Actually, it's because you have a MAC that you're vulnerable. Without one you'd not have a network connection, and you'd be perfectly safe from this attack.
Now, having a Mac would make having a MAC much less risky than under Windows...
Just for fun I cranked up Microsoft Multiplan 1.06 from 1983 for the Commmodore 64 (using the Vice emulator, and the magic calculation (850*77.1) gives the correct answer of 65535.
Though not quite the same, PRINT -65535-1 on the Sinclair Spectrum gives -1E-38, due to a ROM bug!
I don't know how this works with MSI packages, either, because elevation doesn't occur in the same way.
Windows Installer has a service component, which is already running elevated. I guess it just impersonates the user for any normal running, though it also has the new Session 0 restrictions to work around to interact with the desktop.
EMEditor already has a helper application (EEAdmin.exe) that it uses for certain operations (e.g. changing file association) which require elevation.
Maybe just a new RPC interface needed to that, and save/load could be done as you described?:-)
This would be a huge security problem. Threads do not have their own protected memory space, so it is impossible to prevent a lower-privileged thread from screwing with the code or data of an elevated thread.
Hmmm, that's true! I was thinking that VirtualProtect could block untrusted callers from making code areas writeable, but just having writeable data could be enough to cause problems. Not being able to return to filtered token must just be because it's per-process, and could cause all sorts of problems if things started failing due to insufficient rights.
I just wish they'd provided a ShellExecuteEx verb for running non-elevated, to match the "runas" used to elevated. I can't think that would be a security risk, and it would make life so much easier in the cases it is needed.
Now, EMEditor is Vista compatible (certified even), but it would be nice if it could elevate when a write operation fails due to incorrect permissions. Then you could just edit the file as usual, and elevate when you save.
The biggest problem with elevation is that it's not something you can do from an existing process, without launching a completely new process as elevated. Task Manager relaunches itself when you click the Show Processes From All Users, rather than doing any magic. Elevated processes can't return to a limited token either, which causes much grief for setup programs - there's a horrid work around involving scheduling a new task, just so it runs as a normal user with a filtered token again!
The text editor could still support it, but it would need to launch a separate helper application just for saving (and loading?). All data to save/load would have to be passed from the main editor to the helper, probably using RPC. If only the elevation could be done per-thread, it would be so much easier...
Avira AntiVir also reported a virus in my windows-based installer, and a couple of others reported it as suspicious. I reported it to Avira, and they came back fairly quickly with a confirmation that it was a false positive, and that it would be fixed in a future definition update (they didn't say when).
I was using UPS to compress the executable header on an NSIS installer, which seemed to be a combination likely to freak out the "smart" detection of many scanners. Avoiding the use of UPX on the installer cleared everything up for me, tho it was still annoying that I wasted a couple of hours on it and had to convince the reporting user that there really wasn't a problem!
I ended up using http://www.virustotal.com/ to check my new installer against about 25 of the major scanners - very handy free site...
I wonder why it doesn't have gig capable ports on it. Especially as the wireless is now faster than the wired! It's a show stopper for me, and I'm gonna wait until they fix it...
Unregistering is just calling a function inside the DLL which deletes its own registry entries. It needs to be loaded for that to happen so being loaded already is no problem. When both have finished it'll get unloaded, and the lack of registry entries means the browser can't create an instance of it again.
I'm not sure I'd want to create a page to do it tho, even with full permission from the user...
I gotta ask why on earth would you think an ancient compiler would run on the latest OS release? It's a Win32 application running on a Win32 platform, and has no driver requirements to complicate it, so why shouldn't it? After jumping through a few hoops and using Admin rights, it does indeed work. Shame they didn't include it in all the other auto-right-boosting detection they already do for installers.
I need VS6 for legacy reasons too, to build applications with Win95 support. For some reason the VS2005 CRT startup code uses IsDebuggerPresent(), which is a kernel32 export not available on Win95. Except for that one problem my application runs fine, so I'm sticking with the only development tool that does the job. It's also something they could have easily solved with a runtime check for the function too.
VS.NET and 2003 were awfully slow and bloated, and only really useful to C#/.NET programmers. They were a definite step back if you were only developing in C/C++. VS 2005 pulls things back and is definitely better, except for the stupid CRT requirement.
