Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Will Apple show Linux how to put Unix on the desktop?

By theboz in Technology
Fri Mar 23, 2001 at 08:32:40 AM EST
Tags: Software (all tags)
Software

I was reading this article on MacAddict where David Reynolds presents us with the statement, "Linux promises, Apple delivers."


This article explains some of the failures of linux as on the desktop, and how Apple plans to rectify this with OS X. He makes some good points about the lack of a user-friendly interface for things that should be easy such as adding a hard drive, or changing the settings for the gui. All of us that have tried linux have been frustrated one time or another with some of these issues, and even some people that read kuro5hin are dealing with this now.

Another point that is not mentioned in the article that I have thought about is how this could benefit linux. Linux works very well in a server environment. Many of us have run web servers and databases on linux, with better performance and more ease of use (not to mention cheaper) than alternatives such as Windows NT with MS-SQL and IIS. If Apple is able to successfully make a version of unix for the novice users, this could either free up linux to focus more on becoming a better server operating system, or give the people working on kde and gnome a better example to follow than Microsoft or CDE.

All in all, I think that this is a good thing. Apple is known for creating user friendly operating systems, and in this case they will give a desktop OS that is good for the novice and expert alike.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
Do you plan to try out OS X?
o Yes! As soon as it is released. 32%
o No, I will stick with what I use now. 32%
o I plan to wait and see if it sucks first. 35%

Votes: 108
Results | Other Polls

Related Links
o Kuro5hin
o this article
o MacAddict
o this
o Also by theboz


Display: Sort:
Will Apple show Linux how to put Unix on the desktop? | 53 comments (45 topical, 8 editorial, 0 hidden)
Will Apple show Linux how to be a good company? (3.30 / 10) (#2)
by regeya on Thu Mar 22, 2001 at 05:18:18 PM EST

No, for two reasons:

1. Apple certainly has had a checkered past, haven't they? 2. (here's the part where I get flamed for being overly anal)Linux this, Linux that. Linux is a kernel project; different people put a kernel and other software together to make distributions. Some distributions, such as Mandrake, want to be on the desktop, but there's not some grand, unified effort by "the Linux community" to put Linux on every desktop machine, mo matter how much journalists like Doc Searls would like you to believe that. Now, I have nothing against Doc (well, yeah; LJ has turned into a trade pub complete with marketing nonsense and is therefore boring as fuck now, but other than that, no) but some people really don't care if Linux is on every desktop by next week. If something better comes along, by all means use it, but the insult ("the failures of linux as on the desktop") is, well, a little off base. Patrick Volkerding(sp), for one, doesn't seem to be overly concerned about producing a distro for Joe Sixpack, and more power to Patrick. :-)

[ yokelpunk | kuro5hin diary ]

From my perspective. (4.33 / 3) (#3)
by theboz on Thu Mar 22, 2001 at 05:30:23 PM EST

One of the problems I percieve with linux is that there are groups such as mandrakesoft trying to make linux a desktop OS, while others such as Redhat try to set it up to compete with Windows NT, and others yet try to make linux just a good unix clone. The problem is that eventually, these goals counteract each other. Would we necessarily want to have some USB PnP stuff in the kernel if we were running a web server? Not likely, because servers don't usually have a webcam or joystick connected to them. I think there will need to be different kernels for the different purposes. So far the kernel has been growing to serve both purposes, but it's relatively huge and bloated right now. From my experience working with servers, you want to do anything you can in order to optimize the system to run just a little faster. Now, it is true that we could all sit there and go through the kernel ourselves and set it up exactly how we need for our task, but this becomes a burden the larger it gets.

Also, the extra tools and software available for linux is probably not as user friendly and consistant as what Apple will provide. This isn't a fault of linux, just that the developers of the various applications don't have as much consistancy as someone in a company anal about standards.

As far as the comment about linux failing on the desktop, I think it has lost some steam as a desktop recently. The numbers of new users seem to have died down a bit from what I can tell. People in general aren't as excited about technology as they were last year or a few years preceeding. I think now that the hype of the internet is dying, the majority of people are not as interested in computers as they were.

Stuff.
[ Parent ]

USB on webserver (4.50 / 2) (#13)
by wiredog on Thu Mar 22, 2001 at 09:37:38 PM EST

Would we necessarily want to have some USB PnP stuff in the kernel if we were running a web server?

That's what make clean, make xconfig, make, are for. And, hey,what's so user unfriendly about 'rm -rf *' or 'kill -9 -1'?

Unix is user friendly, it's just choosy about its friends.

The idea of a global village is wrong, it's more like a gazillion pub bars.
Phage
[ Parent ]

Kernel bloat (none / 0) (#28)
by stepson3 on Fri Mar 23, 2001 at 12:19:26 PM EST

So far the kernel has been growing to serve both purposes, but it's relatively huge and bloated right now. From my experience working with servers, you want to do anything you can in order to optimize the system to run just a little faster. Now, it is true that we could all sit there and go through the kernel ourselves and set it up exactly how we need for our task, but this becomes a burden the larger it gets.

The kernel source code? Oh yea, its huge, gets bigger all the time. But thats the price of features ...

Now, I have to disagree with you everywhere else though. Just run a make xconfig, or make menuconfig (or even, horrors, make config!), and you can edit out everything you don't want. Running a web server? Don't include USB support, joystick support, and maybe even HID ... but do compile in ipchains, and khttpd. Running a desktop for your brother? Compile in some DRI, agpgart, and USB .. and forget about ipchains (you did set up a firewall for your brother right?). I'll admit, for roughly the same features, my kernel grew about 2-300k when I went from 2.2.17 to 2.4.1 ... but with that, I got a journaling file system, USB support, a stateful firewall that can do 1:1 NAT mapping ... Pretty worth it, IMHO.



[ Parent ]
Boring? (5.00 / 1) (#18)
by lavaforge on Thu Mar 22, 2001 at 11:57:58 PM EST

and is therefore boring as fuck now
You just need to work on your technique and it'll get a whole lot more interesting.
"In theory, there is no difference between theory and practice. But, in practice, there is." -- Jan L.A. van de Snepscheut
[ Parent ]
Reply to your comments (3.28 / 7) (#6)
by tnt on Thu Mar 22, 2001 at 05:43:06 PM EST

You said:

If Apple is able to successfully make a version of unix for the novice users, this could either free up linux to focus more on becoming a better server operating system....

