yeah, well good luck with that. There's ways to get around tables. They're 3x as long, and not as nice. I wrote 5 different themes for my website (used to be xhtml generated from xml through xsl, but then my x key broke so I switched it to regular html output. That sentence is a lie. My xsl parser when I was outputting xml (which is what xhtml is) would shorten things like the textarea tag to , which just screws up some browsers.)
Anyway, 5 different themes, researching table-less design, etc. You know what? I had to put divs all over the place instead of tables. Not all too much cleaner, but I guess better 'logically', but here's the big problem: No resizing.
Example 1: I want to make a form. Forms look nice if the labels are all lined up and all the form elements are lined up. Let's assume I want a page that can have varying size text (which is what people say should be allowed with css).. Ah fuck, can't use divs here. Oh well. I tried. From what I've found, there's no way to link the size of one div to the size of the others in a way that makes logical sense to the document format and flow. I'll welcome any suggestions proving me wrong.
Example 2: Multi-column layouts are annoying as hell. Yeah, they're annoying in tables too, but damn. That's why there's entire web pages devoted to getting away from tables with weird, complicated CSS to do half of the same thing. (This isn't as bad with CSS2 support, min and max width/height are nice, but not supported by all browsers I have to use).
Example 3: Picture page. Captions on the bottom of thumbnails. Again, let's go for optimal use of screen space: float them all, so that they wrap around when there's no horizontal space left. Ah fuck, one of the captions is a line longer than the others! There go all the floats in the next line. Yes, this can't even come close to being done with tables, but it's something that pisses me off anyway. Unless I know exactly which font the user is using, and at exactly which size, I can't make the boxes for each picture to be floated the right height. There's no way for other boxes on the same line to inherit the height of the largest on that line, if they're all floated. This isn't necessarily in the same topic, but wow is it annoying as hell.
Honestly, I don't know why the hell we're taking so long to get AWAY from HTML. HTML as originally designed was a way to structure content logically. Then along came graphical browsers, and it turned in to a presentation language. Then along came the W3C, who try and force it back to a structural language. But that's not what people want! XSL:FO are too verbose, imho, but are much nicer for what I want to do. (Bleh, just cuz I like examples: Give the damned web browser a clue as to what a slashbox is. Don't give it a hierarchy of divs. Don't give it a table. Give it a fucking . Tell it using some other method HOW to render a slashbox. CSS is nice for saying how certain things are rendered, but only fixed attribtues of those certain things. Give me wide support for CREATING my own things, and saying how THOSE are rendered/produced/built.)
Some might argue that that's not a good idea, but holy shit I don't care! Limiting me to such small pieces of crap as divs and spans to build a decent webpage is retarded. Especially wen you want to get fancy and have them look nice (corner/edge graphics on your elements. Imagine slashboxes with curved corners. and a shadow. Oh crap, now half the html document is outputing those slashboxes, and our page is full of presentation again, not structure. You can't get away from presentation, so work with the people to minimize the harm.. CSS won't do it! (again: CSS2 isn't that bad. before and after do help somewhat, btu don't solve it completely)
That's why I use XSL: I dynamically make a page. Tell it what the different parts of the page have to say. the XSL then converts that to HTML, sometimes in a two step process (depends on theme). Step 1: Make everything in
Ok, I was a bit harsh because I'm having problems with egress filtering in a school situation: I live there, and they prevent me from doing my school work and other things by their stupid rules. Places of employment I agree with you a bit more. My fault for saying that it's bad at all times, there are certainly times when it's ok..
I still say egress filtering is a nuicance to people who know what they're doing, but I guess it is a necessary evil against people who think they know what they're doing, and just fark stuff up.
I hope you aren't advocating only allowing certain egress ports, because that right there is the cause of so much headaches for users it's not even funny. Block stupid crap like hotbar, gator, etc. etc.. but PLEASE do NOT make me have to stop my work to go bug the tech person to bug his superior, to bug their superiors to open an egress port. I have to deal with that at my school (work situations are slightly different, but still rather annoying). I have to basically tunnel everything that isn't ftp access on the standard port, telnet on the standard port, or ssh on the standard port. Web access must all go through the proxy (which is completely borked), making working on webpages a pain in the butt.
Egress port blocking = annoying as f*ck. And there's no reason for it that I can tell (unless you're completely anal that only certain programs be allowed, etc. In that case, shouldn't you also have a specific allow list of websites able to be viewed? Even then, it's rather annoying. IM services, etc are very commonly used in work places, and it helps a great deal to be able to IM the person in the next cubicle, or on the other side of the building, than call them up. Less disruptive to thought processes too.)
I'm not quite sure, but maybe if you already have a decent firewire network setup, you can add one as a gateway relatively easily, and therefore not rewire everything for ethernet. Admittedly, this seems rather stupid. However, if everything is setup properly, I believe with firewire you can buy a hub, and share (For example) a digital camera with any mac also connected to the hub. If you have a setup like this, then adding networking to it would be as simple as a configuration change on your gateway, be it connected via modem, ethernet, wireless, etc.
But yeah, especially since most macs don't even need crossover cables for the ethernet, I dunno what the big deal is.
