Avalon Preview Released for XP
CliffH writes "For those that want to play with a preview release of Avalon (the November Community Technology Preview) and the SDK, head on over to this page and download to your heart's delight. It is 261MB+ and is already going slow so be warned."
From the article...
The company warns customers not to use it even on a primary development computer, with there being every likelihood of bugs and a pretty good chance developers will want to reinstall their system once they're done using the Avalon preview.
If Microsoft thinks it's that buggy, I don't think I wanna see it yet.
What really gets me down is the time I spent reading Charles Petzold's book on Win32 programming. 6 months of headscratching, all for nothing. I couldn't even sleep until the brain damage was complete.
Now I have to do the whole freaking thing over...
They must hate us more than we hate them.
it's 255.3mb, and 261,450kb. And what u talkin bout willis? My download started at 1400KB/sec and trickled down to my 5mbit cap!
Just in case it may work on my mac ;-)
Alright, lets see what my Radeon 9200 can do...
Longhorn originally had three major parts. It appears Microsoft has released two of those three now.
.NET framework and allows for integration into Longhorn, Microsoft's new OS.
WinFX is an object-oriented API that uses the
Win -> API
FX -> Framework
WinFS is the vaporware magical file system that includes a new abstraction layer for the files for sorting, searching, indexing, etc.
Monad/MSH is the new command line/shell scipting part of longhorn. It too can be downloaded and used in beta right now. It's probably the most useful aspect of longhorn to the average power-user.
If you are going to play with something that isn't going to scrub your system, I would start with monad. It sits happy on any installed system.
"The goal is give developers a consistent set of APIs," or application programming interfaces, Montgomery said.
And they're doing this by adding ANOTHER set of graphics APIs to Windows, to complement the ones we have now, and the ones we had five years ago, and the ones we had five years before that, and the ones we had five years before THAT?
I don't get it.
If this is anything like the preview they've shown to reviews in the past, this should be a real nightmare. Last time they showcased the Avalon desktop, it ran slow as hell, and started churning the swap file in minutes, and was so slow as to be completely unusable in under ten minutes.
Instead of discussing the technology (which is actually pretty cool...they do have smart engineers at Microsoft), I have a feeling this will be a bunch of +5 Funny Microsoft-bash posts.
:)
One third referencing some obscure GUI from the past where something almost like this has been done already, another third referencing some future project not released yet doing the same, and the rest a bunch of +5 Funny "jokes" rehashing old Microsoft jokes from the last eight years. Okay, I'm generalizing, but that's also what people will be doing about this.
Seriously, it looks like interesting stuff, and I can't wait to not only develop with it, but develop with the competing technologies that will also spring up as a result.
Oh, and for the record, before people say it--OS X does use the 3D card, but only for fast blitting. It is still 2D. Not actual 3D acceleration using hardware triangles like this, where you're dealing with a camera viewport and using meshes.
The main difference is that newer graphics drivers in Longhorn allow for better performance and newer hardware. With Windows XP or Windows Server 2003, users might see slower performance, fewer shades of gray or less 3D animation, Montgomery said.
As we've already learned, Longhorn is going to be graphically intensive (just what you want for the kernel of a server OS, isn't it?). While I agree with the statement in that Longhorn may very well have drivers more appropriate for doing the things that Avalon wants to do and therefore be faster than the current WinXP drivers, I think the chances of performance up to what we expect on currently modern systems with entry-level graphics cards will be pretty much nil with Avalon, Longhorn or no. In fact, we might not even get to see the content as desired on mid-level cards, I'd guess.
*sigh*
This flies in the face of science.
Some screenshots? Anyone have any?
DS vs.
..download through Dijjer?
Simple answer: No.
.NET with the new technologies based on that. They're ditching Win32, though there will be binary compatibility for older apps. The kernel will remain mostly unchanged; it's the overlying technologies that are being rewritten. Not only will this make things much safer (.NET is garbage-collected, type-safe, etc.), but it allows for much easier development (compare MFC to, say, WinForms).
It's all going
Hate Microsoft or not, they're taking a step in the right direction with Longhorn by replacing all that "cruft" (my favorite term for such things). Of course, I still think Apple will just come out with something even better with Longhorn, but at the least, I'll be happy having the majority of people getting their computers into a managed memory environment where I don't have to worry as much about an app taking things down.
I don't know about the rest of you but i would rather bypass downloading a 250MB file and would just like to see some interesting screenshots.
When I read about the evils of drinking, I gave up... reading.-Henny Youngman
Hell, even their previews are months behind schedule. I don't think we'll ever see Longhorn.
Seriously though, anyone have screenshots?
LilMikey.com... I'll stop doing it when you sto
Read this. This is a truthful essay. Of course it will be moderated down as blashphemy by the church of the linux zealots!
./* | grep -v sizeof', I discovered 4 exploitable vulnerabilities in a matter of 15 minutes. More vulnerabilities were found in 2.6 than in 2.4. It's a pretty sad state of affairs for Linux security when someone can find 4 exploitable vulnerabilities in a matter of minutes." - Brad Spengler
In light of the disastrous 2.6 development model that has given sysadmins everywhere a headache by introducing development code into a production line, Linux has signed its own death knell. With more and more people looking to alternatives like FreeBSD 5.x, OS X, and DragonflyBSD, Linux is slowly shovelling the dirt beneath its feet to dig its own grave.
Linux And Windows
Quite simply, the revolution against Windows has run out of steam. While Linux was a viable alternative in the days of Windows 98, when the rallying cry of geeks everywhere was "Down with M$, Linux never crashes," we now have the majority of the Windows userbase running NT-based operating systems. Except in cases of hardware or driver issues, reliability is no longer an issue in the comparison between Linux and Windows.
Eventually, the movement became one of security. In the years after its release, Windows XP was discovered to have several high-profile security flaws. Microsoft underwent a major code audit and released SP2. The rallying cry for OSS was now about security.
However, the community has discovered major flaws in the Mozilla software suite, including bugs marked "confidential" for years at a time. Additionally, major security holes have been appearing in the 2.6 line of Linux kernels, some having existed for years and affecting the 2.4 line. Declaring Linux to be the secure alternative is no longer as true.
