[gregpins]

Look, I've drunk the OS X kool-aid. It is my primary OS, for better or worse (mostly better, but there are certainly trade-offs).

But I do wonder -- how fast, stable and feature-rich would OS 9.x be if development had continued in earnest? How would it compare? Certainly less stable, but would that be balanced by better application performance?

[snerdini]

About as good as OS 9.2.

[pdot]

Originally posted by snerdini:
About as good as OS 9.2.

I concur. Maybe even less stable. You'll probably be wanting 9.6 to get the stability of 9.2.

[KaptainKaya]

Would it have gotten any further than 9.2? What possibly could they have included to make it worthwhile to upgrade? I for one am glad to see dev. stopped on it. It was great while it lasted, but it prob. would have killed apple off (too harsh i know) if anything.

[mitchell_pgh]

It would have been slower, have taken more RAM and still wouldn't multi-task worth a crap... and it would have crashed just as much (some say 9 hardly ever crashed, others said it happened every day... either way, it would have been just about the same crash-wise)

OS 9 seems so fast with some things, but so DOS like with others... Looking back, how did we ever get around without real protected memory, multi-user, multi-threading... and OS X makes others take my system more seriously "I toss out Unix based, and people go Oooooo"

[wataru]

It would have been a thorough waste of time and money.

[TheTraveller]

OS 9.5 (or whatever) would only have added more bloat. I don't think we would have seen anything more in the way of performance or stability. I think Apple had gone about as far as was reasonably possible in those departments with the old architecture, IMO.

[clebin]

You'd be able to change the background every 10 seconds if that's what you really want. You might have a toolbar on each Finder window and maybe bigger icons. But all those Unix programs like Apache, ipfw and Samba, which have really enriched OS X, probably wouldn't be there.

In other words, it would be a pretty sad upgrade (like 8.6 -> 9 was in my opinion) Windows XP would have kicked its arse.

Chris

[Millennium]

Hmm.

Well, they might have finally gotten Guard Pages into it (it's been on the schedule ever since 8.6 but it never made it in). That would have made it a bit more stable, though not to a tremendous degree.

That probably would have been the only worthwhile thing. There just wasn't much more that they could do with it.

[cwasko]

We are gathered here today to say that OS9 is dead, dead, dead. I'm not sure they could have built in enough functionality to warrant a 9.5 version. I'm personally very happy overall with OSX - despite the sluggish interface. Alot of the cool things we have with OSX and will get in OSX are due to its UNIXy nature. There is no way 9 could handle the stuff that OSX can do. So, I think it is safe to say that 9 died before OSX was even available. Poor OS9, we loved her dearly. But, out with the old, in with the new. R.I.P.

[Vanquish]

I can't even remember OS9 after almost a year of 100% OSX

[Terri]

I really miss System 9's open/save dialog boxes and lots of other things from it's interface.

[spiky_dog]

Originally posted by Terri:
I really miss System 9's open/save dialog boxes and lots of other things from it's interface.

Default Folder X has been out for a while...

[gregpins]

Originally posted by spiky_dog:

Default Folder X has been out for a while...

I've never liked Default Folder as much as Action Files. Wish it had been developed for OS X. And I think GoMac had the Dock beat in a lot of ways too.

I understand the points everyone has made. I don't use OS 9 anymore either and there is lots that's great about X. But this performance problem is HUGE. The relevant comparison is not between OS 9 and OS X so much as it is between OS X and Windows XP. XP is just so much faster. It's depressing.

[Terri]

Originally posted by spiky_dog:

Default Folder X has been out for a while...

I tried it and it did make things a little better, but there are still lots of problems.

Can't sort by date, can't view in one large window, file names are shortened.

What I want is what I have under 9 with Action files installed.

I really hate collum view. There should be a button that lets you have the same views as in the finder.

[The Dude]

Originally posted by Millennium:
Hmm.

