That'll just cache the JRE, though. I'm not even sure if Linux is one of the platforms that supports it.
If it works the old UNIX way, that doesn't really fix that you end up with multiple copies of the JRE in memory. You just won't have the disk load delay to read the JRE into memory.
Let me know if I'm off on this, because if it only ever loads one copy of the JRE, that would be really useful!
Since you obviously must have single handedly written all version of Windows, as you purport to know everything about them and be the ultimate authority, you must have suffered major amnesia.
Windows 2000, being based on Windows NT, contained a completely different kernel from Win9x. It included no DOS code, unlike Win9x. It had WOW to run Win16 and MS-DOS code. It featured the HAL for all device interaction. These things are 100% different from Win9x. Task management also shared no code with Win9x, and supported multiple processors. The VM subsystem was also completely separate code.
Additionally, window management went through a different set of libraries, and the GDI libraries was written native to NT. That makes the UI 100% different from Win9x.
Filesystem access were different to support NT extended attributes, NTFS, and encryption. The security model was 100% different as well, seeing that Win9x didn't even have one. The on-disk partitioning was done differently, as well. NT has the ability to do software RAID and dynamic disks, which Win9x has no support for. This was because of completely different partition code.
The network stack was native to NT, and did not share code from Win9x. NT used MS-RPC to do IPC tasks, which was different from Win9x.
All of the management tools were native to NT, sharing no code with Win9x; same for the domain membership code.
NT had full locale support, which Win9x doesn't have.
So, despite your completely incorrect insistance, Win2000 did not share code with Win9x. Of course, the bundled applications shared *some* code, but that is not the operating system by any stretch. Parts of IE, WMP, DirectX, and apps like calculator and wordpad contained shared code. Notepad was new code. IE, WMP, and DirectX for NT had to be maintained somewhat separate from Win9x versions because of things like the NT security model. So the NT version could not run on Win9x.
That would, oddly enough, make the switch from Win9x to WinNT at a much bigger change than WinNT to WinNT. The vast bulk of the difference between XP and Vista are application changes. The change from Win9x to WinNT involved a completely new operation system, large amounts of new API, a suite of new applications, and different code in many of the existing applications. That is a *MUCH* bigger change. Even going from Win3.1 to Win95 didn't involve as much new code, seeing as to how Win95 executed in a DOS extender environment, running with many DOS APIs.
I'd say that you've proven rather effectively that *you* don't have a clue what you're talking about. Why don't you go be a fanatical, screaming, hand-waving zealot somewhere else.
There is more issue than that, actually. In addition to the misunderstanding of combustion efficiency vs. energy transfer efficiency, the MPG number are quite misleading.
Not only were the test conditions not the same, but the manufacturer tauted DOT numbers are usually quite conservative. Many drivers will achieve higher MPG than what is published. It's possible that the acheivement of ~26mpg vs the published ~22.5mpg could've been achieved by conservative throttling, different oil weight and viscosity, different octane fuels, different altitudes, etc.
I just hope that the JREs out there get a lot more streamlined, and become available on more platforms by then. The OS/architecture independant nature is great, but the Java platform isn't available everywhere still.
I think one of the biggest impediments to Java for applications is that you have the load the JRE over again for each app. That's really annoying, especially considering how many apps I tend to have started. I have six apps running right now, and if I had another 20MB per app, that would be just a bit horrid!
Definitely true, and doing it the client way is also really obnoxious to the server, in addition to being slow for the user. I'd love for the IMAP RFC to get cleaned up, and one of those changes would be for multi-folder search.
The only reason I ever use mutt over Pine is because mutt natively supports Maildir. That means that my mail server has mutt installed so that I can read local mailboxes without going over IMAP, just in case something bad happens. For normal use, my text-mode reader of choice does happen to be Pine.
Bub, you don't have to be so hostile just because we're disagreeing! Your method of discussion is not very positive. Looking at your post history, you really should drop the habit of insulting people. I might say that someone is wrong, but I don't call them a moron or an idiot.
I wouldn't consider it to be the most major upgrade simply because 9x->NT was a transition of all parts of the system. There was no shared code between the two products! The intent of the system designs were completely dissimilar. Going from XP->Vista is keeping most of the system the same. Aside from the new APIs, most of the changes are application level. Those APIs are being backported to previous releases, so you can't say that Vista is different on that principle. That leaves the main differences between 2k/XP and Vista in the applications. That's not nearly as big a change.
So, please, calm down on the bloody APIs. I'll have those on my XP desktop if I feel like installing them. I don't need Vista to get them, so if anything, that makes the API changes a big deal, and not that MS is releasing them alongside a new OS revision.
What I said about every other release being the biggest change I guess didn't clarify enough. I didn't mean that *I* thought they were the biggest thing since 95, but that was the marketing spin from MS. They'll always bill the new OS in that sort of light as it helps move product. For the *user*, the biggest change they'll notice is almsot certainly going to be the interface. Considering the amount of debate that the new UI is worse than the old one, I wouldn't necessarily count that as a good thing.
As far as the UI offloading, no, I don't think it's a bad thing. I'm saying it just isn't a big deal, and certainly not a good enough reason to justify all the time and expense that doing the update will be. The reason that UI offloading will improve things so much is because they put so much candy in the UI. The reason other platforms don't talk up this type of thing as such a big deal is because it isn't a big deal. For MS, it's a selling point, for anyone else it's either unnecessary, as their UI doesn't consume noticable resources, or it was an improvement they quietly made so that they could do candy interfaces.
For LUA, the places where you were likely to see it being used was in business. More business apps that you seem to realize are not designed well enough to work as a limited user. I'm serious about the sheer number of these applications. The most that the typical home user will run into that doesn't work as a LUA are games. For the home situation, this will work great. For the business situation, it won't be as a big of an improvement.
You don't seem to realize that for people with non-MS background, the things MS does aren't new. Four of the big five tenants of Cairo that are finally being finished the non-MS world has had for over a decade. We're had RPC since the 80's, messaging since the 70s, and directories since the early 90's (and in some capacity, a fair bit before that). The ideas of powerful search and file metadata aren't anything new, either. You have products like Notes that have done a lot of that for a long time. It just isn't new.
You shouldn't be so surprised that a lot of people that aren't heavy MS people aren't at all excited about Vista. It doesn't bring anything new to the table. Most of these changes bring Windows to where many other OS' have been for a long time.