Worst of all, the Linux kernel developers have no clear process, nor any clear contact person, when it comes to security issues.
Evidence: http://lwn.net/Articles/118251/
Evidence: Long-time shell-provider SDF used Linux until they got hacked into. Now, it's a 64-bit version of NetBSD.
Evidence: PaX discovered the mlockall hole. It was fixed in PaX for two years. Linux just now (2005) caught up.
Evidence: "Using 'advanced static analysis': 'cd drivers; grep copy_from_user -r
The New Linux Development Model
With the 2.6 line of kernels, a new model has been adopted that is considered easier for the kernel developers. Instead of branching a 2.7 line, following the model of odd-numbered version numbers denoting development code, everything is now being thrown into 2.6.
"Not all 2.6.x kernels will be good; but if we do releases every 1 or 2 weeks, some of them *will* be good. The problem with the -rc releases is that we try to predict in advance which releases in advance will be stable, and we don't seem to be able to do a good job of that. If we do a release every week, my guess is that at least 1 in 3 releases will turn out to be stable enough for most purposes. But we won't know until after 2 or 3 days which releases will be the good ones." -- Ted T'So
In other words, this Linux kernel developer believes it is perfectly fine for one in three kernels of the stable line to actually be stable. The new development process is anti-user. "Release early, release often" has outlived its reliability and applicability to the real world.
The excuse given is that Linus is only one man, and there are only 24 hours in a day. If that is true, than Linus needs to address this shortcoming of the process; otherwise, the process is poorly managed.
The Community Has Regurgitated Itself
In a frenzy of newbies, the Linux community has grown, with Slashdot as its rallying center. The cycle of self-feeding groupthink has created a userbase unable to see outside its own perceptions. This leads to unrealistic attitudes about the safety and stability of Linux and its applicability to various solutions.
Didn't microsoft give us some source code?
They're showing the API and methods by releasing what they've got so far and letting people get used to working with it. It's the reason for last year's WinHEC release. They want developers fully fluent in Longhorn's APIs so that when the time comes, apps will be already ported and ready and the transition will be easier. Hopefully even smoother than XP's was (which, looking back, wasn't as rocky as I thought at the time).
Expect alpha preview releases for all sorts of pre-release technologies. Even Visual Studio 2005 is out in an unfinished beta form for you to play with for free.
Too Late... http://www.winexpose.com
.sig
Will this working on Win 2003 server installation under VMware? Or do I have to be the first sucker to try it out?
I guess this is why Longhorn keeps on slipping. Maybe they should let a little code slip through so as not to jeopardize the Duke Nukem Forever bundling agreement.
Help fight continental drift.
wget timing out. Is this being hosted on an XP Avalon box?
Powered by caffeine and sugar; BSD
Do a google search on Entbloess, TopDesk, or WinPlosion. All supposed XP clones of Expose.
If you look at their APIs, they have been getting cleaner and more consistent in the past few years. Just look at their .NET Foundation classes. The Win32 API is an abomination, I would welcome an overhaul.
The last Windows you used must have been Windows 95. With all the Microsoft bashing going on (which I understand and rather enjoy) you should still be honest. A well set-up and firewalled Windows system can be very stable nowadays. Besides, your joe average won't need his system to stay up 24/7. I'm a definite power user, but I turn my system off when I don't need it for extensive amounts of time. The last time Windows crashed on me? I can't even remember it. If you're going to troll, at least bring up some constructive statements.
I wonder if it's really that graphic intensive (not gonna bother downloading it yet though). Seeing things like this get added to XP and things like WinFS taken out of LH, it's making it less and less attractive to ever upgrade to LH. For the first time in my life, I feel like either
1) lots of people will stick to XP, or
2) a lot of people will move on to linux instead.
I've been using pretty only windows in the last few years (ever since I got rid of my atari 1040 and older stuff), coding for it and all... But I'm really loosing interest in the "new" stuff they come out with (like, I got all the themes and such crap all off - "classic" look). It just seems more bloated, and they're trying to put some "nice" (they think it is, anyways) GUI so lusers aren't scared anymore, when in fact, I find it's becoming quite a mess - and an overly bloated one, that is.
I've tried knoppix 3.7 a couple days ago, and I must say it's a VERY viable option for most stuff. Yes, I had a few problems (enabling spdif out on sb lives, xmms wouldn't play mp3's off smb and small things), and it won't run all my usual apps (photoshop, ms office...), but I was very surprised nonetheless. There were some compilers in there, a CAD program (shocked me), OO loaded slow (of course) but it wasn't half bad... It was really easy to pick up and find everything.
Most people I know all love their windows/autocad/photoshop/etc (not that they know how to use it) - but that's mostly because they didn't pay the hefty price tag, but this does the most part, for free (legit). I'm starting to seriously consider "doing the switch", at least on one PC to give it a good try.
I think LH itself is what will make the most people switch to linux (especially combined with all the spyware and other crap most lusers have been crippled with lately). I only see bad in LH - and I'm mostly known as a M$-fanboy... But that's changing lately. I've been starting to convert myself to more open, portable (and perhaps more stable/secure) options (like using LAMP instead of ASP or ASP.Net/IIS/SQL Server like we use at work and such) and I'm liking it, a lot (cheap to host, too). Now if I could find a replacement for most apps (including VS.Net), I think I'd be sold.
To me, that MS-world is just unsustainable. Everybody I know only use it because they can use pirated everything for free. I don't think I know anyone who wants to - or can afford to buy a new windows, office, and everything else license every year (or even for every second version - and who wants to stick to old soft?). I don't mind paying a minimal fee for a good distro or such, but what I use daily on a win box cost me over a few months' salary... How much longer can we keep up with this dream of being afford to use all these apps that cost hundreds of $? (yes, I know, big corps can afford it... whatever).
///<sig
Release early, release often. Now if I only could remember who has said that.. ;)
"It's too bad that stupidity isn't painful." - Anton LaVey
I've tried TopDesk and WinPlosion (along with several others). NONE of them are fast enough to be in any way useful. It takes several seconds for any to respond, generally... and many are crashy.
