Windows Vista Beta 2

They just released Beta 2 to the public if any of you out there have been looking to ty it without being an MSDN subscriber. I'm downloading it now.. I'll post later this weekend with an update I guess.

Rating: D

don't try the beta at all if you can help it.

Pros:

looks fancy

some nifty features

Cons:

3 hour install. I installed it from harddrive (no DVD
invovled) and it took 3 hours form the time it quit requiring my input
to the next time it required input.
Crashes. In the 3 hours I used it my computer blue
screened/crashed without chance of repair at lease 15-20 times.
Yes I updated drivers to vista drivers, no it didn't help
anything. Something would always crash on startup ("system fader"
or something) causes the startup to look like it was done... but
if I did ANYTHING explorer would crash.

3. 50% of the time I did anything explorer would crash. IE didn't work at all, firefox did.
Does not allow Windows XP to be installed over it as an
upgrade. appearantly Vista Beta thinks it's an upgrade to XP and
that XP is a downgrade from it. Very wrong. don't expect to
get back to XP without a reformat or some serious work.

conclusion, if you have nothing better to do this weekend give it a
shot. you'll get to at least see what it is you'll be looking at
for 4 years. If you have a dog/girlfriend/wife/deck
of cards/2 thumbs your better off playing with them/that/her than
trying Vista.

CodeWhisperer

Oof!

What sort of hardware were you running it on? That sounds particularly egregious...I mean...beta 2? That should be able to run for at least 30, 40 minutes without crashing. :)

-cw

Intel 2.6 GhZ w/ 1Gb ram. that's pretty much all that you need to know I guess. I also thought I would be able to open and run programs without restarting two or three times. however, I couldn't. I'm not seeing an early 2007 release at all.

tufty

@tster said:

Intel 2.6 GhZ w/ 1Gb ram. that's pretty much all that you need to know I guess. I also thought I would be able to open and run programs without restarting two or three times. however, I couldn't. I'm not seeing an early 2007 release at all.

You're not alone. Of the few people I know brave and / or foolish enough to try this out, all but one have come back with the comment "Piece of crap, don't fucking bother", even the one person who has thus far had a positive experience slammed explorer and general usability.

I think, however, you will see an early 2007 release, I don't think MS can afford to push it back again. Apple are getting a lot of good press with 10.4, and 10.5 will be out (without "beta" status, natch) around the time Vista comes out, Google spreadsheets and Writely combined threaten to eat a lot of "basic" office customers. MS are in a tight spot, and they know it.

The way I see it, MS have to get this out, or they lose a lot of face; they have already slashed most of the really interesting stuff from its release and delayed repeatedly, thus cracks in what they are showing as the beta will get papered over, it will get pushed out the door, and then "fixed" as time goes on. If what they release bluescreens (or whatever colour they change the screen to), though, they are fucked. People are beginning to realise that computers don't have to crash.

Oh, and (marginally related) http://it.slashdot.org/article.pl?sid=06/06/09/1724239&from=rss

Simon

what this tells me is that I should wait a long while before installing vista once it comes out, even though I'll get it for free from my university.

lpope187

@tster said:

Rating: D

Cons:

3 hour install. I installed it from harddrive (no DVD
invovled) and it took 3 hours form the time it quit requiring my input
to the next time it required input.

Crashes. In the 3 hours I used it my computer blue
screened/crashed without chance of repair at lease 15-20 times.
Yes I updated drivers to vista drivers, no it didn't help
anything. Something would always crash on startup ("system fader"
or something) causes the startup to look like it was done... but
if I did ANYTHING explorer would crash.

3. 50% of the time I did anything explorer would crash. IE didn't work at all, firefox did.

Does not allow Windows XP to be installed over it as an
upgrade. appearantly Vista Beta thinks it's an upgrade to XP and
that XP is a downgrade from it. Very wrong. don't expect to
get back to XP without a reformat or some serious work.

Wow, you had a SHORT install. Try installing Vista on a virtual machine, you'd be lucky to get it done in under 7 hours. I've even heard about installs taking close to 24 hours. The good part is you don't need to reinstall XP when your done - just delete the virtual disk.

