I'm sure I'm about to burn karma with this... but in KDE, Konqueror acts as both web browser and file manager. At least it's entirely userspace, but does anyone know how closely the file managing and web browsing aspects of Konqueror are tied?
IE and Konquerer are basically identical, architecturally speaking.
The two are separated into different kparts. It's a lot of programming jargon, but the fact that you can use something like KHTML in another program (e.g. Safari) without including the file manager and all other integrateable components shows that. Konqueror sort of provides an interface that can be used to combine more functionality than just KHTML.
This is exactly the same relationship Explorer and IE have.
OS X is based on OpenStep which dates back to 1989.
Actually it's NeXTSTEP that dates back to 1988. OPENSTEP was the later API (that was ported to numerous other platforms as well).
There are some fairly significant changes. Different version of Mach, different BSD base, completely new display system, completely new GUI, new APIs (Carbon, Quicktime), major emulation environment (Classic), etc, etc.
The changes from NeXTSTEP -> OS X are *major*. It's not just a few tweaks here and there and a new GUI skin thrown on.
Just like Vista.
Right. Vista will have taken 4 - 6 years (depending on how you want to count) to arrive (and its successor will probably take another 3 - 5), but is undergoing the same "rapid development" that sees OS X releases with major new features every twelve months...
Once again, do you people even *think* before saying stuff like this ?
Rriiigghhtt. That's why Microsoft is dropping APIs in Vista?
Which APIs ?
That's why they're plopping a bunch of 1.0 APIs into the system and reorganizing the kernel?
Development that's been occurring over a period of *6 years*, not *12 months*. The changes from Windows 2003 -> Vista are, ironically enough, similar to the changes from NeXTSTEP -> OS X (although not quite as extensive). The rather important point - the crux of my whole argument - you seem to be missing here is the *timeframe* involved. Once again:
NT - ~13 years old, releases averaging about every 3 - 4 years.
OS X - ~5 years old, releases averaging about every 12 months.
Can you not see the difference, for developers (especially) and users, between OS releases every 12 months vs OS releases ever 3 - 4 years ?
Oh my GOD. The exact same is true for Windows, you troll. Shock and horror, you have to upgrade Norton when you upgrade to a new version of Windows.
Point releases (that would be, say, Windows 2000 -> XP -> 2003) *very* rarely break things on Windows. Even drivers will typically work between them.
Not to mention, *AGAIN*, that these changes in Windows take place over a period of ~5 *years*, not 12 *months*.
Are you really so buried in anti-Microsoft rhetoric you can't even see the difference between having to upgrade software every 12 months as compared to every few years ?
So how does your theory explain new tech like Expose, Spotlight, CoreData, CoreVideo, CoreImage, and such?
Er, my "theory" explains them perfectly. New, fairly major features are being introduced frequently into OS X, because it's still under rapid development. Thank you for supporting my argument.
Your logic also applies to Windows. Unless you're actually going to argue Windows XP doesn't have things that need fixing and implementing.
No. However, OS X has had important problems that need addressing (most visibly performance - which is still sub-par - and parts of the UI, like aspects of the Dock, which were "fixed" with Expose).
For Christ's sake, Windows is still not a true, functioning multi-user system in 2006. That's one broken OS.
Windows NT has been multiuser from day one, and it's had UI niceties to expose this (like "Fast User Switching") longer than OS X has. Where's the Terminal Server equivalent of OS X ?
Vista will be finally catching up to where OS X was April of last year. It's sad and hilarious.
From everything I've read, Vista will be surpassing OS X. NT is already better - or certainly equal - in most lower-level aspects, and even some higher level ones.
(Although, I do wonder if you found it similarly "sad and hilarious" that it took Apple ~7 years (at least) and three attempts, to catch up with Windows NT. Probably not, I'm guessing.)
I believe that NT did not have true multi-user support until Citrix added it into NT 3.5.1 for their WinFrame product.
NT has always been multiuser. Citrix added the ability to run multiple GUI sessions.
Also, *nix based OSs do support more advanced ACLs. The Single Unix Specification *requires* that an OS support ACLs in order to be labelled a UNIX operating system.
Yes, I know. That's why is specifically stated "traditional unix" (which is what 99% of the unix world still uses, even when their OSes are capable of more).
I'm just curious what your basis for that statement is.
The relative frequency, magnitude and timeframe of changes over the last few versions of Windows compared to OS X.
That is, what your source is, and how the term "new" is being used. I'm honestly curious. (I've got little to no interest in these platform-wars. I'm a Free Software nut, so it's all payware to me.) Sure, Mac OS X was a radical departure from System 9, a totally different product, but 13 years ago, Microsoft was still saying classic Win 3.x was a good idea. Windows XP has a massive amount of "new" code. (This says nothing about the *quality* of that code, of course, but that's not what you were talking about.)
Windows XP is NT 5.1. Windows 2000 was NT 5.0. The differences between the two are not large. Windows 2000 was certainly a major upgrade from NT 4.0, but it happened over a period of ~4 years, not 12 months.
Consider OS X. It's averaging a new release about every 12 months, usually with reasonably large changes. Drivers & low level programs are frequently broken by these updates. This *is* changing - right now - but OS X is still in a development phase where new features are being frequently added and backwards compatibility is a lower priority.
Who said OS X was "completely new" or praised it for being such?
OS X's recent arrival and subsequent lack of legacy code (relatively speaking) are *often* touted as advantages (which they are, in certain contexts).