I like the idea so far.
.Net programmer, we have two options for graphics. GDI(+) and DirectX. The latter isn't feasible for web applications, and the former is a gimp pony for anything more complicated then a tetris clone.
I mean, for the
This seems like a good middle man. Though I think most people realize this is MS taking a shot at Flashf, and it's about time.
Though just to vent, I would like to actually see some focus on getting the docs for DX 9.0c SDK for VB.net out the friendly door already.
but that would defeat the purpose of trolling wouldn't it :)
Hmm i can add a printer to my kde in a minute or two.
:)
Then it's smart enough to know that when i need to print something, and i'm not at home, that there are 3 roughly equidistant lasers at work. It'll select the one that's not-broken and least loaded and send it there.
Kde's printing is slick - pick something else to bash
Actually the 3 Pillars of Longhorn were Presentation (Avalon), Data (WinFS) and Communication (Indigo).
Ummmm... that link points to one of those cheezy ISP 404 pages full of Subject links and Search option.
Sadly, it is still one of the most reasoned arguments I've seen in Slashdot Windows thread in days.
=tkk
PS
Bill Gates - Creationist?!?
It is 261MB+ and is already going slow so be warned.
Fortunately, there's no explanation of what the hell "Avalon" is in the text so that ought to help with the download performance. Had someone actually known what the hell you were so excited about, more people would try to download the software.
Good strategy. It's like those morons who put the important part of their comments in the subject line and continue on to their message. I miss important bits of the message since I do not scan for the subjects so miss out on the point they are attempting to make. It sure makes life easier for the reader if we don't know what is going on and do not have to actually get interested in the article or comment enough to read it.
There. Rant over. I feel better now.
Thanks
Simply stating [Citation Needed] does not automatically make you insightful or brilliant.
i'm just a guy working in an office for a supermarket chain. linux is mainstream for us here, but the OS is easy to use and looks like fun so i've been reading/messing around.
here's a link i found on netcraft that was quite interesting:
http://www.business-linux.at/idc/linux.marketshar
another good read:
http://www.itfacts.biz/index.php?id=P1299
not sure i know what you're talking about, the food industry seems to be switching to linux, i don't know about others. they say it's more reliable and very popular. all the techy stuff you're talking about is over my head!
joe user
It's not WinExpose anymore. It's called WinPlosion now. http://www.winplosion.com/
Realistically most people here probably run linux most of the time... or at least they should if they practice what they preach.
:)
I work on solaris, run linux most other places and have a win2k box since my wife needs MS Word and i occasionally have to do VC++ consulting work.
I see kde and gnome every day and see them advancing but since i haven't used a version of windows released in the last 5 years so it's hard to make a direct comparison.
The point with firewalls is that most windows boxes are not firewalled and most linux boxes are... it's a statement more about their userbase than anything else. OSX stands above the crowd here though
let me!
From what I gather, Avalon is Microsoft's version of Quartz. Has anyone compared the two in terms of capabilities, ease of programming etc.?
Worst, Idea, Ever. This sounds like it will be *terrible* for laptops, or any other place power useage is a concern :) Like I really need to run my CPU at 100% so word can lockup that much faster.
Religion is a gateway psychosis. -- Dave Foley
Is the same as the pornsplosion I get when browsing the web?
Any man who afflicts the human race with ideas must be prepared to see them misunderstood. -- H. L. Mencken
After all i've read and seen about Avalon, i still don't get what the fuzz is all about. Yes, it looks flashy, but at the expense of unreasonable processing power (don't forget this is an integral part of the Longhorn kernel). And i consider myself a sucker for eye-candy.
I know it's not comparable, because we're talking windows, but Enlightenment 0.17 will (hopefully) do everything Avalon does, and pretty much everything new Longhorn does as well. Just check the information on the e17 foundation libraries. Amazing stuff.
...you feel.
...that was funny. It made me laugh.
Simply stating [Citation Needed] does not automatically make you insightful or brilliant.
A well set-up and firewalled Windows system can be very stable nowadays.
I believe that -
But how much work is it to get there? I just installed Windows 2000 SP4 under qemu. In the time I had all the important cumulative Patches and Hotfixes applied (there were some 37 to check on windowsupdate) AND fixed all the stuff that got b0rked on the way (Explorer.exe for a basic desktop and msi.dll for the install control in the control center for example) I would happily be running a gentoo stage 1 install (and thus any other distro of choice).
I might have just been unlucky - but still there was some serious tweaking needed to get this one up and running.
That's too bad! :( I actually considered coding something like this once, and came to the conclusion that the most feasible implementation would be slow (screen capture each app, resize, hide all windows, show images). How does OSX achieve the effect? I'm assuming their GUI system has built in capabilities for scaling of application windows.
> Hmm i can add a printer to my kde in a minute or
> two.
Right. I read that, and in order to give you the benefit of the doubt, I tried, yet again, to setup my printer.
I'm pretty experienced, and I'm pretty smart. So for the life of me, I cannot setup the HP Photosmart-1310 that's shared via SMB from a Windows XP host on my TCP/IP, DNS-enabled network, from my Debian (Sid) Linux box with KDE, cups, or anything else for that matter.
From a Windows host, it's easy as can be -- browse to the printer, set as default, there's even a simple way to setup ink saturation, paper size, etc.
From KDE, I can go through the motions of setting up the printer (which does appear via smbtree, etc.), but it won't work. Stuck. Brick wall. Easiest thing in the world to do in Windows, completely lost cause with Linux.
-fb Everything not expressly forbidden is now mandatory.
That's what I'm assuming. I doubt it's difficult when you have access to the code for window management and such, especially when you can use OpenGL to transform them. My guess is that none of these Windows apps use a reasonably efficient method of refreshing their window captures. Instead of grabbing them while a window is up, they seem to restore a window, take a screenshot, then minimize again. If you want to do something like this, I really recommend pinging windows for their status or doing some API hooking. I'm pretty sure a nice imitation, at the least, is possible. If you ever do this, send me an message. I may just be interested in contributing. Or at least becoming a loyal user.