When it comes to MS software, I typically don't deploy until the first service pack. Give them the time to fix the critical bugs. MS has been doing better recently due to their increased use of CTPs and Betas. SQL Server 2005 was extremely stable and robust because of this.

Larry

@Juifeng said:

I recently upgraded Ubuntu Breezy Badger to Ubuntu Dapper Drake over night, and by "upgrade" I mean download and install.

Vista surely looks nice. It will be a nice system to play games on. If it allows games to use some graphic card capacity after it applied all the fancy effecto. :)

Coding..? No, thanks!

Well, seriously. .NET/C# are some nice tools, never expected something that useful from MS. Their IDE seems to be nicer than MonoDevelop too. And XP is 1000 times better than Win 98, and I hope that Vista will be even better.

Yeah. That Dapper upgrade was easy. Only thing I had any problems with was the wpa_supplicant change.

Am I the only one who thinks that Vista looks like it has neat stuff from that company in Redmond, and that it's not another ME, but are actually going to pull this off? That Vista will supply compelling reasons to upgrade from 2k/XP? Maybe even gain some server ground from Linux?

Ok, the user stuff, I'm not sold on. I think that by the time Vista hits, XGL will have gained a lot of support in Linux. Search could be handled just fine by an external application. IE7 won't have enough user share for some time to make it so that web developers don't need to add hacks for IE6. WinFS, Monad, delayed.

That said, I'm excited, or at least curious, about the underlying system. Updated audio and networking stacks, memory manager and IO scheduler. Better privilage seperation. Deadlock detection. Packaging system. Especially WPF and WCF. I'm not happy about a whole new set of proprietary tools [I believe claims of openness from Microsoft like I'd believe John Titor on stock prices, or Infinium labs on their console], but I am interested by the changes.

I heard the memory manager is suppose to "remember" patterns in programs and when it sees you running a program that uses specific pages it preloads the pages that you usually use next. As far as graphics are concerned. the glass effects looked cool, for the 20 seconds that I saw them before making everything look like good old windows 98.

tufty

@TheDauthi said:

Am I the only one who thinks that Vista looks like it has neat stuff from that company in Redmond, and that it's not another ME, but are actually going to pull this off? That Vista will supply compelling reasons to upgrade from 2k/XP? Maybe even gain some server ground from Linux?

Dunno if you're the only one, but I'm certainly not with you.

Assuming they get it out of the door, MS will "pull it off", of course, but only because consumers will have no choice - new machines will come with Vista pre-installed. Compelling reasons to upgrade? I don't see any, to be honest, except that new software will be "vista only"; it will be a forced upgrade path, as per always.

Grabbing server share back from the free OS's is unlikely to happen, and even if it did, would that be a good thing (except for MS shareholders)? Diversity is good for everyone.

Why would someone choose Vista as a serving platform? Much of the "good stuff" has already been ripped out "to be added later". It's unproven, security wise and stability wise. Servers don't need funky graphics; hell they don't need graphics at all. What they need is a proven stable and secure OS running secure and stable software that is patched on a regular basis.

As an aside, there's the old FUD that "<non-windows OS> doesn't get viruses because
there's not enough machines running <non-windows OS>, so nobody bothers". If that's the case, why
would I choose a platform that is ubiquitous and constantly under
attack as my server platform?

For serving, it comes down to costs against benefits. If you're running a server farm, you have several costs:

Hardware costs (purchase cost, running costs and hardware failures)
OS and software licensing costs
Maintenance costs (sysadmins)
Downtime and security breaches

The first is roughly the same across platforms
The second is a no-brainer, but realistically only comes in when you have a number of machines (see next point)
For the third, a good *nix admin costs more than some MSCE cretin, but can do more. A lot more. For a small installation, it possibly makes sense to go MS, as sysadminning is cheaper, but you need generally more admin per machine than with *nix.
The fourth is also a no-brainer. You need machines that don't go down and don't get hacked. If MS get it right, then they should be able to compete on this. Based on previous performance, I'd say that's somewhat unlikely. Also, see my aside above.