That's two applications that you're using. How many apps do you use that aren't written in Java?
To counterpoint: when in Windows, I run BitComet instead, since it is less resource hungry.
My Java use consists of Dell OpenManage (which slams the host machine for memory), and HP WebAdmin (which crashes the JRE a lot). Parts of OpenOffice use Java. The LDAP Browser is written in Java. I'd prefer that OpenOffice was straight C++, and that I would find a better LDAP browser. On Linux I use Azureus, because I haven't found a better native implementation.
There aren't many desktop Java apps that people use. OpenOffice and Azureus are probably the two biggest, and most people don't run either. It isn't that Java apps are bad or anything, but that many of them have better equivalent native implementations. People also don't like the JRE load time.
Java desktop apps aren't dead, people just don't have a need for them. If Linux starts being use as a desktop platform in large numbers, then you'll have Mac, Linux, and Windows. Then I think you'll see a whole lot more Java apps.
IMAP multi-folder searching is actually a vendor extension of UW-IMAP, and I believe Cyrus. It isn't in the IMAP spec and doesn't work against all servers. Pine is written by the guy who does UW-IMAP, so it supports his extensions. It is called a reference implementation, but that is not the case. That is why many IMAP implementations do not support it.
The Java applet sandbox is safe, but this is a Java application. That means it can make unsafe and system modifying calls. You have less to worry about with an application in Java, but you certainly aren't inherently "safe". This email program could still be compromised, though it will be harder to do.
That is to say that even if the JVM cannot be exploited, your application logic still could be.
Only part of running LUA is protecting yourself voluntarily. The other part is locking the machine down so that a user can't do unwanted actions. The main reason that LUA doesn't work right now is that apps don't work without the admin privs. They do things like write to HKLM, %Program Files%, %WINDOWSDIR%, etc. Prompting for a password works for users that *are* the admin. It doesn't fix the major problem of the apps trying to write to admin-only priveleged locations during normal operation. That's the real reason that LUA doesn't work right now, and that's why this feature doesn't fix the problem.
The UI *really really* doesn't take up hardly any CPU time. I feel rather comfortable saying that there would be no difference if you are using the 2k interface. There would be a little difference in the XP/Luna interface, and tremendous difference with Vista/Aero. The reason there is more difference as you go is simply because MS created the problem, and then solved it.
You do realize that *every* MS release is an overdue upgrade, right? They're always fixing something that most of the industry already knew was the wrong way. Often these are things that people have complained about for over a decade. Things that MS ignored fixing because they were too busy adding paper clips or another API. People go off on the nothingness of MS upgrades because they usually don't fix the real problems. They just add more features that probably won't be used, make it look different, do things that will require retraining, introduce piles of new bugs, etc.
Vista isn't the largest upgrade since 95, either. I don't understand why people completely miss this one. MS says it's the biggest thing, but they're trying to market a new product against all their old products that have a massive installed base. The biggest upgrade was going from Win9x to WinNT with the 2000 upgrade. It was a complete and total change of technology. Technologically, it was a bigger change than 3.1->95, seeing as 95 still had all the legacy code that it depended on. How is an incremental bugfix and feature add release the biggest thing?
I seem to remember XP also being the biggest upgrade, and 2000, and Millenium. That's just marketing-speak. It's taken MS five years to get most users off of Win9x, and they're still working on getting people on WinXP. You're right, most people don't care what OS they use, and that's another reason why few people are going to upgrade. If it comes with their computer, then they'll use it, unless it's broken with various unwanted misfeatures. Then they'll ask their children or friend from work or whatever if they can get back their old version.
Vista was supposed to represent the last piece of the Cairo vision. MS has added a piece with every NT based release. Of course, it's taken 16 years to get all the pieces, and Vista won't even include WinFS in the RTM.
My reasoning is just from a different standpoint than yours. I'm looking at this as someone that admins machines, uses machines, and trains users. I know what gives me and my users a hard time, and I'm not too happy about those things not only getting carried forward, as defaults, but becoming even more of an issue.
I'm not critisizing MS for fixing their bugs, I just don't think it's a huge deal that they're doing it. These things should never have survived in their software for as long as they have. I'm glad that they're fixing them, but I hate that they're making fixing their mistakes and bugs into "huge-mega-incredible innovations that will change the world."
BTW, I do believe that proprietary is bad. It fragments software and makes interoperability harder. I'm quite sick and tired of MS pulling that kind of trash, and a whole lot of other people feel the same. It's the reason that so many people are starting to wake up and dump Office. I don't care if their implementation is proprietary, but I do care if the API, format, protocol, etc *is*. MS has proven that they can't be trusted with that sort of thing, and I certainly don't wish to get further burned by their behavior.
You need to have very compelling reasons to convince business users, like me, to upgrade. Especially considering the cost that MS is going to make this particular upgrade. My machines that ran W2k could run XP just fine. My XP machines won't run Vista just fine. What is Vista offering going forward from 2k/XP that I don't already have? That answer is "not very much at all".
It *does* matter what software is out there, too. For example, imaging is already taken care of, and most already have third party software for doing it. The only way that changes is if MS *purposefully* breaks that software. They've done that in the past with various things, so who knows, perhaps they'll do it again.
As I said, my reasoning starts from "Why would I want to upgrade to this?" I then look at the feature list, and see a whole lot of things that I don't care about. Then I see a good number of things that I outright do not want. Along with that, I see a good number of things that I already have from third party software. And finally I see a small number of things that I actually want. So, for a bunch of money, a machine upgrade, and a lot of things that I don't want, I can get a couple of things that MS is backporting to what I already have.
Even if MS put in enough features that I was willing to upgrade, I would wait until SP1. Nobody in their right mind goes with MS software before they do a service release! There's a reason for that: their initial release of anything is bad; it malfunctions, munges data, and presents a massive security threat. That means that even if I wanted to run it, I would be waiting around a year for them to fix it so that it actually works. As an example, I just started moving to XP on my user machines this past spring. It had finally gotten to where it was trustworthy; bugs had been worked out, wierd MS peculiarities had been documented, workarounds existed.
Bottom line, though, is that this release does *not* offer me much, and it does *not* offer business users very much.
---
Volume - My understanding was that it was an API change, not an end-user interface. If it's end-user, then it's more useful. Ideally, this shouldn't be necessary, and I think it'll cause confusion as an end-user tool. I would use this feature, but I'd really hope my users didn't figure out how to do it.