But, but.. the yahoo article linked from slashdot said that it was 2-3weeks afterwards. Also, there's no mention on the IMAX Si..
Ok, what the hell. There WASN'T information on their site the last time I checked. Now there is. Nothing about a release date, besides 'June 2003', which isn't the normal Matrix release day (since that's 5/15, which is a Thursday according to my calendar)
What's the IMAX opening night? I thought it was two-three weeks after the regular opening night? I'm currently out of the country, but if it's 2-3wk delayed for IMAX, that's about when I'm getting home..
Hey yeah, something like this happened in the hitchhiker's guide..
Ignoring that, have three completely separate monitoring programs? Don't use the same code base, and therefore you don't run in to software problems, and minimize the effect of hardware problems (since the software would probably be interacting with the hardware in different ways, if there's a div bug in the CPU, the errors won't necessarily be the same between all three SEPARATE software programs). If they all monitor the same thing and produce the same output (or at least, output that's understandable to the other two), the likelihood of two messing up at the same time, and producing the same wrong output is rather small.
However, it's probably possible to knock two out so they produce DIFFERENT output, so you have one valid source, and two screwed up ones, but screwed up in different ways.. Then you basically have to reboot and hope it fixes them.
Would this work, or am I missing something here? Again, not perfect reliability, but it does tend to make it so that there's a bit of safety from software bugs..
yeah, I can never remember which is length and which is specific.. and I knew the no was superfluous when I re-read it, the parenthesis had screwed me up. *sigh*
Too bad I can't get off with such simple sentences in my tests. I have two finals next week, and I'm probably going to fail both. Fun:)
That question being posed, and assuming the answer to be you want to work on hardware: Good luck getting in to the center of a large Japanese corporation if you're a foreigner. I'm not saying it's impossible, but highly unlikely. Especially if you don't have a perfect command of the language. Foreigners over here are useful for a few things: Teaching drunken businessmen english, staring at, making fun of when they think we can't understand them, and scaring little children.
The best part is when they're talking about you on the train, sitting right next to you, thinking you can't understand what they're saying.. then you say something in japanese to your friend sitting next to you and they shut up for the rest of the train ride. *grin*
I hate to admit it, but you might have a better chance getting in to Microsoft to work on X-Box. Especially if you're doing an internship.. but good luck even for that. I have no personal experience, but I doubt something like X-Box hardware design would use interns much at all. At least there (if you can get in), you get some experience in your native language, and then maybe make the switch to a Japanese company.
Anyway, nan nen kan nihongo benkyou shita? ore.. san nen kan, demo, zenzen benyou shinakatta kara, amari jouzu ja arahen-n ya.. ^_^ Ima, kansai gaikokugo daigaku no gakusei, demo roku gatsu (getsu? I can never remember) no futsuka ni amerika ni kaeru to omoimasu..
Wow romaji is ugly as hell. Especially when you mix in kansai-ben. Bleh.:P
Hopefully the original asker will answer this question, but from what I read, it seems like he's more towards the hardware end of things (working on designing the consoles) than on the software (working on games for them). There's some crossover, but he's in computer engineering: hardware with a bit of software. Nowhere does he mention programming.
So, is it really wanting to be a programmer, or is it wanting to be a designer of the hardware?
Maybe I'm just not understanding completely, but how does that simulate symlinks in linux? It simulates mount points, which are nice, but symlinks are a bit different? Let's say I want C:\Documents and Settings\Spectral\Desktop\Blah to point to D:\Stuff\Blah (A Directory), how can I do that in windows? Sure, I can mount D:\ inside C's directory structure, but how do I get a link that at the FS level makes Blah look like a folder to everything? Or even on the same partition, it doesn't matter.. Hard links wouldn't work here, since they act differently. (I want to be able to delete one, then put another file there with the same name and have the link work. AFAIK, hard links don't work that way. Also, (again, afaik) hard links won't work across partitions/filesystems. I've never used a hard link though, so I might be wrong.)
This is certainly the unix way I guess. There shouldn't need to be a special packaging command to help me find the files though. This filesystem still makes more sense to me. If all things install to their own separate directory tree, then symlink them so everything also appears it's in one spot (like/Configurations.. or just call it/etc since it's shorter;)), we have the best of both worlds. I shouldn't have to depend on some package manager tracking every file that a program needs to run, especially if it's made with scripts afterwards. MacOSX has it right here: Most things are just a 'Package'. A Package is a compressed folder/disk image, and is treated like one by the OS.
Therefore double clicking it will run the program, but you can easily go right in side of it and see all the files, and treat them like files. This works on the command line too. (cd/MacOSXAppsDirectory/CompanyName/Program.Package/e tc will work. Of course those aren't real directory names, but you get the idea.)
There are similar commands for any packager because there needs to be. There's also a command for sorcerer that finds files it's not tracking. When wanting to COMPLETELY remove something, I have to check that list as well. And then hope that IT is complete. Being able to check a directory for a "data" folder, back that up if I want, then blow out the directory would be nice. (Yes, this does screw up symlinks. Therefore it MIGHT be better to have the directory for the program contain the symlinks, as opposed to scattering the symlinks in to the one solid directory. Unless there's a way to reversely traverse a symlink in constant time..)
