Windows Vista successor scheduled for a H2 2009 release?


Recommended Posts

I want an overhaul of all legacy components:

Yeah me too.

I want BIOS to die or to evolve into something. Also, with RAID controller and a bunch of other on board chips, my pre-OS startup time is much longer than OS-load time. And it doesn't matter how fast the CPU is, it still takes forever to get to the stage when the OS starts to load.

Sadly, there's little MS alone can do to improve on that.

I want an overhaul of all legacy components: remove native support (registry, crazy dlls, etc.) for previous generation apps but allow them to be run using virtualization. That way, it'll be easier to build a componentized (sp?) OS and write programs which function correctly with non-adminstrative rights.

I cannot wait for that day to come :D

if they charge another arm and a leg for it then it'll be a failure no matter how good it is. They should at the very least provide a cheap upgrade option for Vista owners if they're going to release it that soon.

Ohh yeah.

Almost $700 for Ultimate is just way too much.

I want an overhaul of all legacy components: remove native support (registry, crazy dlls, etc.) for previous generation apps but allow them to be run using virtualization. That way, it'll be easier to build a componentized (sp?) OS and write programs which function correctly with non-adminstrative rights.

What's wrong with the registry?

Can you suggest a better alternative for storing settings? I can't.

http://blogs.msdn.com/oldnewthing/archive/...26/6523907.aspx

The registry is simply the best option for saving user and machine settings in a fast, flexible, safe, and secure manner, and It's really the only good way to do component registration.

What's wrong with the registry?

Can you suggest a better alternative for storing settings? I can't.

...

The registry is simply the best option for saving user and machine settings in a fast, flexible, safe, and secure manner, and It's really the only good way to do component registration.

Text files? XML files? Text files work with Linux.
What's wrong with the registry?

Can you suggest a better alternative for storing settings? I can't.

http://blogs.msdn.com/oldnewthing/archive/...26/6523907.aspx

The registry is simply the best option for saving user and machine settings in a fast, flexible, safe, and secure manner, and It's really the only good way to do component registration.

I disagree. The registry is contained in only a few files. Lose any of them and it can be a real heartache getting your system back up. I like the linux way of doing it with text files for different services etc.

I doubt they'll drop legacy code. Same reason Intel won't change their architecture to something that makes more sense. It would take a joint effort including MS, a bunch of major software creators, and a bunch of hardware manufacturers - otherwise it's not worth it to MS to brute force their way through the crap they'll take for making such a drastic change. It would be a terrible financial decision if they did it of their own will with no outside support, even though it is their own OS and despite the fact that they could probably manage it all.

-Spenser

Text files? XML files? Text files work with Linux.

Did you even read the link he posted?

Just a few reasons why those don't work:

  1. It's slow
  2. No per-key security
  3. Not thread safe
  4. Prone to Denial-of-Service attacks (or accidental DoS from badly written apps)
  5. Apps can read/write to them directly, limiting options to extend the format.
  6. Manageability is nearly impossible.

There is no way text files could ever fulfill the purpose of the registry.

But yeah, you go ahead an build COM based up on registrations in a text file. Good luck with that.

I disagree. The registry is contained in only a few files. Lose any of them and it can be a real heartache getting your system back up. I like the linux way of doing it with text files for different services etc.

I'm confused. Aren't two files easier to back up than 200?

Plus, it's not like applications are forced to put settings in the registry. The registry is first and foremost for registration. That is, COM coclasses / interfaces, file associations, handlers, protocols, etc. All stuff that is shared by multiple applications. Putting all that stuff in text or XML files would be a nightmare.

I agree with Brandon. :yes: The registy was/is a nice idea for what it was designed and implemented for. Then come all the developers around the world, and start to store all their programs' settings inside the registry -> of course it becomes big and slow. Then to make things worse, everybody tries to write a registry fixer and optimizer to "speed up your computer by 400%" :angry: . Of course it will become a mess.