If you had any CLUE at all about the Mac community, in fact, you'd be amused to know a lot of people stuck with OS 9 for a couple of year after OS X's release.
I knew that already, thanks. Doesn't make any difference to what I said.
"Newness" was certainly not a factor in the Mac community's minds.
Maybe you need to open your other eye. OS X's relatively lack of legacy code is frequently touted as an advantage, particularly when comparing to Windows.
But, hey, if all you've got to defend this three-year-late OS X clone called Vista is to invent crap to mock, have at it.
Funny, here I was thinking all I did was point out comparing the release cycles of a well established 13 year old product to a 5 year old product is silly. I don't recall making any attempt to jump into the irrelevant, petty "who has the newest OS at the moment" rhetoric you're spouting.
Every point you've brought up has been soundly crushed in these discussions, especially your hilarious attempts to call NT "mature" and OS X "immature" despite the age and robustness of the OpenStep APIs compared to the disaster that has been Windows XP.
Such a disaster that it owns most of the market. Damn, I wish I could have disasters like that...
I mean, seriously, Windows XP is mature?
Windows *NT* is mature (heh, and you accuse _me_ of strawman arguments).
The security nightmare that requires registry cleaners and disk defragmenters and antivirus and antispyware and 20 wizards to do everything is "mature?"
Not to mention your FUD...
The OS with the ultra-clunky interface and ridiculous security policies [...]
I don't think anyone advocating the UI train wreck called the Dock is in any position to criticise UI.
[...] (hey, let's run everyone in admin accounts in the year 2006!) [...]
XP was releaased in 2001.
[...] and broken APIs?
Broken how ?
With Vista being restarted back in 2004, it means it's now Windows Vista that is the immature, unstable product with the rapidly changing codebase. All those 1.0 APIs are going to be fun to break.
Vista may be immature at release (by definition) and certainly includes new APIs, but it is built on a well designed, well tested, mature core, still supports numerous existing, mature APIs and still won't be undergoing the rapid changes OS X has been having (I certainly wouldn't complain if the next version of Windows hit in 2008, but I can't see it happening).
You've just bought into the Microsoft marketing machine, and you decided to obsessively respond to every anti-Windows comment you could find to defend your baby. Please, get a life.
Meanwhile, you're just parroting the standard "Windows sucks, OS X rules" party line. At least my arguments have some rationale behind them.
(I can't believe anyone would seriously suggest a platform still undergoing rapid release cycles usually including fairly major changes, is "mature". I mean, do you call the 2.6 kernel "mature" as well ?)
In other words, CentOS is for those who would otherwise use Debian, but feel the need to undermine the business of a company that has supported Linux from long ago.
Or maybe people who have to use Red Hat (eg: for Oracle) for some machines and want the rest to be identical to make systems administration easier.
I believe that people should be responsible for their own actions. You spilled HOT coffee (when you specifically ordered HOT coffee) on your lap after you set a known-weak styrofoam cup between your legs? Guess what? It's stupidity on your part. Don't use that as a lottery ticket to sue a successful corporation because you're a moron), and don't expect gubbament to help you either.
This is OT, but I recently discussed it with someone else and the details are fresh in my mind.
In the case you are referring to, the coffee wasn't just "hot" it was "incredibly hot". Far hotter than it needed to be, and far hotter than any normal person would expect it to be. That, coupled with numerous prior complaints that had been ignored, was the reason why "a successful corporation" was found liable.
It's also important to point out here that the injured person in this case was *not* trying to use their (very serious) injuries as a "lottery ticket", they just wanted their medical costs covered.
In the "McDonalds coffee" case, the bulk of the responsibility is *clearly* McDonald's.
"Personal responsibility" doesn't mean a corporation can sell me Rat Poison advertised as softdrink, then say it's my fault because I didn't have it chemically analysed before drinking. It means that if they sell me Rat Poision and I drink it, I can't blame them for doing so.
Active Directory comes out. We use it today, but it's improved little. I manage it ever hour, and am constantly faced with the awkwardness and inability to do things in it that I could easily do a decade ago in NDS.
No. OS X is still changing quite frequently. It's a relatively immature platform.
Well, I never said anything else was not mature, just that OS X was compartivley mature as well.
Except it's not. Changes are still relatively frequent and major.
So I guess to go back to the comparison game, since I'm only advocating the stability of a platform and you are advoacting for the stability of Windows and Linux and ANYTHING but OS X... who'se the cookie cuttor troll again?
I'm not advocating anything. I'm pointing out that OS X is a relatively immature platform and Windows is a relatively mature platform and as such, comparing their releases schedules is largely meaningless.
As I said elsewhere, if Apple are still pumping out non-trivial updates to OS X every 12 months in 2014, *that* will be impressive.
I think we've heard your tune before.
I've certainly heard yours. The typical Mac Zealot schizophrenia that allows you to turn everything into "OS X rulez".
"Hey, check out OS X, man ! Latest and greatest OS on the market. Completely new. None of that krufty old code like Windows has".
"Unstable ? No, OS X has been around for ages. It's mostly NeXTSTEP, you see, so everything in it has been around the block a few times and is well tested."
Whereas Windows is pretty much the same now as it was 13 years ago?
There's a hell of lot less *new* code in Windows.
(I feel compelled to point out, since it's pretty obvious you haven't put a great deal of thought into it, that this is not necessarily a *good* thing).
So since OS X has now lapped even Vista by havng similar (yet more advanced) features in place years beforehand, does the statement that OS X is an immature product mean that Windows is even less so?