LUA - If users have an admin level password, then you have no control over the machine. This password display is a home-user only feature, since it would be insane to do in a locked down environment. Additionally, the reason that apps do not with in the current LUA implementation is that they weren't properly written. If those vendors actually *fixed* their software, then this problem would be gone. MS isn't fixing the problem, here, they're just changing the probl
MS can do pretty much whatever they want with their software, for all I really care. They're just getting deeper into releasing software that barely considers the customer. In MS' case, it often *is* a problem when they integrate more things into the OS. They're a convicted monopolist, which means they abuse their ability to do exactly that.
My complaint is that they claim "innovations", which only means "new to our software, but existed everywhere else for a decade." Most of the time, the MS implementation is lower quality than the existing implementations, but prevails because of their monopoly.
What is changing is so riculous that it doesn't matter to users. I say that there is no difference, because what is different is irrelevant. I don't care about changes that *aren't* useful, so if that's what the changes are, then little has changed to justify a switch of platform. In that regard, it offers no more than WinXP/2k already do.
You have to consider the market for the product. Basically you have end-users, and corporate users. Many of the features don't matter *at all* for end users, and there are as many bad changes as good for corporate users. MS looks to have managed to make Vista harder to administer than 2k/XP.
I've given quite a few examples, but to give you examples from your own points:
APIs - So now we have Win16 and Win32 [relagated to WoW on Win64], Win64 where applicable, the kernel API, MFC (various),.NET, WFC, and probably others that I'm missing. They all have security issues because they never finishing fixing any of them. You have apps written to all of them. You do have backwards compatibility. You don't have a compelling reason to use one over another (except for avoiding MFC). It bugs me because it is poorly done.
Per-app volume control - trivial to do yourself
Network stack - security is always good; are the other changes going to be used? Probably not.
Indigo - implementing something that will almost certainly be proprietary, and that is already done elsewhere.
Avalon/Aero - you say there's nothing else comparable; I say that you're wrong. Also, most of the features that they're showcasing with Aero make the system harder to use.
Search - This might be a real, tangible improvement. It's been implemented elsewhere, but not to the same scale.
Mini-apps - irrelevant, already possible. The smaller footprint isn't going to be realized anyway.
Task Scheduler - Maybe. Fixing the current implementation is good, but it isn't used that much anyway.
LUA - still won't work right until third-party devs fix their broken apps. If that happened, LUA would work anyway. Giving users a password for admin-level changes defeats the purpose of locking the machine down.
App jailing - or they could fix their implementation to not do stupid things. Having the feature is good, though. This will encourage bad habits.
App crashes - apps shouldn't crash anyway. App code needs to be fixed. Improving the handling is good, but will encourage bad habits.
Transactional FS - this has been available elsewhere for quite some time. Its addition is welcome, but not generally useful for end users on a consumer OS. Apps should handle the conditions that this protects, already, NTFS handles the rest.
Transactional Registry - the registry is a bad bad idea as is. Hacking transactions into it doesn't change this. This wouldn't be a problem if it was done in a sane manner to begin with, especially in conjunction with the FS transactions.
(System Restore - this is a function that could be implemented on most any other OS. It is not currently implemented on other platforms, to my knowledge.)
Aux display - this is irrelevant. Devices with status displays already have drivers for them, and would still require the same for this.
Improved performance - always good. App loading improvement isn't really necessary, especially on the systems this will run on.
You're right, they are doing a lot of new things. Too bad all of them fall into a few easy catagories: things already done, things nobody wants, things that don't matter, things that are good.
Many of the changes are in the "things already done" catagory. A few of the big ones are "things nobody wants", which is to say that nobody=customers, which are who should matter. The majority are "things that don't matter". A few are good, like LUA, the network working right, etc. Of course, those in the last catagory are how it should already have been.
Now a more realistic look at things. MS is changing APIs around, *again*. That is highly annoying, as apps need to be ported again, tested again, etc. There isn't really a benefit from doing it, aside from supporting the newest MS stuff that will invariably be abandoned by them in the next release. They're adding all sorts of support nightmares, like the default encrypted filesystem. They're changing the GUI *again*, forcing users to need retraining. They're changing the driver model *again*, forcing more rewrites and testing.
The most useful changes that they're making seem to be in fixing things that they broke to begin with. Their bad network policy, sleep implementation, service defaults, browser design, email program, user permissions, malware, search, etc. Much of this will require application changes to get to work right. Yay.
Most of all, all of the stuff that really is talked about will be available on non-Vista platforms. If you actually *read* that winsupersite link, you can see that most of the things listed are either minor, or the same feature.
In short: No, there isn't much else. Vista is XP with a shinier interface to disable, a few hotfixes, and new versions of a couple of apps. Absolutely nothing to justify a new OS.
Spot on with what you said. I'm glad there are other people willing to speak out on exactly that. Teacher unions are unreal and absolutely horrid to work with. The way most unions now do the pay negotiation completely leaves out any concept of merit pay. So forget earning promotions, you're entitled to them by union contract based solely on seniority and such.
And people wonder why work ethic in the USA is so bad...
Second, I didn't say humans were the most likely to survive, just that they are the most adaptable. Cockroaches don't need to adapt, they are simply versatile. No other species can adapt the environment to themselves, and that makes humans something special.
Can you think of any event, that isn't likely to wipe out life on earth, that would wipe out humanity? Sure, a comet impact would kill off humans, but would global warming, an ice age, a supervolcano, etc? We'd probably manage to survive even if the atmosphere took on a poisonous gas mixture.
It is not the first time Intel took a gamble on a new ISA and lost. Hell, the very existence of x86 was as a filler move between non-x86 product lines. The ISA is one of the worst out there, and it *can't* get better.
It's a real loss to everyone that the industry won't just drop the damned architecture. MIPS64, PowerPC, IA64, Alpha, and UltraSPARC are vastly superior to x86. One of the reasons that all of those architectures have lower clock speeds than current x86 is because they still execute more instructions per cycle than Intel and AMD do on x86.
I can't say that I wanted Itanium to prevail, but I would've been happy with not running x86. At least us UNIX types have the option of not being stuck on that archaic architecture. Unfortunately, the majority of people out there run Windows, and they *are* stuck.