I don't think many would see things that way. [I hope this doesn't start a little flame war about licenses, open source, or free software again, but] Apple's OSX is not free (while Linux is); and even though the kernel that runs OSX is open source, it is not free (as in freedom) software. Many people want a free (as in freedom) desktop environment, and OSX does not bring this.

Later you said:

If Apple is able to successfully make a version of unix for the novice users, this could... give the people working on kde and gnome a better example to follow than Microsoft or CDE.

This I agree with. Not really because OSX is based on a Unix environment, but because Apple has a history of producing and developing really great User Interface concept, ideas, & technologies.



--
     Charles Iliya Krempeaux, B.Sc.
__________________________________________________
  Kuro5hin user #279

Splitting hairs. (4.25 / 4) (#8)
by harb on Thu Mar 22, 2001 at 06:00:09 PM EST

There are two parts of Mac OSX and Mac OSX Server. The underlying technology is called Darwin, which is basically a FreeBSD hack. It's a true UNIX system.

Quartz (which Aqua is but the 'feel' of) sits on top of that. It's the GUI.

Now, the Darwin team has worked very closely, and openly, with the fBSD team.. if that continues, I don't see any reasons for all the good aspects of FreeBSD, or really, any BSD, to suddenly have a GUI interface.

And Darwin is free.

But yes, as you mention, Mac OSX is rather expensive. And, at this point, only runs on Apple-happy hardware (which sucks!).

With things like Linux and BSD, you get freedom of enviroment, fees, and licensing.

Ups and downs. *shrugs*

harb.

bda.
[ Parent ]

Close, but not quite (4.00 / 1) (#14)
by Rhamadanth on Thu Mar 22, 2001 at 10:03:42 PM EST

>and even though the kernel that runs OSX is open >source, it is not free (as in freedom) software. >Many people want a free (as in freedom) desktop >environment, and OSX does not bring this.

You're part way there, but you're mostly wrong.
Most people don't give 2 cents about how free an operating system or desktop is. The majority of users in the world are on a Windows box. When you say 'many', you're right, but the word 'people' should not be misused when what you really mean is 'geek'.

I think that OSX stands a chance with 'ordinary' users, *and* geeks. It has the interface that many users like and want, and it has the power that many geeks can harness. Apple is shipping *all* the developer tools with it, too. When I was growing up, the NeXT developer tools cost my Dad an effing arm and a leg. Free? For tools of that calibre? A bargain. And if you want to write free software (free as in freedom), you can do that, too. OS X has something for everybody. And this time it looks like it has enough to make both sides (users and geeks) happy, as opposed to being watered down all around, and making both sides sad.

Let me add that Steve Jobs is a nut. He's completely off his rocker. But he's done a good job getting rich, and he knows how to make money. Pixar still turns a fair profit from what I hear, and Apple is doing no worse in this 'economic slowdown' than anyone else.

Ah, one last thing, while it's in my head. Yes, people will buy Macs with OSX despite the cost. Remember the market that it's primarily marketed at. They'll buy an iMac, because its no more expensive than a Dell or a Compaq. Never mind that I could build them a cheaper intel/AMD system if I did it piecemeal.


-- The /bin/truth is out there.
[ Parent ]
Re: Close, but not quite (4.00 / 1) (#25)
by tnt on Fri Mar 23, 2001 at 11:15:59 AM EST

In reply to when I said

and even though the kernel that runs OSX is open source, it is not free (as in freedom) software. Many people want a free (as in freedom) desktop environment, and OSX does not bring this.

you said:

You're part way there, but you're mostly wrong. Most people don't give 2 cents about how free an operating system or desktop is. The majority of users in the world are on a Windows box. When you say 'many', you're right, but the word 'people' should not be misused when what you really mean is 'geek'.

You are correct, most people "don't give 2 cents about how free an operating system or desktop is". But I never said they did!!! As I've said before here, most people use Windows because it came with their computer. To them, it was free because it came with their computer (... they don't see or maybe care about the hidden costs). To them, the free, as in freedom, doesn't matter; to them, the free, as in free cake, matters.

And I do know that the word many is a subjective term; and the exact definition of many, when I use it, is defined by me (... and will still likely be different depending on the context in which I use it). Now, more than likely, this definition of many will be different from yours. But I don't see why you are nit-picking over that... I assumed everyone would see that as being pretty obvious.

Now, just to make things clear (as to what I mean by many) I do not mean just amoung the population of geeks*. I've encountered a concern about free (as in freedom) in other places too... namely I encountered it amoung the indie movie culture. This seems to be spured by their use of the GIMP. (I was actually surprised the first time I heared someone, from the indie movie culture, talk to me about the GIMP [... usually first asking me if I've ever heard about it...], and even talking to me about Linux sometimes, after they find out that I have a Comp Sci degree.)

You later said:

I think that OSX stands a chance with 'ordinary' users, *and* geeks. It has the interface that many users like and want, and it has the power that many geeks can harness. Apple is shipping *all* the developer tools with it, too. When I was growing up, the NeXT developer tools cost my Dad an effing arm and a leg. Free? For tools of that calibre? A bargain. And if you want to write free software (free as in freedom), you can do that, too. OS X has something for everybody. And this time it looks like it has enough to make both sides (users and geeks) happy, as opposed to being watered down all around, and making both sides sad.

Agreed, OSX does stand a good chance with 'ordinary' users, *and* (as you call them) geeks. I for one have been interested in HCI, useability, and UI design/research so I have an interest in it, from that point of view too. (Not to mention that I think it looks cool.) Also, about the writing of free (as in freedom) software for OSX, I've been keeping that in mind -- other platforms besides Linux, like: OSX, Windows, etc -- with the free (as in freedom) I (and others) are working on. I have alot of (artist) friends that work exclusively on the Mac, and they have an interest in seeing free (as in freedom) software on it too. [For them, they don't see moving over to Linux as very free, when all the programs they already use, and already payed for are on Mac... or Windows.]

