Windows 7

[Riptides]

What is the winsxs folder looking like for others using Win7?

5.4GB so far.

I have it (64bit) running Virtual Box as well.

I used my Vista (user/name) folder and copied it over to Win7 and all I had to do was install VirtualBox and fire it up and within a few minutes had a shared folder setup for transferring my files between my WinXP and Win7 installs. I didn't need to mess with anything else download or compatibility wise.

The superbar is fine, but I had been using a W7 visual style for Vista so I was at least used to how different it looks. Functionality-wise, apart from the shutdown and search textbox it makes the start menu redundant

Agree about the start button, but would like to see a separate section I can add and organize quickbar icons, that would expand to larger ones when I use them, then go back small again, instead of either/or Large or Small across the bar. Also having my Show Desktop button at the edge on the right side fucking sucks, I want to be able to move it to the left. The main thing about the superbar I'm not liking right now is it's lack of overall customization.

a) Explorer navigation bar is now a mix of Vista's favourites and filesystem. I much prefered just the favourites as this is too busy/cluttered.

I just want my UP button back so I don't have to stop using my mouse in order to hotkey Alt+CursorUp just to go up a level. Don't care the reasoning behind why it went missing in Vista, I want it back for productivity. Also agree with your b. c. and d. points being steps back.

if you're already using Vista there's very little reason to go out and buy this.

I strongly disagree with this statement, I've been using Vista since it's retail release date and am ready to toss it into the closet on the top shelf next to my dusty WinME box. Vista was a next step in OS's for Microsoft, but it's just so broken on so many levels, and I'm ready to get past that and working with an OS that works and keeps settings set more often than not.

I'm seeing so many better things in this OS, which is inherently still Vista, that I can't see where anyone should hold on to Vista beyond the first Win7 Service Pack at the most and at the very least it would be worth an Upgrade box purchase for most people using Vista.

 

[Richard]

Your usage pattern is not average usage pattern. I double click in empty space for no reason (and I know other ppl who do the same). Assigning action to this behavior would be devastating assuming that my anecdotal argument corresponds to usage patterns of at least several percent of users.

Option?

Because you can't host scripts in thin air?

I guess the emoticon didn't tip you off I wasn't being literal. What I meant was that all the people complaining about Vista's sidebar and now raving about W7 having gadgets without the sidebar don't know a gradient can be disabled. I personally don't care either way as I don't use gadgets.

From what I've seen so far I will buy Win7 on day one while I wouldn't be able to justify purchase of boxed copy of Vista on its release. So who's right?

Since when this was about being right? Can't we both be right? In fact, not only do I agree with you that Vista on its release date wasn't worth a retail purchase I'll do you one better and say that it's still not worth it now that W7 is (nearly) here.

If functionality and usage patterns match yours, there's no reason to buy Win7, that's the answer. Perfect example of it is how you described superbar comparing it to Vista + new visual style. Win7 superbar UX is nothing like Vista, reskinned or not. But apparently whatever is new doesn't appeal to you.

Maybe you need to re-read that part, especifically the part where I'm comparing the visuals of the superbar to the Vista VS as in: it wasn't the ooh-aah surprise you get reading some of the reviews. In fact, if you scroll down a bit you'll see me praising some of the features made possible BY the superbar.

Take it easy. Not everyone is out to get you.

 

[Dominik D]

Option?

"The Old New Thing" by Raymond Chan explains well why in-box options are not the best idea (most of the time). Your dbl click functionality can be achieved via 3rd party app, so there is an option for you.

I guess the emoticon didn't tip you off I wasn't being literal. What I meant was that all the people complaining about Vista's sidebar and now raving about W7 having gadgets without the sidebar don't know a gradient can be disabled. I personally don't care either way as I don't use gadgets.

It's not just gradient. Win7 gadgets can stick to each other and to the edges of the screen and they are not always on top. This is a change in behavior from Vista. Again - it's probably about your usage pattern. If you keep them in one place like on Vista, changes in behavior mean nothing to you and the only visible "novelty" is the lack of gradient.

Since when this was about being right? Can't we both be right?