Again, a pipe dream I'm sure, and I'll admit there are certain things about the linux/unix file system that are nice. Configs mostly in one place, etc. But yikes it's certainly a mess. Something like this would help a great deal, I think.
(Another example: at a konsole, hit k, then hit tab. How many things come up? How many of those are kde programs? Are those ALL the kde programs? Probably not. What if you want to see all the executables that are part of the 'kde distribution' ? I guess you're off checking your package manager: Make a list of what you consider the kde distribution, run that list through the package manager, dump that in to a tiny little thing that sees if they're executable, etc.
Not too much different than just using ls/find/grep/bash/whatever, but what if your package DB gets corrupt? If bash/ext2 gets corrupt you have a bit more to worry about I'd think.)
FS/OS support for links makes it so easy to do such cool stuff that's essentially impossible in some other operating systems (Shortcuts are files that are treated specially by the shell in Windows. Not by the OS's FS layer. Therefore, they're nowheres comparable.) Why don't we use that support to make a FS structure that makes sense to everyone, and kicks ass? You can keep the old layout, and have a nice new layout too. Best of both worlds.
Where the hell are you people getting this BS about it being in the windows directory?! It wasn't there in w2k. I should know, I'm running it right now. It's not there in XP (Professional). I should know, the computer next to me is running it.
That being said, the linux file system structure SUCKS! Windows isn't much better, but christ.. especially with the distros. Where is your config file for samba? Well, I don't quite know. It's somewhere in the/etc directory I'm sure. Is it in it's own subdirectory? Possibly! Let's go and see.
Having all the stuff AT LEAST symlinked from some common directory would be SO NICE. (cd/Programs/XFree86/4.3.. oh look, everything X installed.) Yeah, that could get confusing. Therefore there might be a bin directory, a config directory, and a data directory. They can all be symlinks, I don't care, but if I had to come up with where KDE stores it's default menu, I would have no f*cking clue. Somewhere in/usr I guess? Might depend on the distro.. Agh.
XP includes Zip-As-Folder support. an annoyingly limited subset of functionality is provided for it compared to a real folder though, which screws up the metaphor. Anything that looks like an Explorer window better f*cking act like one. Agh.
I meant remote via software/internet. I knew about grabbing RF from monitors, didn't think keyboards were that bad (less power, I thought it'd have been slightly more secure than the massive CRTs:)). I should have been clearer. Though you bring up interesting topic: the PGP thing, how does that work? is there a screen shot to see what they did to make it harder for RF receivers?
that's why the 3d accelerated one would work. the secure windows are textures that are marked as secure on the video card. grab a screen capture, and it doesn't grab the secure textures. Easier than invaldiating areas of the screen that might not necessarily be rectangular (think overlapping windows, secure on the bottom, you have to chop out at least two rectangles of 'secret' area. If the video card knows that the surface indicating the window is private, when grabbing framebuffer, it won't show that surface. Doesn't matter if zbuffer killed part of it, the rest is invalidated automatically.)
Or TEMPEST style radiation detection, etc. Yeah, there are ways to do it remotely, but I don't think (though I don't know for sure) that they're nearly as easy as a direct hardware tap would be.
Though I meant 'remotely' as in "Someone on the other side of the world through the Internet" type of remotely. They can't do that I'm almost certain. Lock down any remote threats, and assume people are smart enough to lock down physical local threats, if it really matters that much what they're doing.
However, your comment on encrypting the data sent between the card and monitor is NOT new. There were several talks of this before, I don't know where they went.. but a next gen connection that was encrypted and digital. (Or at least, certain parts would be encrypted: DRM protected stuff.) Therefore, no access to the DRM material. Could be 'elevated' in use to be useful for all material that you/microsoft/content manufacturers don't want copied.
Keeping things digital/encrypted as long as possible, making it harder to get to the analog signal lead to 1) slightly improved quality, 2) Much harder to crack.
They'll take a while in replacing the video cameras, but the loss of quality/difficulty of use is rather prohibitive. Expect the MPAA/RIAA to push for brain jacks to combat the rampant piracy of screeners and people with *gasp* microphones.
I don't know.:) You would think that the video driver would be able to stop this, but maybe not? Certainly older systems with memory mapped video would make it easy to grab the contents (if it is read/write..), I don't know how it's done anymore (I assume everything must go through GDI in windows. But maybe there are ways in their driver model to grab stuff from a window/screen buffer.)
I really don't know for sure, but I assumed that was the problem they were mentioning. Maybe the problem arises with DMA? Again, I'm not much in to the internals of Operating Systems and stuff, but if there's a way to link the hard drive and the video card, maybe another card on the system can do the same and link itself to the video card and grab its memory. Highly unlikely, but.. Maybe technically possible? (though, it's most likely up to the OS to initiate a DMA request, so again, it could probably be managed in software.)