Well, they might have finally gotten Guard Pages into it (it's been on the schedule ever since 8.6 but it never made it in). That would have made it a bit more stable, though not to a tremendous degree.

That probably would have been the only worthwhile thing. There just wasn't much more that they could do with it.

What? Guard Pages? Do elaborate...

[booboo]

Yes those guard pages - an attempt to implement some degree of memory protection - I can't help but feeling that could have been done, but wasn't for 'political' reasons... or maybe I'm being too cynical, perhaps too many 3rd party app's would have been broken, and Apple didn't want developers spending time on continued 9 support...

I think that 9 had gone beyond critical mass, and every new feature seemed to add to instability. I know some people don't really have a problem with crashes in 9, but as someone who tended to run a few app's at once, and then expect to be able to open one more, 9 had become something like a house of cards for me. A lot of OS X app's aren't particularly stable, but at least I'm not having to hear the chord of doom 2 or 3 times a day...

Last great 'classic' Mac OS for me was 8.5.1 - then we got 'bloat-creep...'

[cwasko]

Originally posted by booboo:
Yes those guard pages - an attempt to implement some degree of memory protection - I can't help but feeling that could have been done, but wasn't for 'political' reasons... or maybe I'm being too cynical, perhaps too many 3rd party app's would have been broken, and Apple didn't want developers spending time on continued 9 support...

LOL. Yea, right... Apple concerned with pissing off a developer? roflmao.

I'm sure the descision not to include the GuardPages was purely political. Why would Apple spend anytime on an OS that they had full intentions of killing? Honestly, I don't blame. The only have a certain amount of resources. I'm happy they dumped them all into OSX.

[cwasko]

Originally posted by The Dude:
What? Guard Pages? Do elaborate...

I'm sure this is an over simplification of it and is probably technicaly wrong in most aspects, but I think it is 'close enough'.

Guard Pages, were supposed to be memory buffers put between apps. This way, if an app overwrote its space it would be writing to a space that nothing cares about. There was probably some OS level stuff going on that check the 'buffer' and terminated the app. But, that is only speculation. But, they were essentialy memory buffers around the application.

[SoClose]

Originally posted by The Dude:


What? Guard Pages? Do elaborate...

Guard pages were essentially 4K blocks of memory on either side of the application memory partition. These would help prevent applications writing into someone else's memory space. However, Classic Mac OS memory management isn't very clean and the initial implementations of guard pages didn't make it past the development (d) builds of Mac OS 8.6.

[Anomalous]

I think Apple stopped development on Classic not only because of Mac OS X, but because 9.2 is the "ultimate" version of Classic and there isn't much more they could do to improve it without doing a total rewrite (which is what they're doing now). Honestly, I don't think there were any really revolutionary changes since Mac OS 8.6.

[Graymalkin]

The only way for Classic to have improved itself would be if you rewrote the entire system and essencially made it into Carbon running on a microkernel. It would then cease to be Classic and thus no longer the quicky OS that wouldn't friggin die.

Classic has gained far too much cruft in its long years of service to be a viable candidate for development funding. The interface on the surface might still be superior to X but its internals are seriously messy.

The sort of work Apple touts its systems as being able to do is done very poorly under OS9. Say you're in a networked prepress environment (like the one I work in). You've got your page layout program, your networked pagination system, and your Fetch-like client like Portfolio. You've just finished your page layout and want to print a proof to send off to a customer. You send your page off to your RIP for a composite proof. The more graphics you've got on the page the longer it is going to take to send to the printer. It'd be nice if you could send to the printer in the background but nope OS9 isn't about to let you do that. (at least not with this particular layout app). Then you want to save your multiple megabyte layout onto your network share. You've got to wait for it to save before you can do anything else. Then finally you can check the job in to the pagination system and check out the next job.