Ok, I finished downloading it, and installed everything in the ISO file. How can I get a demo of this thing in action, now? Anyone?
Try clicking that link. How does any living human manage such a typo?
I'm speechless!
Don't thank God, thank a doctor!
Ever seen the fast user switching 'cube' effect in Mac OS X? Ever heard of Quartz Extreme? That stuff is all done on the 3D portion of the video card. They use the 3D portion for fast compositing too.
And they're doing this by adding ANOTHER set of graphics APIs to Windows, to complement the ones we have now, and the ones we had five years ago, and the ones we had five years before that, and the ones we had five years before THAT?
That may be true in the high performance DirectX area (D3D and DDraw revisions frequently made large changes to the API), but in the normal application area we've been stuck with GDI since Windows was conceived. Only relatively recently has GDI+ come into play, although at a high level it's simply an OO wrapper around GDI, and likewise, MFC graphics classes are also GDI wrappers. At the core, Windows basically supports GDI for normal applications and DirectDraw/Direct3D for high performane graphics, and so the situation then isn't as complex as you make it to be. GDI itself currently is very underpowered when you compare it to things the Mac OS can do, so it makes sense to finally revise the API after 20 years of usage.
I can't possibly describe Avalon's capabilities here, but as a simple example, in GDI you draw rectangles, lines, etc., whereas in Avalon you define visual objects and Avalon automatically renders them as needed. In computer games and other applications that need a deeper level of control Avalon won't be that appropriate (although IIRC you can do simple 3D in Avalon), but for normal applications I think it'll be awesome. Death to GDI!
Of course after rereading my post it does seem like Windows has a bunch of graphics APIs. Just remember that all that application-side ones are reducible to the ancient and horribly underpowered GDI.
This is, I think, the typical "I-Tried-Installing-Linux-2-Years-Ago, Tried-Playing-with-it-for-10minute, And-found-it-suckz-because-none-of-my-1337-windows -apps-runs-on-it" user.
The user only remember a few surface stuff he noticed 5 years ago and doesn't stop complaining about them.
Most notably in the "A Decade Later
They both also succeeded replacing Windows-based server whose administrator got fed up with microsoft's products.
We see more Windows-to-Linux thant Linux-to-Windows server migration.
Linux and BSD are also used a lot in academics.
And Linux IS used on Dekstop even if it isn't as visible as it's other uses.
[CTRL] [ALT] [+] and [CTRL] [ALT] [-] since I installed my first distribution.
(You should have paid more attention to the manual).
Meanwhile, you had to use some hack to avoid rebooting Windows 95 in order the effect to take place...
Then FreeDesktop.org doesn't exist, I think...
It's not desktop's purpose to implement installations. (Just like it's not DirectX's job either).
...)
...) should watch and learn. They could win a lot of clients if they had a single point for software acquisition/update like this...
I think it's the exact opposite.
Almost all linux distributions have a package managment system (YaST, apt-get, emerge, drakrpm, yum
Unless you want to use new version of a sfotware that isn't available yet in your distribution, you got a SINGLE place to uninstall unneeded packages, install new softwares that are optimised for YOUR distribution, and you can easily get updates for them.
Compare to windows where you have your Installation CD, Windows Update, separate installer that you must download from separate website for each software you want.
You must track updates alone for every single software you installed (do you remembre that small plug-in you installed 6 months ago in WinAmp and for which there's now a patch against a buffer overflow ?)
I really think PC providers (like Dell, HP,
There's not only one, but a few of them. :
Notable one
- OpenGL : So good for 3D graphics that it's also used under Windows for games like thoses from ID software.
- SDL : 2D GFX/Audio library that is also used by windows programms (like emulators).
Most of the base of the design is borrowed from older Unices which where available long time before Windows.
KDE got most of them from the begining.
Even the parts that are inspired by Windows are much more configurable than windows.
1. Mono is not the only V
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
I'm constantly amazed that this kind of uninsightful crap that you consistently trot out over and over again gets modded up as insightful. Probably because of the fact that you are a drooling MS fanboy, you seem to forget perhaps the most important feature that this "coolness" comes with: Next Generation Secure Computing Base.
Yes boys and girls, underneath all the ooo, shiny is that wonderful bit of technology: Trusted Computing. You know, the kind of trust where your computer doesn't trust you? But I suppose you wouldn't want to yell too loudly about that particular feature of Longhorn, now, would you, since it paints MS in a less than favorable light?
So you can have your fucking spinning Notepads and videos looping in the background of windows--to me the price that comes with that technology is simply too high.
At the end of the day, you just have to face the fact that foo bar baz.
Of course, I still think Apple will just come out with something even better
Where? How? There is nothing on Apple's roadmap. Objective-C is neither garbage collected nor type-safe, yet it is still what Apple is pushing. And while Apple kind of inherited a scalable graphics engine and toolkit with NeXTStep, that is technology from the 1980's; their competitors are designing with the benefit of hindsight and with knowledge of today's needs and requirements.
The people most likely to come out with something "even better" are the open source community: between the X11 enhancements and Mono and Mozilla as application development platforms, they are on track to ship something better than Longhorn before Longhorn even ships.
Can someone mod this up a little bit? I used my mod points earlier and its sad to see that even if this makes a few interesting points, its rated at -1.
I actually considered coding something like this once, and came to the conclusion that the most feasible implementation would be slow (screen capture each app, resize, hide all windows, show images
Well, I don't know about Windows or OS X, but on X11, you could enable backing store for each window, which would keep a complete bitmap on the server for each such window. In newer versions of X11, you should be able to display a scaled version of that bitmap on the server, with no client traffic.
You may say that it's expensive to keep the complete window content for each window on the server and you would be right. I believe OS X has that content around anyway (which is why its graphics system is so memory intensive) and they are making lemonade out of lemons by making a bit of visual glitz out of it. On GUIs that actually have a choice (like X11 or Windows), people have been reluctant to impose many megabytes of extra memory footprint merely to give users quick iconic previews of window contents.