Who knows. Typical linux troll response: They want to get everything under their control, to lock us out. I somewhat doubt this explanation, but it might be correct.:)
Not many computers would anymore, but a lot of older computers used a type of static ram. Therefore it would keep the memory without being refreshed, and could keep it for several hours, maybe days. The reasons were speed, but the cost is prohibitive for the amount of memory being used. Some memory technologies use small pockets of SRAM as a type of cache, before moving to the DRAM (D=Dynamic. Must be refreshed often, which is why it loses its contents after poweroff)
At least, I'm pretty sure that's correct, and also pretty sure that's why I've seen this effect.
The problem is that the memory in graphics cards aren't wrapped in to the security model. Therefore, anything with access to reading some memory from the video card (not too uncommon I'd imagine), can go and grab everything. Including the current screen contents.
Video cards therefore need to be modified to be secure and support access control on their memory, the way the CPU's privileged mode (and whatever hardware they're going to tack on to make palladium work 'better') allows the OS to control what can access certain parts of main memory.
USB: Handled by the OS. Easy to deal with. Monitors: would require hardware tapping, much harder to do (Especially remotely). Keyboards: Again, MOSTLY handled by the OS. (Windows passes most every key combination through hooks, except ctrl+alt+del. They'll probably change this so that if a secure window is on top, no hooks grab the data.)
Etc. etc. I don't argue that it's a bad idea (that there needs to be changes to the video card hardware to support this properly), it's just very poorly worded in the article.
I wonder if maybe the 3d support being used helps this? If you define everything as a texture, then you only need to secure certain textures (the secure ones), not the entire screen. Therefore things running unmanaged still work just fine. 2d accelerated blits might also do the trick, but probably not as well. (Do the 2d accel blits have a concept afterwards of what a window is, and therefore to hide a certain one? Probably not..)
Humans are a security problem, because they contain their own pool of memory too. Let's get rid of them. Deleting a person's memory is easier than the video card's too: One click of the trigger is all it takes. Just Point and Click.
I'd have no clue how to wipe out my video card's memory. (No, shutting off the computer won't do it. I've seen plenty that when they turn back on, the last screen visible is there for a split second.)
Probably because it's personalized, it's harder to spoof the window. Password boxes using data that only the OS knows and personalized for that computer are better. At least, if all dialog boxes looked one way, then up came a popup that looked compeltely different, it's pretty damned obvious it's a fake, and you don't want to put sensitive stuff in it.
Ok, this can be done and (mostly) automated if you have your network set up properly. It requires a bit of knowledge with batch scripts (that I don't have), or just write a little program to do it for you. Requirements: windows file sharing enabled (and that you know the computer names or IP addresses. Can run a sniffer for these), desktops running an NT based operating system, preferrably all the same, and hopefully 2k+.
Look for something called 'pstools', it is half of what you need. the others are available in other places, but the primary ones are also available in the 'fastpush' package to put VNC on computers remotely. You don't need to do that, but it can also help from a support point of view.
the pstools let you do a lot of things remotely. The other ones that come with fastpush (Specifically regini) also help with this.
From there, it's a simple matter of making a bat file to copy over all the files you need, and shoving something in the 'run once' registry key, then doing a reboot on them (regini for reg editing, psshutdown for rebooting them). Obviously best to do this if no one's on the computer (or just send them a message telling them to reboot their computer if there's a risk of this. Or hell, use VNC and do it the hard way). Then loop through all your computer names.
I'm assuming you're competent enough to know about the $ shares, net use, some batch scripting, etc., and foolish enough to think this'll work well.:) Automated upgrades of this magnitude seem to break a lot of things. Do a lot of testing on your scripts first!
This way you can copy the files over to each computer automatically, at least. By putting it in to the run once, they'll run without you having to go to each computer. Hopefully you're smart enough to set up install scripts for IE and stuff, so you won't have to do that all too.
Windows remote administration isn't all THAT difficult, provided the right tools..
Of course, you'll want to test this before you do it to the entire corporation. Breaking everything remotely will make it so you have to go to each one, one by one, and getting shit from everywhere else about hwo they can't do anything.
Or, take the lazy route and do as the parent suggests. It's less work for you, and will get most every computer in the place, since people are stupid.;)
As an aside: Does anyone know equivalents of these tools for linux? (psexec, regini, etc.).. they'd make it so much easier for me, not having to reboot all the time in to windows to use them. Yes, VMWare, etc. etc. Not an option.
Yes, that's why Microsoft goes and reinvents the interface for everythign all the time. First it was ReallyLongFunctionName, then along came ReallyLongFunctionNameEx. Then they decided that subsystem was too limited, and made a new one and hyped that. Then tried to deny that that was ever a good idea (After they introduced a new one).
There's a reason why Longhorn is getting rid of such a large number (what was it, like 75%) of its function calls. They're old, not needed (Except for legacy support), and introduce more places for bugs.
Open Source does the same. Wait, everyone does. It's just that [Open Source Product] doesn't have the market clout to be able to FORCE manufacturers to code for the new platform/API. (Though, being Open Source, most programs can be changed by other people sufficiently motivated by a need for that program. Which is why some people slam nvidia for releasing binary drivers: when either X or the kernel change their driver interfaces, they have to wait for new drivers, as opposed to just fixing them to work with the new interface on their own.)