Depending on your particular setup (the one I work with isn't the most efficient I could envision) there's a ton of downtime associated with relatively simple tasks that OS9 chokes on because of its inherent design problems. It is a cooperatively multitasking OS hacked to run multiple apps at once. Most of the time you can't interface with anything while another app is doing something even remotely laborious (like a network save over a slow network). That is simply not a design paradigm condusive to effective workflow, especially when you've got people like me that are used to preemptive environments where the OS better divies up processor time. Not that I like using Windows but with NT4 even I can send a large file to a RIP and save all in the background while working in another app. OSX has pretty much the same ability if not better.

OS9 may have had a little more Mac-ish interface but its underlying problems make me glad for OSX despite my problems with it. I can look forward to OSX only improving in speed and ability, all I had to look forward to in 9 was more waiting around while it did stuff.

[Ron Goodman]

OS 9.5 probably wouldn't have had column view, which is far and away my favorite part of the OS X interface.

[mrtew]

I liked OS9 as much as the next guy (actually 8.6 - I never even got OS9 - jumped right to OSX but I can't tell the diff between the two) but now when I boot up in OS9 it feels so primitive that I almost feel like I am working in Windoze. I don't even understand the point of this question. It's like asking, "If you never learned to walk, how well do you think you would be crawling by now?" Pretty well I am sure, but there'd really be no comparison to running.

[Brien]

Originally posted by pdot:


I concur. Maybe even less stable. You'll probably be wanting 9.6 to get the stability of 9.2.

I think they milked that cow for as long as they could. It was time for a much needed new OS. I mean, 6.0, 7.0, 7.5, 8.0 and 8.5 were major upgrades, but who really cared about 7.6, 8.6, or any of the 9.x's? Sure, they were bug fixes - But no major features. System 6 had the multifinder, 8 had platinum, sherlock, and a crap load of other features, 8.5 had 32 bit icons, but 9.x was mainly, at least after 9.0.2, just to keep compatibility with classic. The fact remains that they were adding newer code to outdated code, and left 680XX code in their, which led to some performance issues and took up a lot of space, as well as made the code sloppy.

EDIT - 8.1 should be important, because, as subzerodiesel pointed out, it introduced HFS+.

[lookmark]

I've got a hunch that trying to continue development of the Classic MacOS into a even moderately modern operating system with all the buzzword-compliant features that OS X has would've probably been a quixotic task at best.

But for an interesting, alternate view of the what-ifs check out Dave Every's article here.

The story, as always, is more complicated than black and white.

[Subzero Diesel949]

I think the last worthwhile Classic OS was 8.1, when HFS+ was introduced. It ran lightning fast on my PowerBook 1400. 8.5/8.6 were nice to have upgrades but nothing very compelling (I won't argue against 8.6 being very stable, but only because it fixed 8.5x's issues). It was such a RAM hog too. 9.0 added a lot of bloat on top of 8.6 to make it very unstable. 9.1 and 9.2 don't do much to help other than Classic Environment support. I still get the same frequent crashes when I am booted into 9.2. I don't think there's anything further Apple could do other than play more Jenga with the code. I'm glad to see it on its death bed.

[wataru]

Originally posted by Graymalkin:
a design paradigm condusive to effective workflow

Wow, do you write mission statements for major corporations?

[dru]

Oh no, not an Every fan. UGH.

9.5 wouldn't have been any fun although it might not require me to use SmoothType for decent anti-aliased text and might not lockup hard so often or make me suffer interminable waits while DiskFirstAid checks the Volume's post-crash state.

I do wish X would've been more 9-like on the user experience end. Apple already did so much usability work with Copland (the UI's codename slips my mind offhand) that it's a pity it was Steve'd in favour of the NeXT way.

Really, there's not a huge difference from an abstract view from what X brought to what (post) Gershwin would've had. Carbon on a microkernel. And the (Taligent) CommonPoint as Cocoa (but *truly* crossplatform). We'd probably not have Display PDF in the form of Quartz but something a step beyond the developer rejected QuickDraw GX. Aqua might just be a theme. We'd have clones from IBM, Motorola, Power Computing and others running Mac OS or Linux or even Windows for PPC. We'd be better off on the file management side with better metadata support (not just Labels but file history revision tracking, keywords, etc.) and Viewers (live/saved Finds). And drawers. Full implementation of spring-loaded folders (like 8.5+ not the half-baked Jaguar version). And so on.