You know since you like old software, perhaps you can run OS/2 Warp 3. Screw windows. I do have a thing for old software/hardware, though that's away from my work requirements. I did try and install OS/2 Warp 3 a few months back, but it chokes on the large cylinder count on my HDD. I do have BeOS R5 Pro installed on another partition too:)
Actually, both Visual Studio 6 and the very latest Visual Studio 2005 give warn of incompatibilities when you install/run them under Vista. Both must be run with Admin rights for full functionality, and VS6 requires the post-boot continuation installer to be cancelled and re-run as Admin to finish correctly. There are no official solutions for these issues yet, and I'd guess that VS6 may never be fixed.
Isn't it great when MS can't even get their own products to be compatible?
Fair enough, donate Hawking's issues of Private Eye to a library or something. Though it might be best for everyone if Professor Thorne keeps his used Penthouse subscription.
It's a new print path running in parallel to the compatibility GDI print path, to produce a XPS document. It's exactly what OS X does for producing PDF documents, and since XPS is effectively Microsoft's version of PDF, they're reinventing something else Apple did 5 years ago.
The labels are a cool idea, and provide flexibility that you can't get with folders. But it's still super-handy to be able to dump all mail, from, say, a mailing list into a particular folder.
I do just that, with different labels for each mailing list. Looking at the sidebar I can see how many unread messages are in each one, leaving anything not matching a label rule in my Inbox.
Alternatively, a way to view only messages without labels would be helpful.
It would, and I'm sure it'll get added at some point. Using my Inbox for unhandled messages is generally as good as viewing unlabelled messages in my case. As soon as I've processed them I Archive, and the search function makes it trivial to find anything anywhere too.
The trouble is that SF3 is very difficult to crack, so there aren't many no-CD cracks available. The couple that I have found are for the original release only, so you can't install any patches.
I think it'd definitely be worth me contacting each of the companies to ask about 64-bit support. If enough of us do it they maybe they'll actually fix it - I'm not very hopeful tho, as it'll probably cost them money to do.
Last time I checked, most (all?) games using the Starforce 3 copy protection system wouldn't run under x64. It uses a driver for low-level CD access, and there was no 64-bit version available.
I'd quite like to upgrade my home PC to make 64-bit development and testing easier, but I'm gonna hold off if it'll break the 5-6 SF3-protected game titles I own. The games themselves will run fine I'm sure, so it's insane the copy protection is the only obstacle. Who was it that said DRM was not an issue for legal owners?
Opera did go through a patch with version 7 where there would be random crashes. They were a bit irritating, even tho the session recovery would restore all your open tabs+history when you opened it next.
Thankfully, crashes are long gone now and it's as fast and smooth as ever. Still not quite up to Firefox-level compatibility, but display problems are pretty rare. About the worst is the odd site that refuses to work with it, even though fiddling the user-agent will show it works fine.
If you're not a current registered subscriber, grab your free key while you can. It's definitely worth another look - certainly until Firefox can be speeded up (it's frustratingly sluggish after Opera!).
In recent years the same thing has been done with Sinclair Spectrum games, recorded using an emulator: http://www.rzxarchive.co.uk/. My favourites replays are Jet Set Willy and Head Over Heels (the latter is VERY fast/tight).
They're in a replay format which includes the game code, and all input needed to replay the file. You'll need a supported emulator to play them, as detailed on the site itself.
One of the "toy" Linux servers in work was compromised, using an Apache SSL exploit. It had a rootkit installed on it and was doing SSL scans for other vulnerable machines. This was a few years ago now...
The connections came from Romanian dial-up accounts, and I reported it to the ISP (obviously, nothing was ever done). I also extracted IRC information from a bot on the compromised server, and joined the channel. I found a handful of other bots in there, and mailed the admins of each I could see (with some responding to thank me).
I sat in there all day and some real users joined - though I was named like a bot I decided to chat. They came across as being very young, highly likely just script kiddies rather than major scammers. I guess things might have changed now there's money to be made!
I'd guess 95%+ of Windows apps use MessageBox to prompt for simple confirmation, where the order is fixed (Yes/No/Cancel, OK/Cancel, etc.). Anything creating a custom dialog in the Cancel/OK order just doesn't look/feel right to me, and certainly stands out as non-standard.
As a fairly recent Mac convert I'm a fan of the text-based buttons, that don't require you to carefully read the sentence above them to make sure you hit the right one. It's been a while since I've used GNOME, so I didn't realise they'd gone that way.
