Discussions

@SteveRichter:

I'd be happy if the standard ATX supplies would have a way of communicating that "oops, main power disconnected, you have about 0-5 seconds to flush data to disk depending on how much juice there is in my capacitors).

Technically it could be made such that it

1) sends a message that mains power is out

2) after certain standardized amount of milliseconds, 12v output is cut, this gives enough time to spinning HDD to decide to say "OK main power cut and no time to write more data to disk, I'm going to try to send this data in my cache back to the OS which can then send it to a 3 or 5 volt flash memory device for store until such time power is restored

3) there could be motherboard input for a battery that is used to keep RAM powered after the PSU caps run out and there was not enough time to flush data to SSD. maybe a driver could be made to write the cpu and RAM to SSD with this battery power at minimal CPU clock speed while GPU's etc buses are powered off

4) apps could request, with user approval, access to the "mains power cut" message, so if user wanted the data of a particularly important app saved with priority over other apps, that could be done.

If this is about that story I read earlier, it has to do with attempting similar sort of "acceleration" as the Amazon browser does. Technically the idea is sound since if you do most of the web page requests with lower latency and then push them to the phone in one roundtrip, that should make web loading 10-1000x faster depending on the latencies involved. For HTTPS though, it should be option for the user to "opt-in" whether they want to get similar acceleration for the secured sites.

The only way banks could *try* to enforce that there is no MITM, is by shipping/physically handing out some sort of USB device where the access to the bank information was en/decrypted but unless that device plugs between your keyboard and the computer, it would still be quite vulnerable for sniffing (they should just hand out a custom smartphone when you open account). I know one bank which requires users to install Java runtime (while experts have for years complained that it's full of exploits, requiring additional measures to attempt to secure it if you have to install it) that they use to run a custom en/decryption at the client computer. This is much better than using the Windows API's which are easily sniffed but the fact they made it with Java makes it stink since most users won't know or bother even the most basic level of securing JRE once installed (namely, turning it off for everything but the bank that needs it or installing it in a VM).

When it comes to real security, you need a way for the system to inform that an attacker is studying it. If the security system is not "obscurity based" then there's likely less need to study it in order to break it and you may not get any warning before the attack takes place. This and ability to isolate the system during the reverse engineering phase, is what makes even the most "sophisticated" security systems worth nothing. The only good security is the kind that the attacker cannot study without getting noticed and there has to be reason to study and that's why I prefer security by obscurity in addition to the "mathematically sound security", which really is only secure if you assume the attacker is some empty pockets thief that can't afford hundred thousand custom designed chips (or cloud compute time) to crack your stuff.

I was trying to post a comment on this review, without registering a new account since they offer options to sign in with existing account.

http://www.huffingtonpost.com/frank-schaeffer/the-debacle-that-is-the-m_b_2360368.html

So I tried the various options for signing up with existing account and:

If you sign in with:

Google, you give away "Google Contacts access", and it's not stated if this access is given on a permanent basis.

Hotmail/Live/AOL, you apparently don't give any access, but reading through the Live URL suggests that you do, since it contains stuff like "%2Fauthorize%3Fscope%3Dwl.basic%252Cwl.contacts_emails%252Cwl.offline_access" but the Live sign-in doesn't say what that entails atleast at the point where you enter your Live password.

Twitter- "This application will be able to:

Again, it must be assumed Huffington Post wants access to do all of that in your name in Twitter for as long as they want. If it doesn't mean that, it should be worded otherwise.

Yahoo- "Allow sharing of your Yahoo! profile and connections info with The Huffington Post."

Clearly, the pattern is that some of these say what they will give access to, Live doesn't say, unless you read the URL and make your own conclusions, but that's not very convenient.

I would like if C9 could contact Hotmail/Live executives for comment on how they plan to improve the disclosure policies. (no point asking WHEN since we all know it takes Microsoft atleast 2+ years to implement the most trivial of changes in any product)

These 64 bit replies should be truncated as this was about using 32 bit OS.

I have both 64  & 32 bit installs and I use the latter for stuff that I have found to not work satisfactorily in 64 bit Windows 7. I just find that it starts to crawl after I have lot of IE's opened and it's paging things. One iexplore.exe, 7 million page faults...

Like I said in the op, when writing to pagefile in 32 bit OS, write to both the inaccessible (over 4GB) memory, and disk. Only read from the ram disk. When the ram disk is full, then purge least frequently used pages from the ram disk and read them back from disk only when needed. Also you could provide some notification that the ram disk is full and further access will be slower.

This would be great improvement over the current situation with 32 bit OS where most of the memory is not used and the system crawls after opening couple dozen IE windows, leaving them open and then coming back to them later which I have verified to be due to pagefile access. I very rarely have needed more than 8 GB of memory during using 64 bit Win 7 so I know this issue is solved well enough with ramdisk pagefile for this use case.

re: the Problem of programs being limited to what they can access in Windows 7 32 bit.