I was referring to your "if you're already using Vista there's very little reason to go out and buy this." Third person does not refer to your tastes, it's advice and as such is wrong if it's biased. And it is since you're passing totally subjective judgement. It's prob detail irrelevant to 99% of people but I tend to see these kind of tiny things (probably to my own demise).

Maybe you need to re-read that part, especifically the part where I'm comparing the visuals of the superbar to the Vista VS as in: it wasn't the ooh-aah surprise you get reading some of the reviews. In fact, if you scroll down a bit you'll see me praising some of the features made possible BY the superbar.

I just did. Your initial statement and overal tone of the post color your superbar feature enumeration as just that - enumeration. I previously found no indication of "excitement" other than with explorer copy progress. But since you pointed that out - yes, this fragment alone is very positive and if was supposed to be taken as such, I clearly misinterpreted it, sorry.

Take it easy. Not everyone is out to get you.

If it was that simple.

-*-*-

As for the WinSxS directory...
The idea is that every time you get new version of DLL in the System32 directory, the old one is being archived (to rollback changes if needed or aid in DLL hell in case two applications use the same DLL but require it in different version). So for the most part files in WinSxS are hard links to files in System32. WinSxS grows with applications and patches installed (I'm not sure if it shrinks when you remove software - I hope it does). To get the size of files with only one link (i.e. to get the extra space files from this dir consume by excluding files that are part of System32 already) you can run this small tool:

#include <windows.h>
#include <strsafe.h>
#include <stdio.h>

LONGLONG GetFileSize(WCHAR *wszDir, WCHAR *wszFile)
{
    const size_t        cchBuf          = 2000;
    WCHAR               wszBuf[cchBuf]  = {0};
    LONGLONG            llSize          = 0;
    HANDLE              hFile           = INVALID_HANDLE_VALUE;
    BOOL                fResult         = FALSE;
    FILE_STANDARD_INFO  FileStdInfo     = {0};
    HRESULT             hr              = S_OK;

    // Concatenate string and produce path to the file
    hr = StringCchPrintfW(wszBuf, cchBuf, L"%s%s", wszDir, wszFile);
    if (S_OK == hr)
    {
        hFile = CreateFileW(
            wszBuf,
            GENERIC_READ,
            FILE_SHARE_READ | FILE_SHARE_WRITE,
            NULL,               // security
            OPEN_EXISTING,
            FILE_ATTRIBUTE_NORMAL,
            NULL);              // template
        if (INVALID_HANDLE_VALUE != hFile)
        {
            fResult = GetFileInformationByHandleEx(
                hFile,
                FileStandardInfo,
                &FileStdInfo,
                sizeof(FILE_STANDARD_INFO));
            if (FALSE != fResult)
            {
                if ((FALSE == FileStdInfo.Directory) && (1 == FileStdInfo.NumberOfLinks))
                {
                    llSize = FileStdInfo.AllocationSize.QuadPart;
                    wprintf(L"(file) %s\n", wszBuf);
                }
                else
                {
                    wprintf(L"(link) %s\n", wszBuf);
                }
            }
            CloseHandle(hFile);
            hFile = INVALID_HANDLE_VALUE;
        }
    }

    return llSize;
}

LONGLONG GetDirectorySize(WCHAR *wszDir);
LONGLONG GetDirectorySize(WCHAR *wszDir, WCHAR *wszSubdir)
{
    const size_t    cchBuf          = 2000;
    WCHAR           wszBuf[cchBuf]  = {0};
    LONGLONG        llSize          = 0;
    HRESULT         hr              = S_OK;

    // Concatenate string and produce new path ending with backslash
    hr = StringCchPrintfW(wszBuf, cchBuf, L"%s%s\\", wszDir, wszSubdir);
    if (S_OK == hr)
    {
        llSize = GetDirectorySize(wszBuf);
    }

    return llSize;
}

LONGLONG GetDirectorySize(WCHAR *wszDir)
{
    const size_t    cchBuf          = 2000;
    WCHAR           wszBuf[cchBuf]  = {0};
    WIN32_FIND_DATA FindFileData    = {0};
    LONGLONG        llSize          = 0;
    HANDLE          hFind           = INVALID_HANDLE_VALUE;
    HRESULT         hr              = S_OK;