Well, uh, yes, given that "even less immature" would mean it is more mature.
Of course, I guess you could mean "mature" in the sense of "dottering".
"Mature", as in stable, proven technology. Like the 2.4 kernel is "mature". Like Solaris is "mature".
Yes, I guess I could see that but you should probably try to be clearer next time you post.
My meaning is perfectly clear to people who aren't standard-issue, cookie-cutter, anti-Microsoft trolls. Do you have similar difficulties understanding the meaning of:
"The problem with your comparison is that the 2.2 kernel is a mature product, whereas the 2.6 kernel is not." ?
OS X is pretty obviously an extension of many of the technical principles and innovations first developed for NeXTStep, which started in the late '80s and therefore predates the genesis of Win32 by several years. Factor in that much of the BSD foundation upon which NeXTStep relied had existed for years before that, and he should feel embarassed.
And NT is largely "an extension of many of the technical principles and innovations" in VMS, started in the mid '70s.
Neither of these two tidbits of historical interest is particularly relevant to OS X and Windows NT which, as discrete products, have been on the market and under development for ~5 and ~13 years, respectively. Most of the code in OS X as it exists today, was *not* in NeXTSTEP.
Which is true if you limit yourself to just the kernel and immediate surrounds. OS X includes a tad more code than that, however, and rather large portions of it were done from scratch.
On Windows, the home and end keys take me to the beginning and end of the line I'm editing, consistantly in every different app in which an editing window appears. Not so the Mac. Here I end up using a modifier key with the arrow keys... sometimes. Bizarre. The list goes on.
You are simply describing a different UI standard. In OS X, the Home and End keys move the pageview to the start and end of the document, respectively, without moving the insertion point. This is the standard OS X behaviour, and if some program isn't doing that it's because it has been specifically written not to.
I personally dislike this behaviour as well, but it's fairly simple to change and make it like Windows, if you prefer. It's certainly no more significant a change than OS X window widgets being on the left and Windows widgets being on the right.
They really have to get away from the single menu at the top of the page thing as well - it only adds to the nostalgia for Z-80 based single-board computers.
There are numerous UI studies proving it to be a superior placement of the menu bar.
Excuse me... but Windows patches *do* break thing regularly.
Statistics ?
There's a reason that MS has had to pull so many of their service packs.
Even the most well publicised "service pack breaking stuff" - XP's SP2 - only broke a tiny percentage of machines.
Service packs and hotfixes breaking things is _extremely_ uncommon. It is, however, extremely widely reported on. You should not conflate the two.
In both my experience, and in the experience of many admins responsible for Windows servers, it is often dangerous to apply Windows hotfixes, too. Even with all the precautions of test environments, and staged updates... all too often the patch breaks everything anyway.
Sounds more like your staging process is broken, if you don't find out something doesn't work until you actually get to the production systems.
It can takes weeks to figure out *why* this happened, since MS doesn't exactly tell you what they changed in more than vagueries.
Have you ever bothered to, you know, *ASK* ?
You also demonstrated that different operating systems do things differently.
My point being you were criticising Windows for something that unix does much more of.
I don't see DirectX available on anything other than some versions of Windows. A Windows application doesn't run on Solaris, or MacOS, or anything else. What's your point?
Your implication that there's some "standard unix API" that someone can write to and have it work everywhere is, at best, ignorant and, at worst, deceptive.
Thus, your criticism of Windows for not having the equivalent, was hypocritical.
There is far more standards following and compatability across various UNIX implementations than there is between Windows and *anything* else.
Yes, just enough so people like you say there's nothing to worry about, but still enough so people like me find out there still is a lot of things to worry about.
So sound doesn't always work... so what, it's only sound. Mission critical applications don't even use that. As for UI things... Xlib and POSIX will take you quite far. There are other options, such as all of those the GP mentioned.
That sound you can barely hear is the point flying way, way over your head.
My point about the MS APIs were that they were the "way it will be done" according to MS. Then they were abandoned in favor of new toys to do the same thing.
You say this like its something that happens frequently, like it happens unexpectedly, like it happens suddenly and like it doesn't happen on every other platform, when none of these things are true.
MS suffers horribly from the lack of foresight. It is obvious through all of their APIs and products. They are a marketing company, not a software one, and it shows.
How does it show ? By supporting 20+ year old APIs ? By offering 5 - 10 year transition periods ? By having one of the best records for legacy support in the industry ?
You also seem to assume that UNIX = Linux.
No, I do not. Nor am I ever likely too, seeing how Linux is one of my least favourite unix implementations.
You won't believe how amny times some of my less computer-knowledgable friends or clients are afraid to apply a Windows Update patch because they (rightfully) fear that it will break more things than it might fix.
With you "advising" them, I certainly could believe. The fact is, however, Windows patches break things quite rarely. Much less frequently, in my experience, than updates to the average Linux distribution.
(You could believe, if you stopped blindly ignoring every argument against your favorite defective OS.)
I have no "favourite OS". And if you want me to stop poking holes in your arguments, come up with some better ones that involve more facts and less FUD.
You can choose any of QT, GTK+, WXWidgets, Motif, and more. All of them will work, so pick whichever one you like best for whatever reason. For audio, you could use either ALSA or OpenAL for precise control, or Xine or GStreamer if you need multimedia features. Again, all will work. For drivers, I do not have enough knowledge to tell you (not that you want to hear me), but there's a good chance the situation is similar there too.