X is good but X can be even better but not outright rejecting all that the post-Jobs Macintosh System had developed. I can only hope current hardware will 'live' (usefully) to see this (even if Unsanity or other 3rd parties have to step up to the plate).

*sigh*

[lookmark]

Originally posted by dru:
Oh no, not an Every fan. UGH.

9.5 wouldn't have been any fun although it might not require me to use SmoothType for decent anti-aliased text and might not lockup hard so often or make me suffer interminable waits while DiskFirstAid checks the Volume's post-crash state.

I do wish X would've been more 9-like on the user experience end. Apple already did so much usability work with Copland (the UI's codename slips my mind offhand) that it's a pity it was Steve'd in favour of the NeXT way.

Really, there's not a huge difference from an abstract view from what X brought to what (post) Gershwin would've had. Carbon on a microkernel. And the (Taligent) CommonPoint as Cocoa (but *truly* crossplatform). We'd probably not have Display PDF in the form of Quartz but something a step beyond the developer rejected QuickDraw GX. Aqua might just be a theme. We'd have clones from IBM, Motorola, Power Computing and others running Mac OS or Linux or even Windows for PPC. We'd be better off on the file management side with better metadata support (not just Labels but file history revision tracking, keywords, etc.) and Viewers (live/saved Finds). And drawers. Full implementation of spring-loaded folders (like 8.5+ not the half-baked Jaguar version). And so on.

X is good but X can be even better but not outright rejecting all that the post-Jobs Macintosh System had developed. I can only hope current hardware will 'live' (usefully) to see this (even if Unsanity or other 3rd parties have to step up to the plate).

*sigh*

For someone who dismisses Every out of hand, you sure seem to agree with a lot of the things he says.

I'm far from a fan of everything Dave Every says (his comments on UI, f'r instance, I find particularly uninspired). He's a smart guy, though, and more often than not pretty interesting; the article on Copland and NeXT that I linked to (here) provides a fairly thought-provoking look at whether a "9.5" could have ever been possible.

The comments posted after the article, from several people involved in the Copland/Maxwell project are just as interesting, by the way -- perhaps even more so.

[Avon]

Originally posted by dru:

I do wish X would've been more 9-like on the user experience end. Apple already did so much usability work with Copland (the UI's codename slips my mind offhand) that it's a pity it was Steve'd in favour of the NeXT way.

I love the new interface more than sliced bread. It's the greatest thing in the world! I can't say I miss OS 9 at all. I did for a little while until I understood the directory structure, but that was then and this is now and now is a very happy time.

I thought the dock was so silly at first. Yeah, until I used it. Its simply awesome.

When you compare it to FreeBSD, apple has done a very good job at making it easy to use... example: Users are in the /User directory not the /usr/home directory and hard disks are in the /Volumes directory instead of the /mnt directory and are shown on the finder or the computer.

Very cool....

I can't even imagine having to wait for a file to copy on an AppleTalk network to work on something else. If I can't do 15 things as once while installing another 5, then i wont be happy with the OS. Die Windows.

Good riddance OS 9. I�m glad the decrepit sick old dog has been put out of its misery.

[Millennium]

Originally posted by dru:
Really, there's not a huge difference from an abstract view from what X brought to what (post) Gershwin would've had. Carbon on a microkernel.

Bzzzt!

Copland/Gershwin had no concept of anything remotely like Carbon. This is a common misconception, given that the early Copland screenshots happened to show an interface which would be later made -more or less- into the default theme for OS8 (not entirely, however; Copland used Espy Sans as the system font rather than the hideous Charcoal).