I suppose there's nothing stopping MS changing the standard order, which will change most apps, but they'll still be simple Yes/No/Cancel type buttons that are no better. Maybe they need a MessageBoxExEx that accepts button labels for Longhorn...
Slashdot Mirror
Actually, it's because you have a MAC that you're vulnerable. Without one you'd not have a network connection, and you'd be perfectly safe from this attack.
Now, having a Mac would make having a MAC much less risky than under Windows...
Just for fun I cranked up Microsoft Multiplan 1.06 from 1983 for the Commmodore 64 (using the Vice emulator, and the magic calculation (850*77.1) gives the correct answer of 65535.
Though not quite the same, PRINT -65535-1 on the Sinclair Spectrum gives -1E-38, due to a ROM bug!
Windows Installer has a service component, which is already running elevated. I guess it just impersonates the user for any normal running, though it also has the new Session 0 restrictions to work around to interact with the desktop.
Maybe just a new RPC interface needed to that, and save/load could be done as you described?
Hmmm, that's true! I was thinking that VirtualProtect could block untrusted callers from making code areas writeable, but just having writeable data could be enough to cause problems. Not being able to return to filtered token must just be because it's per-process, and could cause all sorts of problems if things started failing due to insufficient rights.
I just wish they'd provided a ShellExecuteEx verb for running non-elevated, to match the "runas" used to elevated. I can't think that would be a security risk, and it would make life so much easier in the cases it is needed.
The biggest problem with elevation is that it's not something you can do from an existing process, without launching a completely new process as elevated. Task Manager relaunches itself when you click the Show Processes From All Users, rather than doing any magic. Elevated processes can't return to a limited token either, which causes much grief for setup programs - there's a horrid work around involving scheduling a new task, just so it runs as a normal user with a filtered token again!
The text editor could still support it, but it would need to launch a separate helper application just for saving (and loading?). All data to save/load would have to be passed from the main editor to the helper, probably using RPC. If only the elevation could be done per-thread, it would be so much easier...
Avira AntiVir also reported a virus in my windows-based installer, and a couple of others reported it as suspicious. I reported it to Avira, and they came back fairly quickly with a confirmation that it was a false positive, and that it would be fixed in a future definition update (they didn't say when).
I was using UPS to compress the executable header on an NSIS installer, which seemed to be a combination likely to freak out the "smart" detection of many scanners. Avoiding the use of UPX on the installer cleared everything up for me, tho it was still annoying that I wasted a couple of hours on it and had to convince the reporting user that there really wasn't a problem!
I ended up using http://www.virustotal.com/ to check my new installer against about 25 of the major scanners - very handy free site...
Unregistering is just calling a function inside the DLL which deletes its own registry entries. It needs to be loaded for that to happen so being loaded already is no problem. When both have finished it'll get unloaded, and the lack of registry entries means the browser can't create an instance of it again.
I'm not sure I'd want to create a page to do it tho, even with full permission from the user...
If this is to be a tera_BIT_ drive then I believe the headline should read "Tb" rather than "TB".
:-)
And ironically, there's only 1 bit difference between those 2 statements
http://retrospec.sgn.net/game-overview.php?link=a
The Telnet client is missing from the default install too, tho there's an option somewhere to add it back in.
I need VS6 for legacy reasons too, to build applications with Win95 support. For some reason the VS2005 CRT startup code uses IsDebuggerPresent(), which is a kernel32 export not available on Win95. Except for that one problem my application runs fine, so I'm sticking with the only development tool that does the job. It's also something they could have easily solved with a runtime check for the function too.
VS
Actually, both Visual Studio 6 and the very latest Visual Studio 2005 give warn of incompatibilities when you install/run them under Vista. Both must be run with Admin rights for full functionality, and VS6 requires the post-boot continuation installer to be cancelled and re-run as Admin to finish correctly. There are no official solutions for these issues yet, and I'd guess that VS6 may never be fixed.
Isn't it great when MS can't even get their own products to be compatible?
Though only by:
- Signing it with an approved code signing certificate for $300-500
- Connecting a kernel debugger to the machine
- Pressing F8 to disable digital signature enforcement on every boot
- Patching the kernel to disable the enforcement (until MS update and fix it)
and none of those are going to be particularly attractive to most people. Which method were you thinking of for bypassing?Fair enough, donate Hawking's issues of Private Eye to a library or something. Though it might be best for everyone if Professor Thorne keeps his used Penthouse subscription.
