I've never had any trouble hot-plugging PS/2 keyboards, apart from the keyboard autorepeat being reset to a slow speed. Perhaps you're using nasty plasticky keyboards (as opposed to the One True PS/2 Keyboard, the Model M) or flaky klone motherboards (again, real PS/2s don't have a problem, so why should the clones?).
Yeah, you're right, in some cases having two half-speed CPUs is better than one full-speed because of caching. To guarantee equally good performance with all apps, you'd need to make your single processor have not only twice the clock speed, but also twice the size of cache, and make sure your OS's context switches didn't happen often enough for frequent pipeline flushes to affect performance.
Hang on a bit, you say the L1 and L2 caches have to be flushed on each context switch? That seems crazy. The caches deal with physical addresses not logical ones, right?
One database query per page is not too bad. You can make that scalable and it's certainly a lot less effort than trying to track large amounts of data _outside_ the DB.
You have a problem when a single page view takes hundreds of database queries (as happened with a certain web toolkit I used to develop on).
I know that in the server market you often go for tried-and-tested rather than latest-and-greatest, and that the Pentium III still sees some use in new servers. But 1.26GHz with PC133 SDRAM? Surely they'd have got better performance from a single 2.8GHz Northwood with Rambus or DDR memory, and it would have required less cooling and fewer moving parts. Even a single Athlon 2200+ might compare favourably in many applications.
SMP isn't a good thing in itself, as the article seemed to imply: it's what you use when there isn't a single processor available that's fast enough. One processor at full speed is almost always better than two at half the speed.
Computer hardware is so fast relative to the amount of traffic coming to almost any site that any web server is a high-performance web server, if you are just serving static pages. A website made of static pages would surely fit into a gigabyte or so of disk cache, so disk speed is largely irrelevant, and so is processor speed. All the machine needs to do is stuff data down the network pipe as fast as possible, and any box you buy can do that adequately. Maybe if you have really heavy traffic you'd need to use Tux or some other accelerated server optimized for static files.
With dynamically generated web content it's different of course. But there you will normally be fetching from a database to generate the web pages. In which case you should consult articles on speeding up database access.
In other words: an article on 'building a fast database server' or 'building a machine to run disk-intensive search scripts' I can understand. But there is really nothing special about web servers.
I'm afraid it's true. Open Spectrum beats current wireless networking technologies hands down.
The clock frequency of the Z80 processor is well-optimized for generating long- and short- range RF waves. With the Spectrum case Open, the highly advanced rubber keys act as efficient waveguides, particularly helped by the printed-on aerials (square shapes with one half or one quarter cut out, depending on the phase of the waveform).
The only snag is that due to an unfortunate typo, the Sinclair Microwave communication device ended up as the bizarre Sinclair Microdrive.
If the X server does all its root-needing stuff at startup, then surely it can drop root priveleges immediately afterwards and so a suid-root X server is not such a security risk as often seems to be implied. Of course having no root priveleges at all and relying on a systrace policy is still a step up...
Yeah I kinda meant 'something like IMAX' and not necessarily any one particular patent-encumbered standard dominated by a single monopolistic comany. Er, in other words append the standard Slashdot orthodoxy to whatever I wrote:-).
Reduced seating is a disadvantage but isn't the trend towards smaller theatres anyway? Perhaps that is to fit more screens into a multiplex rather than because they deliberately want to shrink capacity, however.
'The client software hasn't been widely distributed' - what about Samba? I've used the smbclient program to send messages to Windows PCs.
Fortunately this annoyance is easy to fix - just turn off the useless Messenger service in Windows. In fact, turn off all the Windows services except for OLE and Spooler.
Couldn't something like IMAX restore the cinema's advantage over television? Even the real early adopters don't have an IMAX screen in their living room. A pity that more films aren't made in a suitable large-screen format.
I opened loads of browser windows one time at school and the shared net connection, not being intelligent enough to allocate bandwidth per-user, pretty much came to a halt for everyone else on the network! Actually I think it was lots of small FTP downloads from Simtel in parallel.
If you can configure Moz to (a) have tabs open in the background and (b) have 'open in new tab' bound to a single mouse click then I might do that. Actually, I knew already that (b) was possible but I haven't taken the time to learn the XUL or other voodoo necessary to configure it.
After discovering Dillo I hardly use Mozilla any more. I might try Phoenix if it is a bit snappier (how does it compare to Galeon?).
The style of web browsing I use is to load all the links I want to read in new windows by clicking on them with the middle button. Then they can be loading in the background while I read the first part of the article. It forms a kind of queue of pages to read, so when I've finished reading the first page I just close that window and go on to the next (which is ready instantly). The result is up to a hundred browser windows open at once - but I know that I'm not the only person who browses like this. Of course, it helps to have a browser which can open lots of windows without thrashing and slowing the machine to a crawl (like Dillo) or one that has tabbed browsing.