* This is a little of topic but I've always had a problem with this (relatively new) label of geek, being attributed to anyone with a Comp Sci background, or who can code. They define a geek as someone who is not-athletic, and who is socially awkward... and then for some reason attach this label to everyone with a Comp Sci background, or who can code.

I am extremely athletic (being a pro-athlete was one of the options I looked at when going to University) and have no problems in a social environment; yet because I'm educated in Comp Sci some people try to attach that label.



--
     Charles Iliya Krempeaux, B.Sc.
__________________________________________________
  Kuro5hin user #279

[ Parent ]
Geeks 'n stuff (4.00 / 1) (#30)
by Rhamadanth on Fri Mar 23, 2001 at 04:10:15 PM EST

This is a little of topic but I've always had a problem with this (relatively new) label of geek, being attributed to anyone with a Comp Sci background, or who can code. They define a geek as someone who is not-athletic, and who is socially awkward... and then for some reason attach this label to everyone with a Comp Sci background, or who can code.
I am extremely athletic (being a pro-athlete was one of the options I looked at when going to University) and have no problems in a social environment; yet because I'm educated in Comp Sci some people try to attach that label.
This is a little far afield, but that's okay. :) Anyway, the term 'geek' is rightfully applied to anyone that's excited and commited to their subject of study. I know engineering geeks, I know CS geeks, I know math geeks...I even know a nursing geek. Being socially well adjusted has nothing to do with it. I race mountain bikes, and used to be a competitive swimmer. I have a girlfriend, and I can converse at parties about things other than computers. I'm still very much a geek, though. I enjoy computers, computing science, graph theory, and I can hack through C source code with the best of them. Being a geek is a badge of honor that can even be bestowed to the people that used to be excluded, that don't have social skills and aren't all the things that normally make you 'popular'. Being a geek is an equal opportunity field.

By the same token, I know a lot of people in CS and Engineering that do it, but aren't geeks at all. 'Geek' isn't a social condition, it's a mental one. And you can take that any way that you like. :)

Anyway, the point is that I was unclear, myself. The term 'geek' in this case was short form for 'the subset of geeks that are interested in computing'. I was being lazy....but I've never had problems with being lazy before. :)

-- The /bin/truth is out there.
[ Parent ]
What we need is not Linux or OSX (none / 0) (#51)
by cable on Mon Mar 26, 2001 at 04:32:58 PM EST

But something with the freedom of Linux and the ease of use and install of OSX. OSX has that BSD License, Linux has the GNU License. Not the same.

You want OSX On Intel? File a petition with the rest of them to get it on the Intel platform. That is if Apple isn't so paranoid that it will kill their hardware sales. Why buy a Mac to run OSX if it runs on almost any generic WINTEL box? OSX so far, is limited to Macs. Any plans to at least support CHRP/POP PowerPC boxes? Or do the systems have to have that Apple logo on them to run? I heard that the Mac Clones, or at least some Motorola Starmax clones won't run OSX. So did Apple cut out the Mac Clones as well?

One step forward, and two steps back. That is what I feel on OSX. If Apple can qaush the bugs, provide better third party hardware support, get in laptop and CDRW/CDR features, and avoid the Unix gotchas that other Unix variants suffer from, then more power to them. If not, better keep that copy of MacOS 9.1 or earlier handy just in case OSX doesn't work out for you. OSX might become a "Not ready for prime time" OS like BeOS, OS/2, Desqview, and many others (Nextstep/NextOS anyone? A/UX? Unixware? SCO Unix/Xenix? Minix?) the only thing that might save it is the ability to emulate OS 9.1 systems and run the GNU utilities and POSIX programs. Let's hope that ARDI or someone else doesn't write a PowerMac emulator for Unix or something that can also run OS 9.1 stuff?



------------------
Only you, can help prevent Neb Rage!
[ Parent ]

I have often felt (3.80 / 5) (#7)
by weirdling on Thu Mar 22, 2001 at 05:59:58 PM EST

If any *nix could become as easy to use as MacOS 9, I'd be very impressed. I'm frankly worried about MacOS X, because it is a *much* more complex operating system base, and that complexity is bound to show up in the UI sooner or later. Much of the features found in traditional MacOS are actually features of the underlying OS and file system, both of which are quite unique, as anyone who has ever had to code for a Mac will testify.
Anyway, all that being said, I'll give X a try, and if it doesn't work, I'll go back to using Yellow Dog to run OS9 in a VM, which gives me the best of both worlds...

I'm not doing this again; last time no one believed it.
Let's not claim defeat yet (3.33 / 6) (#9)
by RangerBob on Thu Mar 22, 2001 at 08:45:03 PM EST

There's a difference between Linux and Apple. Linux is still growing in popularity, it's getting easier to use as distributions mature. Redhat 7 actually set X up correctly for my monitor and card for the first time. Previous versions didn't bother or didn't come close. That's an improvement for me. Heck, KDE 2 with antialiasing turned on is actually a pretty good GUI IMHO. My wife can use it, and that's one of my test cases if something makes sense to a non-geek.

Apple was at the top and lost it. They lost it because of screwy management and the IBM syndrome. Plus, I still think Jobs is eventually the kiss of death for anything he's involved in. Making ATI "pay" is still a sure sign of the good old arrogance. Yeah, he does some things right, but everything he's touched has eventually folded or come close to folding. The only real difference between people like Gates, Ellison, and Jobs is that Gates is more efficient in screwing people and companies over.

Plus, not all of us go nuts over the Mac interface and think that the KDE/Gnome/CDE/Windows interfaces are evil incarnate. No, they're not perfect, but it's not like people loose sleep and billions of dollars in lost productivity a year over them.

I also disagree with the freeing up bit. For one, the kernel developers will have the same jobs be it for a server or a desktop. They want it to perform as best as it can reguardless of what it's on or used for. Different kernel developers are concerned about different things. That's the beauty of being able to compile your own kernel. If you want it to be a server, add some things in. If you prefer desktop, add different things. The KDE people have their own concerns, the GNOME folks theirs. I don't think there'd be one bit of freeing up going on because open source developers are always going to carry their labors of love.