"That's crazy... throughout most of the country teachers earn upwards of $45,000. They also work only 8 months of the year and in a job that doesn't require any form of manual labor. That would make the average teacher salary above $65,000, if they taught a full 12 months. Consider that the average salary across the country is around $35,000 for 12 months of work, the average teacher is making nearly double the average wage. They also get benefits and often receive stipends, which are not figured as part of the salary."
So you would get that teacher's salary, plus your IBM pension. You'd also still have benefits, since teachers get those. That looks like a good deal to me!
That's crazy... throughout most of the country teachers earn upwards of $45,000. They also work only 8 months of the year and in a job that doesn't require any form of manual labor. That would make the average teacher salary above $65,000, if they taught a full 12 months. Consider that the average salary across the country is around $35,000 for 12 months of work, the average teacher is making nearly double the average wage. They also get benefits and often receive stipends, which are not figured as part of the salary.
Since that isn't enough to attract teachers, and paying more certainly would, then we should look at why they don't get paid more. Try to get a complete copy of some school's yearly budget, and a copy of their annual expenditures. Then you'll see how poorly schools tend to spend their money. That is why they *can't* get paid more. Why they *don't* is probably because there hasn't been a shortage of teachers yet. I suppose you could raise taxes more, and then justify why that tax revenue isn't much more needed elsewhere, and then give it to the schools. If you manage those three, then you might see at least some of it going to teachers.
The problem really is not politicians, per se. It is the politicians that spread drivel like what you're repeating. Engineers have specialized knowledge, teachers do not. Physicians have both specialized knowledge and a very large liability; teachers have neither. Why would you pay the three jobs with the same pay? Teachers obviously have the lowest requirement, then engineers, then doctors. Coincidentally, that is how the pay breaks down.
While there is nothing that guarantees that humanity will persevere, it is most likely that we will. As a species, humans are more adaptable than any plant or animal species in history.
The biggest threat to our species is our species doing something like nuking itself out of existence. That's not something that you have any way to fix before everyone dies from it.
If the climate keeps heating, we will alter the environment to be suitable for ourselves. That will mean cooling and shielding indoor areas while altering the atmosphere to reduce temperatures. When that's overdone and it becomes too cold, we'll do the opposite. Eventually we'll be decent enough at it to keep it roughly where it is now. If food becomes a problem, then we'll create places to grow/raise food, and control the environments there, though on a much smaller scale.
Now, I'm not saying we won't royally screw up the environment in the process. Honestly, we're going to worry about people before plants and animals. We'll do what is necessary to insure the survival of our species.
That's why *I* think life will continue to include homo sapiens (unless we blow ourselves up).
The Sahara was a bad example.;-) It used to be quite small, but grew to the current tremendous size without human intervention. Much of that land was inhabitable even 5,000 years ago, but by 1,500 years ago it was mostly considered wasteland.
OK now, when I take consider how ridiculous the GP post was, it was almost certainly tongue-in-cheek. I have an odd feeling that *your* post was serious. On with the show...
No, one of the worst things you could do is let the UN have control like this. It is even unconstitutional in the US, and probably something equivalent in many other places, for that to happen. I would never respect any form of tax the UN tried to impose, and would actually support withdrawing or the dissolving of the organization over it. It is not, should not, and hopefully will never be, the place of the UN to do something like that.
The best way to push up the price would be for OPEC/etc. to decide they want more profit. Second best way is for individual countries to place excises or tariffs on oil. The EU has found that that approach makes them quite a lot of money, and the US is just catching on to that revenue source now. You notice that it doesn't cut consumption too much; that's because it was only intended to get revenue for the governments.
Worst case is that your tax would start another world war. Best case is that people use less fossil fuels. Most likely is that countries withdraw from the UN and do as they were before. Those "points of production" would either ignore the UN, or withdraw.
Realistically, the vast majority of people don't want oil to be expensive. That would mean that they have to spend more to get their petroleum products. The right way is to have the alternate form in advance, and make it a better product than the petroleum based products. There are no other forms of portable energy product that are better than oil products, and that exist today.
You propose forcing people to give their money to some organization that they have no power over, and then using that money for fuel research. A better way is to convince them to conserve, convince them to donate their money, and convince businesses that it is a good area for investment.
Someday people will learn that you can't fix things with taxes and regulation. Most problems are social in nature, and *people* need to change to fix them. Making something illegal does not stop it, and neither does taxing it. This issue is no different. Conservation programs try to fix it in the social aspect, and occasionally offer incentives for that conservation.
At least your "third/easy way" makes sense. It probably wouldn't work, but might work is a decent reason. As long as there is no force to try to coerce people into only having one child. Logic would say that less people equals less resource consumption, but I think people would fill the gap. You have a lot of third world countries that are industrializing, so the effect of negative population growth would not likely be felt beside that.
The last product associated with the Cairo name was NT5/Win2000.
In 1991, Cairo was announced to include (in real product names): MS-RPC, WinFS, AD, Explorer, Exchange, IIS/MSN Search. MS-RPC happened in NT 3.1; Explorer in in NT 4; AD in NT 5/2000. Obviously Exchange and IIS/MSN Search came out as applications instead of OS functions. WinFS is now finally getting finished.
So really, it'll be 16 years to get WinFS, 12 years for MSN Search, 9 years for AD, approx 5 years for Exchange/IIS, and 2 years for MS-RPC.
That's a pretty horrible timeline for a feature set that was supposed to be a unified product.
One of the reasons that Slackware doesn't come with PAM is because of security issues. A lot of people feel that PAM is inherently insecure, and don't want anything to do with it. Not including it also keeps the distribution a lot more simple.
OTOH, I've added PAM to a Slackware system before, and it didn't require all that much effort.
I love Slackware because it lets me easily configure things without using some distro-specific tool, and without breaking any of the same by doing it manually. I also prefer the BSD boot process to SysV.
I find that I use Slackware on boxes that need special things, like some of the nicer filesystems, since OpenBSD doesn't support those. I'm not very happy with the way that Linux is headed. It is becoming much heavier every day, and I have noticed that stability is going down along with that. At least you can still strip a lot of that out. Four years ago, I would've preferred Linux for all my servers, but now I'm using BSD based things more and more. Slackware is hands down my choice for Linux in the situations that I do use it.
All in all, I'm still happy to be a Slackware user, ten years later.
Slashdot Mirror
That'll just cache the JRE, though. I'm not even sure if Linux is one of the platforms that supports it.
If it works the old UNIX way, that doesn't really fix that you end up with multiple copies of the JRE in memory. You just won't have the disk load delay to read the JRE into memory.