This style of following links can also work well with offline browsing and a proxy server designed for offline use like WWWOFFLE. If you go online briefly and click on all the links you want to load, the proxy remembers to download them. Then a few minutes later you can go online again and all the pages will be loaded ASAP. Once they've loaded you can disconnect again and continue browsing. This makes the most sense for people whose internet access is metered (hmm, I wonder if something like this could work for palmtops).
But what I'd really like to see in a browser is an explicit 'to read' queue. When you click on a link with the middle button, it doesn't immediately open in a new window or tab but instead is added to the queue and starts downloading in the background. On the browser's toolbar there is a 'next page' button which goes to the next URL you have marked for reading.
Automatic prefetching of all links from a page, la wget -r, would be crazy for many heavily-linked sites. But you could have heuristics for it or specify particular sites where the link following should be more aggressive.
It has to be incredibly fast! How could a chip called '970' not be really quick and powerful?
Expect to see a slightly cheaper version called the 940, and a low-end chip called the 880. Well in fact there won't be any real difference between those two chips, but you pay for the extra prestige the extra 60 gives you.
Yup, I'm thinking of something a bit like systrace. But with a nicer interface.
Imagine a GUI library with 'save' and 'load' mechanisms. (Preferably, ones like those in ROX, which is the only intuitive and non-ugly way to load or save files; but I digress.) The loading and saving user interface is actually provided by a separate process, which will grant the application read or write access to the file chosen, and only that file.
In this way a lot of scripting holes could be avoided, and also nastiness with malware which sends data back to HQ or saves files in silly locations. But it's debatable whether this kind of setup is useful other than for paranoia value, since it's a much better answer to just not have executable code embedded in documents (or if it is, sandbox it in the application) and not run software for which you don't have source code.
I think I remember the 'MS VP disses Windows security' story but I don't really take much notice of what Microsoft execs claim. This filtering applies equally whether they are promoting or disparaging their own software:-).
The OS can compartmentalize resources so that if one app makes an illegal memory access, it doesn't crash the machine. The OS can limit access so that if one server is compromised, it can only screw up its own files and not the others on the box. NT does both these things (the latter with the ability to run a server as a particular user). However, no OS can do anything about deliberately stupid applications which choose to execute scripts stored in documents, for example.
Well, I suppose it would be possible to run Outlook under its own user account or with a reduced set of permissions, so that it could access only its own mail spool and not the rest of the user's files. But that would really get in the way of typical usage. Perhaps if there were some way to allow small extensions of permissions a la Java ('Outlook is trying to save a file c:\foo.doc. Do you wish to allow this?' and press Yes if it's something you asked for, No if it looks like a worm doing something nasty). But AFAIK no desktop OS has ever done anything like this; all desktop apps run with the uid of the current user and have full access to his files.
When developers make moronic decisions like auto-executing scripts in documents, it is not fair to blame the operating system. It is not so much Windows as the crap which festers around it (albeit coming from the same company). You don't hear about too many exploits in the Windows FTP server program (although surely there are some). Why not? Because FTP is a standard protocol and Microsoft haven't been able to set their monkeys loose on it and add insecure extensions.
'Piece of shit' is a bit strong to describe CVS. 'Crude but effective' is closer to the mark.
Or perhaps 'piece of shit, but one that has been lying around in the hot sun and baked so hard that it actually makes quite an effective tool for banging in nails, provided you don't mind the smell'.
Bitkeeper sounds good technically but c'mon, the stink from McVoy's licensing surely outweighs the small amount of nose-holding you have to do when using CVS. When the choice is between something technically crude (but 80% good enough) and something that does 100% of the job but has odious licensing policies, surely the difference in licences outweighs the 20% of extra functionality you're getting.
Maybe you didn't have to look at the manual to understand the GUI, but most adults who hadn't used a computer before would certainly need guidance. It's pointless to compare Blender to some mythical computer interface that can be picked up first time, because any UI has a learning curve. Even with web browsing, it can take some time for non-computer-literate adults to grasp the idea of 'if it's underlined, you can click on it'. (No, really.)
A better comparison would be to say: for those already used to computers, Blender's interface still has a steep learning curve, whereas it would be possible to give it an interface that could be picked up quickly by children or by computer-literate adults.
Slashdot Mirror
I've never had any trouble hot-plugging PS/2 keyboards, apart from the keyboard autorepeat being reset to a slow speed. Perhaps you're using nasty plasticky keyboards (as opposed to the One True PS/2 Keyboard, the Model M) or flaky klone motherboards (again, real PS/2s don't have a problem, so why should the clones?).