I get the feeling someone's taking Cockney Rhyming Slang a bit too far...
So if you are concerned about this, just turn it off.
But why the heck is this turned ON by default??
The labels are a cool idea, and provide flexibility that you can't get with folders. But it's still super-handy to be able to dump all mail, from, say, a mailing list into a particular folder.
I do just that, with different labels for each mailing list. Looking at the sidebar I can see how many unread messages are in each one, leaving anything not matching a label rule in my Inbox.
Alternatively, a way to view only messages without labels would be helpful.
It would, and I'm sure it'll get added at some point. Using my Inbox for unhandled messages is generally as good as viewing unlabelled messages in my case. As soon as I've processed them I Archive, and the search function makes it trivial to find anything anywhere too.
The trouble is that SF3 is very difficult to crack, so there aren't many no-CD cracks available. The couple that I have found are for the original release only, so you can't install any patches.
I think it'd definitely be worth me contacting each of the companies to ask about 64-bit support. If enough of us do it they maybe they'll actually fix it - I'm not very hopeful tho, as it'll probably cost them money to do.
Last time I checked, most (all?) games using the Starforce 3 copy protection system wouldn't run under x64. It uses a driver for low-level CD access, and there was no 64-bit version available.
I'd quite like to upgrade my home PC to make 64-bit development and testing easier, but I'm gonna hold off if it'll break the 5-6 SF3-protected game titles I own. The games themselves will run fine I'm sure, so it's insane the copy protection is the only obstacle. Who was it that said DRM was not an issue for legal owners?
Opera did go through a patch with version 7 where there would be random crashes. They were a bit irritating, even tho the session recovery would restore all your open tabs+history when you opened it next.
Thankfully, crashes are long gone now and it's as fast and smooth as ever. Still not quite up to Firefox-level compatibility, but display problems are pretty rare. About the worst is the odd site that refuses to work with it, even though fiddling the user-agent will show it works fine.
If you're not a current registered subscriber, grab your free key while you can. It's definitely worth another look - certainly until Firefox can be speeded up (it's frustratingly sluggish after Opera!).
In recent years the same thing has been done with Sinclair Spectrum games, recorded using an emulator: http://www.rzxarchive.co.uk/. My favourites replays are Jet Set Willy and Head Over Heels (the latter is VERY fast/tight).
They're in a replay format which includes the game code, and all input needed to replay the file. You'll need a supported emulator to play them, as detailed on the site itself.
One of the "toy" Linux servers in work was compromised, using an Apache SSL exploit. It had a rootkit installed on it and was doing SSL scans for other vulnerable machines. This was a few years ago now...
The connections came from Romanian dial-up accounts, and I reported it to the ISP (obviously, nothing was ever done). I also extracted IRC information from a bot on the compromised server, and joined the channel. I found a handful of other bots in there, and mailed the admins of each I could see (with some responding to thank me).
I sat in there all day and some real users joined - though I was named like a bot I decided to chat. They came across as being very young, highly likely just script kiddies rather than major scammers. I guess things might have changed now there's money to be made!
Strangely, I was thinking more along the lines of:
kid: Mommy, can I have an ice-cream?
mom: No
kid: Mommy, can I have an ice-cream?
mom: No
kid: Mommy, can I have an ice-cream?
mom: No
kid: Mommy, can I have an ice-cream?
mom: No
kid: Mommy, can I have an ice-cream?
mom: No
kid: Mommy, can I have an ice-cream?
mom: Yes, if you'll be quiet then!
I'd guess 95%+ of Windows apps use MessageBox to prompt for simple confirmation, where the order is fixed (Yes/No/Cancel, OK/Cancel, etc.). Anything creating a custom dialog in the Cancel/OK order just doesn't look/feel right to me, and certainly stands out as non-standard.
As a fairly recent Mac convert I'm a fan of the text-based buttons, that don't require you to carefully read the sentence above them to make sure you hit the right one. It's been a while since I've used GNOME, so I didn't realise they'd gone that way.
I suppose there's nothing stopping MS changing the standard order, which will change most apps, but they'll still be simple Yes/No/Cancel type buttons that are no better. Maybe they need a MessageBoxExEx that accepts button labels for Longhorn...