Let me know if I'm off on this, because if it only ever loads one copy of the JRE, that would be really useful!
Since you obviously must have single handedly written all version of Windows, as you purport to know everything about them and be the ultimate authority, you must have suffered major amnesia.
Windows 2000, being based on Windows NT, contained a completely different kernel from Win9x. It included no DOS code, unlike Win9x. It had WOW to run Win16 and MS-DOS code. It featured the HAL for all device interaction. These things are 100% different from Win9x. Task management also shared no code with Win9x, and supported multiple processors. The VM subsystem was also completely separate code.
Additionally, window management went through a different set of libraries, and the GDI libraries was written native to NT. That makes the UI 100% different from Win9x.
Filesystem access were different to support NT extended attributes, NTFS, and encryption. The security model was 100% different as well, seeing that Win9x didn't even have one. The on-disk partitioning was done differently, as well. NT has the ability to do software RAID and dynamic disks, which Win9x has no support for. This was because of completely different partition code.
The network stack was native to NT, and did not share code from Win9x. NT used MS-RPC to do IPC tasks, which was different from Win9x.
All of the management tools were native to NT, sharing no code with Win9x; same for the domain membership code.
NT had full locale support, which Win9x doesn't have.
So, despite your completely incorrect insistance, Win2000 did not share code with Win9x. Of course, the bundled applications shared *some* code, but that is not the operating system by any stretch. Parts of IE, WMP, DirectX, and apps like calculator and wordpad contained shared code. Notepad was new code. IE, WMP, and DirectX for NT had to be maintained somewhat separate from Win9x versions because of things like the NT security model. So the NT version could not run on Win9x.
That would, oddly enough, make the switch from Win9x to WinNT at a much bigger change than WinNT to WinNT. The vast bulk of the difference between XP and Vista are application changes. The change from Win9x to WinNT involved a completely new operation system, large amounts of new API, a suite of new applications, and different code in many of the existing applications. That is a *MUCH* bigger change. Even going from Win3.1 to Win95 didn't involve as much new code, seeing as to how Win95 executed in a DOS extender environment, running with many DOS APIs.
I'd say that you've proven rather effectively that *you* don't have a clue what you're talking about. Why don't you go be a fanatical, screaming, hand-waving zealot somewhere else.
There is more issue than that, actually. In addition to the misunderstanding of combustion efficiency vs. energy transfer efficiency, the MPG number are quite misleading.
Not only were the test conditions not the same, but the manufacturer tauted DOT numbers are usually quite conservative. Many drivers will achieve higher MPG than what is published. It's possible that the acheivement of ~26mpg vs the published ~22.5mpg could've been achieved by conservative throttling, different oil weight and viscosity, different octane fuels, different altitudes, etc.
I just hope that the JREs out there get a lot more streamlined, and become available on more platforms by then. The OS/architecture independant nature is great, but the Java platform isn't available everywhere still.
I think one of the biggest impediments to Java for applications is that you have the load the JRE over again for each app. That's really annoying, especially considering how many apps I tend to have started. I have six apps running right now, and if I had another 20MB per app, that would be just a bit horrid!
Definitely true, and doing it the client way is also really obnoxious to the server, in addition to being slow for the user. I'd love for the IMAP RFC to get cleaned up, and one of those changes would be for multi-folder search.
The only reason I ever use mutt over Pine is because mutt natively supports Maildir. That means that my mail server has mutt installed so that I can read local mailboxes without going over IMAP, just in case something bad happens. For normal use, my text-mode reader of choice does happen to be Pine.
Bub, you don't have to be so hostile just because we're disagreeing! Your method of discussion is not very positive. Looking at your post history, you really should drop the habit of insulting people. I might say that someone is wrong, but I don't call them a moron or an idiot.
I wouldn't consider it to be the most major upgrade simply because 9x->NT was a transition of all parts of the system. There was no shared code between the two products! The intent of the system designs were completely dissimilar. Going from XP->Vista is keeping most of the system the same. Aside from the new APIs, most of the changes are application level. Those APIs are being backported to previous releases, so you can't say that Vista is different on that principle. That leaves the main differences between 2k/XP and Vista in the applications. That's not nearly as big a change.
So, please, calm down on the bloody APIs. I'll have those on my XP desktop if I feel like installing them. I don't need Vista to get them, so if anything, that makes the API changes a big deal, and not that MS is releasing them alongside a new OS revision.
What I said about every other release being the biggest change I guess didn't clarify enough. I didn't mean that *I* thought they were the biggest thing since 95, but that was the marketing spin from MS. They'll always bill the new OS in that sort of light as it helps move product. For the *user*, the biggest change they'll notice is almsot certainly going to be the interface. Considering the amount of debate that the new UI is worse than the old one, I wouldn't necessarily count that as a good thing.
As far as the UI offloading, no, I don't think it's a bad thing. I'm saying it just isn't a big deal, and certainly not a good enough reason to justify all the time and expense that doing the update will be. The reason that UI offloading will improve things so much is because they put so much candy in the UI. The reason other platforms don't talk up this type of thing as such a big deal is because it isn't a big deal. For MS, it's a selling point, for anyone else it's either unnecessary, as their UI doesn't consume noticable resources, or it was an improvement they quietly made so that they could do candy interfaces.
For LUA, the places where you were likely to see it being used was in business. More business apps that you seem to realize are not designed well enough to work as a limited user. I'm serious about the sheer number of these applications. The most that the typical home user will run into that doesn't work as a LUA are games. For the home situation, this will work great. For the business situation, it won't be as a big of an improvement.
You don't seem to realize that for people with non-MS background, the things MS does aren't new. Four of the big five tenants of Cairo that are finally being finished the non-MS world has had for over a decade. We're had RPC since the 80's, messaging since the 70s, and directories since the early 90's (and in some capacity, a fair bit before that). The ideas of powerful search and file metadata aren't anything new, either. You have products like Notes that have done a lot of that for a long time. It just isn't new.
You shouldn't be so surprised that a lot of people that aren't heavy MS people aren't at all excited about Vista. It doesn't bring anything new to the table. Most of these changes bring Windows to where many other OS' have been for a long time.
That's two applications that you're using. How many apps do you use that aren't written in Java?
To counterpoint: when in Windows, I run BitComet instead, since it is less resource hungry.