The best thing (3.50 / 2) (#10)
by spacejack on Thu Mar 22, 2001 at 08:54:03 PM EST

Is that you will finally be able to run a lot of mainstream apps missing from *nix, either new versions or via Carbon. The reason *nix isn't practical for me is because I can't do anything with it.

I for one am very curious about OS X. I'm gonna try it as soon as I can and if I like it, I'll be gettin a Mac for the first time in 6 years.

OS X (3.00 / 4) (#12)
by YelM3 on Thu Mar 22, 2001 at 09:35:28 PM EST

I rated this to the front page because I am very curious as to what people's opinions are about the potential of OSX.

Licensing issues aside, it seems to me that if Apple does this right, it could make a huge impact on Free Software and UNIX in general. Remember, even if the core is not Free the apps that run on it can be.

And how about OSX for Intel? Apple releases OSX/Intel -> Microsoft has some serious OS competition -> Hundreds of thousands of users try OSX -> UNIX users and exposure to UNIX in general grow exponentially. Could happen.

OS (Linu)X ? (2.66 / 3) (#15)
by korc on Thu Mar 22, 2001 at 11:23:22 PM EST

Any kernel hackers out there want to let me know how complicated the relationship between OS X's kernel and upper layers is, and whether shoving Mach 3.0 out of the way is a possibility? How numerous are the hooks?

Ignorantly yours,
korc
Why? (none / 0) (#31)
by nstenz on Fri Mar 23, 2001 at 04:39:10 PM EST

I'm just wondering why you'd want to ditch the Mach kernel. Do you want to swap it out for the Linux 2.4 kernel, or what? I suppose Linux is more popular and has better hardware support at the moment... Yes/no?

[ Parent ]
Cocoa and Mach (none / 0) (#34)
by sinclair on Fri Mar 23, 2001 at 05:56:18 PM EST

Perhaps this link to Apple's Cocoa documentation answers your question? (Look at the grey box, if you have a graphical browser.)

[ Parent ]

I'm not a kernel hacker... (none / 0) (#40)
by DJBongHit on Sat Mar 24, 2001 at 12:12:56 AM EST

... but I play one on Kuro5hin.

Any kernel hackers out there want to let me know how complicated the relationship between OS X's kernel and upper layers is, and whether shoving Mach 3.0 out of the way is a possibility? How numerous are the hooks?

It's not really necessary - while generally Mach-based kernels have performance issues, Apple put a lot of work into making this not the case with Darwin. I'm not sure the specifics, but they did a lot of optimization work on it.

~DJBongHit

--
GNU GPL: Free as in herpes.

[ Parent ]
does unix belong on the desktop? (2.75 / 8) (#16)
by rebelcool on Thu Mar 22, 2001 at 11:35:28 PM EST

Since when is it a good idea that the mass consumer (read: technically incompetant) use a 25 year old design that was intended to run massive client-server systems? Seems rather like giving every driver on the road an 18 wheeler to drive around.

The vast majority simply don't need it. They won't benefit from it. Companies should spend their time thinking up better interfaces, if they're aiming at the mass market.

COG. Build your own community. Free, easy, powerful. Demo site

Because everything is Unix (4.75 / 4) (#20)
by Miniluv on Fri Mar 23, 2001 at 03:19:37 AM EST

In case you weren't aware, Unix is the dominant model in operating system design these days. The only major contender that's not Unix themed is Windows, and even that is becoming more and more unix-like. As far as giving the users "too much" operating system? That's bullshit and everybody knows it. A well designed interface, with ease of use built in by design, can be easily incorporated into a unix environment. Ask the KDE developers how it's done, because they've obviously been taking lessons.

OSX appears to be the first Unix that sucessfully cloaks its Unixish features in a cloud of useability. The worst thing they can do though is lock the CLI away and draw a line in the sand beyond which their users cannot grow. The flexibility provided by the abstraction inherent in a Unix architecture means that upgrade paths will be easy in OSX, especially for power users.

I'll also point out that Apple picked the most user friendly of the free unixes to port to their hardware, and they're being awfully good sports about sharing some code with the community. Darwin is awful close to being useable on the x86, and maybe we can pressure/encourage Apple into releasing a cheap-ware version of their GUI for Darwin users?

The best thing OSX is doing for the Unix world is showing us that it can be done. SGI showed the world once that you can make a nice graphical computing environment built around Unix with IRIX and Iris, and now Apple is reminding us all that it doesn't have to be painful. Rather than focusing on obscure features first, Apple has made useability priority 1. Rather than coding and recoding the kernel Apple brought it to an acceptable level of stability and shifted gears towards promoting other goals. In essence, they ran their developement team like they intended to ship a product, and they tailored their design goals towards their users, instead of their coders. These are all things the open source movement lacks more often than has, and that's why Linux still isn't ready for the desktop.

"Its like someone opened my mouth and stuck a fistful of herbs in it." - Tamio Kageyama, Iron Chef 'Battle Eggplant'
[ Parent ]

looking beyond (none / 0) (#27)
by rebelcool on Fri Mar 23, 2001 at 11:58:16 AM EST

I think mass-consumption computers could be made *alot* easier. The current crop of OS's are simply ridiculous. it should be easy enough that someone's mother could take the computer home, take it out of the box, plug it in, turn it on and be able to do something productive with it immediately.

Try doing that with someone who has little or no computer experience.

COG. Build your own community. Free, easy, powerful. Demo site
[ Parent ]

either that (none / 0) (#33)
by spacejack on Fri Mar 23, 2001 at 05:35:38 PM EST

or casual users won't need PCs. Maybe cel phones, consoles, mp3 players, DVDs and PDAs will eventually cover most people's needs. Maybe eventually, the only people using computers will be "power users", whether they're animators, programmers, artists, musicians, etc. (that's a lot of people to be sure, but it won't be everyone).

There may be a risk in dumbing the interface down so much to accomodate the masses that it ceases to be useful to a power user. Right now, *nix is only useful to a small subset of power users. Apple may strike the balance just right for everyone except those very casual home users.