I don't know if anybody here knows the story behind the notification area next to the clock. It is the same story :whistle: : it was designed to notify the user about events, then later all the developers had the idea to minimize their application onto the notification area, or create a quickstarter icon there. And now, what do you see there on an average computer: antivirus, firewall, bandwidth monitor, quicktime, nero, adobe, openoffice, nvidia and ati and other tray icons. Was the idea good? Yes. Can you use easily a PC with 15-20 tray icons next to the clock? No.

^^^ Microsoft themselves encouraged software developers to do this. And Microsoft led by example with their apps.

I prefer smaller files, in human-readable (and fixable) text format, that are accessed/read by the system only when needed. System files kept out of user write areas. Per user files kept inside each user's home folder. Clean. Orderly. Maintainable.

Sure, there are benefits to a binary database containing all entries from all apps and the OS. But surely you can see the disadvantages, too. ;)

I hope it's true as well. There may be a lot of annoyed Vista users though.

All Vista users are annoyed already

Even the ones like myself who never have any issues - but dont like the bloated overhead.

Im eagerly awaiting something different - Im getting tired of dual-booting.

Did you even read the link he posted?

Just a few reasons why those don't work:

  1. It's slow
  2. No per-key security
  3. Not thread safe
  4. Prone to Denial-of-Service attacks (or accidental DoS from badly written apps)
  5. Apps can read/write to them directly, limiting options to extend the format.
  6. Manageability is nearly impossible.

There is no way text files could ever fulfill the purpose of the registry.

But yeah, you go ahead an build COM based up on registrations in a text file. Good luck with that.

Why aren't *nix based systems (Linuxn BSD, etc.) a nightmare? Or slow? or less safe than Windows (ha!)? Kudos to them I guess right? Why so many applications do not use the registry and use text or xml or ini files to store settings? And why do many of us have to format Windows and reinstall because the registry (among other things) is full of crap?

Microsoft shouldn't allow people to write to the registry unless they follow proper guidelines.

Well I'd bet we won't see it before H2 2009. But I'm pretty certain Microsoft will make sure it's out before 2011 hits. They said they don't want no more 5 year releases and I agree. But 2 years seems to soon to me. 3 years sounds like a sweet spot, maybe with a better upgrade price too :D

Why aren't *nix based systems (Linuxn BSD, etc.) a nightmare? Or slow? or less safe than Windows (ha!)? Kudos to them I guess right? Why so many applications do not use the registry and use text or xml or ini files to store settings? And why do many of us have to format Windows and reinstall because the registry (among other things) is full of crap?

Microsoft shouldn't allow people to write to the registry unless they follow proper guidelines.

If you bothered to look at the link I posted, it would have spelled in great detail for you the problems posed by these kind of files, and why the registry is "better" than config files in many cases.

Some of the problems are specific to INI files, but a lot of the listed issues are problems that any kind of configuration file can have:

XML manages to sidestep many of the problems that INI files have, but only if you promise only to read from them (and only if everybody agrees to use a standard-conforming parser), and if you don't require security granularity beyond the file level. Once you write to them, then a lot of the INI file problems return.

Mio, some of those issues pointed out by Microsoft to validate Microsoft's direction are specific to Microsoft products only. The very first one talks about the Windows API limitation at the time. The second one poses the "administrator" values versus the "user" values. This would be handled by two different files, as any app with specific settings to work in that corporate environment would be out of reach of the users. And users would have their own settings that are read (local drive or user's 'home' folder off network) so they can make their own preference adjustments (UI settings and such). The third one is purely implementation issue, if two separate tasks are allowed to both open the file in write mode and both attempt writes to it. That's a strawman - it is nonsense, so just don't do it!

Some of the items are true - it takes fractionally longer to read and parse a file that is in ASCII text.

It is just weighing the pros and cons of large binary database versus smaller text files that are only read when needed (and are kept in memory during application run, so aren't constantly re-read and parsed).

I know where I put my money. :)

It would make perfect sense for the new version to come that soon, considering what a huge failure Vista has been.

It's a failure only in the eyes of the AMB trolls, check out CDW's latest report. Doubling OS X's market share in a year is a pretty good indicator.

Now before the fanboys scream "TROLL!", there is not one business I know of switching to Vista, it would be like suicide. The reasons are too numerous to go into now, but it's just not happening. I work in an IT dept and we have been evaluating Vista and found it simply not fit for purpose.