My Java use consists of Dell OpenManage (which slams the host machine for memory), and HP WebAdmin (which crashes the JRE a lot). Parts of OpenOffice use Java. The LDAP Browser is written in Java. I'd prefer that OpenOffice was straight C++, and that I would find a better LDAP browser. On Linux I use Azureus, because I haven't found a better native implementation.
There aren't many desktop Java apps that people use. OpenOffice and Azureus are probably the two biggest, and most people don't run either. It isn't that Java apps are bad or anything, but that many of them have better equivalent native implementations. People also don't like the JRE load time.
Java desktop apps aren't dead, people just don't have a need for them. If Linux starts being use as a desktop platform in large numbers, then you'll have Mac, Linux, and Windows. Then I think you'll see a whole lot more Java apps.
Q: How do I advertise on Slashdot?
;-P
A: Submit an article.
Q: How do I guarantee that my ad gets posted?
A: Post a comment that gets modded as funny.
Q: How is my spelling?
A: It's pretty bad.
That would be the license on the JRE that was mentioned, I believe.
s e.txt
http://java.sun.com/j2se/1.5.0/jdk-1_5_0_05-licen
Just FYI:
IMAP multi-folder searching is actually a vendor extension of UW-IMAP, and I believe Cyrus. It isn't in the IMAP spec and doesn't work against all servers. Pine is written by the guy who does UW-IMAP, so it supports his extensions. It is called a reference implementation, but that is not the case. That is why many IMAP implementations do not support it.
The Java applet sandbox is safe, but this is a Java application. That means it can make unsafe and system modifying calls. You have less to worry about with an application in Java, but you certainly aren't inherently "safe". This email program could still be compromised, though it will be harder to do.
That is to say that even if the JVM cannot be exploited, your application logic still could be.
Only part of running LUA is protecting yourself voluntarily. The other part is locking the machine down so that a user can't do unwanted actions. The main reason that LUA doesn't work right now is that apps don't work without the admin privs. They do things like write to HKLM, %Program Files%, %WINDOWSDIR%, etc. Prompting for a password works for users that *are* the admin. It doesn't fix the major problem of the apps trying to write to admin-only priveleged locations during normal operation. That's the real reason that LUA doesn't work right now, and that's why this feature doesn't fix the problem.
The UI *really really* doesn't take up hardly any CPU time. I feel rather comfortable saying that there would be no difference if you are using the 2k interface. There would be a little difference in the XP/Luna interface, and tremendous difference with Vista/Aero. The reason there is more difference as you go is simply because MS created the problem, and then solved it.
You do realize that *every* MS release is an overdue upgrade, right? They're always fixing something that most of the industry already knew was the wrong way. Often these are things that people have complained about for over a decade. Things that MS ignored fixing because they were too busy adding paper clips or another API. People go off on the nothingness of MS upgrades because they usually don't fix the real problems. They just add more features that probably won't be used, make it look different, do things that will require retraining, introduce piles of new bugs, etc.
Vista isn't the largest upgrade since 95, either. I don't understand why people completely miss this one. MS says it's the biggest thing, but they're trying to market a new product against all their old products that have a massive installed base. The biggest upgrade was going from Win9x to WinNT with the 2000 upgrade. It was a complete and total change of technology. Technologically, it was a bigger change than 3.1->95, seeing as 95 still had all the legacy code that it depended on. How is an incremental bugfix and feature add release the biggest thing?
I seem to remember XP also being the biggest upgrade, and 2000, and Millenium. That's just marketing-speak. It's taken MS five years to get most users off of Win9x, and they're still working on getting people on WinXP. You're right, most people don't care what OS they use, and that's another reason why few people are going to upgrade. If it comes with their computer, then they'll use it, unless it's broken with various unwanted misfeatures. Then they'll ask their children or friend from work or whatever if they can get back their old version.
Vista was supposed to represent the last piece of the Cairo vision. MS has added a piece with every NT based release. Of course, it's taken 16 years to get all the pieces, and Vista won't even include WinFS in the RTM.
My reasoning is just from a different standpoint than yours. I'm looking at this as someone that admins machines, uses machines, and trains users. I know what gives me and my users a hard time, and I'm not too happy about those things not only getting carried forward, as defaults, but becoming even more of an issue.
I'm not critisizing MS for fixing their bugs, I just don't think it's a huge deal that they're doing it. These things should never have survived in their software for as long as they have. I'm glad that they're fixing them, but I hate that they're making fixing their mistakes and bugs into "huge-mega-incredible innovations that will change the world."
BTW, I do believe that proprietary is bad. It fragments software and makes interoperability harder. I'm quite sick and tired of MS pulling that kind of trash, and a whole lot of other people feel the same. It's the reason that so many people are starting to wake up and dump Office. I don't care if their implementation is proprietary, but I do care if the API, format, protocol, etc *is*. MS has proven that they can't be trusted with that sort of thing, and I certainly don't wish to get further burned by their behavior.
You need to have very compelling reasons to convince business users, like me, to upgrade. Especially considering the cost that MS is going to make this particular upgrade. My machines that ran W2k could run XP just fine. My XP machines won't run Vista just fine. What is Vista offering going forward from 2k/XP that I don't already have? That answer is "not very much at all".
It *does* matter what software is out there, too. For example, imaging is already taken care of, and most already have third party software for doing it. The only way that changes is if MS *purposefully* breaks that software. They've done that in the past with various things, so who knows, perhaps they'll do it again.
As I said, my reasoning starts from "Why would I want to upgrade to this?" I then look at the feature list, and see a whole lot of things that I don't care about. Then I see a good number of things that I outright do not want. Along with that, I see a good number of things that I already have from third party software. And finally I see a small number of things that I actually want. So, for a bunch of money, a machine upgrade, and a lot of things that I don't want, I can get a couple of things that MS is backporting to what I already have.
Even if MS put in enough features that I was willing to upgrade, I would wait until SP1. Nobody in their right mind goes with MS software before they do a service release! There's a reason for that: their initial release of anything is bad; it malfunctions, munges data, and presents a massive security threat. That means that even if I wanted to run it, I would be waiting around a year for them to fix it so that it actually works. As an example, I just started moving to XP on my user machines this past spring. It had finally gotten to where it was trustworthy; bugs had been worked out, wierd MS peculiarities had been documented, workarounds existed.
Bottom line, though, is that this release does *not* offer me much, and it does *not* offer business users very much.