[ Parent ]
absolutely right. (none / 0) (#35)
by rebelcool on Fri Mar 23, 2001 at 07:17:05 PM EST

the majority of folks out there, simply dont need a full computer they sell today.

Off the top of my head, here's the most useful programs needed to date:

Word processing, for writing letters. Granted, it should be a simplified word processor...grandma doesnt need vbscript, macros and all that crap in word.

Internet browser.

E-mail.

A few other minor applications, like calculator and so on.

Do away with the mouse completely..it confuses many people. Just have a touch screen. The only thing a keyboard should be used for is typing, of course. No silly shortcut keys.

COG. Build your own community. Free, easy, powerful. Demo site
[ Parent ]

magilla (none / 0) (#36)
by cpt kangarooski on Fri Mar 23, 2001 at 07:37:05 PM EST

touch screens have been tried before. the ergonomics are truly awful - the description of what it feels like to use them after a while is 'gorilla arms.'

a combination of keyboard, voice recognition, and tablet (with handwriting recognition -- loopy bits, spiky bits, yep, that's handwriting ;) with the ability to add a mouse or joystick or whatever seems like it might work out.

but it's so far out there that we can safely can it, and get back to trying to at least design a good keyboard/mouse ui.

--
All my posts including this one are in the public domain. I am a lawyer. I am not your lawyer, and this is not legal advice.
[ Parent ]
sexy devices (none / 0) (#37)
by spacejack on Fri Mar 23, 2001 at 08:34:24 PM EST

I saw this really sexy looking cel/pda combo from Nokia today in a local weekly rag. Let's see if they've got a pic online... ... yup here it is.

[ Parent ]
not if the interface is done properly (none / 0) (#38)
by rebelcool on Fri Mar 23, 2001 at 09:52:51 PM EST

you dont get gorilla arms by toggling buttons and things in your car do you? Of course not. Because you dont toggle things that often. With a well designed UI, you can use it for as little as possible.

COG. Build your own community. Free, easy, powerful. Demo site
[ Parent ]

false dichotomy (none / 0) (#43)
by crank42 on Sat Mar 24, 2001 at 03:22:42 PM EST

In case you weren't aware, Unix is the dominant model in operating system design these days. The only major contender that's not Unix themed is Windows, and even that is becoming more and more unix-like.

In other words, nobody's done commercial work for many years on anything other than UNIX or DOS (MacOS being, since System 7, a series of bags hanging off the side of System 7), so you have to pick either UNIX or DOS. What nonsense.

It's too bad, for instance, that BeOS couldn't get the Mac heads to buy in at the beginning (or sell out to Apple): it actually had all the consumer-oriented stuff Apple is only just getting around to providing, slowly, through the endless series of betas of OS X. My money's still on Apple missing their deadline. Be, of course, was building something utterly unlike either DOS or UNIX in conception. I can think of other examples, like PalmOS. There are lots of alternatives to the UNIX/DOS dichotomy, and many of them are much better choices than either UNIX or DOS for future consumer-target systems. Linux is probably going to win in the embedded market, because it's free and flexible; that doesn't mean development needs to go that way, or even that Linux is a technically better choice.

SGI showed the world once that you can make a nice graphical computing environment built around Unix with IRIX and Iris, and now Apple is reminding us all that it doesn't have to be painful.

SGI did that by creating more suid root programs than the entire rest of the world combined (I exaggerate, but not by much). I think this is a good example, actually, of why UNIX is probably too much horsepower for most consumers: it's easy for wizards to make it do tricky things, and most consumers don't know how to prevent that. The result is a great whack of needless vulnerabilities for systems which didn't need the potential to be so vulnerable in the first place. Let's hope that Apple doesn't end up in the awkward position SGI once was, or that MS is now in: security vulnerability of the week. The potential for OS X to turn into an excellent cracker target, though, is pretty high. Me, I'd spend a long time examining OS X before I started encouraging people to use it on my network.

[ Parent ]
Setuid and inherent danger (none / 0) (#45)
by Miniluv on Sat Mar 24, 2001 at 10:29:49 PM EST

The problem with setuid isn't that it gives root privileges, it is actually that the Unix world has put all their eggs in one basket when it comes to the root account. Capabilities and mandatory access controls would alleviate this problem to a great extent.

Yes, they add complexity. Yes, they're hard to configure in their present form. To me that doesn't mean we shouldn't use them, but instead that we need to start putting resources on the problem of making them more flexible, easier to use and more trustworthy. The OS isn't there to protect the user from their mistakes, that's what good documentation is supposed to do.

SGI made the mistake of assuming their users were more technical than they were, and thus not providing a safer default setup and documentation on how to keep their setup safe. Security is one of those things where everybody passes the buck, especially when it comes to education. Nobody wants to write the documentation to walk an inexperienced user through proper setup of a system with setuid root binaries. Then again, with a good capabilities based access control system you wouldn't need setuid root anymore, because you could grant access to the extremely specific function that the binary needs.

Here's an example of a stupid setuid situation that capabilities could help with:

KDE2 screensavers allow you to set a requirement of supplying the password for that account before removing the screensaver. To do this they need access to the /etc/shadow file. This file traditionally is read-only root, with no group or world permissions. Their solution, the only one available in traditional Unix, is to make the appropriate binaries setuid, so they can read /etc/shadow.
The KDE team figured out the only solution available to them to get the feature users want included. This is a failing of the Unix model in that no one is willing or able to extend it. This doesn't mean the Unix model is flawed, but instead that it is time for people to realize that part of the Unix philosophy allows for adaptability. The overall Unix model of bitstreams, everything is a file, and small programs doing one job well is a good one. It works in an extraordinarily large number of scenarios, but there's more to the OS picture than just those three things. We can evolve Unix without losing the original strengths.

"Its like someone opened my mouth and stuck a fistful of herbs in it." - Tamio Kageyama, Iron Chef 'Battle Eggplant'
[ Parent ]
I couldn't agree more (none / 0) (#47)
by crank42 on Sun Mar 25, 2001 at 01:23:39 PM EST

The overall Unix model of bitstreams, everything is a file, and small programs doing one job well is a good one. It works in an extraordinarily large number of scenarios, but there's more to the OS picture than just those three things. We can evolve Unix without losing the original strengths.