Slashdot Mirror
yeah, well good luck with that. There's ways to get around tables. They're 3x as long, and not as nice. I wrote 5 different themes for my website (used to be xhtml generated from xml through xsl, but then my x key broke so I switched it to regular html output. That sentence is a lie. My xsl parser when I was outputting xml (which is what xhtml is) would shorten things like the textarea tag to , which just screws up some browsers.)
Anyway, 5 different themes, researching table-less design, etc. You know what? I had to put divs all over the place instead of tables. Not all too much cleaner, but I guess better 'logically', but here's the big problem: No resizing.
Example 1: I want to make a form. Forms look nice if the labels are all lined up and all the form elements are lined up. Let's assume I want a page that can have varying size text (which is what people say should be allowed with css).. Ah fuck, can't use divs here. Oh well. I tried. From what I've found, there's no way to link the size of one div to the size of the others in a way that makes logical sense to the document format and flow. I'll welcome any suggestions proving me wrong.
Example 2: Multi-column layouts are annoying as hell. Yeah, they're annoying in tables too, but damn. That's why there's entire web pages devoted to getting away from tables with weird, complicated CSS to do half of the same thing. (This isn't as bad with CSS2 support, min and max width/height are nice, but not supported by all browsers I have to use).
Example 3: Picture page. Captions on the bottom of thumbnails. Again, let's go for optimal use of screen space: float them all, so that they wrap around when there's no horizontal space left. Ah fuck, one of the captions is a line longer than the others! There go all the floats in the next line. Yes, this can't even come close to being done with tables, but it's something that pisses me off anyway. Unless I know exactly which font the user is using, and at exactly which size, I can't make the boxes for each picture to be floated the right height. There's no way for other boxes on the same line to inherit the height of the largest on that line, if they're all floated. This isn't necessarily in the same topic, but wow is it annoying as hell.
Honestly, I don't know why the hell we're taking so long to get AWAY from HTML. HTML as originally designed was a way to structure content logically. Then along came graphical browsers, and it turned in to a presentation language. Then along came the W3C, who try and force it back to a structural language. But that's not what people want! XSL:FO are too verbose, imho, but are much nicer for what I want to do. (Bleh, just cuz I like examples: Give the damned web browser a clue as to what a slashbox is. Don't give it a hierarchy of divs. Don't give it a table. Give it a fucking . Tell it using some other method HOW to render a slashbox. CSS is nice for saying how certain things are rendered, but only fixed attribtues of those certain things. Give me wide support for CREATING my own things, and saying how THOSE are rendered/produced/built.)
Some might argue that that's not a good idea, but holy shit I don't care! Limiting me to such small pieces of crap as divs and spans to build a decent webpage is retarded. Especially wen you want to get fancy and have them look nice (corner/edge graphics on your elements. Imagine slashboxes with curved corners. and a shadow. Oh crap, now half the html document is outputing those slashboxes, and our page is full of presentation again, not structure. You can't get away from presentation, so work with the people to minimize the harm.. CSS won't do it! (again: CSS2 isn't that bad. before and after do help somewhat, btu don't solve it completely)
That's why I use XSL: I dynamically make a page. Tell it what the different parts of the page have to say. the XSL then converts that to HTML, sometimes in a two step process (depends on theme). Step 1: Make everything in
Ok, I was a bit harsh because I'm having problems with egress filtering in a school situation: I live there, and they prevent me from doing my school work and other things by their stupid rules. Places of employment I agree with you a bit more. My fault for saying that it's bad at all times, there are certainly times when it's ok..
I still say egress filtering is a nuicance to people who know what they're doing, but I guess it is a necessary evil against people who think they know what they're doing, and just fark stuff up.
I hope you aren't advocating only allowing certain egress ports, because that right there is the cause of so much headaches for users it's not even funny. Block stupid crap like hotbar, gator, etc. etc.. but PLEASE do NOT make me have to stop my work to go bug the tech person to bug his superior, to bug their superiors to open an egress port. I have to deal with that at my school (work situations are slightly different, but still rather annoying). I have to basically tunnel everything that isn't ftp access on the standard port, telnet on the standard port, or ssh on the standard port. Web access must all go through the proxy (which is completely borked), making working on webpages a pain in the butt.
Egress port blocking = annoying as f*ck. And there's no reason for it that I can tell (unless you're completely anal that only certain programs be allowed, etc. In that case, shouldn't you also have a specific allow list of websites able to be viewed? Even then, it's rather annoying. IM services, etc are very commonly used in work places, and it helps a great deal to be able to IM the person in the next cubicle, or on the other side of the building, than call them up. Less disruptive to thought processes too.)
I'm not quite sure, but maybe if you already have a decent firewire network setup, you can add one as a gateway relatively easily, and therefore not rewire everything for ethernet. Admittedly, this seems rather stupid. However, if everything is setup properly, I believe with firewire you can buy a hub, and share (For example) a digital camera with any mac also connected to the hub. If you have a setup like this, then adding networking to it would be as simple as a configuration change on your gateway, be it connected via modem, ethernet, wireless, etc.
But yeah, especially since most macs don't even need crossover cables for the ethernet, I dunno what the big deal is.
But, but.. the yahoo article linked from slashdot said that it was 2-3weeks afterwards. Also, there's no mention on the IMAX Si..