---
Volume - My understanding was that it was an API change, not an end-user interface. If it's end-user, then it's more useful. Ideally, this shouldn't be necessary, and I think it'll cause confusion as an end-user tool. I would use this feature, but I'd really hope my users didn't figure out how to do it.
LUA - If users have an admin level password, then you have no control over the machine. This password display is a home-user only feature, since it would be insane to do in a locked down environment. Additionally, the reason that apps do not with in the current LUA implementation is that they weren't properly written. If those vendors actually *fixed* their software, then this problem would be gone. MS isn't fixing the problem, here, they're just changing the probl
MS can do pretty much whatever they want with their software, for all I really care. They're just getting deeper into releasing software that barely considers the customer. In MS' case, it often *is* a problem when they integrate more things into the OS. They're a convicted monopolist, which means they abuse their ability to do exactly that.
.NET, WFC, and probably others that I'm missing. They all have security issues because they never finishing fixing any of them. You have apps written to all of them. You do have backwards compatibility. You don't have a compelling reason to use one over another (except for avoiding MFC). It bugs me because it is poorly done.
My complaint is that they claim "innovations", which only means "new to our software, but existed everywhere else for a decade." Most of the time, the MS implementation is lower quality than the existing implementations, but prevails because of their monopoly.
What is changing is so riculous that it doesn't matter to users. I say that there is no difference, because what is different is irrelevant. I don't care about changes that *aren't* useful, so if that's what the changes are, then little has changed to justify a switch of platform. In that regard, it offers no more than WinXP/2k already do.
You have to consider the market for the product. Basically you have end-users, and corporate users. Many of the features don't matter *at all* for end users, and there are as many bad changes as good for corporate users. MS looks to have managed to make Vista harder to administer than 2k/XP.
I've given quite a few examples, but to give you examples from your own points:
APIs - So now we have Win16 and Win32 [relagated to WoW on Win64], Win64 where applicable, the kernel API, MFC (various),
Per-app volume control - trivial to do yourself
Network stack - security is always good; are the other changes going to be used? Probably not.
Indigo - implementing something that will almost certainly be proprietary, and that is already done elsewhere.
Avalon/Aero - you say there's nothing else comparable; I say that you're wrong. Also, most of the features that they're showcasing with Aero make the system harder to use.
Search - This might be a real, tangible improvement. It's been implemented elsewhere, but not to the same scale.
Mini-apps - irrelevant, already possible. The smaller footprint isn't going to be realized anyway.
Task Scheduler - Maybe. Fixing the current implementation is good, but it isn't used that much anyway.
LUA - still won't work right until third-party devs fix their broken apps. If that happened, LUA would work anyway. Giving users a password for admin-level changes defeats the purpose of locking the machine down.
App jailing - or they could fix their implementation to not do stupid things. Having the feature is good, though. This will encourage bad habits.
App crashes - apps shouldn't crash anyway. App code needs to be fixed. Improving the handling is good, but will encourage bad habits.
Transactional FS - this has been available elsewhere for quite some time. Its addition is welcome, but not generally useful for end users on a consumer OS. Apps should handle the conditions that this protects, already, NTFS handles the rest.
Transactional Registry - the registry is a bad bad idea as is. Hacking transactions into it doesn't change this. This wouldn't be a problem if it was done in a sane manner to begin with, especially in conjunction with the FS transactions.
(System Restore - this is a function that could be implemented on most any other OS. It is not currently implemented on other platforms, to my knowledge.)
Aux display - this is irrelevant. Devices with status displays already have drivers for them, and would still require the same for this.
Improved performance - always good. App loading improvement isn't really necessary, especially on the systems this will run on.
You're right, they are doing a lot of new things. Too bad all of them fall into a few easy catagories: things already done, things nobody wants, things that don't matter, things that are good.
Many of the changes are in the "things already done" catagory. A few of the big ones are "things nobody wants", which is to say that nobody=customers, which are who should matter. The majority are "things that don't matter". A few are good, like LUA, the network working right, etc. Of course, those in the last catagory are how it should already have been.
Now a more realistic look at things. MS is changing APIs around, *again*. That is highly annoying, as apps need to be ported again, tested again, etc. There isn't really a benefit from doing it, aside from supporting the newest MS stuff that will invariably be abandoned by them in the next release. They're adding all sorts of support nightmares, like the default encrypted filesystem. They're changing the GUI *again*, forcing users to need retraining. They're changing the driver model *again*, forcing more rewrites and testing.
The most useful changes that they're making seem to be in fixing things that they broke to begin with. Their bad network policy, sleep implementation, service defaults, browser design, email program, user permissions, malware, search, etc. Much of this will require application changes to get to work right. Yay.
Most of all, all of the stuff that really is talked about will be available on non-Vista platforms. If you actually *read* that winsupersite link, you can see that most of the things listed are either minor, or the same feature.
In short: No, there isn't much else. Vista is XP with a shinier interface to disable, a few hotfixes, and new versions of a couple of apps. Absolutely nothing to justify a new OS.
Spot on with what you said. I'm glad there are other people willing to speak out on exactly that. Teacher unions are unreal and absolutely horrid to work with. The way most unions now do the pay negotiation completely leaves out any concept of merit pay. So forget earning promotions, you're entitled to them by union contract based solely on seniority and such.
And people wonder why work ethic in the USA is so bad...
First, algae is not plant or animal.
Second, I didn't say humans were the most likely to survive, just that they are the most adaptable. Cockroaches don't need to adapt, they are simply versatile. No other species can adapt the environment to themselves, and that makes humans something special.
Can you think of any event, that isn't likely to wipe out life on earth, that would wipe out humanity? Sure, a comet impact would kill off humans, but would global warming, an ice age, a supervolcano, etc? We'd probably manage to survive even if the atmosphere took on a poisonous gas mixture.
It is not the first time Intel took a gamble on a new ISA and lost. Hell, the very existence of x86 was as a filler move between non-x86 product lines. The ISA is one of the worst out there, and it *can't* get better.
It's a real loss to everyone that the industry won't just drop the damned architecture. MIPS64, PowerPC, IA64, Alpha, and UltraSPARC are vastly superior to x86. One of the reasons that all of those architectures have lower clock speeds than current x86 is because they still execute more instructions per cycle than Intel and AMD do on x86.
I can't say that I wanted Itanium to prevail, but I would've been happy with not running x86. At least us UNIX types have the option of not being stuck on that archaic architecture. Unfortunately, the majority of people out there run Windows, and they *are* stuck.