I'm not totally sure, but usually i go into the CUPS web interface and just use socket://address.of.printer and do it that way.
Of course that only works for network aware printers.
I've definitely entered smb:// urls into cups and recall not having any problems, but dont have anything like that running now.
3m:50s into the demo in "daniel_lehenbauer_avalon_3d_MBR.wmv" there's an Amiga poster hanging on the guy's wall.
look, i admit that i am not a s/w developer-- i actually work in digital media. long story short, i installed this (all of it), and didn't notice anything different. did i just install a whole new set of programming tools expecting some eye candy, or is there something i'm missing?
How is this? If the old API never goes away, how are you not adding another layer of "cruft"? Consider DirectX -- even with the DirectX 9 SDK, I can still compile apps that use Directdraw surfaces (which was "removed" from the API years ago). I'm not sure how you can maintain compatibility but ensure against writing new apps that use old api's.
This is going to sound like a troll, but I'm honestly curious. Is Microsoft simply removing the documentation for Win32 and nothing more?
That's what OS X's next release, Tiger, has with Core Graphics - a set of routines which are hardware accelrated to do image manipulation. Your standard filters like blur and so forth and other manipulations.
The cool (?) thing is that since it's all accellerated, you can even apply it to video...
In theory you could build a half-decent Photoshop clone in Tiger that would mostly make use of the system libraries.
I think they only apply to 8-bit graphics though, so Photoshop will be around for a while.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Apple can keep the OS they have now. It is great and will probably be able to meet or beat Longhorn in the graphics and usability dept.
Everytime Apple introduces a new OS... Oh. they really only did that once.
OS X.
Why would I waste my time playing with Avalon when running my Mac I can use Aqua now, or try and get my hands on Looking glass from Sun. One of which is current real world kickass and the other is a research project by sun. 3d desktops are going to be more interresting than the next MS fad.
".NET is garbage-collected"
:)
Duh, learn your punctuation! It should be ".NET is garbage, collected".
If no icon will ever be smaller than 128x128, you have a point. Not otherwise.
A great deal of work goes into crafting small icons that convey exactly the right visual information. All manner of interesting optical illusions come into play when you are doing 16x16 or 32x32 or 48x48. For example: colour doesn't stay in one pixel but "spreads" to neighbouring ones; a line drawn through a 1-pixel-in-4 checkerboard pattern [ie. line 1: XOXOXO line 2: OOOOOO] changes colour depending on its position relative to the pattern...
If it had really been true, the "all you need is vector graphics" argument would have made font design really easy. No need for all that skilled work adding hints to make OpenType/TrueType fonts work at small sizes.
Except in cases of hardware or driver issues, reliability is no longer an issue in the comparison between Linux and Windows.
Apart from the requirement to have hardware and drivers to run a system.... Windows without these would be pretty boring. This also neglects the phenomental virus/worm/security issues for Windows over the past couple of years.
Linux will remain a niche. It's over ten years later, and Linux is still just a marginal server OS beneath BSD.
50% of all new blade servers are shipped with Linux, and it is the pre-installed system on around 20% of ALL other new servers. Strange use of the word 'niche'.
In the desktop market, it has barely made a dent. Before Google Zeitgeist removed its OS numbers, Linux was at a mere 1%.
No, Linux has around 2.5-3% of desktop market share. Due to the nature of Linux, this is not a measure of installed base. I'm sure most organisations that use Linux on the desktop do what we did: Buy PCs with pre-installed Windows (as that is where the bios update and hardware analysis tools are installed), then re-partition and dual-boot Linux. Installed base of a free OS is very hard to measure, but its certainly going to be larger than the purchased volume.
Watching the video with Daniel L., it seems like programming
Avalon is about as simple as programming OpenGL.
Except it runs very slow. w00t
The world is everything that is the case
It is not replacing that Cruft. Since Microsoft roots lots of the api back into Win32 it is more like putting another easier accessable OO Layer on top of it and mixing lots of C# code into it as well. Apple has been there since the good ole NeXT days and KDE also is there, as well as there was BeOS. What Microsoft currently is doing is, they kick out COM which was overly complicated and lousy and replace it with another component technology which is easier to use and easier to access (and even more bound to Windows than Com was) Apple has had that since the good ole NeXT days with OpenStep (now called cocoa) it is pretty much the same except that the main implementation language for the objects is ObjectiveC, but currently you can use java as well to access and or implement those (see the similarity) KDE has that with KParts and bindings in both directions into many languages although C++ is the preferred language of choice there. Ditto for Beos. Microsoft tried it in the early to mid nineties with Com and basically failed because they shot over the top with their implementation and never really could get it that stable to base the whole API on top of it. (The original plan was that Windows should become a set of many components where you only had to glue your app together, like NeXT showed that it can work and later KDE and even before Microsoft OpenOffice/StarOffice (yes StarOffice really is that old) showed. IBM and Apple back then wanted to go that way with Taligent and failed due to whatever reasons and Microsoft failed because they shot themselves into the foot with their programming model they chose for Com (and never admitted that) So what Microsoft is doing currently is catching up to most other systems in this regard by moving the API into a component layer which is easier to access and basically tying everything stronger than ever into windows than they used to (they basically force you to move your development process to C# and make the app unportable into other systems)
Objective-C is dynamically typed, and therefore is by definition type-safe. It is also garbage collected, although it is both reference counted and manual. But I know what you meant. You meant it isn't type-checked at compile time and automatically garbage collected. Which like any language feature debate has both pros and cons. Neither are key issues that you can judge the superiority of a whole OS on.
Apple leapfrogged Windows with OS X in 1999. Not until Longhorn will Microsoft have caught up on the technology. And given that there will likely be 2 OS X releases before Longhorn (Tiger and the next one) it would be crazy to decide now who will be ahead at that point.
Only judging by how far ahead apple is *right now* I'd dare to doubt this is gonna change with longhorn.
It was a very smart move by apple to move their OS to the stable darwin (BSD-like) core. Now they're building on a stable and proven foundation (developement speed has likely increased) where MS still fiddles their legacy mess with no way out.