Thank you for demonstrating why your previous relative criticism about the quantity of APIs for Windows is groundless.
It's better to have a low-leve API that does little but does it so well that it won't need to be obsoleted by a fix 2 years from now.
Which Windows APIs are you thinking of that have been unexpectedly obseleted 2 years after introduction ?
That's why the above-mentioned GUI and audio APIs are so stable as well.
Stable, you say ? How old of a version of QT will support the current release of KDE ? What's the oldest version of KDE you can run with today's QT ?
The Windows APIs are extremely stable. Windows apps written to them 10+ years ago work in todays Windows 2003 systems.
Autoconf and make are an automated system to adjust a generic package to the specifics of a system that is potentially different from the one next to it.
Note that with a "standard" and "stable" "unix API", this should not be necessary.
These differences are not an inherent flaw. They stem from something called "choice", which I'm afraid you may not be able to grasp.
Funny how last post you were criticising Windows for the choice is was offering in "no less than six APIs that I can think of, just off the top of my head".
This is not because having alternative APIs, which all work together, is bad.
No, it's because having dozens of APIs that all do basically the same thing is bad (well, wasteful at the very least).
It's because most IT businesses think in Windows-terms - vendor lock-in and no choice whatsoever.
Most IT business think in terms of economics. And having to support multiple implementations of the same functionality (be it from the developer or end user perspective) costs a lot of money with few reciprical benefits.
You (along with a large proportion of the Linux community, so don't feel too badly about it) seem to be missing one of the main reasons why so little commercial software is produced for Linux - it's because there aren't any standards (or because there are too many, depending on your perspective) and because the APIs tend to be relatively unstable (as they are developed under the assumption that software using them will be open source and thus can just be recompiled when things change).
Lets say Adobe wanted to port Photoshop to Linux. Which package management scheme should they use ? Which distro should they target ? Which GUI should they write for ? Which version ?
With Windows, these questions are easy to answer. Hell, most of them are barely even worth asking, because there's only one answer/
"Unix" systems, as you call them, are not a replacement for Widnows, but an alternative, and consequently do certain things differently. Try this page for a better explanation.
You don't see any calendar functions in MSN messenger, do you?
Actually, if you're running the whole shebang - AD, Exchange, Outlook and an internal MSN server - when you do things that are scheduled in your calendar (eg: go to a meeting, go to lunch) your IM presence is automatically updated to reflect that. It's quite nice.
The unix security model is far superior and flexible... It just requires more effort up front
You have this completely arse-about-face. (Traditional) Unix's permissions system is quite basic and primitive, with no better than per-group granularity and the inherent weakness of an all-powerful superuser. The per-user ACLs in Windows are far, far more fine-grained and flexible, (which thus requires more management effort for non-trivial scenarios).
Windows's permissions model is a superset of (traditional) unix's.
Correct me if I'm wrong, but are you implying that comparing business card designs *isn't* geeky?
I've yet to see any geeks get hard over the texture of the paper/cardboard their business card is made out of, or whether it using imperial bronze or golden sunrise coloured gilding.
By what standard is Windows a mature product compared to any other OS?
Windows NT = ~13 years old.
OS X = ~5 years old.
They release updates and fixes every MONTH and there are like 15-umpty "versions" of Windows in the wild right now. The MSKB is FULL of "bug reports" and stupid workarounds for things that SHOULD have been fixed long ago.
Rrrrright... Both of which are indicators of a more mature product...
OS X is still undergoing relatively rapid feature development and major changes, with things like API stability and backwards compatibility of secondary concern. Windows NT development, OTOH, is more focussed on refinements and improvements to the existing codebase with things like API stability and backwards compatibility considered much more important.
It hasn't been at all uncommon for OS X's major point releases to break applications - particularly applications delving more into the lower levels of the OS.
OS X development is roughly where Windows NT development was ca. 1998. It's in a transitional period from rapid, signficant changes to more gradual, refined changes. Hence, for example, the reason the OS X kernel API has only officially been deemed "stable" with the latest OS X release.
*That* is why OS X releases have been more frequent, and more significant. To put it bluntly, they've had more things that needed fixing and implementing.
Since when is OS X an OSS project? And even if you can somehow claim it is, Apple is a fairly large commercial enterprise with all those pressures and responsibilities you mention.
The poster I was replying to, in his prior post, commented on the "release early, release often" philosophy commonly seen in OSS projects.
I didn't say that Vista hasn't had a lot of internal changes, just that they don't matter in that way. They don't enable new applications or technologies; they fix shortcoming in the previous implementation.
So... Just like very other update to a mature platform, then ?
I'm not as much talking about vendors as technologies. MS comes up with their own versions of things, pushes everyone to use them, and then they drop it from something shinier.
So... They're just like everyone else ?
The non-MS part of the world has been using things called "standards", and they have been doing so far longer than Windows has existed, let alone been used.
Really ? What's the standard API for a "unix" GUI application ? How about using audio devices ? Which API should I use to make sure my hardware driver compiles on Linux, FreeBSD, Solaris and OS X without modifications or special cases ?
We have POSIX as an API standard, and that's been around for a long time.
And is basically useless (not to mention largely ignored) for anything except trivial command line applications.
Heck, it's not at all uncommon to find trivial open source "unix" applications that only work on x86 Linux machines with particular versions of glibc.
Most "cross platform" unix source code doesn't compile on a wide range of platforms because of "standards", it does so because of the amount of work done by things like autoconf and make.