This is one of those observations that is so exactly right that, I'm afraid, no-one will accept it. A couple of UNIX implementations have included ACLs, of course, and they are somewhat more secure for it. But they're not perfect, partly because so much of the legacy software knows nothing about them.

That said, ACLs and other such mechanisms are hardly a magic bullet. Win NT includes what are, in my view, pretty good capabilities restrctions and extremely flexible access controls. Yet NT + IIS has become a favourite, easily-exploitable target. That is at least partly because of MS's desire to make the platform easy to use. People who haven't bothered to read the documentation, or to set up the system correctly, are left vulnerable. Even though the system could be more secure, it usually isn't. The same thing happened with SGI, and I worry (as I did in my original argument, above) that Apple will go down the same road, for largely the same reasons. Apple users are, if anything, even more hostile to careful attention to the documentation than are NT admins. If Apple doesn't have two or three security wizards auditing their products (or, better still, thousands -- but they already decided to go the proprietary route), they run a serious risk of opening the same kinds of holes we keep seeing elsewhere. Poweful systems lend themselves to powerful exploits, and Apple's UNIX base will be no exception.

[ Parent ]

Unix WAS intended to be a desktop OS (none / 0) (#41)
by dabadab on Sat Mar 24, 2001 at 10:39:48 AM EST

"a 25 year old design that was intended to run massive client-server systems?"

And to run on mainframes, eh?
When will you people get your facts right? Unix was developed as a "desktop OS", because that was what the guys at Bell Labs needed for themselves.
--
Real life is overrated.
[ Parent ]
How Mom benefits from Unix (none / 0) (#49)
by scruffyMark on Sun Mar 25, 2001 at 08:45:01 PM EST

Regardless of it original purpose, Unix was designed well enough to run massive client-server systems, and still be responsive under heavy load. These benefits carry over:
  • Robustness - When a buggy app crashes, the whole operating system doesn't die with it.
  • Multiple users - When Dad changes the preferences for his e-mail client, the changes don't mess up Mom's carefully tweaked setup. When Mom empties the trash, she doesn't trash files Dad wasn't quite sure he wanted to get rid of yet.
  • Powerful servers - Mom wants to make a simple web page for family snapshots. All she has to to is click the "web server" button in the Internet control panel, and she has Apache running (not some lame server like IIS/WebStar)
  • Ease of development - Mom probably isn't going to write a lot of software, but the fact that it is miles easier to write and debug software means that she gets a better selection of better-written programs to choose from
That was just off the top of my head, there's probably lots more benefits.

[ Parent ]
It might, but... (3.50 / 4) (#17)
by Mr. Piccolo on Thu Mar 22, 2001 at 11:53:19 PM EST

There's a huge obstacle in their path.

That obstacle is that MacOS X only runs on Apple computers.

I'd love to try out MacOS X, but I, like 90 percent of computer users, have an Intel-compatible computer. In other words, to even try it I have to buy a new computer, and even Microsoft's operating systems don't cost that much for one copy. (I don't consider playing around with one at Sears trying it, BTW.)

So, to successfully put U**X on the desktop, Apple has to get people to buy their computers. They really are trying -- seen those commercials lately? Will it succeed? Kind of depends on how many new and repeat buyers they can get, as I doubt many current WIntel users will want to replace all their current applications to make the switch.

The BBC would like to apologise for the following comment.


Sure (2.00 / 1) (#24)
by Bob Abooey on Fri Mar 23, 2001 at 10:25:21 AM EST

Apple is still a hardware company. They need the OS to help them sell the hardware. They are similar to Sun with Solaris. Solaris on X86 is pretty lame and Sun gives it very little attention, because they make their money by selling sparc hardware. But sun needs solaris for the sparcs to run like a screaming train so they can sell tons of big dollar sparc servers.

I hope OSX is a huge success. I might even buy one (a Mac with OSX) at some point down the road. I do have to say that I think it's silly to make such a big deal about the fact that there is a *nix based kernel under the hood though. Why should this be such a big deal. The only reason we have never had a killer desktop for *nix is because nobody built one, not because it's impossible.


-------
Comments on politics from a man whose life seems to revolve around his lunch menu just do not hold weight. - Casioitan
[ Parent ]
You're wrong about Solaris/x86 (4.00 / 1) (#32)
by Mr. Piccolo on Fri Mar 23, 2001 at 05:00:32 PM EST

On what grounds would you call Solaris/x86 "lame"? It seemed to work pretty well for me, even on a pretty minimal configuration. In fact, I'd probably still be using it if I hadn't bought the wrong network card. :-P Sure, there's no games, but that's what dual-booting into Windows is for.

I think the reason Solaris/x86 gets a bad rap is that AFAIK it doesn't handle IDE drives well, running them in PIO mode. Not a good thing, but use SCSI and performance should be significantly better. It also tends to use a lot of RAM, but now that 256MB is standard, that shouldn't be a problem (minimum is 64MB).

Plus, if Sun didnt't care about it, why do they bother putting out patches?

About hoping OS X succeeds: I agree. I really want to give Mac OS X a try myself. The problem is having to buy a Mac to do it.

I'll be frank: I'm not sure why putting *nix on the desktop is a big deal, except if it happens to be your favorite operating system.

I do know that what Joe Sixpack wants is a computer that is as easy or easier to use as Windows, and it would be nice if it didn't crash and you didn't have to reboot every time you install a program. Using *nix as the kernel and building a decent desktop on top of it is just one way to do it, though probably a very good way.

The BBC would like to apologise for the following comment.


[ Parent ]
What value? (none / 0) (#48)
by chill633 on Sun Mar 25, 2001 at 07:03:14 PM EST

On the grounds that there is almost no 3rd party apps available for it, outsite of GNU and Sun. Solaris is used to run the big business apps -- with things like SAP, Oracle, etc. None of these run on Solaris X86 (but they do on Linux).