We all know how buggy MS software is and if longhorn turns out like any of their previous OS releases they might have a harder time forcing their userbase into yet another paid Beta-test than the last times.
After all the kids growing up today are more computer-savvy than ever
and and not that easy to fool anymore. Even more so with the alternatives
(OS X, maybe even linux in a few years) getting better and better.
Comparing the polished OS X desktop (springloaded-folders, responsiveness, integration, drag'n'drop software install/uninstall, eyecandy) to the windows crapshow (bugs, bugs, bugs, crashes, security holes, registry mess, spyware) is almost not fair anymore.
people have been reluctant to impose many megabytes of extra memory footprint merely to give users quick iconic previews of window contents.
Sad. When it's possible it should be an option (on/off switch in the user-preferences).
Not that I'm a big expose-fan but I've heard some people like it.
I, personally, would be more interested in window-zoom that could be implemented using the same mechanism. I imagine a feat. that zooms a window to it's original size when it gets mouse-focus and smoothly shrinks it by 50% (or whatever the user desires) when it loses focus.
That way I could easily arrange my dozen or so windows on the desktop (they'd be only about windowmaker-icon-size) and always see at a glance what's there.
That would be so much better (and more intuitive) than all the iconify and roll-up kludges we've come up with over the years...
Linux is not Windows
Bound to Windows? WTF. MS has published .NET interface and C#. Linux (works with BSD also) has Mono-project going on and I think there is some stuff on Mac also. .NET programs written and even compiled on Windows runs nicely on Linux. And you can use a whole bunch of programming languages to use .NET (C++, C#, Perl, Ruby comes to mind). Maybe you should learn a bit more about .NET and C# before you come here to spread false information.
You don't know what you don't know.
For better or for worse, the current release of X11 has a complete set of primitives built in to do everything Aqua is doing, including window translucency and window rescaling and warping (in addition to fully scalable anti-aliased text and graphics).
You meant it isn't type-checked at compile time and automatically garbage collected.
No, I meant that it is unsafe. You can free free objects, you can access past the end of an array, you can use a pointer of one type as a pointer of another accidentally, etc.
Apple leapfrogged Windows with OS X in 1999. Not until Longhorn will Microsoft have caught up on the technology.
Apple didn't "leapfrog" anybody: they are shipping NeXTStep with a new theme and some performance hacks. You're dreaming if you think that that's going anywhere other than down. People have come to like safer languages, and C#/.NET and Mono/Gnome are delivering. Apple has nothing on the horizon.
has anyone considered how digital media artists (particularly 3D artists) are going to benefit from this? xml and xaml are easy enough to learn scripting languages. software developers would no longer have to consider the front-end... leave that to us DM ppl whose job is to create the pretty and interactive. developers create the programs, and everyone. could be happy... (here come the flames again...) or do s/w developers actually enjoy making the front-end?
Windows 2000 introduced window alpha blending through the use of the new Layered Windows APIs which allow the window manager to retain a copy of the contents of a window so it can composite them on the desktop. Any acceptable expose' clone for Windows would need to hack into this mechanism rather than capturing the contents of each window manually.
"Taligent is still pure vapor. Maybe they'll be the last who jumps up on Openstep... "
Wow, you were using Windows NT graphics APIs in January 1990? Wow!!!! You must be pretty darn connected to get copies of OSs before they were released or even writen!
Then you don't actually know what type-safe means. Dynamically typed languages such as objective C are by definition type-safe.
You can free free objects
Not a type-safety issue. But objects are only deleted when the last reference goes. That's what reference counting means.
you can access past the end of an array
No. If you try it you get an exception.
you can use a pointer of one type as a pointer of another accidentally, etc.
You're confused. You can have pointers which will point to objects of any class, or you can have pointers which are specific to a single class of objects. The former is type-safe because of dynamic typing, the latter gives an exception if you fuck up.
Sorry to bust your bubble.
Apple didn't "leapfrog" anybody: they are shipping NeXTStep with a new theme and some performance hacks. You're dreaming if you think that that's going anywhere other than down. People have come to like safer languages, and C#/.NET and Mono/Gnome are delivering. Apple has nothing on the horizon.
How's about you actually learn about it before you dismiss it? Heck even learning what type-safe means would be a start. Then you might be taken more seriously.
| Read this. This is a truthful essay. Of course it will be moderated down as
| blashphemy by the church of the linux zealots!
|
| In light of the disastrous 2.6 development model that has given sysadmins
| everywhere a headache by introducing development code into a production line,
| Linux has signed its own death knell. With more and more people looking to
| alternatives like FreeBSD 5.x, OS X, and DragonflyBSD, Linux is slowly
| shovelling the dirt beneath its feet to dig its own grave.
I think "disastrous" is a little sensationalistic here.
| Linux And Windows
|
| Quite simply, the revolution against Windows has run out of steam. While Linux
| was a viable alternative in the days of Windows 98, when the rallying cry of
| geeks everywhere was "Down with M$, Linux never crashes," we now have the
| majority of the Windows userbase running NT-based operating systems. Except in
| cases of hardware or driver issues, reliability is no longer an issue in the
| comparison between Linux and Windows.
I believe that given the huge amount of spyware installed on most computers,
reliability is still an issue. I've seen computers ground to a hault becuase of
this.
| Eventually, the movement became one of security. In the years after its
| release, Windows XP was discovered to have several high-profile security flaws.
| Microsoft underwent a major code audit and released SP2. The rallying cry for
| OSS was now about security.
SP2 has created quite a performance hit for the few people I've spoken with.
The "security rallying cry" has been a consistent argument for OSS. Also
consider that OSS != linux.
| However, the community has discovered major flaws in the Mozilla software
| suite, including bugs marked "confidential" for years at a time. Additionally,
| major security holes have been appearing in the 2.6 line of Linux kernels, some
| having existed for years and affecting the 2.4 line. Declaring Linux to be the
| secure alternative is no longer as true.