This chilled fog could be useful for overclocking! Imagine a Beowulf cluster of these!
Yeah, you're right, in some cases having two half-speed CPUs is better than one full-speed because of caching. To guarantee equally good performance with all apps, you'd need to make your single processor have not only twice the clock speed, but also twice the size of cache, and make sure your OS's context switches didn't happen often enough for frequent pipeline flushes to affect performance.
Hang on a bit, you say the L1 and L2 caches have to be flushed on each context switch? That seems crazy. The caches deal with physical addresses not logical ones, right?
One database query per page is not too bad. You can make that scalable and it's certainly a lot less effort than trying to track large amounts of data _outside_ the DB.
You have a problem when a single page view takes hundreds of database queries (as happened with a certain web toolkit I used to develop on).
I know that in the server market you often go for tried-and-tested rather than latest-and-greatest, and that the Pentium III still sees some use in new servers. But 1.26GHz with PC133 SDRAM? Surely they'd have got better performance from a single 2.8GHz Northwood with Rambus or DDR memory, and it would have required less cooling and fewer moving parts. Even a single Athlon 2200+ might compare favourably in many applications.
SMP isn't a good thing in itself, as the article seemed to imply: it's what you use when there isn't a single processor available that's fast enough. One processor at full speed is almost always better than two at half the speed.
Computer hardware is so fast relative to the amount of traffic coming to almost any site that any web server is a high-performance web server, if you are just serving static pages. A website made of static pages would surely fit into a gigabyte or so of disk cache, so disk speed is largely irrelevant, and so is processor speed. All the machine needs to do is stuff data down the network pipe as fast as possible, and any box you buy can do that adequately. Maybe if you have really heavy traffic you'd need to use Tux or some other accelerated server optimized for static files.
With dynamically generated web content it's different of course. But there you will normally be fetching from a database to generate the web pages. In which case you should consult articles on speeding up database access.
In other words: an article on 'building a fast database server' or 'building a machine to run disk-intensive search scripts' I can understand. But there is really nothing special about web servers.
I'm afraid it's true. Open Spectrum beats current wireless networking technologies hands down.
The clock frequency of the Z80 processor is well-optimized for generating long- and short- range RF waves. With the Spectrum case Open, the highly advanced rubber keys act as efficient waveguides, particularly helped by the printed-on aerials (square shapes with one half or one quarter cut out, depending on the phase of the waveform).
The only snag is that due to an unfortunate typo, the Sinclair Microwave communication device ended up as the bizarre Sinclair Microdrive.
The punishment should be that she really is forced to switch from the Mac to Windows XP.
If the X server does all its root-needing stuff at startup, then surely it can drop root priveleges immediately afterwards and so a suid-root X server is not such a security risk as often seems to be implied. Of course having no root priveleges at all and relying on a systrace policy is still a step up...
Yeah I kinda meant 'something like IMAX' and not necessarily any one particular patent-encumbered standard dominated by a single monopolistic comany. Er, in other words append the standard Slashdot orthodoxy to whatever I wrote :-).
Reduced seating is a disadvantage but isn't the trend towards smaller theatres anyway? Perhaps that is to fit more screens into a multiplex rather than because they deliberately want to shrink capacity, however.
'The client software hasn't been widely distributed' - what about Samba? I've used the smbclient program to send messages to Windows PCs.
Fortunately this annoyance is easy to fix - just turn off the useless Messenger service in Windows. In fact, turn off all the Windows services except for OLE and Spooler.
It doesn't matter, people who don't like Apple's attitude can just switch to alternative suppliers of Macintoshes and buy OS X from someone else.
Couldn't something like IMAX restore the cinema's advantage over television? Even the real early adopters don't have an IMAX screen in their living room. A pity that more films aren't made in a suitable large-screen format.
I opened loads of browser windows one time at school and the shared net connection, not being intelligent enough to allocate bandwidth per-user, pretty much came to a halt for everyone else on the network! Actually I think it was lots of small FTP downloads from Simtel in parallel.
If you can configure Moz to (a) have tabs open in the background and (b) have 'open in new tab' bound to a single mouse click then I might do that. Actually, I knew already that (b) was possible but I haven't taken the time to learn the XUL or other voodoo necessary to configure it.
After discovering Dillo I hardly use Mozilla any more. I might try Phoenix if it is a bit snappier (how does it compare to Galeon?).
The style of web browsing I use is to load all the links I want to read in new windows by clicking on them with the middle button. Then they can be loading in the background while I read the first part of the article. It forms a kind of queue of pages to read, so when I've finished reading the first page I just close that window and go on to the next (which is ready instantly). The result is up to a hundred browser windows open at once - but I know that I'm not the only person who browses like this. Of course, it helps to have a browser which can open lots of windows without thrashing and slowing the machine to a crawl (like Dillo) or one that has tabbed browsing.