So it comes down to a tradeoff of software costs against admin costs. And in my experience thus far, except for one or two machine installations, or cases where the software cost can be pushed "elsewhere" (by which, of course, I mean "onto the users without them noticing"), it makes vastly more sense to have a decent admin running machines with open source software running on a platform that is known to be stable and secure.

If you want serious "enterprise level" (and here I'm not using the ironic "enterprisey" meaning, but the real "you can trust your business to this" meaning) serving then you're better served with Solaris or *BSD than anything else, Linux included.

For servers at least, the OS is irrelevant, it's a commodity item, it has a "value" of zero. It's getting that way for a lot of other software, and MS (along with a lot of the other "dinosaurs") are being very slow on the uptake. Open source software is catching up, and in many cases far outstripping, the best that "commercial" software can deliver. Commercial software has to deliver something far ahead of what open source can deliver to even get a foot in the door for new installations, and for existing installations it's more a question of inertia than anything else that they are still there.

I could go on at length about this, but I'll stop here.

Simon

@tufty said:

@TheDauthi said:
Am I the only one who thinks that Vista looks like it has neat stuff from that company in Redmond, and that it's not another ME, but are actually going to pull this off? That Vista will supply compelling reasons to upgrade from 2k/XP? Maybe even gain some server ground from Linux?

Dunno if you're the only one, but I'm certainly not with you.

Assuming they get it out of the door, MS will "pull it off", of course, but only because consumers will have no choice - new machines will come with Vista pre-installed. Compelling reasons to upgrade? I don't see any, to be honest, except that new software will be "vista only"; it will be a forced upgrade path, as per always.

Grabbing server share back from the free OS's is unlikely to happen, and even if it did, would that be a good thing (except for MS shareholders)? Diversity is good for everyone.

grins You're... preaching to the choir, to say the least. I switched over to linux instead of upgrading to windows 95. I do keep one dual-boot box around for gaming, when I can't get something running under Cedega. That leaves 3 computers running various unices. At work, I have to deal with both, but I ensure that everything I write will work in Linux. I've been slowly working on the large-scale conversion of a fortune 500's client PCs, and I have backing from our security teams, my management, and different parts of IT. Unfortunately, not our programmers.

Everything I know about windows, I learned at my job. However, I'm of the mind that the competition in the market itself is good, and that Vista is Microsoft's first sign that it knows it IS competing. As well, since Microsoft has, for the forseeable future, achieved a great deal of vendor lock-in, I would like to see the platform that I, as a programmer, will have to deal with to be as stable and programmer-friendly as possible. This is better for the users of that platform. I do not believe that their lock-in is sustainable in the long-term.

@tufty said:

Why would someone choose Vista as a serving platform?
Much of the "good stuff" has already been ripped out "to be added later". It's unproven, security wise and stability wise. Servers don't need funky graphics; hell they don't need graphics at all. What they need is a proven stable and secure OS running secure and stable software that is patched on a regular basis.

Yes. Which Microsoft looks like they're trying to fix. The GUI is not the only thing of interest in Vista; in fact, I discounted it as largely unimportant in my earlier post. In fact, I pretty much discount anything they aim at users; I tend to find that Microsoft doesn't have a clue about what users want, they deliver what they want the user to want. Henry Ford could pull that off, but Microsoft hasn't managed yet. The stuff aimed at developers and admins are useful... and that's what Microsoft's platform is based on.

@tufty said:

As an aside, there's the old FUD that "<non-windows OS> doesn't get viruses because
there's not enough machines running <non-windows OS>, so nobody bothers". If that's the case, why
would I choose a platform that is ubiquitous and constantly under
attack as my server platform?

Ok, we won't discuss why it's FUD. There are other reasons other platforms do better than Windows. I hear this one about once a week.