But there was nothing resembling Carbon even planned for Copland. That came entirely out of developer reaction to Rhapsody, which was basically "give us a. upgrade path for our code or we're out of here".

And the (Taligent) CommonPoint as Cocoa (but *truly* crossplatform).

The only thing stopping Cocoa from being cross-platform is that Apple doesn't want to distribute the Yellow Box for Windows, except as part of WebObjects. It is part of WebObjects, however, and some Cocoa apps can actually be compiled to run on Windows if WebObjects is installed.

We'd probably not have Display PDF in the form of Quartz but something a step beyond the developer rejected QuickDraw GX. Aqua might just be a theme.

GX was the planned graphics architecture for Copland and Gershwin. No "one step beyond" or anything like that. Then again, GX was actually quite badass on its own. Shame the developers didn't like it, but this was partly Apple's own fault. It suffered the same fate as OpenDoc: Apple dragged its feet in putting out adequate documentation, and in integrating the technology into its own apps (to say nothing of the Finder).

We'd be better off on the file management side with better metadata support (not just Labels but file history revision tracking, keywords, etc.) and Viewers (live/saved Finds).

I believe we'll be seeing this coming soon enough; likely 10.3. The Search item in Jaguar's Finder toolbar was the first step.

And drawers. Full implementation of spring-loaded folders (like 8.5+ not the half-baked Jaguar version). And so on.

If you ask me, 8.5's Drawers were pretty half-based too. What I'd have liked to see there were Drawers that you could stick to the side of the screen.

What's so half-baked about Jaguar's spring-loading, anyway?

[CharlesS]

Originally posted by Millennium:
What's so half-baked about Jaguar's spring-loading, anyway?

I think he's referring to the absence of the "click-and-a-half" ability. I think that its functionality is more than replaced by the column view, and SLF is there for what it's good at - allowing you to drag files through multiple levels of folders.

[dru]

Originally posted by Millennium:

Bzzzt!

Copland/Gershwin had no concept of anything remotely like Carbon.

Copland ran 7.x apps together, IIRC, not unlike Classic (conceptually mind you, not in that it ran one OS over another for truly pathetic performance). Copland apps would be conceptually akin to Carbon in that it's cleaned up and able to take advantage of the modern underpinnings. I don't think this is a complete falsehood in that Copland-looking technologies were already introduced to developers. QuickDraw GX being one example. OpenDoc being another. Thread and Appearance managers and so on.

But there was nothing resembling Carbon even planned for Copland. That came entirely out of developer reaction to Rhapsody, which was basically "give us an upgrade path for our code or we're out of here".

Developer reaction to "having" to move to Cocoa, sure. I can't believe it would be reasonable to think none of the R&D from Copland with profiling apps as they attempted, failingly, to make certain things reentrant, offered nothing to what became Carbon. Some Carbon things, like use of Appearance Manager clearly came from the road to Copland.

The only thing stopping Cocoa from being cross-platform is that Apple doesn't want to distribute the Yellow Box for Windows, except as part of WebObjects. It is part of WebObjects, however, and some Cocoa apps can actually be compiled to run on Windows if WebObjects is installed.

I'm well aware it's a political not technological thing. The difference being Taligent's tech was *always* to be crossplatform, a reality that was out of Apple's control. It did ship, despite what some think. IBM shipped CommonPoint 1.0 on AIX and a late beta for OS/2. Just like they shipped OpenDoc integrated into OS/2. Word from an Apple engineer at the time was that when they got the CommonPoint code and put it on uKernel performance was horrifying. Mind you, it required (IIRC) over 100MB RAM on AIX which was no trivial requirement in the early/mid 90's.

GX was the planned graphics architecture for Copland and Gershwin. No "one step beyond" or anything like that. Then again, GX was actually quite badass on its own. Shame the developers didn't like it, but this was partly Apple's own fault. It suffered the same fate as OpenDoc: Apple dragged its feet in putting out adequate documentation, and in integrating the technology into its own apps (to say nothing of the Finder).