Solaris/x86 sole purpose in life is to provide a training ground for people to learn/develop Solaris but can't spend a wad on SPARC boxes. It is also a hedge against the day SPARCs dwindle into obscurity.

[ Parent ]
Please... It's Apple we are talking about... (3.66 / 3) (#26)
by sergio on Fri Mar 23, 2001 at 11:27:01 AM EST

Dear All,

The day that apple produces an OS that can run in the comodity hardware that we are all used to (read: cheap Intel/AMD boxes) we should tehn compare. They had a chance and didn't make it there in '95.

I used to have a NeXT in '93. Nice machine... crazy marketing. The same thing with Apple. On some NeXT boxes I used to have to mount the filesystem by hand to clean up the dynamic swap file as it grew too large and it wouldn't boot anymore... tons of problems like that!

At home I support my wife's Mac: It is a nightmare when things go wrong. Not to mention that you have to buy specialized drivers to make it play nice with a vanilla ULTRA2 scsi disk... Wouldn't expect less from the fine folks at Apple.

The bottom line: comparisons with Linux are really not very useful in this case. Linux will get to the desktop, slowly but surely!

Apple issues (none / 0) (#46)
by dvNull on Sun Mar 25, 2001 at 05:07:51 AM EST

Actually if you buy a Mac and dont get the 249.99 Applecare support package you are not entitled to any phone support (read: even RMA numbers) after 90 days unless you are willing to keep calling higher and higher up the tech ladder.

One of my colleagues had a Mac which kinda died and it was pretty obviously a hardware problem, took me 2 hours and 4 people on tech support before I was able to get a RMA # so I can get it replaced. I refused to pay $50 for tech call and $100 for local care center diagnostics for a machine which wouldnt even power up. Especially when this happened 133 days after purchase.

Apple has to change its support techniques since its starting to piss more and more people off


If you can see this, then the .sig fell off.
[ Parent ]
The reference to Reynolds (4.00 / 3) (#29)
by j on Fri Mar 23, 2001 at 03:58:17 PM EST

All in all an interesting submission, but I have one problem with it. Your claim that
This article explains some of the failures of linux as on the desktop, and how Apple plans to rectify this with OS X
is rather inaccurate. IMnsHO, he doesn't explain a thing. He just makes some broad statements without really being able to substantiate any of them. Reynolds just repeats three of the most common accusations against Linux (disk management, printing and X Windows configuration) without going into any detail.
Likewise, he just claims that Apple will do better in that you you simply run the Mac OS X installer--no formatting required. No details here, either. What it boils down to is: Reynolds has seen Mac OS X and he likes it. He is not currently able to give us any conclusive information on how it is better than Linux, but we'll believe him that it is. OK, works for me.
As for his allegations about Linux, I'll just say this:
  • Adding a hard drive to Linux is trivial; integrating it into your file system structure can be tricky. Luckily, distributions like RedHat do the partitioning and formatting automatically during the installation process.
  • Whoever complains about printing under Linux obviously hasn't tried CUPS yet.
  • I hear that setting up X can be a painful experience. My distribution (Debian) got XFree86 4.0.1 up and running with minimal intervention, so I'm not qualified to comment on this.

As to the potential success of Mac OS X: Who knows. I have a feeling that it is going to end up in a niche market. Too dumbed down to be interesting for real nerds, not quite intuitive enough to win over many Windows users. But we'll see.

My OS X experiences (4.00 / 3) (#39)
by DJBongHit on Fri Mar 23, 2001 at 11:35:00 PM EST

My roommate (a HUGE mac fun) just bought a titanium G4, so I've been playing around with his old PowerBook G3. I installed OS X beta on it and I've been using it on and off over the past month.

First of all, the interface: it's nice, but it's not for power users, by any stretch of the imagination. It feels less like Unix than, say, Win2k with Cygwin installed. It's not terribly customizable (still haven't found a way to make it do sloppy focus, which I can't live without). I found an application (Space) which makes it use virtual desktops, but there's no way to set up hotkeys to switch between them, and it's relatively buggy - if you switch too fast between desktops, windows will sometimes follow you. I much prefer WindowMaker or, lately, KDE 2, which is phenomenal.

Second of all, the underlying Unix: it's a very nice Unix (BSD based, built on Mach) with all the nice utilities preinstalled - OpenSSH, Perl, wget, ncftp, etc, etc.... I was very surprised that I didn't have to spend all day downloading and building the utilities I use, because I have to do this for other free Unices I install (Mandrake being the one exception here). And you can do most system configuration stuff from the command line if you wish - for example, instead of going through Apple's control panel to turn on the FTP server, I simply edited /etc/inetd.conf and turned on the server and then "killall -HUP inetd". Apple's utilities are smart about this, too - they'll notice any changes you made manually and still work fine. However, it does get irritating - you can't start OS X apps from the command line, since apps are stored as directories with certain files inside (called "Bundles"). This is nice in some ways, because there's no longer the issue of the resource fork causing problems when you copy of file between a Mac and a non-Mac.

A nice thing, for non-power users, is that you NEVER have to touch the command line, ever. In fact, IIRC, it doesn't even install the terminal application by default. But you still get the stability and performance of BSD with a glitzy interface.

All in all, my perspective is this: OS X is fantastic for people who aren't hardcore UNIX geeks, but the rest of us will be left wanting something more. If somebody would write a new window manager for OS X with more features that power users use (virtual desktops with hotkey switching, sloppy focus, etc...), it would be a different story. But for now, I'm sticking with FreeBSD and Mandrake.

~DJBongHit

--
GNU GPL: Free as in herpes.

Opening apps from the command-line (none / 0) (#42)
by beppu on Sat Mar 24, 2001 at 11:35:31 AM EST

However, it does get irritating - you can't start OS X apps from the command line, since apps are stored as directories with certain files inside (called "Bundles").

On my old-school NeXT slab, there's a utility called open (or was it openapp?). I forget, but anyway, one could do: $ openapp Edit.app and have start an application from the command-line. I'd be surprised if OS X didn't have something like this.