@tufty said:

For serving, it comes down to costs against benefits. If you're running a server farm, you have several costs:
Hardware costs (purchase cost, running costs and hardware failures)
OS and software licensing costs
Maintenance costs (sysadmins)
Downtime and security breaches
The first is roughly the same across platforms
The second is a no-brainer, but realistically only comes in when you have a number of machines (see next point)
For the third, a good *nix admin costs more than some MSCE cretin, but can do more. A lot more. For a small installation, it possibly makes sense to go MS, as sysadminning is cheaper, but you need generally more admin per machine than with *nix.

All true. You are missing several points, however. Chief among them is that word that Ballmer liked so much. If Microsoft provides better development tools, libraries, etc, some developers will choose their platform over the others that are available. In a mixed shop, like mine, if a developer says he needs a windows server for the software, he'll get it. If this is for interal software, security is not going to question the decision much.
The admins won't care [of course, most of our admins are admins for the Microsoft platform].

I don't think there would be as many IIS servers if .NET weren't useful.
@tufty said:

The fourth is also a no-brainer. You need machines that don't go down and don't get hacked. If MS get it right, then they should be able to compete on this. Based on previous performance, I'd say that's somewhat unlikely.

The quality of Microsoft's software has noticably improved since ME. The big exploits are fewer in number. After enough million times of saying that Microsoft is going to need to re-architect different subsytems, they look like they're finally doing exactly that. I'm reserving my final judgement, but if they deliver on half of their promises, it looks like a big step forward. And delivering about half of their promises is quite believable, based on prior experience.
@tufty said:

Also, see my aside above.

So it comes down to a tradeoff of software costs against admin costs. And in my experience thus far, except for one or two machine installations, or cases where the software cost can be pushed "elsewhere" (by which, of course, I mean "onto the users without them noticing"), it makes vastly more sense to have a decent admin running machines with open source software running on a platform that is known to be stable and secure.

If you want serious "enterprise level" (and here I'm not using the ironic "enterprisey" meaning, but the real "you can trust your business to this" meaning) serving then you're better served with Solaris or *BSD than anything else, Linux included.

And I say that I want to use the best product possible for my needs. Open source doesn't always meet those; and no non-Microsoft platform implements the full windows API. When OSS meets my needs, I use it. When it doesn't, I use a closed product.

@tufty said:

For servers at least, the OS is irrelevant, it's a commodity item, it has a "value" of zero. It's getting that way for a lot of other software, and MS (along with a lot of the other "dinosaurs") are being very slow on the uptake. Open source software is catching up, and in many cases far outstripping, the best that "commercial" software can deliver. Commercial software has to deliver something far ahead of what open source can deliver to even get a foot in the door for new installations, and for existing installations it's more a question of inertia than anything else that they are still there.

I could go on at length about this, but I'll stop here.

Simon

Inertia is not a discountable market force. Yes, OSS is catching up, and passing, closed solutions. As someone who uses OSS often, I'm glad. I want better hardware support for X, I want better wireless. I want more applications. I don't want to have to check up on drivers for hardware before I buy it for a Linux box. I want Microsoft's death-grip on the market to dissappear. But not for the same reasons you do, I think: I simply want better software, whether it's open or closed. Competition. That's what I want to see. Being open is only one data point to me. Being useful is another one.

tufty

Only replying to stuff I disagree with here...

@TheDauthi said:

I've been slowly working on the large-scale conversion of a fortune 500's client PCs, and I have backing from our security teams, my management, and different parts of IT. Unfortunately, not our programmers.

Then, to a certain extent, you may have the wrong programmers. If someone can't make the step between one platform and another (with a certain amount of retraining), they they are no good. Programmers don't (or, at least shouldn't) drive what a business wants to do, except in some very restricted circumstances.

@TheDauthi said:

All true. You are missing several points, however. Chief among them is that word that Ballmer liked so much. If Microsoft provides better development tools, libraries, etc, some developers will choose their platform over the others that are available. In a mixed shop, like mine, if a developer says he needs a windows server for the software, he'll get it. If this is for interal software, security is not going to question the decision much.
The admins won't care [of course, most of our admins are admins for the Microsoft platform

Do they provide "better" tools and libraries, though? Or do they just provide new and improved versions of what they have provided before, to keep their locked in, "I only know MS" programmers thinking that they are working with the best? I don't know, personally, in 20 years in the business I've managed to never code for Windows, so hey. I don't even particularly like IDEs, I prefer a good text editor and a bunch of terminal windows.