Ok, what the hell. There WASN'T information on their site the last time I checked. Now there is. Nothing about a release date, besides 'June 2003', which isn't the normal Matrix release day (since that's 5/15, which is a Thursday according to my calendar)
What's the IMAX opening night? I thought it was two-three weeks after the regular opening night? I'm currently out of the country, but if it's 2-3wk delayed for IMAX, that's about when I'm getting home..
Hey yeah, something like this happened in the hitchhiker's guide..
Ignoring that, have three completely separate monitoring programs? Don't use the same code base, and therefore you don't run in to software problems, and minimize the effect of hardware problems (since the software would probably be interacting with the hardware in different ways, if there's a div bug in the CPU, the errors won't necessarily be the same between all three SEPARATE software programs). If they all monitor the same thing and produce the same output (or at least, output that's understandable to the other two), the likelihood of two messing up at the same time, and producing the same wrong output is rather small.
However, it's probably possible to knock two out so they produce DIFFERENT output, so you have one valid source, and two screwed up ones, but screwed up in different ways.. Then you basically have to reboot and hope it fixes them.
Would this work, or am I missing something here? Again, not perfect reliability, but it does tend to make it so that there's a bit of safety from software bugs..
yeah, I can never remember which is length and which is specific.. and I knew the no was superfluous when I re-read it, the parenthesis had screwed me up. *sigh*
:)
Too bad I can't get off with such simple sentences in my tests. I have two finals next week, and I'm probably going to fail both. Fun
That question being posed, and assuming the answer to be you want to work on hardware: Good luck getting in to the center of a large Japanese corporation if you're a foreigner. I'm not saying it's impossible, but highly unlikely. Especially if you don't have a perfect command of the language. Foreigners over here are useful for a few things: Teaching drunken businessmen english, staring at, making fun of when they think we can't understand them, and scaring little children.
:P
The best part is when they're talking about you on the train, sitting right next to you, thinking you can't understand what they're saying.. then you say something in japanese to your friend sitting next to you and they shut up for the rest of the train ride. *grin*
I hate to admit it, but you might have a better chance getting in to Microsoft to work on X-Box. Especially if you're doing an internship.. but good luck even for that. I have no personal experience, but I doubt something like X-Box hardware design would use interns much at all. At least there (if you can get in), you get some experience in your native language, and then maybe make the switch to a Japanese company.
Anyway, nan nen kan nihongo benkyou shita? ore.. san nen kan, demo, zenzen benyou shinakatta kara, amari jouzu ja arahen-n ya.. ^_^ Ima, kansai gaikokugo daigaku no gakusei, demo roku gatsu (getsu? I can never remember) no futsuka ni amerika ni kaeru to omoimasu..
Wow romaji is ugly as hell. Especially when you mix in kansai-ben. Bleh.
Hopefully the original asker will answer this question, but from what I read, it seems like he's more towards the hardware end of things (working on designing the consoles) than on the software (working on games for them). There's some crossover, but he's in computer engineering: hardware with a bit of software. Nowhere does he mention programming.
So, is it really wanting to be a programmer, or is it wanting to be a designer of the hardware?
Maybe I'm just not understanding completely, but how does that simulate symlinks in linux? It simulates mount points, which are nice, but symlinks are a bit different? Let's say I want C:\Documents and Settings\Spectral\Desktop\Blah to point to D:\Stuff\Blah (A Directory), how can I do that in windows? Sure, I can mount D:\ inside C's directory structure, but how do I get a link that at the FS level makes Blah look like a folder to everything? Or even on the same partition, it doesn't matter.. Hard links wouldn't work here, since they act differently. (I want to be able to delete one, then put another file there with the same name and have the link work. AFAIK, hard links don't work that way. Also, (again, afaik) hard links won't work across partitions/filesystems. I've never used a hard link though, so I might be wrong.)
This is certainly the unix way I guess. There shouldn't need to be a special packaging command to help me find the files though. This filesystem still makes more sense to me. If all things install to their own separate directory tree, then symlink them so everything also appears it's in one spot (like /Configurations .. or just call it /etc since it's shorter ;)), we have the best of both worlds. I shouldn't have to depend on some package manager tracking every file that a program needs to run, especially if it's made with scripts afterwards. MacOSX has it right here: Most things are just a 'Package'. A Package is a compressed folder/disk image, and is treated like one by the OS.
/MacOSXAppsDirectory/CompanyName/Program.Package/e tc will work. Of course those aren't real directory names, but you get the idea.)
Therefore double clicking it will run the program, but you can easily go right in side of it and see all the files, and treat them like files. This works on the command line too. (cd
There are similar commands for any packager because there needs to be. There's also a command for sorcerer that finds files it's not tracking. When wanting to COMPLETELY remove something, I have to check that list as well. And then hope that IT is complete. Being able to check a directory for a "data" folder, back that up if I want, then blow out the directory would be nice. (Yes, this does screw up symlinks. Therefore it MIGHT be better to have the directory for the program contain the symlinks, as opposed to scattering the symlinks in to the one solid directory. Unless there's a way to reversely traverse a symlink in constant time..)
