsoft in the mid to late 80s.
Both of these versions of OS/2 were going to replace the DOS+Windows combo.
Then Windows 3.0 - and later 3.1 - were unexpected runaway successes. Microsoft started wanting to make "OS/2 NT" more like Windows than older versions of OS/2.
Then Microsoft and IBM split. Since "OS/2 NT" was basically all Microsoft's work, they took it with them to continue development, eventually releasing Windows NT - which lives on today as Windows XP and 2003.
IBM went back to the OS/2 2.X code (which at the time was their responsibility, although Microsoft still owned parts like HPFS) and proceeded to released OS/2 3.x and 4.x. "eComstation" developed from the codebase.
So NT isn't really a "descendent" of OS/2, because they don't have any common heritage, apart from a name and a failed business partnership. They don't share any sort of architectural similarities and really the only thing remotely in common is some "optional" APIs for legacy support.
NT was always supposed to replace the DOS-based versions of Windows. Windows 95 was the "NT lite" bridge product that was supposed to get everyone off DOS-based Windows and onto NT. There was never supposed to be a Windows 98 or Me. The problem was people kept demanding their low-level DOS and Windows compatibility - particularly with old hardware devices - that NT simply cannot provide. So the timetable got stretched out by 5 - 10 years.
As opposed to the application-centric approach of Windows; first with the Program Manager, then with the Start Menu. Compare:
OS/2: Navigate to Folder -> Open File/Template
Windows: Start Application -> New/Open File
This is how lots of people do it, but it is not, in fact, how the Windows 95 interface was *supposed* to be used. It was _supposed_ to be like the OS/2 GUI - you would navigate to some folder then right click -> New -> "Object".
It really amazed me just how much OS/2 resembled NT4 in a lot of ways, only with a better GUI and much more reliable.
Not really. You'd be hard pressed to find many ways OS/2 and NT are[/were]at all similar.
The fact that a lot of banks used OS/2 for a long time, indicates just how well made OS/2 was at the time when compared to DOS/Win3.1, Win9x and early WinNT.[...]
Well, one must remember that *at the time* OS/2 was supposed to be the successor of DOS/Windows. You'd hope it was a lot better.
I can't agree OS/2 [Warp] was better than NT4, however. Somewhat lighter on hardware, certainly - but then again it was doing less, so you'd expect that.
I think Microsoft, kind of, caught up to OS/2 with Windows 2000 SP3 in terms of reliability.
Interesting you say that, because it was the release of NT4 back in 1996 that dragged me away from OS/2 - primarily because OS/2 (while better than DOS+Windows 3.x) wasn't as stable.
But MS still doesn't seem to "get" the concept of a proper Object Oriented desktop. OS/2 did. NeXTSTEP did. And of course, Mac OS X does.
OS/2's GUI was its main strength (after about 1994 - 95). However, it was also quite complex to really use (ie: benefit from the OO part). I wouldn't agree that OS X's GUI is anything close to PM with regards to OO - I wouldn't even say it's as good as Windows.
I'm not the type of guy to bash Microsoft, but I must say I was quite surprised when spyware of some sort infected IE on a fresh and updated install of WinXP. www.google.com was redirected to another site offering spyware removal (What a joke)
Slashdot Mirror
To some of us, this is an exceptionally useful feature.
When was the last time anyone came up with a "major IT breakthrough" ?
IE managed to make a significant impact on Navigator's marketshare before it was "bundled"...
OTOH, the Internet community back then was arguably more tech-savvy, so that might have had something to do with it.
In name only. They don't have a common codebase from which both have been derived. One need look no further than the complete lack of anything substantial in common in their designs to see that.
The only reason I will mock it is because it isn't multiuser.
Windows 2000 came on a bootable CD.
The vast bulk of "compatibility problems" (indeed, the vast bulk of problems in general) with Windows software are because the applications are poorly written.
Or the features (hence the lack of "bloat").
FYI, XP comes with the older, plainer Media Player 2 - you'll just need to manually change the file associations.
You have a uterus? Oh. Then you have no real say.
Why should the father, who may well have equal emotional stakes in the child[ren] (and will probably have equal stakes in the cost of the child[ren], even if he _doesn't_ want it/them) have no input on the matter ?
It takes two to tango, as they say.
This is like saying "not free, needs a computer".
It is trivial to get something to start on login (~/.bashrc, etc).
True enough. But then, there can be no security if the user doesn't know what s/he is doing.
A point I continually try and make, but no-one seems to grasp. This is the fundamental problem - users have to decide what to run and what not to run and the vast majority of them are quite happy to run _anything_.
Hopefully the distro. It's simple enough and effective enough that such a system should be provided by the more user friendly distributions.
Your idea sounds very similar to System Restore on Windows, only expanded to cover user files as well as system files.
Given that the old boys network never lets anyone into a major CEO role until they're already "worth a lot of money", it's kind of hard to know if anyone can do the same job for less...