    // Concatenate string and produce new path ending with wildcard
    hr = StringCchPrintfW(wszBuf, cchBuf, L"%s*", wszDir);
    if (S_OK != hr)
        goto _exit;

    // Start iterating through the files in the directory
    hFind = FindFirstFileW(wszBuf, &FindFileData);
    if (INVALID_HANDLE_VALUE != hFind)
    {
        do
        {
            // Skip '.' and '..'
            if (L'.' != FindFileData.cFileName[0])
            {
                if (0 != (FindFileData.dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY))
                {
                    llSize += GetDirectorySize(wszDir, FindFileData.cFileName);
                }
                else
                {
                    llSize += GetFileSize(wszDir, FindFileData.cFileName);
                }
            }
        } while (0 != FindNextFile(hFind, &FindFileData));
    }

    // Exit and cleanup
_exit:
    if (INVALID_HANDLE_VALUE != hFind)
    {
        FindClose(hFind);
        hFind = INVALID_HANDLE_VALUE;
    }

    return llSize;
}

int __cdecl wmain(
    __in                int     argc,
    __in_ecount(argc)   WCHAR   *argv[]
)
{
    int         iResult = 1;
    int         iLength = 0;
    LONGLONG    llSize  = 0;

    // Parameter count
    if (2 != argc)
        goto _exit;

    // Parameter ends with backslash
    iLength = wcslen(argv[1]);
    if (argv[1][iLength -1] != L'\\')
        goto _exit;

    // Calculate
    wprintf(L"\nDirectory: %s\n\n", argv[1]);
    llSize = GetDirectorySize(argv[1]);
    wprintf(L"Size: %I64d\n", llSize);

    // Exit and cleanup
    iResult = 0;
_exit:
    if (0 != iResult)
    {
        wprintf(L"\nSyntax:\n\tdirsize [directory ending with \\]\n\n");
    }

    return iResult;
}

"dirsize c:\windows\winsxs\" in my case resulted in ~1.5GB used, which is much less than dir properties show (~6GB). That means ~4.5GB of files are in fact in System32.

 

[Kaotik]

I just want my UP button back so I don't have to stop using my mouse in order to hotkey Alt+CursorUp just to go up a level. Don't care the reasoning behind why it went missing in Vista, I want it back for productivity. Also agree with your b. c. and d. points being steps back.

Assuming the folder names aren't extremely long, it takes max 1 mouse click more on Vista to go up folder(s), in many cases when you need to go up several folders, it takes less clicks from the mouse.
In case you haven't noticed, no matter how long Vista has been out there, clicking on a folder in the adressbar takes you there.

Example:
Folder you're in
C:\Windows\System32\Drivers\etc\
and you want to go back to C:\Windows\
On XP, you press the "up" button 3 times to get there
On Vista/7 you see this:
<< Windows > System32 > Drivers > etc
And just click on the "Windows" and you're there, 1 click

Assuming you're somewhere really deep in the folder jungle, you need total of 2 clicks max compared to x clicks of XP & earlier, just click on the "<<" part and pick from the list the folder you want to go to

 

[Richard]

"The Old New Thing" by Raymond Chan explains well why in-box options are not the best idea (most of the time). Your dbl click functionality can be achieved via 3rd party app, so there is an option for you.

That argument can be used for a lot of other new options W7 introduces too. And the fact that there are 3rd party apps is a weak point as that never stopped MS from putting functionality similar to existing 3rd party apps in Windows before.

It's not just gradient. Win7 gadgets can stick to each other and to the edges of the screen and they are not always on top. This is a change in behavior from Vista. Again - it's probably about your usage pattern. If you keep them in one place like on Vista, changes in behavior mean nothing to you and the only visible "novelty" is the lack of gradient.

The snapping is new but Vista gadgets aren't always on top either. The novelty isn't that the gradient is gone. It's that you interact with gadgets directly in W7 whereas in Vista half of the time you're interacting with the sidebar. That's much more important than a gradient, even though it's lost on many commentators.