[ Parent ]

it does (none / 0) (#53)
by 31: on Wed Mar 28, 2001 at 12:07:02 AM EST

open blahblahblah

I was playing with OS X today at work... makes me have one of those moments where i wish i had mac hardware instead of intel... it's got all the power stuff i use unix for, while looking good.

i spent the whole time on the command line though, and got pathetically lost the second i wanted to use the gui.

i really liked sshing into the computer, and starting up aps remotely... now if they just had it so it would display aps on the computer you're sitting at, i'd be just about to ditch everything else but openbsd...

-Patrick
[ Parent ]
Bundles and resource forks (none / 0) (#44)
by Potsy on Sat Mar 24, 2001 at 05:30:49 PM EST

The package layout has changed since Public Beta, but if you want to launch an app from the command line, just do this (on OSX final, that is):
bash$ MyApp.app/Contents/MacOS/MyApp
The point of the bundles is not to elminate resource forks. The point is to allow multi-localization. For any given app, just look in a directory called MyApp.app/Contents/Resources/, and inside that are several smaller bundles called "English.lproj", "French.lproj", "Japanese.lproj", etc. Each of these contains localized strings for a given language. When you launch the application, it chooses which "lproj" bundle to use based on which lanaguage you have picked in the "International" prefs panel. You can list several languages in order of preference, and if an app doesn't have your first choice, it will go with the next one down the list. In general, most apps that ship with the system have localized strings for English, Japanese, French, Spanish, German, etc.

Also contained in the "lproj" directories (at least in the case of Cocoa apps) are the .nib files. These are the files that define the UI elements (menus, buttons, checkboxes, etc.). These can be edited using InterfaceBuilder without having to recompile the application. I cannot empasize enough just how cool that is. Don't like the menu keystroke equivalents in an application? Feel like rearranging the buttons and other controls in a dialog? Just edit the .nib file and there you go! No recompile needed.

Resource forks are a tricky thing on OSX. When it comes to filesystems, OSX supports both Apple's HFS+ and BSD-style UFS. Any files that have resource forks on HFS+ are stored them the way they always have been, as a single file. But on UFS, if you have a file with a resource fork, it shows up as two files: "filename" and "._filename" (where "._filename" contains the resource fork). If you just use the UNIX "cp" command on an HFS+ volume, you don't have to worry about the resource forks, since everything shows up as a single file, and the resource forks get carried along when copying/moving files. If you are copying files on UFS however, you either have to manually watch out for the "._" files, or use the command-line tool "/Developer/Tools/CpMac", which will preserve resource forks when copying, even on UFS, and even when going back and forth between UFS and HFS+.

In general, applications and files that are carry-overs from NeXT (such as Cocoa apps) have no resource forks, while applications and files that are carry-overs from the older MacOS do have resource forks. Also note that in addition to "CpMac", when copying moving files in the Finder GUI, the resource forks are preserved, even when on UFS volumes.

As for the sloppy focus and virtual desktops, I'm sure it's only a matter of time before a 3rd-party addon will provide that functionality, especially now that the final retail version of OSX has been released, and it comes with development tools in the box.

[ Parent ]

I smell... (none / 0) (#50)
by jxqvg on Mon Mar 26, 2001 at 04:21:21 PM EST

(warning: the following may be a rant)

a feeble attempt at your typical sensational marketing from the king of "it'll make your kids' teeth whiter". I'm sorry, I don't want to step on any toes by not posting a detailed analysis, but this whole thing just smells like Yet Another Feeble Marketing Ploy, and you know it if for no other reason than that it's Jobs. I mean, come on, what's with all these Yupppie appeal commercials on TV lately? "Giving power back to the people", you might as well throw in a "do it for World Peace" while you're at it. "If you buy it, Barry White and hoards of other famous musicians will be at your beck and call".

Don't get me wrong. If they continued playing the "Computing for the masses" tune, and they were actually hitting the "Computing for the masses" price point, I'd say more power to you, Jobs. That kind of thing has yet to be achieved by Microsoft, *nix, or any other major flavor of O/S.

But that just isn't what it's about. It's the same game Apple's been playing since their inception: Proprietary O/S, Proprietary Hardware, Licensed servicing at licensed dealers only, and all at a stiff premium. That kind of strong-arming just never will work for Apple until they regain some legitimate market share. Even Microsoft stands to lose a huge part of their user base if XP doesn't fly, and that's just software by MS's rules.

So don't let Apple fool you when they say, "We're making UNIX for the masses". All they want is to draw some of the (more foolhardy) blood from the Linux phenomenon, in the hopes that it'll eventually help put that pesky old "cheap generic PC hardware" genie back in the bottle.

[sig]

Knocking down the straw man (none / 0) (#52)
by tres on Tue Mar 27, 2001 at 03:39:52 AM EST

Can you remember the first time that working on your computer made you smile? Can you remember that golden feeling the first time you wrote a program that worked? For me the fun is what it was all about.

Isn't that what this is really all about--fun? I'm sorry to generalize here, but it seems like anymore all I hear from the GNU/Linux community is license-babbling zealots and wannabe marketers trying to mimick Redmond. I often wonder what happened to the giving culture that Eric S. Raymond wrote about.

Reading through the reactions to this article, both here and on Slashdot have been all too predictable--imaginary marketers scared that they may lose their imaginary market share.

If market share is what you want, go for it... but you'll be humping the wrong leg. Much of that +90% of the market that Microsoft holds need someone to tell them what they want; they don't want quality, they don't care what works better. Most of them got on this ride just for the promise of some magic box that was going to make their life more fun and meaningful. They'll get off the ride when the X-Box is released because Microsoft will tell them its what they need.

But GNU/Linux zealots seem more interested these days in luring the lemmings away than they are in the possiblities that Mac OS X represents for the Linux community. Really, it's ironic that Slashdot, on the day of the Mac OS X release, had nothing about it, but the next day, there was an article posted about a review of a beta of a Microsoft operating system on Slashdot. I haven't seen anything from the GNU/Linux community about Mac OS X except disdain and discomfort.

Setting this article up as a straw man, the Linux community got its kicks in before they ever even gave OS X half a chance.

I guess I shouldn't have been suprised.

Will Apple show Linux how to put Unix on the desktop? | 53 comments (45 topical, 8 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!