Jeez. One of the biggest hurdles to wider commercial adoption of Linux is the sheer volume of different APIs (many of which all do essentially the same thing), and you're here trying to say there's no such problem at all ?
MS has had no less than six APIs that I can think of, just off the top of my head.
And "unix" has dozens (if not hundreds). Your point ?
They tried to have their own networking protocols, their own email formats, APIs, and on and on.
So... Just like every other commercial vendor ?
They have all been problematic, and largely dropped for the standards that were already there. In that regard, yes, I can think of "vendors" that it doesn't apply to.
Such as ? Certainly not Apple, Novell or IBM. Maybe Sun, but the intersection of markets between Solaris and Windows is vanishingly small.
You act like Microsoft come up with something, then run away from it the first chance they get just to screw everyone over. Yet things like Win32, MFC and DirectX have been around for over a decade, and will *still* be in legacy support 5 years down the track, if not longer. Heck, Vista will still support Win16 on 32-bit x86, an API that's around twenty years old.
Slashdot Mirror
IE and Konquerer are basically identical, architecturally speaking.
Yes ! That would be great ! They could call them something like Task Manager and tasklist...
This is exactly the same relationship Explorer and IE have.
OS X is based on OpenStep which dates back to 1989.
Actually it's NeXTSTEP that dates back to 1988. OPENSTEP was the later API (that was ported to numerous other platforms as well).
There are some fairly significant changes. Different version of Mach, different BSD base, completely new display system, completely new GUI, new APIs (Carbon, Quicktime), major emulation environment (Classic), etc, etc.
The changes from NeXTSTEP -> OS X are *major*. It's not just a few tweaks here and there and a new GUI skin thrown on.
Just like Vista.
Right. Vista will have taken 4 - 6 years (depending on how you want to count) to arrive (and its successor will probably take another 3 - 5), but is undergoing the same "rapid development" that sees OS X releases with major new features every twelve months...
Once again, do you people even *think* before saying stuff like this ?
Rriiigghhtt. That's why Microsoft is dropping APIs in Vista?
Which APIs ?
That's why they're plopping a bunch of 1.0 APIs into the system and reorganizing the kernel?
Development that's been occurring over a period of *6 years*, not *12 months*. The changes from Windows 2003 -> Vista are, ironically enough, similar to the changes from NeXTSTEP -> OS X (although not quite as extensive). The rather important point - the crux of my whole argument - you seem to be missing here is the *timeframe* involved. Once again:
NT - ~13 years old, releases averaging about every 3 - 4 years.
OS X - ~5 years old, releases averaging about every 12 months.
Can you not see the difference, for developers (especially) and users, between OS releases every 12 months vs OS releases ever 3 - 4 years ?
Oh my GOD. The exact same is true for Windows, you troll. Shock and horror, you have to upgrade Norton when you upgrade to a new version of Windows.
Point releases (that would be, say, Windows 2000 -> XP -> 2003) *very* rarely break things on Windows. Even drivers will typically work between them.
Not to mention, *AGAIN*, that these changes in Windows take place over a period of ~5 *years*, not 12 *months*.
Are you really so buried in anti-Microsoft rhetoric you can't even see the difference between having to upgrade software every 12 months as compared to every few years ?
So how does your theory explain new tech like Expose, Spotlight, CoreData, CoreVideo, CoreImage, and such?
Er, my "theory" explains them perfectly. New, fairly major features are being introduced frequently into OS X, because it's still under rapid development. Thank you for supporting my argument.
Your logic also applies to Windows. Unless you're actually going to argue Windows XP doesn't have things that need fixing and implementing.
No. However, OS X has had important problems that need addressing (most visibly performance - which is still sub-par - and parts of the UI, like aspects of the Dock, which were "fixed" with Expose).
For Christ's sake, Windows is still not a true, functioning multi-user system in 2006. That's one broken OS.
Windows NT has been multiuser from day one, and it's had UI niceties to expose this (like "Fast User Switching") longer than OS X has. Where's the Terminal Server equivalent of OS X ?
Vista will be finally catching up to where OS X was April of last year. It's sad and hilarious.
From everything I've read, Vista will be surpassing OS X. NT is already better - or certainly equal - in most lower-level aspects, and even some higher level ones.
(Although, I do wonder if you found it similarly "sad and hilarious" that it took Apple ~7 years (at least) and three attempts, to catch up with Windows NT. Probably not, I'm guessing.)
NT has always been multiuser. Citrix added the ability to run multiple GUI sessions.
Also, *nix based OSs do support more advanced ACLs. The Single Unix Specification *requires* that an OS support ACLs in order to be labelled a UNIX operating system.
Yes, I know. That's why is specifically stated "traditional unix" (which is what 99% of the unix world still uses, even when their OSes are capable of more).
I'm just curious what your basis for that statement is.
The relative frequency, magnitude and timeframe of changes over the last few versions of Windows compared to OS X.
That is, what your source is, and how the term "new" is being used. I'm honestly curious. (I've got little to no interest in these platform-wars. I'm a Free Software nut, so it's all payware to me.) Sure, Mac OS X was a radical departure from System 9, a totally different product, but 13 years ago, Microsoft was still saying classic Win 3.x was a good idea. Windows XP has a massive amount of "new" code. (This says nothing about the *quality* of that code, of course, but that's not what you were talking about.)