I was referring to your "if you're already using Vista there's very little reason to go out and buy this." Third person does not refer to your tastes, it's advice and as such is wrong if it's biased. And it is since you're passing totally subjective judgement. It's prob detail irrelevant to 99% of people but I tend to see these kind of tiny things (probably to my own demise).

You can say that again. You're human therefore inherently biased. Using your argument, you're wrong too. Heck, according to you, every single advice given by every human for the duration of our stay on Earth is wrong.

 

[Florin]

Example:
Folder you're in
C:\Windows\System32\Drivers\etc\
and you want to go back to C:\Windows\
On XP, you press the "up" button 3 times to get there
On Vista/7 you see this:
<< Windows > System32 > Drivers > etc
And just click on the "Windows" and you're there, 1 click

That's all fine and good, but you have to read and think about what you click. That actually slows you down. The 'up button' on the other hand is just something distinct and yellow that you hammer until you see a familiar directory view.

Everyone understands how Vista's method works. But understanding it isn't necessarily agreeing that it is better.

 

[Kaotik]

If someone really enjoyed the sidebar rather than just the gadgets, I can try to dig up the way to get that sidebar there, I know it's possible and has been done already, it's just a matter of digging the right post on another forum (with the worst custom board software ever with even worse search functions)

 

[suryad]

You can whine all you want for better backwards compatibility, you wont get it

Am I the only one who does NOT want any backwards compatibility? Isnt that the main thorn in the side of MS and their products and part of a reason why making progress is so hard for them? Look at the jump from Mac OS 9 to OS X...as an example.

 

[Richard]

Am I the only one who does NOT want any backwards compatibility? Isnt that the main thorn in the side of MS and their products and part of a reason why making progress is so hard for them? Look at the jump from Mac OS 9 to OS X...as an example.

Well that's not a particularly good example. The first version of OS X was pretty much broken, so much so that Apple gave away 10.1 for free .

Anyway, while I do believe there's some BC that could be moved to virtualisation (I love Virtual PC), the truth is that most Windows users either don't care or don't want to use a dedicated virtualisation app, install a new OS inside their OS, and _then_ run their critical apps. I think MS is striking a good balance between having BC and not letting it hold back inovation. It's a complex problem to solve.

EDIT: And a bit of sad news: MS has qualified "bring back column headers for all views, not just detail" as Closed (Won't Fix). That's very unfortunate.

 

[Florin]

Am I the only one who does NOT want any backwards compatibility? Isnt that the main thorn in the side of MS and their products and part of a reason why making progress is so hard for them? Look at the jump from Mac OS 9 to OS X...as an example.

Is there another reason than backwards compatibility for Windows' dominance?

 

[suryad]

Is there another reason than backwards compatibility for Windows' dominance?

Good point but it games are a good selling point, also audio visual software like Photoshop and video editing suites...and also that OS is just easy to use and familiarity....

 

[pjbliverpool]

Am I the only one who does NOT want any backwards compatibility? Isnt that the main thorn in the side of MS and their products and part of a reason why making progress is so hard for them? Look at the jump from Mac OS 9 to OS X...as an example.

It depends what you mean by backwards compatable.

Do I want it to play all the games and run all the applications that were released before it for XP and Vista? Hell yes!

How far back do I want that compatability to extend? To be honest, i want it to extend all the way back to DOS. If I get the urge to fire up Doom 2 I just wanna be able to do it without worrying about compatability.

Still, if its seriously holding performance back on modern applications, I wouldn't mind dropping native backwards compatability for say the pre-XP era. Especially if they could run older apps via emulation.

 

[DJ12]

Am I the only one who does NOT want any backwards compatibility? Isnt that the main thorn in the side of MS and their products and part of a reason why making progress is so hard for them? Look at the jump from Mac OS 9 to OS X...as an example.

They could start with ditching x86 support.

 

[swaaye]

x64 (XP, Vista, 7?) does dump 16-bit app support, meaning lots of old installers from even the Win9x era stop working. And DOS apps don't work because the NTVDM is gone. Fine with me as I use DOSBOX anyway. And the few games that use 16-bit installers usually have online workarounds now.