To quite myself from elsewhere:
"That's crazy... throughout most of the country teachers earn upwards of $45,000. They also work only 8 months of the year and in a job that doesn't require any form of manual labor. That would make the average teacher salary above $65,000, if they taught a full 12 months. Consider that the average salary across the country is around $35,000 for 12 months of work, the average teacher is making nearly double the average wage. They also get benefits and often receive stipends, which are not figured as part of the salary."
So you would get that teacher's salary, plus your IBM pension. You'd also still have benefits, since teachers get those. That looks like a good deal to me!
That's crazy... throughout most of the country teachers earn upwards of $45,000. They also work only 8 months of the year and in a job that doesn't require any form of manual labor. That would make the average teacher salary above $65,000, if they taught a full 12 months. Consider that the average salary across the country is around $35,000 for 12 months of work, the average teacher is making nearly double the average wage. They also get benefits and often receive stipends, which are not figured as part of the salary.
Since that isn't enough to attract teachers, and paying more certainly would, then we should look at why they don't get paid more. Try to get a complete copy of some school's yearly budget, and a copy of their annual expenditures. Then you'll see how poorly schools tend to spend their money. That is why they *can't* get paid more. Why they *don't* is probably because there hasn't been a shortage of teachers yet. I suppose you could raise taxes more, and then justify why that tax revenue isn't much more needed elsewhere, and then give it to the schools. If you manage those three, then you might see at least some of it going to teachers.
The problem really is not politicians, per se. It is the politicians that spread drivel like what you're repeating. Engineers have specialized knowledge, teachers do not. Physicians have both specialized knowledge and a very large liability; teachers have neither. Why would you pay the three jobs with the same pay? Teachers obviously have the lowest requirement, then engineers, then doctors. Coincidentally, that is how the pay breaks down.
While there is nothing that guarantees that humanity will persevere, it is most likely that we will. As a species, humans are more adaptable than any plant or animal species in history.
The biggest threat to our species is our species doing something like nuking itself out of existence. That's not something that you have any way to fix before everyone dies from it.
If the climate keeps heating, we will alter the environment to be suitable for ourselves. That will mean cooling and shielding indoor areas while altering the atmosphere to reduce temperatures. When that's overdone and it becomes too cold, we'll do the opposite. Eventually we'll be decent enough at it to keep it roughly where it is now. If food becomes a problem, then we'll create places to grow/raise food, and control the environments there, though on a much smaller scale.
Now, I'm not saying we won't royally screw up the environment in the process. Honestly, we're going to worry about people before plants and animals. We'll do what is necessary to insure the survival of our species.
That's why *I* think life will continue to include homo sapiens (unless we blow ourselves up).
The Sahara was a bad example. ;-) It used to be quite small, but grew to the current tremendous size without human intervention. Much of that land was inhabitable even 5,000 years ago, but by 1,500 years ago it was mostly considered wasteland.
OK now, when I take consider how ridiculous the GP post was, it was almost certainly tongue-in-cheek. I have an odd feeling that *your* post was serious. On with the show...
No, one of the worst things you could do is let the UN have control like this. It is even unconstitutional in the US, and probably something equivalent in many other places, for that to happen. I would never respect any form of tax the UN tried to impose, and would actually support withdrawing or the dissolving of the organization over it. It is not, should not, and hopefully will never be, the place of the UN to do something like that.
The best way to push up the price would be for OPEC/etc. to decide they want more profit. Second best way is for individual countries to place excises or tariffs on oil. The EU has found that that approach makes them quite a lot of money, and the US is just catching on to that revenue source now. You notice that it doesn't cut consumption too much; that's because it was only intended to get revenue for the governments.
Worst case is that your tax would start another world war. Best case is that people use less fossil fuels. Most likely is that countries withdraw from the UN and do as they were before. Those "points of production" would either ignore the UN, or withdraw.
Realistically, the vast majority of people don't want oil to be expensive. That would mean that they have to spend more to get their petroleum products. The right way is to have the alternate form in advance, and make it a better product than the petroleum based products. There are no other forms of portable energy product that are better than oil products, and that exist today.
You propose forcing people to give their money to some organization that they have no power over, and then using that money for fuel research. A better way is to convince them to conserve, convince them to donate their money, and convince businesses that it is a good area for investment.
Someday people will learn that you can't fix things with taxes and regulation. Most problems are social in nature, and *people* need to change to fix them. Making something illegal does not stop it, and neither does taxing it. This issue is no different. Conservation programs try to fix it in the social aspect, and occasionally offer incentives for that conservation.
At least your "third/easy way" makes sense. It probably wouldn't work, but might work is a decent reason. As long as there is no force to try to coerce people into only having one child. Logic would say that less people equals less resource consumption, but I think people would fill the gap. You have a lot of third world countries that are industrializing, so the effect of negative population growth would not likely be felt beside that.
The last product associated with the Cairo name was NT5/Win2000.
In 1991, Cairo was announced to include (in real product names): MS-RPC, WinFS, AD, Explorer, Exchange, IIS/MSN Search. MS-RPC happened in NT 3.1; Explorer in in NT 4; AD in NT 5/2000. Obviously Exchange and IIS/MSN Search came out as applications instead of OS functions. WinFS is now finally getting finished.
So really, it'll be 16 years to get WinFS, 12 years for MSN Search, 9 years for AD, approx 5 years for Exchange/IIS, and 2 years for MS-RPC.
That's a pretty horrible timeline for a feature set that was supposed to be a unified product.
One of the reasons that Slackware doesn't come with PAM is because of security issues. A lot of people feel that PAM is inherently insecure, and don't want anything to do with it. Not including it also keeps the distribution a lot more simple.
OTOH, I've added PAM to a Slackware system before, and it didn't require all that much effort.
I love Slackware because it lets me easily configure things without using some distro-specific tool, and without breaking any of the same by doing it manually. I also prefer the BSD boot process to SysV.
I find that I use Slackware on boxes that need special things, like some of the nicer filesystems, since OpenBSD doesn't support those. I'm not very happy with the way that Linux is headed. It is becoming much heavier every day, and I have noticed that stability is going down along with that. At least you can still strip a lot of that out. Four years ago, I would've preferred Linux for all my servers, but now I'm using BSD based things more and more. Slackware is hands down my choice for Linux in the situations that I do use it.
All in all, I'm still happy to be a Slackware user, ten years later.