3551
General Software Discussion / Re: Start Panicking! ... and polishing those tin foil hats.
« Last post by f0dder on April 26, 2009, 06:48 AM »Word, Ehtyar.
Recent Posts
I'm still wondering where all the 386s went. It's like the Gremlins scarfed 'em all up and hid 'em in a landfill before people found out you could run Linux on 'em. You'd think you should be able to buy one for $20 or something. Is the scrap value really more than a running PC? Strange.Does linux still support 386?-MilesAhead (April 25, 2009, 06:02 PM)
I'd have opted to simply call it an MRU list, but that's just me. And after all, the name doesn't detract from it's usefulnessYou should probably change the name - when seeing cache, I half expected some file/disk caching stuff rather than "just" a list of used foldersWhen I see Disk Cache I expect a disk cache. When I see folder cache, I expect a folder cache. Sorry to disappoint.-f0dder (April 23, 2009, 06:15 PM)-MilesAhead (April 23, 2009, 07:32 PM)
| Memory consumption | Run an OS on a machine with too little memory, and it's going to be dog slow no matter which OS it is - linux included. Vista might require more memory than XP (although a lot of people don't understand filesystem caching and shout their mouths off wrongly). I haven't used it on hardware with less than 2GB of RAM, but from watching my own usage stats I'd expect it to run smoothly with 1GB except when playing recent games. |
| Disk overhead | A lot of people bitch at Vista's disk indexer - but that can be turned off. Other than that, I'd say that Vista feels smoother than XP, probably because of the more aggressive prefetcher and cache system. Might also have something to do with Vista being able to I/O requests larger than 64KB... and I/O prioritization is a nice feature as well. |
| GUI overhead | On able hardware, the Vista GUI is a lot smoother than XP. It's probably a bit unfortunate that MS decided to implement the acceleration purely with shaders, though, since that leaves some otherwise capable fixed-function hardware without acceleration. |
(oh, how I wish somebody would make a NT version of "vramdir"...)Maybe I am missing something, but can't this be accomplished by just editing the permissions to prevent deletion?Beat me to it-steeladept (April 23, 2009, 03:41 AM)
- simply set NTFS permissions and you're good to go. Notice that you'll have to find the "advanced permissions" to see the Delete token.OK, but in practice I find editors tend not to enable access to STDIN etc. to user tools.I can't recall seeing STDIN support (and never needed it), but the editors I've used that supports user tools have been able to capture STDOUT.
Good point: I grew up with persistent selections. From your last sentence, I presume those features must be present in NPP, so I suppose I'd better go and look for themNope, I've gotten used to living without them
- I'm contemplating coding a "mark-begin, mark-end" plugin though ).Re. using a flash card for you page file - bad idea. It will be slower than a fast SATA drive (by a large margin) because it effectively uses a USB bus and you will wear out your flash card rapidly.Won't use USB if you use an IDE->CF adapter, but it's still going to be dog slow-Carol Haynes (April 22, 2009, 07:02 PM)
- dunno about the amount of write/erase cycles - depends on media quality, wear-leveling algorithm and pagefile activity. I wouldn't do it myself, though.12GB ha! XP (32bit) can only use 3GB of Ram and most of the laptops I have seen with Vista pre-loaded only come with 3GB as well.32bit client Windows OSes can use 4GB physical addresses - because of memory mapped devices, this usually ends up at max ~3.5GB, though, depending on BIOS, chipset and the cards you've installed in your computer. Prior to SP1, XP supported 4GB of physical RAM, regardless of whether regions were above the 4GB physical mark - this has been possible ever since the Pentium Pro. The change was because of "driver incompatibilities" (sloppy driver coders using LowPart instead of QuadPart of the PHYSICALADDRESS datatype, I bet).-SchoolDaGeek (April 22, 2009, 01:31 PM)
AIUI there are few performance gains moving from 32 to 64 bit (in some cases it appears they are slower than 32 bit - which is ludicrous) the only advantage of 64 bit is being able to add more memory and if you have 4Gb you will probably get to use 3.5Gb of it (depending on your graphics card) so to me it isn't really worth the effort.Why is that ludicruous? Pointers double in size, and the native integer size does as well. Code size increases a bit... all that means higher CPU cache pressure. Don't port your applications to 64bit just for the heck of it-Carol Haynes (April 21, 2009, 04:30 AM)
Having said that Windows 7 should really be called Vista SP3 so it shouldn't be too much of a problem from the off in terms of compatibility.IMHO it falls somewhere between a service pack and a "full new OS". There's enough kernel changes that simply calling it a SP is unfair, and there's enough UI changes as well. It's not as much of a change as Vista was from XP (UAC and all that), though.-Carol Haynes (April 21, 2009, 04:30 AM)
I just bought a laptop with 32 bit OS. Does that mean I will be unable to run windows 7 on it when it comes out? Will a 64 bit OS run on a machine with 32 bit? Is it the chip that makes it a 32/64 bit machine or the programs?AFAIK, Win7 will still come in a 32bit version.-Kimitwo (April 21, 2009, 02:03 AM)
And as far as the os being 64-bit, how big are those letters you type, how many equations do you have in your spreadsheet, how much imperceptibly smoother does that movie have to play, and are you really so good at that fps that you have to have lightening fast graphics to make those incredibly subtle moves that keep you in the top 10 of players?Moving to 64bit doesn't give much speed advantage in the general case, and can even cause slowdowns if developers are careless about porting their code. But memory demands are growing, and to access lots of memory 32bit just isn't enough. Games are getting closer and closer to the edge, and it's not just about bad programming - datasets are becoming huge.-DDRAMbo (April 21, 2009, 12:57 AM)
I fail to see which advances USB flashdrives would benefit from (the next several years, anyway), considering they're limited by slow flash memory speeds rather than anything that has to do with the interface?(Talking about USB3...)
Is that really going to be much of an issue? USB2 is 480mbit, and enve though it's hard reaching the full 60mbyte/s sustained through USB, isn't that plenty bandwidth for pendrives? How fast are the fastest now, anyway?-f0dder (April 20, 2009, 06:11 PM)
It depends on the ultimate published spec. If it is purely a faster interface and that is all, then you are right. However, if there are other technical advances, these MAY make a difference. I do know either way it will be designed for backward compatibility, so if the advances are worth it, waiting for the USB3 spec designs may allow you to have that capability where available and the USB2 or USB 1.1 specs where it isn't. Just like with a USB2 device now.-steeladept (April 20, 2009, 10:58 PM)