So, as this is a "beta" test, how does MS get feedback? What is the point of this beta test? Is there a way to submit comments? I haven't bothered to get 7 though but I'm wondering if there's any point other than "super exciting preview of yet another MS OS"?

 

[Richard]

x64 (XP, Vista, 7?) does dump 16-bit app support, meaning lots of old installers from even the Win9x era stop working. And DOS apps don't work because the NTVDM is gone. Fine with me as I use DOSBOX anyway. And the few games that use 16-bit installers usually have online workarounds now.

Yeah, DOSBox is fine for old games. Also, I should note that Windows Server 2008 R2 (also in beta) lets admins dump x86 support altogether.

So, as this is a "beta" test, how does MS get feedback? What is the point of this beta test? Is there a way to submit comments? I haven't bothered to get 7 though but I'm wondering if there's any point other than "super exciting preview of yet another MS OS"?

I don't know how the public beta works but TechNet and MSDN subscribers can submit feedback and we can track the (public) reports, vote on the severity, validate them, comment and post workarounds. MS sometimes comments and then classifies the issue from Active to Resolved and Closed. Some issues I've seen have been closed with "Won't Fix". If you've used Connect before it's the same deal.

The impression I get, is that the public beta users will be most useful for the aggregate telemetry data, rather than explicit feedback reported. Compatibility reports, installation failures, performance bottle-necks and so on. W7 comes with several troubleshooting wizards that let you see your particular instance of some of the data MS has access to.

 

[suryad]

It depends what you mean by backwards compatable.

Do I want it to play all the games and run all the applications that were released before it for XP and Vista? Hell yes!

How far back do I want that compatability to extend? To be honest, i want it to extend all the way back to DOS. If I get the urge to fire up Doom 2 I just wanna be able to do it without worrying about compatability.

Still, if its seriously holding performance back on modern applications, I wouldn't mind dropping native backwards compatability for say the pre-XP era. Especially if they could run older apps via emulation.

I can see and and agree with what you are saying but I would like MS to be a bit bolder and just dump backwards compatibility completely. Use an old OS if you want to use your old stuff. A change like that in my most humble opinion is needed...sort of like spring cleaning Brand new efficient frameworks and APIs to develop new software on is the way to go...but that could be the software geek inside me that is wanting that and forgetting about practicality

 

[nutball]

I can see and and agree with what you are saying but I would like MS to be a bit bolder and just dump backwards compatibility completely. Use an old OS if you want to use your old stuff. A change like that in my most humble opinion is needed...sort of like spring cleaning Brand new efficient frameworks and APIs to develop new software on is the way to go...but that could be the software geek inside me that is wanting that and forgetting about practicality

Well plenty of others have tried this and lost.

 

[swaaye]

I can see and and agree with what you are saying but I would like MS to be a bit bolder and just dump backwards compatibility completely. Use an old OS if you want to use your old stuff. A change like that in my most humble opinion is needed...sort of like spring cleaning Brand new efficient frameworks and APIs to develop new software on is the way to go...but that could be the software geek inside me that is wanting that and forgetting about practicality

I think that thorough backwards compatibility is one of the reasons that MS dominates, actually. Home users may be happy to run out and buy new hardware every year or so, but there are MS customers that aren't so happy about it or interested in doing it.

There may not even be any real advantage in dumping that backwards compatibility anyway. And as I said earlier, they did dump 16-bit Windows and DOS app support way back with XP x64 and with every x64 variant since. Dumping 32-bit x86 support would be suicide. And AFAIK, cuz CPUs happily run 32-bit software almost completely natively in x64 mode.

 

[Davros]

On one hand i dont want to loose backwards compatability
but on the other if you take beos as an example the performance to be gained from starting totally from scatch is pretty amazing

 

[Humus]

I can see and and agree with what you are saying but I would like MS to be a bit bolder and just dump backwards compatibility completely. Use an old OS if you want to use your old stuff. A change like that in my most humble opinion is needed...sort of like spring cleaning Brand new efficient frameworks and APIs to develop new software on is the way to go...but that could be the software geek inside me that is wanting that and forgetting about practicality

Well, considering all the crap MS has gotten for not implementing DX10 on XP it's not hard to understand that they usually opt for backward compatibility.