I'd really have to see references to these "confidential" bugs. It's possible
that they were marked this way for nonsecurity reasons and the person who
submitted the bug forgot about it --- this is very conspiracy theoryish. I
acknowlege that there have been security flaws which were found that date back
a few kernel versions. The author fails to mention that once found they are
fixed and patches are released. I think it's worth mentioning that these flaws
are not discovered on a weekly basis. Also most (all) security related bugs
require local access to the machine -- as opposed to flaws which simply require
the attacker to have network access.
| Worst of all, the Linux kernel developers have no clear process, nor any clear
| contact person, when it comes to security issues.
I thought the linux kernel mailing list was the defacto method for disclosing
linux kernel issues. I also thought submitting security alerts to places like
CERT and security focus were the way the industry delt with security issues.
| Evidence: http://lwn.net/Articles/118251/
|
| Evidence: Long-time shell-provider SDF used Linux until they got hacked into.
| Now, it's a 64-bit version of NetBSD.
Were they up to date and patched? Also worth mentioning, NetBSD is opensource
software.
| Evidence: PaX discovered the mlockall hole. It was fixed in PaX for two years.
| Linux just now (2005) caught up.
According to the page above:
http://lwn.net/Articles/118251/
December 15th: I send Linus a mail with a subject line of
"RLIMIT_MEMLOCK bypass with locked stack"
December 27th: The PaX team sends Linus a mail with a subject line of
"2.6.9+ mlockall/expand_down DoS by unprivileged users"
January 2nd: The PaX team resends the previous mail to Linux and Andrew
Morton...
The PaX tem sent Linus an email --- amazingly they knew who to
-- john
The ECMA standard only supports 10% of the core APIs of .net, Microsoft is holding several patents on other API parts so it is questionable on how long Mono can be able to clone the parts outside of ECMA.
Jeez. It's type-safe. You can't run a operation on an object that isn't an operation for that object. That's it. There is no more. That is type-safe. Period.
Most people expect type-safe to mean that the program won't crash, either. And most people consider a program halting with an exception to be equivalent to a crash, even if it hasn't actually dumped core.
So you haven't heard of catching exceptions either. The mention of an exception was in regard to an attempt to index beyond the end of an array. You'd have to catch that exception whether you were in Objective-C or C#, there's no difference.
As to not-crashing, let me assure you that it's all to easy to create C# applications that crash. You have a non-issue there.
Therefore, most people do not consider dynamically-typed languages to be type-safe. Some of them are, by the most academically strict definition of the term, but not everyone here's an academic...
Most people my ass. Don't judge the majority by the gaps in your own knowledge. Most people on here have probably done or are doing a CS degree, or are professional software engineers. They will probably accept the correct definition of what type-safe means, rather than your mis-use of the term.
... and report back.
The operations will still work, they'll just not be hardware accelerated. Weither or not it is still usable at that point, depends on what the application is doing...
I agree though that it would be good for them to strive to get it to work with a higher percentage of existing systems, I think they will make some efforts along those lines.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Dynamically typed languages such as objective C are by definition type-safe.
Not all pointers in Objective-C are dynamically typed; in particular, none of the C pointers are. Furthermore, only Objective-C's method dispatch checks the dynamic type of pointers--other operations (member access etc.) don't. And Objective-C's method selectors are not checked across compilation boundaries (so you can declare an "doit:(int*)" in one header file and a "doint:(float*)" in another).
But objects are only deleted when the last reference goes. That's what reference counting means.
Reference counting isn't garbage collection. But even if it were, Objective-C on Cocoa doesn't have automatic reference counting; reference counting is implemented through explicit calls, i.e. manual storage management. C++ has a better claim to being "garbage collected" because in C++, at least you can automate the reference counting.
No. If you try [to access past the end of an array] it you get an exception.
No, you don't. You may get an exception if you access past the end of an NSArray, but you don't get an exception if you access past the end of a C array, which are very much still a part of Objective-C.
How's about you actually learn about it before you dismiss it? Heck even learning what type-safe means would be a start. Then you might be taken more seriously.
Objective-C is a superset of C, including all the unsafe features of C, and they are used in every Objective-C program. The consequences of this are serious; for example, if you load a bundle (say, a plug-in) and the bundle contains a bug, it can crash the entire application (or worse, just mess up your data) and there is nothing you can do in Objective-C to guard against that. In C# and Java, on the other hand, you can guarantee that dynamically loaded components don't crash or mess up the whole application.
so you can declare an "doit:(int*)" in one header file and a "doint:(float*)" in another)
Pardon my typo; that should have been: "doit:(int*)" in one and "doit:(float*)" in the other. Neither the compiler nor the runtime ever checks that you are using the right argument type for the object you are invoking the method on.
While the post you referring to is simply blathering, in 1990 NT was a good two years into development.
It's all done on the GPU. That's why Expose is blazingly fast even though most of the OS X GUI feels like it's covered in treacle.
On the reference counting issure refer back, I clarified your original statement that there is no garbage collector, to there is no automatic garbage collector. You've partly come round, though you are still denying that reference counting is a type of garbage collection. Again, if you go back to your computer science definitions, you'll find that it is.
Note that I'm making no claim that Objective-C is as robust as C#. Just that you painted it as if it was no more robust than C++, and it is. Use late binding instead of v-tables, and you get a whole lot more stability.
In practical terms the distinctions between C# and Objective-C isn't significant enough to favour one OS over another. There are other far more important issues.
Just that you painted it as if it was no more robust than C++, and it is.
Both are intrinsically unsafe languages; what minor distinctions you want to make doesn't really matter.
it was about it's use in programming for Apple computers, i.e. Cocoa, and you are therefore using NSArrays, and pointers to Objective-C objects.
That argument doesn't work: the ANSI C subset of Objective C isn't just something one drops down occasionally, it's a crucial and unavoidable part of Cocoa programming.
Objective C might have a future if Apple could figure out how to achieve what you suggest: make the Objective C language complete enough so that people do not need ANSI C constructs except in code explicitly declared unsafe. The Tom language was an attempt at updating Objective C. But the way it is, the language has stagnated for the past 20 years.