Again, a pipe dream I'm sure, and I'll admit there are certain things about the linux/unix file system that are nice. Configs mostly in one place, etc. But yikes it's certainly a mess. Something like this would help a great deal, I think.
(Another example: at a konsole, hit k, then hit tab. How many things come up? How many of those are kde programs? Are those ALL the kde programs? Probably not. What if you want to see all the executables that are part of the 'kde distribution' ? I guess you're off checking your package manager: Make a list of what you consider the kde distribution, run that list through the package manager, dump that in to a tiny little thing that sees if they're executable, etc.
Not too much different than just using ls/find/grep/bash/whatever, but what if your package DB gets corrupt? If bash/ext2 gets corrupt you have a bit more to worry about I'd think.)
FS/OS support for links makes it so easy to do such cool stuff that's essentially impossible in some other operating systems (Shortcuts are files that are treated specially by the shell in Windows. Not by the OS's FS layer. Therefore, they're nowheres comparable.) Why don't we use that support to make a FS structure that makes sense to everyone, and kicks ass? You can keep the old layout, and have a nice new layout too. Best of both worlds.
This is, actually, C:\Documents and Settings\user
/etc directory I'm sure. Is it in it's own subdirectory? Possibly! Let's go and see.
/Programs/XFree86/4.3 .. oh look, everything X installed.) Yeah, that could get confusing. Therefore there might be a bin directory, a config directory, and a data directory. They can all be symlinks, I don't care, but if I had to come up with where KDE stores it's default menu, I would have no f*cking clue. Somewhere in /usr I guess? Might depend on the distro.. Agh.
Where the hell are you people getting this BS about it being in the windows directory?! It wasn't there in w2k. I should know, I'm running it right now. It's not there in XP (Professional). I should know, the computer next to me is running it.
That being said, the linux file system structure SUCKS! Windows isn't much better, but christ.. especially with the distros. Where is your config file for samba? Well, I don't quite know. It's somewhere in the
Having all the stuff AT LEAST symlinked from some common directory would be SO NICE. (cd
XP includes Zip-As-Folder support. an annoyingly limited subset of functionality is provided for it compared to a real folder though, which screws up the metaphor. Anything that looks like an Explorer window better f*cking act like one. Agh.
I meant remote via software/internet. I knew about grabbing RF from monitors, didn't think keyboards were that bad (less power, I thought it'd have been slightly more secure than the massive CRTs :)). I should have been clearer. Though you bring up interesting topic: the PGP thing, how does that work? is there a screen shot to see what they did to make it harder for RF receivers?
that's why the 3d accelerated one would work. the secure windows are textures that are marked as secure on the video card. grab a screen capture, and it doesn't grab the secure textures. Easier than invaldiating areas of the screen that might not necessarily be rectangular (think overlapping windows, secure on the bottom, you have to chop out at least two rectangles of 'secret' area. If the video card knows that the surface indicating the window is private, when grabbing framebuffer, it won't show that surface. Doesn't matter if zbuffer killed part of it, the rest is invalidated automatically.)
Or TEMPEST style radiation detection, etc. Yeah, there are ways to do it remotely, but I don't think (though I don't know for sure) that they're nearly as easy as a direct hardware tap would be.
Though I meant 'remotely' as in "Someone on the other side of the world through the Internet" type of remotely. They can't do that I'm almost certain. Lock down any remote threats, and assume people are smart enough to lock down physical local threats, if it really matters that much what they're doing.
However, your comment on encrypting the data sent between the card and monitor is NOT new. There were several talks of this before, I don't know where they went.. but a next gen connection that was encrypted and digital. (Or at least, certain parts would be encrypted: DRM protected stuff.) Therefore, no access to the DRM material. Could be 'elevated' in use to be useful for all material that you/microsoft/content manufacturers don't want copied.
Keeping things digital/encrypted as long as possible, making it harder to get to the analog signal lead to 1) slightly improved quality, 2) Much harder to crack.
They'll take a while in replacing the video cameras, but the loss of quality/difficulty of use is rather prohibitive. Expect the MPAA/RIAA to push for brain jacks to combat the rampant piracy of screeners and people with *gasp* microphones.
I don't know. :) You would think that the video driver would be able to stop this, but maybe not? Certainly older systems with memory mapped video would make it easy to grab the contents (if it is read/write..), I don't know how it's done anymore (I assume everything must go through GDI in windows. But maybe there are ways in their driver model to grab stuff from a window/screen buffer.)
:)
I really don't know for sure, but I assumed that was the problem they were mentioning. Maybe the problem arises with DMA? Again, I'm not much in to the internals of Operating Systems and stuff, but if there's a way to link the hard drive and the video card, maybe another card on the system can do the same and link itself to the video card and grab its memory. Highly unlikely, but.. Maybe technically possible? (though, it's most likely up to the OS to initiate a DMA request, so again, it could probably be managed in software.)
Who knows. Typical linux troll response: They want to get everything under their control, to lock us out. I somewhat doubt this explanation, but it might be correct.