When I meant "one step beyond" I meant in as much as eventually QDGX would've seen an upgrade. It wasn't bad tech. Part of its problem was early requirements were "high". Apple doomed that entire era's tech with feet dragging of one sort of the other.

I believe we'll be seeing this coming soon enough; likely 10.3. The Search item in Jaguar's Finder toolbar was the first step.

If you ask me, 8.5's Drawers were pretty half-baked too. What I'd have liked to see there were Drawers that you could stick to the side of the screen.

One problem with using the sides is they'll certainly be covered by new windows if they're on the left and on the right they'd interfere with the default location of mounted icons and the trash (Damn Apple for the Dock partly covering desktop icons in the X Finder ). Hard to say if it would've been worthwhile to eventually allow a user option in the System itself or leave that to 3rd parties. Apple wisely believes when it comes to the basic system and useability less is often more in sharp contrast to the kitchen sink approach at Microsoft.

I figure Drawers will come to X only through a 3rd party (like the Drop Drawer's folks). Apple seems to think the Dock is superior and sufficient. I don't agree but I'm just one user.

What's so half-baked about Jaguar's spring-loading, anyway?

There's no click-and-a-half "drilling" support. And the Dock is braindead.

[dru]

Originally posted by lookmark:
For someone who dismisses Every out of hand, you sure seem to agree with a lot of the things he says.

From his bitter MacKido "pieces", I found him worthy of being taken with a very large grain of salt. So as not to hijack the thread into a flame war I'll just let it lay.

The comments posted after the article, from several people involved in the Copland/Maxwell project are just as interesting, by the way -- perhaps even more so.

Yes, the comments are interesting. I'd love to see a book in time which discusses the Copland mess, what "might" have been and the "birth" of Mac OS X, AQUA included.

[dru]

Originally posted by Avon:
I love the new interface more than sliced bread. It's the greatest thing in the world!

There's a certain elegance to AQUA in terms of dialog layout, fonts, sheets and so on that make it *clearly* the top dog. None of that couldn't be part of an "Appearance Manager 2.0" in a post-Copland era.

The stability of X makes it highly desireable vs. 9x. So too does the smooth user interaction vs. the stuttering of 9.x.

The new X Finder (spacially brain dead, drawers gone, etc.) and some other bits of the user experience, OTOH, dismiss perfectly good and well R&D'd concepts along with a nice chunk of muscle memory. I'm glad "Haxies" exist.

[Sven G]

... From a recent MOSR page:

"[...] In a near-future version of Classic (most likely introduced with 10.3 or 10.4, well after Apple has moved all new Macs to OS X-only booting), integrated with a corresponding update to the old OS -- probably 9.3 -- Classic applications will take on an Aqua-ified appearance as well as limited OS X-native features like direct access of some hardware devices through the OS X kernel. OS 9.3 would not be a bootable operating system but would instead be specifically designed to run only under Classic.
Additionally, direct access to Classic's menubar for use of Control Panels and the like would be eliminated in favor of a new OS X Preferences panel that would integrate most if not all necessary settings functions to replace such Classic-specific controls. [...]"

Very interesting, if it is reasonably true.

[york28]

Somehow I doubt that Apple's really going to put that much work into Classic. I'm not terribly up to date on applications, but other than pro audio, what's left to move to OS X?

The only possibility is that they could release a theme for OS9 that is based on Aqua, but it would look like crap because of things being the wrong size and shape and in the wrong places. It would just confuse users anyhow.

Personaly I'd like to see them further develop Aqua than update classic any further. Just let it die.

[mrtaber]

Not good enough. I'm more than willing to put up with tradeoffs to have a modern, stable system.

9 was starting get that overripe, baroque look...too much time stuck on a given step. Too many restarts for me, too much worrying about running out of memory. Too much instability. I'm glad it's gone. I'm fortunate that I don't even have to run Classic now.

I don't miss it.

Mark