This style of following links can also work well with offline browsing and a proxy server designed for offline use like WWWOFFLE. If you go online briefly and click on all the links you want to load, the proxy remembers to download them. Then a few minutes later you can go online again and all the pages will be loaded ASAP. Once they've loaded you can disconnect again and continue browsing. This makes the most sense for people whose internet access is metered (hmm, I wonder if something like this could work for palmtops).
But what I'd really like to see in a browser is an explicit 'to read' queue. When you click on a link with the middle button, it doesn't immediately open in a new window or tab but instead is added to the queue and starts downloading in the background. On the browser's toolbar there is a 'next page' button which goes to the next URL you have marked for reading.
Automatic prefetching of all links from a page, la wget -r, would be crazy for many heavily-linked sites. But you could have heuristics for it or specify particular sites where the link following should be more aggressive.
They're not porting it to XENIX? This is an outrage!
It has to be incredibly fast! How could a chip called '970' not be really quick and powerful?
Expect to see a slightly cheaper version called the 940, and a low-end chip called the 880. Well in fact there won't be any real difference between those two chips, but you pay for the extra prestige the extra 60 gives you.
Yup, I'm thinking of something a bit like systrace. But with a nicer interface.
Imagine a GUI library with 'save' and 'load' mechanisms. (Preferably, ones like those in ROX, which is the only intuitive and non-ugly way to load or save files; but I digress.) The loading and saving user interface is actually provided by a separate process, which will grant the application read or write access to the file chosen, and only that file.
In this way a lot of scripting holes could be avoided, and also nastiness with malware which sends data back to HQ or saves files in silly locations. But it's debatable whether this kind of setup is useful other than for paranoia value, since it's a much better answer to just not have executable code embedded in documents (or if it is, sandbox it in the application) and not run software for which you don't have source code.
I think I remember the 'MS VP disses Windows security' story but I don't really take much notice of what Microsoft execs claim. This filtering applies equally whether they are promoting or disparaging their own software :-).
The OS can compartmentalize resources so that if one app makes an illegal memory access, it doesn't crash the machine. The OS can limit access so that if one server is compromised, it can only screw up its own files and not the others on the box. NT does both these things (the latter with the ability to run a server as a particular user). However, no OS can do anything about deliberately stupid applications which choose to execute scripts stored in documents, for example.
Well, I suppose it would be possible to run Outlook under its own user account or with a reduced set of permissions, so that it could access only its own mail spool and not the rest of the user's files. But that would really get in the way of typical usage. Perhaps if there were some way to allow small extensions of permissions a la Java ('Outlook is trying to save a file c:\foo.doc. Do you wish to allow this?' and press Yes if it's something you asked for, No if it looks like a worm doing something nasty). But AFAIK no desktop OS has ever done anything like this; all desktop apps run with the uid of the current user and have full access to his files.
When developers make moronic decisions like auto-executing scripts in documents, it is not fair to blame the operating system. It is not so much Windows as the crap which festers around it (albeit coming from the same company). You don't hear about too many exploits in the Windows FTP server program (although surely there are some). Why not? Because FTP is a standard protocol and Microsoft haven't been able to set their monkeys loose on it and add insecure extensions.
'Structurally more secure'? What, with a single root account and no ACLs or capabilities?
NT by *design* is much more secure than Unix, it's just the implementation and the apps (IIS, IE, Outlook, Office) which are royally screwed up.
'Piece of shit' is a bit strong to describe CVS. 'Crude but effective' is closer to the mark.
Or perhaps 'piece of shit, but one that has been lying around in the hot sun and baked so hard that it actually makes quite an effective tool for banging in nails, provided you don't mind the smell'.
Bitkeeper sounds good technically but c'mon, the stink from McVoy's licensing surely outweighs the small amount of nose-holding you have to do when using CVS. When the choice is between something technically crude (but 80% good enough) and something that does 100% of the job but has odious licensing policies, surely the difference in licences outweighs the 20% of extra functionality you're getting.
Maybe you didn't have to look at the manual to understand the GUI, but most adults who hadn't used a computer before would certainly need guidance. It's pointless to compare Blender to some mythical computer interface that can be picked up first time, because any UI has a learning curve. Even with web browsing, it can take some time for non-computer-literate adults to grasp the idea of 'if it's underlined, you can click on it'. (No, really.)
A better comparison would be to say: for those already used to computers, Blender's interface still has a steep learning curve, whereas it would be possible to give it an interface that could be picked up quickly by children or by computer-literate adults.
Please name a GUI that doesn't require people to get used to it.