Windows XP is NT 5.1. Windows 2000 was NT 5.0. The differences between the two are not large. Windows 2000 was certainly a major upgrade from NT 4.0, but it happened over a period of ~4 years, not 12 months.
Consider OS X. It's averaging a new release about every 12 months, usually with reasonably large changes. Drivers & low level programs are frequently broken by these updates. This *is* changing - right now - but OS X is still in a development phase where new features are being frequently added and backwards compatibility is a lower priority.
Who said OS X was "completely new" or praised it for being such?
OS X's recent arrival and subsequent lack of legacy code (relatively speaking) are *often* touted as advantages (which they are, in certain contexts).
If you had any CLUE at all about the Mac community, in fact, you'd be amused to know a lot of people stuck with OS 9 for a couple of year after OS X's release.
I knew that already, thanks. Doesn't make any difference to what I said.
"Newness" was certainly not a factor in the Mac community's minds.
Maybe you need to open your other eye. OS X's relatively lack of legacy code is frequently touted as an advantage, particularly when comparing to Windows.
But, hey, if all you've got to defend this three-year-late OS X clone called Vista is to invent crap to mock, have at it.
Funny, here I was thinking all I did was point out comparing the release cycles of a well established 13 year old product to a 5 year old product is silly. I don't recall making any attempt to jump into the irrelevant, petty "who has the newest OS at the moment" rhetoric you're spouting.
Every point you've brought up has been soundly crushed in these discussions, especially your hilarious attempts to call NT "mature" and OS X "immature" despite the age and robustness of the OpenStep APIs compared to the disaster that has been Windows XP.
Such a disaster that it owns most of the market. Damn, I wish I could have disasters like that...
I mean, seriously, Windows XP is mature?
Windows *NT* is mature (heh, and you accuse _me_ of strawman arguments).
The security nightmare that requires registry cleaners and disk defragmenters and antivirus and antispyware and 20 wizards to do everything is "mature?"
Not to mention your FUD...
The OS with the ultra-clunky interface and ridiculous security policies [...]
I don't think anyone advocating the UI train wreck called the Dock is in any position to criticise UI.
[...] (hey, let's run everyone in admin accounts in the year 2006!) [...]
XP was releaased in 2001.
[...] and broken APIs?
Broken how ?
With Vista being restarted back in 2004, it means it's now Windows Vista that is the immature, unstable product with the rapidly changing codebase. All those 1.0 APIs are going to be fun to break.
Vista may be immature at release (by definition) and certainly includes new APIs, but it is built on a well designed, well tested, mature core, still supports numerous existing, mature APIs and still won't be undergoing the rapid changes OS X has been having (I certainly wouldn't complain if the next version of Windows hit in 2008, but I can't see it happening).
You've just bought into the Microsoft marketing machine, and you decided to obsessively respond to every anti-Windows comment you could find to defend your baby. Please, get a life.
Meanwhile, you're just parroting the standard "Windows sucks, OS X rules" party line. At least my arguments have some rationale behind them.
(I can't believe anyone would seriously suggest a platform still undergoing rapid release cycles usually including fairly major changes, is "mature". I mean, do you call the 2.6 kernel "mature" as well ?)
Or maybe people who have to use Red Hat (eg: for Oracle) for some machines and want the rest to be identical to make systems administration easier.
This is OT, but I recently discussed it with someone else and the details are fresh in my mind.
In the case you are referring to, the coffee wasn't just "hot" it was "incredibly hot". Far hotter than it needed to be, and far hotter than any normal person would expect it to be. That, coupled with numerous prior complaints that had been ignored, was the reason why "a successful corporation" was found liable.
It's also important to point out here that the injured person in this case was *not* trying to use their (very serious) injuries as a "lottery ticket", they just wanted their medical costs covered.
In the "McDonalds coffee" case, the bulk of the responsibility is *clearly* McDonald's.
"Personal responsibility" doesn't mean a corporation can sell me Rat Poison advertised as softdrink, then say it's my fault because I didn't have it chemically analysed before drinking. It means that if they sell me Rat Poision and I drink it, I can't blame them for doing so.
Canyou give some examples ?
No. OS X is still changing quite frequently. It's a relatively immature platform.
Well, I never said anything else was not mature, just that OS X was compartivley mature as well.
Except it's not. Changes are still relatively frequent and major.
So I guess to go back to the comparison game, since I'm only advocating the stability of a platform and you are advoacting for the stability of Windows and Linux and ANYTHING but OS X... who'se the cookie cuttor troll again?
I'm not advocating anything. I'm pointing out that OS X is a relatively immature platform and Windows is a relatively mature platform and as such, comparing their releases schedules is largely meaningless.
As I said elsewhere, if Apple are still pumping out non-trivial updates to OS X every 12 months in 2014, *that* will be impressive.
I think we've heard your tune before.
I've certainly heard yours. The typical Mac Zealot schizophrenia that allows you to turn everything into "OS X rulez".
"Hey, check out OS X, man ! Latest and greatest OS on the market. Completely new. None of that krufty old code like Windows has".
"Unstable ? No, OS X has been around for ages. It's mostly NeXTSTEP, you see, so everything in it has been around the block a few times and is well tested."
There's a hell of lot less *new* code in Windows.
(I feel compelled to point out, since it's pretty obvious you haven't put a great deal of thought into it, that this is not necessarily a *good* thing).
Well, uh, yes, given that "even less immature" would mean it is more mature.
Of course, I guess you could mean "mature" in the sense of "dottering".