Once again, check out the CDW report, many businesses are now starting their vista migrations, my own employer included.

Windows me was only released 7 months after win2k.

Windows ME was not a successor to 2000 it was a successor to 98, and it was being developed concurrently with 2000 so that feat is not really as impressive as you make it sound.

Yeah me too.

I want BIOS to die or to evolve into something. Also, with RAID controller and a bunch of other on board chips, my pre-OS startup time is much longer than OS-load time. And it doesn't matter how fast the CPU is, it still takes forever to get to the stage when the OS starts to load.

Sadly, there's little MS alone can do to improve on that.

Wait till we get better EFI support from motherboard makers :D

I disagree. The registry is contained in only a few files. Lose any of them and it can be a real heartache getting your system back up. I like the linux way of doing it with text files for different services etc.

Linux is not immune to problems, I have had application installs b0rk in linux that could only be fixed with a complete reinstall.

Why aren't *nix based systems (Linuxn BSD, etc.) a nightmare? Or slow? or less safe than Windows (ha!)? Kudos to them I guess right? Why so many applications do not use the registry and use text or xml or ini files to store settings? And why do many of us have to format Windows and reinstall because the registry (among other things) is full of crap?

Microsoft shouldn't allow people to write to the registry unless they follow proper guidelines.

Not true at all, in fact I personally have found that on every computer I have ever used, applications actually take longer to launch on Linux than on Windows. What the registry gets loaded with is mainly down to vendors writing bad uninstall programs, if you ran windows how it came out of the box, and never installed any applications on it then you would never have to format it, the general rule of thumb is that the fault lies with software vendors.

The actual reality is that the windows registry is rarely if ever used as a vector of attack, it does not make Windows inherently unsafe.

Also, something you folks seem to be failing to consider is that part of the excessive technical complexity of Linux comes down to the way it manages application installs, you have to perform different edits for different apps, whereas with the Windows Registry everything is centralised and editable from a single location, making it much easier to configure and tweak.

Windows ME was not a successor to 2000 it was a successor to 98, and it was being developed concurrently with 2000 so that feat is not really as impressive as you make it sound.

XP was released october 2001. Which is 1? year after win2k, and 1 year after Win ME.

What Microsoft needs to do is start making new OS branches. One that supports legacy, one that gives the whole thing the much needed overhaul.