The main reason I discounted development cost in my previous post was that, again, over 20 years, I haven't seen the cost of software development come down appreciably. What I have seen, time and time again, is a "new technology" project (C++, Java) being used as a springboard to a consultancy career, with the inevitable high staff turnover and poor understanding of the code resulting in late delivery (if at all) of a poorly implemented system. I can't see .NET being any different to that.

@TheDauthi said:

The quality of Microsoft's software has noticably improved since ME. The big exploits are fewer in number. After enough million times of saying that Microsoft is going to need to re-architect different subsytems, they look like they're finally doing exactly that. I'm reserving my final judgement, but if they deliver on half of their promises, it looks like a big step forward. And delivering about half of their promises is quite believable, based on prior experience.

Yes, they are getting better. For the moment, yes, the big exploits are fewer in number. But they are still there, and still cropping up regularly enough, to be a major worry. It doesn't even have to be a big exploit to get you rooted, a combination of small exploits can work just as well. Of course, MS are far from alone in having to deal with this, but they have a lamentable track record thus far, in terms of pure number and severity of exploits, willingness to accept that they exist, and time to fix. Still, they are getting better.

Part of their problem is a "need" to be backwards compatible. Does anyone still actually need to run those Win3.1 or DOS programs from 19mumblety? Or even from 5 years ago? They should take a leaf from Apple's book, IMO. In the same timescale as Win3.11 -> Vista, Apple went from a single tasking OS to a co-operative multitasking OS to a full-blown unix-alike system with plenty of dead ends and 2 hardware platform changes along the way. You can no longer run a lot of "Classic" software without a separate emulator; indeed the "Classic" environment no longer exists. It's dead.

@TheDauthi said:

@tufty said:
If you want serious "enterprise level" (and here I'm not using the ironic "enterprisey" meaning, but the real "you can trust your business to this" meaning) serving then you're better served with Solaris or *BSD than anything else, Linux included.

And I say that I want to use the best product possible for my needs. Open source doesn't always meet those; and no non-Microsoft platform implements the full windows API. When OSS meets my needs, I use it. When it doesn't, I use a closed product.

Oh, I agree entirely. I'm not some RMS open source "it has to be free" zealot. But why does no OSS solution implement the full windows API? Because MS don't want you to be able to interoperate. They want to keep their monopoly. If you can interoperate, you can migrate away. If you can migrate away, there's no reason to stay with MS unless they start providing something really innovative. And despite all their hype about innovation, when was the last time they actually did innovate?

When it comes down to it, you have to make a value judgement. Does paying to have the full windows API available deliver you real business benefit, and does that benefit beat any benefit you could get by using another solution that avoids vendor lock-in?

@TheDauthi said:

I want Microsoft's death-grip on the market to dissappear. But not for the same reasons you do, I think: I simply want better software, whether it's open or closed. Competition. That's what I want to see. Being open is only one data point to me. Being useful is another one.

No, I think we want the same thing. I use plenty of "closed" software; hell, my main working environment since I stopped using Linux on the desktop in 2001 has been OSX. My preferred editor is closed source. If it's better, and the benefits outweigh the costs, I'll use it.

As for MS, I have no great hatred for them. They are a dinosaur, in the same way that Oracle and Sybase are, and in the same way that IBM were at the tail end of the 80s. If they want to avoid some years in the desert, they are going to have to do one of two things:

Reinvent themselves, massively
Fight tooth and nail to keep their monopoly

They are doing the latter while trying to look like doing the former, and that's not going to work IMO. There's a lot of people gunning for them, for every product on their lineup there's a multitude of arguably as good or better products out there. Inertia is keeping them in place, but inertia will fail them.

Anyway, need to go and put some client software live now

Simon