"Mature", as in stable, proven technology. Like the 2.4 kernel is "mature". Like Solaris is "mature".
Yes, I guess I could see that but you should probably try to be clearer next time you post.
My meaning is perfectly clear to people who aren't standard-issue, cookie-cutter, anti-Microsoft trolls. Do you have similar difficulties understanding the meaning of:
"The problem with your comparison is that the 2.2 kernel is a mature product, whereas the 2.6 kernel is not." ?
And NT is largely "an extension of many of the technical principles and innovations" in VMS, started in the mid '70s.
Neither of these two tidbits of historical interest is particularly relevant to OS X and Windows NT which, as discrete products, have been on the market and under development for ~5 and ~13 years, respectively. Most of the code in OS X as it exists today, was *not* in NeXTSTEP.
Which is true if you limit yourself to just the kernel and immediate surrounds. OS X includes a tad more code than that, however, and rather large portions of it were done from scratch.
You are simply describing a different UI standard. In OS X, the Home and End keys move the pageview to the start and end of the document, respectively, without moving the insertion point. This is the standard OS X behaviour, and if some program isn't doing that it's because it has been specifically written not to.
I personally dislike this behaviour as well, but it's fairly simple to change and make it like Windows, if you prefer. It's certainly no more significant a change than OS X window widgets being on the left and Windows widgets being on the right.
They really have to get away from the single menu at the top of the page thing as well - it only adds to the nostalgia for Z-80 based single-board computers.
There are numerous UI studies proving it to be a superior placement of the menu bar.
Statistics ?
There's a reason that MS has had to pull so many of their service packs.
Even the most well publicised "service pack breaking stuff" - XP's SP2 - only broke a tiny percentage of machines.
Service packs and hotfixes breaking things is _extremely_ uncommon. It is, however, extremely widely reported on. You should not conflate the two.
In both my experience, and in the experience of many admins responsible for Windows servers, it is often dangerous to apply Windows hotfixes, too. Even with all the precautions of test environments, and staged updates... all too often the patch breaks everything anyway.
Sounds more like your staging process is broken, if you don't find out something doesn't work until you actually get to the production systems.
It can takes weeks to figure out *why* this happened, since MS doesn't exactly tell you what they changed in more than vagueries.
Have you ever bothered to, you know, *ASK* ?
You also demonstrated that different operating systems do things differently.
My point being you were criticising Windows for something that unix does much more of.
I don't see DirectX available on anything other than some versions of Windows. A Windows application doesn't run on Solaris, or MacOS, or anything else. What's your point?
Your implication that there's some "standard unix API" that someone can write to and have it work everywhere is, at best, ignorant and, at worst, deceptive.
Thus, your criticism of Windows for not having the equivalent, was hypocritical.
There is far more standards following and compatability across various UNIX implementations than there is between Windows and *anything* else.
Yes, just enough so people like you say there's nothing to worry about, but still enough so people like me find out there still is a lot of things to worry about.
So sound doesn't always work... so what, it's only sound. Mission critical applications don't even use that. As for UI things... Xlib and POSIX will take you quite far. There are other options, such as all of those the GP mentioned.
That sound you can barely hear is the point flying way, way over your head.
My point about the MS APIs were that they were the "way it will be done" according to MS. Then they were abandoned in favor of new toys to do the same thing.
You say this like its something that happens frequently, like it happens unexpectedly, like it happens suddenly and like it doesn't happen on every other platform, when none of these things are true.
MS suffers horribly from the lack of foresight. It is obvious through all of their APIs and products. They are a marketing company, not a software one, and it shows.
How does it show ? By supporting 20+ year old APIs ? By offering 5 - 10 year transition periods ? By having one of the best records for legacy support in the industry ?
You also seem to assume that UNIX = Linux.
No, I do not. Nor am I ever likely too, seeing how Linux is one of my least favourite unix implementations.
My apologies in my other post, where I have assumed you were the poster I originally replied to.
As I have stated before, if a machine reboot negatively affects your service availability, then your architecture is broken.
You won't believe how amny times some of my less computer-knowledgable friends or clients are afraid to apply a Windows Update patch because they (rightfully) fear that it will break more things than it might fix.
With you "advising" them, I certainly could believe. The fact is, however, Windows patches break things quite rarely. Much less frequently, in my experience, than updates to the average Linux distribution.
(You could believe, if you stopped blindly ignoring every argument against your favorite defective OS.)
I have no "favourite OS". And if you want me to stop poking holes in your arguments, come up with some better ones that involve more facts and less FUD.
You can choose any of QT, GTK+, WXWidgets, Motif, and more. All of them will work, so pick whichever one you like best for whatever reason. For audio, you could use either ALSA or OpenAL for precise control, or Xine or GStreamer if you need multimedia features. Again, all will work. For drivers, I do not have enough knowledge to tell you (not that you want to hear me), but there's a good chance the situation is similar there too.
Thank you for demonstrating why your previous relative criticism about the quantity of APIs for Windows is groundless.
It's better to have a low-leve API that does little but does it so well that it won't need to be obsoleted by a fix 2 years from now.
Which Windows APIs are you thinking of that have been unexpectedly obseleted 2 years after introduction ?
That's why the above-mentioned GUI and audio APIs are so stable as well.
Stable, you say ? How old of a version of QT will support the current release of KDE ? What's the oldest version of KDE you can run with today's QT ?
The Windows APIs are extremely stable. Windows apps written to them 10+ years ago work in todays Windows 2003 systems.