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • Microsoft launches Godot Sample to streamline Xbox PC game development on the engine by Pulasthi Ariyasinghe Microsoft today announced a new endeavor that aims to make it simpler for Godot developers to get their products into the Xbox PC ecosystem. Dubbed the "XBOX Godot Sample," this is a new public reference for developers using the open-source engine. This is set to serve as an example of how Microsoft GDK, Xbox Services, and PlayFab can be integrated into their projects. The sample is available now on GitHub as a working example. This covers key features in gaming projects that developers may need to release their projects on Xbox PC, with everything from matchmaking and game sign-in to gamepad compatibility with Godot being covered. This release is being called the first step in giving Godot developers the tools to bring their games over to Xbox PC, with more changes to come based on feedback and issue reports. However, the company was clear that this is not related to bringing Godot projects to Xbox consoles. The engine's open development model stops it from accessing console SDKs due to the requirement of NDAs and legal contracts. Here's how it explained this Godot sample project's focus: This is a source-only sample, not a product. It's MIT-licensed at the wrapper layer; the GDK and PlayFab dependencies still require their own installs and license acceptance, consistent with our other XBOX samples. There is no set update cadence for support or maintenance. We’ll watch the repo, monitor issues, and iterate where it makes sense, but this isn't a commercial release. That said, we’re excited to hear your feedback and see any community PRs, as we evolve this together. This is the first step in bringing Godot for XBOX on PC. We plan to evolve it over time based on what the community tells us is most valuable. This sample is built specifically for XBOX on PC. It doesn’t include support for XBOX Series X|S or XBOX One. If you’re already building for XBOX Series X|S or XBOX One, please talk with your XBOX representative. If not, you can get started by signing up here. Game developers can find the XBOX Godot Sample by heading to GitHub over here. Documentation on how to get started with Godot for building an Xbox PC project can be seen here.
    • I don't understand the vision. Do people really want to buy a new computer from Dell with 6 browsers installed? We all keep asking for Microsoft to stop having so much junk on their OS, and adding a bunch of browsers seems to go against that. Ideally, we would just be asked what browser we want during OOBE but Google is just going to pay Dell a bunch of money to include Chrome. Additionally, would you want your phones to start including all the browsers too when you get them? The only thing I ever wanted was to be able to uninstall IE or edge and I believe you are now able to. I do agree that microsoft needs to chill with their "are you sure you don't want to try edge before you install chrome" ads when going to download chrome.
    • It is notable that around 70% of web browser users choose Google Chrome. However, it is puzzling why anyone on Windows would opt for Chrome when Microsoft Edge is often superior in many aspects and comes pre-installed. Edge collects less data, uses less RAM, and is more optimized for Windows as a native Microsoft product. While some may point to bloat in Edge, much of it can be removed with simple tools, requiring no more effort than installing Chrome. Meanwhile, Chrome reportedly downloads large amounts of AI data (4 GB) without explicit consent. I'm sure you Chrome users love that, or? Here is one example of a tool that doesn't even need to be installed to be able to use: https://github.com/TheBobPony/MSEdgeTweaker Although Microsoft’s aggressive promotion of Edge may be questionable, the browser’s current advantages make it a preferable choice over Chrome today, even if Chrome may have been better in the past.
    • JetBrains rolls out IntelliJ IDEA update with Markdown preview fixes and more by David Uzondu Image via JetBrains IntelliJ 2026.1.3 from JetBrains has landed, bringing several highly requested bug fixes that target common UI glitches and terminal rendering issues. If you run tmux inside the integrated terminal, the IDE no longer renders the cursor above the active line. The Markdown preview bug, which was fixed in this release, had annoyed developers for quite some time, as the preview pane failed to render images saved outside the project directory. Instead of displaying the actual image, the IDE simply showed a broken image icon, a problem that stuck around for two years before this update. Over on Windows, developers running WSL can now use wsl.exe to spin up their environments without losing terminal functionality. In previous builds, launching a terminal shell with something like wsl.exe -d ubuntu inside a Windows-based project broke both shell integration and active process detection. Other bug fixes in this release include: An issue where Gradle sync incorrectly reported success as a failure on WSL when using Gradle 9.5.0. A syntax highlighting bug that flagged valid Java for-loop initialization blocks with multiple statements as incorrect. A warning bug that triggered a false non-null local variable alert when using JSpecify annotations. A database generation bug that hid the option to use a DELETE statement instead of a TRUNCATE checkbox. A Kotlin highlighting failure where an assertion error in the Gradle redundant library inspection broke error highlighting. A UI bug where the ComboBox popup lacked a maximum height restriction. A Snowflake syntax error where DataGrip failed to support the "create temp" command. A Svelte syntax parsing failure that incorrectly flagged quotes inside inline expressions. A VCS repository manager deadlock that triggered thread pool exhaustion. A memory leak where the LazyTree component kept all previous versions of a tree in memory. IntelliJ 2026.1.3 is the third bug fix release for the IntelliJ 2026.1 series. The first one landed back in April with a fix for the WSL Python interpreter freeze, another fix for guest participants using Emmet abbreviations, and corrected WildFly server deployment errors.
  • Recent Achievements

    • Collaborator
      Asgardi earned a badge
      Collaborator
    • Conversation Starter
      mobandz earned a badge
      Conversation Starter
    • Apprentice
      fernan99 went up a rank
      Apprentice
    • One Month Later
      nothanks earned a badge
      One Month Later
    • One Month Later
      B2Proxy earned a badge
      One Month Later
  • Popular Contributors

    1. 1
      +primortal
      468
    2. 2
      PsYcHoKiLLa
      243
    3. 3
      Skyfrog
      79
    4. 4
      FloatingFatMan
      73
    5. 5
      Michael Scrip
      60
  • Tell a friend

    Love Neowin? Tell a friend!