This isn't really that big problem. Few programs I have use more than 2 GB of memory. And those that do, can use said memory without problems with either a 32 bit server kernel or by patching the Windows 7 32 bit kernel by changing one or two bytes, enabling larger allocations. I've seen couple games that crash in consumer Windows and work in server Windows because of their texture allocation being simplistic. Though it may be also that ASLR introduction has something to do with this, but I'll test that theory ASAP to see if it ends the random crashes on level loads of same map. (eg. Run game, load map, crash. Run game again, load same map, no crash)

edit: I found on SuperUser comments that DataRam RamDisk and SuperSpeed RamDisk Plus can both allow use of the un-available ("upper") memory in 32 bit os as ram disk for page file. Setting ClearPagefileAtShutdown may be needed to prevent errors on boot.

Quick googling found an answer but it may be brand specific. This page doesn't say it but I heard from another forum that this does exactly what I suggested. Too bad I don't have ASRock mb. Atleast now I know what brand to get for my Haswell setup. I'm getting a brand whose engineers say "can be done":

http://www.asrock.com/feature/xfast/xfastram/index.asp

XFast RAM is designed to allow users to manage the wasted memory space under Windows® 32-bit OS*, improve web surfing experiences, boost application's speed 5 times faster and reduce the frequency of accessing SSDs or HDDs to extend their lifespan.

evildictaitor > ... can't be done.

How to run 32 bit Windows with 8 GB so that PageFile is located in upper half of 8 GB?

I have 32 bit Windows installed for some specific reasons but I find that without pagefile the memory runs out too soon and I don't want the pagefile on SSD and I don't want it on HDD either. I want it in the un-used 4 GB that's after the 32 bit address space.

How do I do that?

e: I don't mind if the page file is also written to the HDD simultaneously as long as it isn't read from there as that's when I find the system to crawl, not when Windows writes to it.

@Dr Herbie: That might be satisfactory if the desktop experience was practically identical to Windows 7. Even in 7 you could configure it to resemble more XP/Vista style by setting small icons + combine when taskbar full and you can even add quick launch icons back just as it was in XP with little work.

I currently pin often launched document based stuff (excel, VS etc) on the start menu and only very frequently used stuff in the taskbar (IE, notepad replacement etc). The more rarely used things I launch from either desktop or the Run (win+r). I prefer the more often used or text input based stuff in Run history than desktop as I can start then start and run them purely without moving my hand to the mouse, where as mouse based stuff can be on the desktop as icons. I don't launch stuff using Start Menu search at all since I always disable indexing services and as Windows unindexed search is unusably slow I use NTFS search as that handles millions of files in second. I don't think the MSFT search guys can even dream up that kind of perf, or if they can, it will involve some background indexer service that happens to run when I'm using a MIDI keyboard and increase jitter...

re: release cycle

That would make sense. I'm tempted to here think about all the crap I've read (I don't own their devices, have only tried out) about Apple and Google cycle/updates. As Microsoft has experience here and should know what doesn't work well in the competitors models, I'll just wait and see if this would actually work out better than what the competition is doing.

@Sven Groot:

That much is obvious but getting there by creating a negative sentiment for the early adopters doesn't seem like the best strategy. They could have at minimum made the Modern UI a "default" that desktop users can easily change to the Windows 7 style experience from a checkbox.

 re: The history lesson,

I don't this is quite comparable. I used similar self made boot menus and config.sys etc as most gamers had to, we had many boot options for specific games as well. I could have DR-DOS 6 installed and Win 3.1 did not mess around with the DR-DOS 6 environment. Win 8 going to Modern and then messing up a great deal with the Win 7 style experience on top of that pretty much created what you just said. I installed Windows 8 on alternate boot, took a good look and haven't bothered going back, particularly  after finding some biggest most popular games had problems in their installer in Win8, which leaves one to question about the compatibility for all the obscure stuff I have. (I did manage to get the popular game installed but only to find out that it said it won't support anything newer than Windows 7 when running it and compatibility mode didn't help because this message was from the drivers the game installs - maybe this has been fixed by now though for this popular title)

If you count Windows versions from 1.0 to 98, then add that count to the Modern UI version, giving atleast say a year for each version, we might be closer to when the Modern UI is actually usable. 

Making the Modern UI as "in your face" for desktop users at its "v1 stage" was a big mistake. I support the idea of a tablet/laptop hybrid that can also easily hook to larger monitors but this is comparable to making computers boot to Win 1.0 and then needing to exit that in order to use the apps you really want to use, and do this crap until Windows 98 came around.

My hope is that most people are smarter than to install Win 8 even if they got it free, as that will send a clear signal that the bean counters will understand. Yes. I would not even install Win 8 if you gave me it for free. And infact, the low ball offers MS made for Win 8 upgrades just suggest that it's real value is closer to worthless than Win 7 which was not so heavily discounted.