Autoconf and make are an automated system to adjust a generic package to the specifics of a system that is potentially different from the one next to it.
Note that with a "standard" and "stable" "unix API", this should not be necessary.
These differences are not an inherent flaw. They stem from something called "choice", which I'm afraid you may not be able to grasp.
Funny how last post you were criticising Windows for the choice is was offering in "no less than six APIs that I can think of, just off the top of my head".
This is not because having alternative APIs, which all work together, is bad.
No, it's because having dozens of APIs that all do basically the same thing is bad (well, wasteful at the very least).
It's because most IT businesses think in Windows-terms - vendor lock-in and no choice whatsoever.
Most IT business think in terms of economics. And having to support multiple implementations of the same functionality (be it from the developer or end user perspective) costs a lot of money with few reciprical benefits.
You (along with a large proportion of the Linux community, so don't feel too badly about it) seem to be missing one of the main reasons why so little commercial software is produced for Linux - it's because there aren't any standards (or because there are too many, depending on your perspective) and because the APIs tend to be relatively unstable (as they are developed under the assumption that software using them will be open source and thus can just be recompiled when things change).
Lets say Adobe wanted to port Photoshop to Linux. Which package management scheme should they use ? Which distro should they target ? Which GUI should they write for ? Which version ?
With Windows, these questions are easy to answer. Hell, most of them are barely even worth asking, because there's only one answer/
"Unix" systems, as you call them, are not a replacement for Widnows, but an alternative, and consequently do certain things differently. Try this page for a better explanation.
You don't see any calendar functions in MSN messenger, do you?
Actually, if you're running the whole shebang - AD, Exchange, Outlook and an internal MSN server - when you do things that are scheduled in your calendar (eg: go to a meeting, go to lunch) your IM presence is automatically updated to reflect that. It's quite nice.
You have this completely arse-about-face. (Traditional) Unix's permissions system is quite basic and primitive, with no better than per-group granularity and the inherent weakness of an all-powerful superuser. The per-user ACLs in Windows are far, far more fine-grained and flexible, (which thus requires more management effort for non-trivial scenarios).
Windows's permissions model is a superset of (traditional) unix's.
I've yet to see any geeks get hard over the texture of the paper/cardboard their business card is made out of, or whether it using imperial bronze or golden sunrise coloured gilding.
Windows NT = ~13 years old.
OS X = ~5 years old.
They release updates and fixes every MONTH and there are like 15-umpty "versions" of Windows in the wild right now. The MSKB is FULL of "bug reports" and stupid workarounds for things that SHOULD have been fixed long ago.
Rrrrright... Both of which are indicators of a more mature product...
OS X is still undergoing relatively rapid feature development and major changes, with things like API stability and backwards compatibility of secondary concern. Windows NT development, OTOH, is more focussed on refinements and improvements to the existing codebase with things like API stability and backwards compatibility considered much more important.
It hasn't been at all uncommon for OS X's major point releases to break applications - particularly applications delving more into the lower levels of the OS.
OS X development is roughly where Windows NT development was ca. 1998. It's in a transitional period from rapid, signficant changes to more gradual, refined changes. Hence, for example, the reason the OS X kernel API has only officially been deemed "stable" with the latest OS X release.
*That* is why OS X releases have been more frequent, and more significant. To put it bluntly, they've had more things that needed fixing and implementing.
Since when is OS X an OSS project? And even if you can somehow claim it is, Apple is a fairly large commercial enterprise with all those pressures and responsibilities you mention.
The poster I was replying to, in his prior post, commented on the "release early, release often" philosophy commonly seen in OSS projects.
So... Just like very other update to a mature platform, then ?
I'm not as much talking about vendors as technologies. MS comes up with their own versions of things, pushes everyone to use them, and then they drop it from something shinier.
So... They're just like everyone else ?
The non-MS part of the world has been using things called "standards", and they have been doing so far longer than Windows has existed, let alone been used.
Really ? What's the standard API for a "unix" GUI application ? How about using audio devices ? Which API should I use to make sure my hardware driver compiles on Linux, FreeBSD, Solaris and OS X without modifications or special cases ?
We have POSIX as an API standard, and that's been around for a long time.
And is basically useless (not to mention largely ignored) for anything except trivial command line applications.
Heck, it's not at all uncommon to find trivial open source "unix" applications that only work on x86 Linux machines with particular versions of glibc.
Most "cross platform" unix source code doesn't compile on a wide range of platforms because of "standards", it does so because of the amount of work done by things like autoconf and make.
Jeez. One of the biggest hurdles to wider commercial adoption of Linux is the sheer volume of different APIs (many of which all do essentially the same thing), and you're here trying to say there's no such problem at all ?
MS has had no less than six APIs that I can think of, just off the top of my head.
And "unix" has dozens (if not hundreds). Your point ?
They tried to have their own networking protocols, their own email formats, APIs, and on and on.
So... Just like every other commercial vendor ?
They have all been problematic, and largely dropped for the standards that were already there. In that regard, yes, I can think of "vendors" that it doesn't apply to.
Such as ? Certainly not Apple, Novell or IBM. Maybe Sun, but the intersection of markets between Solaris and Windows is vanishingly small.
You act like Microsoft come up with something, then run away from it the first chance they get just to screw everyone over. Yet things like Win32, MFC and DirectX have been around for over a decade, and will *still* be in legacy support 5 years down the track, if not longer. Heck, Vista will still support Win16 on 32-bit x86, an API that's around twenty years old.