Why not ? Think about the most popular vectors for malicious code for the last ten years, then explain which of those a program running as a regular user couldn't do.
[...] or at the very least it won't be able to modify any executables, so a simple restart will fix it.
This hasn't been a particularly common method of malware propogation for a _very_ long time. It's not like OS-level binary executable files are shared between machines very often these days.
Not to mention all those wonderful "user friendly" frontends to sudo banging around these days - conveniently giving $RANDOM_BINARY an easy way to prompt the user to elevate its permissions.
Plus a simple cron job can be designed to back up user files to a non user-writeable directory.
Who is going to set it up and manage it ?
Now, assuming this is a typical end user PC (to make the Windows comparison valid), why does this matter ?
No, you don't. This blatant FUD.
Indeed, Microsoft specifically have a licensing scheme that allows employees to have a copy of something like Excel at home and use it, without having to purchase an additional copy.
Because the easiest way to differentiate between a "high end" (ie: large, critical computer-based business processes, more able and more willing to pay more $$$) and a "low end" (ie: small, business is probably not completely dependent on computers, not a lot of $$$ to throw around) business is the presence of multiprocessor hardware.
Basically it's just a crude way of identifying rich customers from poor customers and charging both as much as possible.
This is not strictly true. The product that became Windows NT was going to be the "high end" version of OS/2 - portable, multiuser, heavily multithreaded, SMP capable, microkernel[ish], etc, etc. This was the OS/2 Microsoft was working on 1988 - onwards.
The product that was (at the time, yet to be released) OS/2 2.x was going to be the "consumer" version of OS/2. Basically the OS/2 equivalent of Windows 9x. It wasn't portable, wasn't SMP, was single user, etc, etc. This was the OS/2 IBM were working on with Microsoft in the mid to late 80s.
Both of these versions of OS/2 were going to replace the DOS+Windows combo.
Then Windows 3.0 - and later 3.1 - were unexpected runaway successes. Microsoft started wanting to make "OS/2 NT" more like Windows than older versions of OS/2.
Then Microsoft and IBM split. Since "OS/2 NT" was basically all Microsoft's work, they took it with them to continue development, eventually releasing Windows NT - which lives on today as Windows XP and 2003.
IBM went back to the OS/2 2.X code (which at the time was their responsibility, although Microsoft still owned parts like HPFS) and proceeded to released OS/2 3.x and 4.x. "eComstation" developed from the codebase.
So NT isn't really a "descendent" of OS/2, because they don't have any common heritage, apart from a name and a failed business partnership. They don't share any sort of architectural similarities and really the only thing remotely in common is some "optional" APIs for legacy support.
NT was always supposed to replace the DOS-based versions of Windows. Windows 95 was the "NT lite" bridge product that was supposed to get everyone off DOS-based Windows and onto NT. There was never supposed to be a Windows 98 or Me. The problem was people kept demanding their low-level DOS and Windows compatibility - particularly with old hardware devices - that NT simply cannot provide. So the timetable got stretched out by 5 - 10 years.
OS/2: Navigate to Folder -> Open File/Template Windows: Start Application -> New/Open File
This is how lots of people do it, but it is not, in fact, how the Windows 95 interface was *supposed* to be used. It was _supposed_ to be like the OS/2 GUI - you would navigate to some folder then right click -> New -> "Object".
Actually IBM went back to the 2.x code and worked from there.
Yeah, and "desktop printing" was one of those things notorious for weird and wonderful problems on MacOS...
Not really. You'd be hard pressed to find many ways OS/2 and NT are[/were]at all similar.
The fact that a lot of banks used OS/2 for a long time, indicates just how well made OS/2 was at the time when compared to DOS/Win3.1, Win9x and early WinNT.[...]
Well, one must remember that *at the time* OS/2 was supposed to be the successor of DOS/Windows. You'd hope it was a lot better.
I can't agree OS/2 [Warp] was better than NT4, however. Somewhat lighter on hardware, certainly - but then again it was doing less, so you'd expect that.
I think Microsoft, kind of, caught up to OS/2 with Windows 2000 SP3 in terms of reliability.
Interesting you say that, because it was the release of NT4 back in 1996 that dragged me away from OS/2 - primarily because OS/2 (while better than DOS+Windows 3.x) wasn't as stable.
But MS still doesn't seem to "get" the concept of a proper Object Oriented desktop. OS/2 did. NeXTSTEP did. And of course, Mac OS X does.
OS/2's GUI was its main strength (after about 1994 - 95). However, it was also quite complex to really use (ie: benefit from the OO part). I wouldn't agree that OS X's GUI is anything close to PM with regards to OO - I wouldn't even say it's as good as Windows.
Were you running as an Administrator ?
Since I assume you're talking about IE, may I also assume you're unaware of how it works and thus, unable to give any sort of accurate analysis ?
When OS X (and Linux) have market share and end-user demographics similar to Windows, this is an assetion you can make. Not before.