Not many computers would anymore, but a lot of older computers used a type of static ram. Therefore it would keep the memory without being refreshed, and could keep it for several hours, maybe days. The reasons were speed, but the cost is prohibitive for the amount of memory being used. Some memory technologies use small pockets of SRAM as a type of cache, before moving to the DRAM (D=Dynamic. Must be refreshed often, which is why it loses its contents after poweroff)
At least, I'm pretty sure that's correct, and also pretty sure that's why I've seen this effect.
The problem is that the memory in graphics cards aren't wrapped in to the security model. Therefore, anything with access to reading some memory from the video card (not too uncommon I'd imagine), can go and grab everything. Including the current screen contents.
Video cards therefore need to be modified to be secure and support access control on their memory, the way the CPU's privileged mode (and whatever hardware they're going to tack on to make palladium work 'better') allows the OS to control what can access certain parts of main memory.
USB: Handled by the OS. Easy to deal with. Monitors: would require hardware tapping, much harder to do (Especially remotely). Keyboards: Again, MOSTLY handled by the OS. (Windows passes most every key combination through hooks, except ctrl+alt+del. They'll probably change this so that if a secure window is on top, no hooks grab the data.)
Etc. etc. I don't argue that it's a bad idea (that there needs to be changes to the video card hardware to support this properly), it's just very poorly worded in the article.
I wonder if maybe the 3d support being used helps this? If you define everything as a texture, then you only need to secure certain textures (the secure ones), not the entire screen. Therefore things running unmanaged still work just fine. 2d accelerated blits might also do the trick, but probably not as well. (Do the 2d accel blits have a concept afterwards of what a window is, and therefore to hide a certain one? Probably not..)
Humans are a security problem, because they contain their own pool of memory too. Let's get rid of them. Deleting a person's memory is easier than the video card's too: One click of the trigger is all it takes. Just Point and Click.
I'd have no clue how to wipe out my video card's memory. (No, shutting off the computer won't do it. I've seen plenty that when they turn back on, the last screen visible is there for a split second.)
Probably because it's personalized, it's harder to spoof the window. Password boxes using data that only the OS knows and personalized for that computer are better. At least, if all dialog boxes looked one way, then up came a popup that looked compeltely different, it's pretty damned obvious it's a fake, and you don't want to put sensitive stuff in it.
nope. Living in Japan at the moment, going to uni. The reverse lookup will be our firewall, with a .jp address, so not me. :)
Ok, this can be done and (mostly) automated if you have your network set up properly. It requires a bit of knowledge with batch scripts (that I don't have), or just write a little program to do it for you. Requirements: windows file sharing enabled (and that you know the computer names or IP addresses. Can run a sniffer for these), desktops running an NT based operating system, preferrably all the same, and hopefully 2k+.
:) Automated upgrades of this magnitude seem to break a lot of things. Do a lot of testing on your scripts first!
;)
Look for something called 'pstools', it is half of what you need. the others are available in other places, but the primary ones are also available in the 'fastpush' package to put VNC on computers remotely. You don't need to do that, but it can also help from a support point of view.
the pstools let you do a lot of things remotely. The other ones that come with fastpush (Specifically regini) also help with this.
From there, it's a simple matter of making a bat file to copy over all the files you need, and shoving something in the 'run once' registry key, then doing a reboot on them (regini for reg editing, psshutdown for rebooting them). Obviously best to do this if no one's on the computer (or just send them a message telling them to reboot their computer if there's a risk of this. Or hell, use VNC and do it the hard way). Then loop through all your computer names.
I'm assuming you're competent enough to know about the $ shares, net use, some batch scripting, etc., and foolish enough to think this'll work well.
This way you can copy the files over to each computer automatically, at least. By putting it in to the run once, they'll run without you having to go to each computer. Hopefully you're smart enough to set up install scripts for IE and stuff, so you won't have to do that all too.
Windows remote administration isn't all THAT difficult, provided the right tools..
Of course, you'll want to test this before you do it to the entire corporation. Breaking everything remotely will make it so you have to go to each one, one by one, and getting shit from everywhere else about hwo they can't do anything.
Or, take the lazy route and do as the parent suggests. It's less work for you, and will get most every computer in the place, since people are stupid.
As an aside: Does anyone know equivalents of these tools for linux? (psexec, regini, etc.).. they'd make it so much easier for me, not having to reboot all the time in to windows to use them. Yes, VMWare, etc. etc. Not an option.
Yes, that's why Microsoft goes and reinvents the interface for everythign all the time. First it was ReallyLongFunctionName, then along came ReallyLongFunctionNameEx. Then they decided that subsystem was too limited, and made a new one and hyped that. Then tried to deny that that was ever a good idea (After they introduced a new one).
There's a reason why Longhorn is getting rid of such a large number (what was it, like 75%) of its function calls. They're old, not needed (Except for legacy support), and introduce more places for bugs.
Open Source does the same. Wait, everyone does. It's just that [Open Source Product] doesn't have the market clout to be able to FORCE manufacturers to code for the new platform/API. (Though, being Open Source, most programs can be changed by other people sufficiently motivated by a need for that program. Which is why some people slam nvidia for releasing binary drivers: when either X or the kernel change their driver interfaces, they have to wait for new drivers, as opposed to just fixing them to work with the new interface on their own.)