In practical terms the distinctions between C# and Objective-C isn't significant enough to favour one OS over another. There are other far more important issues.
That's Apple's bet, and I think it's a losing bet: in practical terms, the distinctions mean the difference between being able to build large, robust componentized software or not.
With choosing Objective C, Apple got part of the equation right: they got a dynamically typed language with some reflection capabilities. In fact, as a dynamically typed language, it's nicer than C# or Java. But its not nice enough to make up for its serious deficiencies in safety.
Depends what you mean. It's perfectly possible and indeed normal to program with only type-safe Cocoa and Cocoa derived objects. You have to go out of your way to produce C style type-unsafe code. Which is why it's an academic rather than practical distinction.
That's Apple's bet, and I think it's a losing bet: in practical terms, the distinctions mean the difference between being able to build large, robust componentized software or not.
Come back when there's some evidence to support your case. As it is, OSX is far more robust than Windows and has great components to use. Longhorn is still just an unstable piece of vapourware. Remember the warnings attached to this very Avalon beta (alpha?).
Depends what you mean. It's perfectly possible and indeed normal to program with only type-safe Cocoa and Cocoa derived objects
.NET gives people the ability to write C#-based components today. And Linux has Mono with Gnome bindings for the same purpose. Both support anti-aliasing and scalable graphics. Neither Linux nor Windows need to catch up with OS X, they are already ahead. What's going on under the covers with .NET/MFC and Mono/Gnome may not be pretty (bindings to C/C++-based components), but neither is what's going on under the covers with Cocoa (lots of C code, DisplayPDF, etc.).
No, it's not: every Objective C program uses C pointers, if not for anything else, for member access, and every Objective C program requires manual reference counting. And, as I pointed out, not even the Objective C object system is type-safe due to its buggy implementation of method pointers.
As it is, OSX is far more robust than Windows and has great components to use.
Windows is written in C/C++. And in my experience (I use all three of them), Windows XP, OS X, and Linux are all about equally (un-)stable these days. Yes, Macintoshes crash, and so do Macintosh applications, with about the same frequency as any big C/C++ program. Stability at that level just isn't a selling point anymore, and no system built in a language without garbage collection and bullet-proof runtime safety is going to do better.
Longhorn is still just an unstable piece of vapourware.
Longhorn doesn't matter for the purposes of this discussion.
Again, it's a marketing illusion to think that Apple has any technological advantage with Objective C.
But that's bullshit, dude.
You obviously haven't used Mac OS X. First of all you can adjust "blurriness" of the text. The least strong option looks pretty much exactly as ClearType (except of course it was released WAY before XP).
Second of all, W2K does NOT have realtime compositing, no matter the hardware.
Flicker free drawing is available by using double buffering only and that's why it's dog slow and no one uses it.
Resolution independent graphics is not available - Mac OS X draws pretty much everything as vectors using Adobe Display PDF.
And you must be joking about advanced typography. Apple's typography is head and shoulders above Microsoft's, as it has always been. MS typography APIs don't even know what "kerning" is, fer chrissakes.
There's nothing flicker-free about OS X graphics, either. The OS double buffers (one reason OS X is such a pig for RAM) for you so you can skip some of the optimizations, but double buffering isn't magic and won't stop all flicker. You use exactly the same techniques for no-flicker drawing on Windows as you would on OS X. Double buffering works fine on Windows, and the blit is hardware accelerated, people don't use it because of the memory cost (it's not worth doubling your RAM usage to prevent corner-case flicker).
As for tyography, you've got your head up your ass. GDI does and has supported all sorts of advanced text rendering, including kerning. GDI+ does it a lot better.
Pretty much the only thing in your post which is correct is that Windows doesn't use vector graphics. Yay. That's not the same as not being available, but they aren't built into the system.
Could you point me to some place in MSDN which describes how to output text with kerning in GDI/GDI+? Because if you can't (and I bet you can't, because it's not there) then you're the one with the head you know where.
:0)
DirectX does compositing, but in its current design it only works well with one drawing thread.
And I don't see any flicker in Mac OS X. None. Zero. Double buffering or triple buffering - I don't give a shit. If I need to spend $50 on additional RAM, it's fine with me. It's hard to use Windows and (especially) Linux after Mac OS X.
The ExtTextOut function is the one you want. I believe, although I'm not 100% certain and can't check right now, that the basic GDI+ text drawing functions support kerning. That's real kerning, based on OpenType fonts. Fonts without kerning information will obviously not be kerned.
DirectX does compositing, but in its current design it only works well with one drawing thread. :0)
Does OSX really allow you to draw from more than one thread? Most OSes don't do this because it's a performance hit. You can if you really want to in Windows, but most people don't.
And I don't see any flicker in Mac OS X. None. Zero. Double buffering or triple buffering - I don't give a shit. If I need to spend $50 on additional RAM, it's fine with me. It's hard to use Windows and (especially) Linux after Mac OS X.
See, now you're redefining your claims. You can do flicker-free drawing on Windows if you want to. It's a design choice based on trade-offs with RAM usage and performance vrs. memory usage. A lot of the native windows controls aren't double buffered because the memory usage would be extreme. Windows does have a couple glitches that make flickering common, like the seperate WM_ERASE and WM_PAINT messages, which I won't defend, but you can work around those. OS X chose to sacrifice performance and RAM in the interests of global double buffering. Windows didn't, which makes far more sense for it's market. It's not a case of "Haha M$ sucks they can't even double buffer", it's an informed judgment call. I've seen OS X apps that flicker, by the way. It's not hard - you just need to do naive Windows-style drawing, where you erase and then paint.
It was in joint developement by Microsoft, DEC and others for almost a decade, but there were not versions available to the public until years later, when Microsoft took over the project completely and released Windows NT 3.2 (I think that was the first public version, I know 3.5 and 4.0 sold much better).
First public release was 3.1 in 1993. 3.5, 3.51, 4.0, 5.0 (Win2k), 5.1 (XP) and 5.2 (Win2k3) came in 1993, 1994, 1995, 1996, 2000, 2001 and 2003 respectively.