The real use of learning assembly is to validate your compiler output. It's rare to want to code much from the ground up in assembly anymore, save hardware-banging code, but knowing assembly still makes all the difference.
Yes, most compilers will perform 1,001 optimizations, but the final bit of optimization you can eck out demands that you profile, examine the hot points, and replace them with assembly or restructure your C with a good idea of the code you really want it to generate.
You'll still find things that the compiler engineers haven't thought of, or were lazy about implementing, or operations which implement unneccessary logic/precision.
For example, sometimes it's best to hoist an expensive calculation -into- a loop because it fits where the processor otherwise stalls, but most compilers take the opposite approach. It's tough to force a C compiler to do this, because most want to hoist code -out- of loops when possible, assuming it's cheapest.
Dependency on non-native signedness, float format or type sizes can mean that shifts, transcendental math, type conversion, etc is being done in software when a single opcode is what you expected.
Understanding the code will tell you if something unusual is happening, such as non-native sized bools or less than optimal variables being turned into register variables.
Some operations are simply not available to the C-only programmer. Subsequent lookups to sin/cos operations can be combined into a single opcode, but I've yet to see it done by any compiler. It's only available in assembly.
The PowerPC offers a rough inverse/division approximation which is faster than a real divide, and is good enough for operations only requiring low precision or an approximation... also completely inaccessible to C code.
Knowing a variable is guaranteed to be within a certain range or of a fixed set of values at a certain point can let you get away with all kinds of murderous assumptions which it's impossible to express in C.
That said, I haven't seen a single PPC assembler reference that was half as good as just looking at the code. Look at code, look at a lot of code, and past that, just look at system implementors' documentation. x86 through Pentium III aside, most current assembler books are just fluff, are wrong about half the pipelining, omit a million useful optimzations and don't cover the real story at all. It's really gotten to the point where the only real way to learn is by doing and doing and doing.
I find something worth having Flash for maybe once every other month or so. Certainly not enough to compensate for the rest of the crap that's dumped on me.
Just kill your extensions. If a site needs fancy songs and zooming images to keep your interest, what it's trying to tell you probably isn't very interesting.
If you really can't bear to live without Flash/Java/etc from time to time, set up a second browser, and use the clean one as your default.
If a company is advertising [...] in a banner ad on a web page I like [...] they're starting out with an even trade. A little support for my interests in exchange for a little bit of my attention.
Well said. And shame on those who favour software that blocks banner ads, and expect their favourite web sites to provide them a service without any revenue
I'll meet you half way. I still block out banner ads from companies that request cookies with the ad. Unfortunately, that's most of them these days. I'll allow unobtrusive ads supporting a site I like, except when they engage in secretive tracking of my browsing behaviors.
An advertiser keeping tabs on me without first asking me if that's okay steps way over the line of trust.
I've been there. I've seen the complaining and bitching, but the fact remains that the people who repeatedly go back and get familiar enough with the system to participate in polls, etc are largely DSL enthusiasts.
No. I wouldn't pay a penny per page so long as similar material was out there for free. And I would go out of my way to look for free material under a penny-a-page payment model.
Call me a pessimist, but my belief is that businesses are incapable of handling this kind of thing responsibly. The moment we go to penny-per-page, we'll start to see things artificially segmented across a dozen pages, and all kind of fluff and noise between the front page and any useful pages.
Make it a penny/nickel/dime a day for access to a whole domain, depending on the quantity and nature of the content within, and I might be interested.
dslreports was running this as a story as well, basically most of their users came to the conclusion that the general populous would "sell their grandmas" before returning to a modem. Non-techies don't want to wait for their information, this is the only thing that brought them to the 'net. at least I _hope_ it wasn't for the ads...
People who go to a site to talk about how exciting DSL is come to the conclusion that DSL is better than dialup?
IE 6 allows you to turn off sounds in web pages, kill dancing baloney, manage trust relationships with websites & control cookies with fine granularity.
This doesn't stop Shockwave, etc from making sound. The audio settings only affect the standard html and javascript-controlled sound. You'd have to remove all your plugins to stop the sounds in IE.
Of course, this will only have an effect if a large number of people act in a relatively concerted way. I have frequently thought that there ought to be some sort of centralized database-based "blacklist" of which companies use various irresponsible marketing techniques.
It's already had an effect. The kinds of places that spam are the kinds of places that don't give a fuck about anything but growing the business. By avoiding spammers, I'm increasing the probability that I'm dealing with a company that has some measure of a concern about the consumer.
The rest will sort itself out. The worse spam gets, the more likely it is to get legislated into a black hole, or the more likely it is that enough ISPs will start deploying cooperative spam circumvention devices to make it effective.
It'll get there, through law or common sense. In this day and age, the problem just has to become expensive to big business first.
Sometimes, you get something through the snail mail that actually interests you. Occasionally those carefully targetted advertizing campaigns actually hit the mark.
Now I am not trying to apologize for irresponsible spammers, but you have to admit, some spam is actually interesting. I am talking about the mailings from corporations whose product you have bought in the past, and who you might be interested in dealing with again.
I actively avoid supporting marketers who annoy me, and I urge you to do the same. If a company is advertising in a magazine I like, in a banner ad on a web page I like, or on a TV show I like, they're starting out with an even trade. A little support for my interests in exchange for a little bit of my attention.
If they spam, throw junk mail at me, pop up annoying windows all over my desktop, stick business cards or flyers on my windshield or otherwise inconvenience me, I ensure that I never ever support them. I've stopped dealing with companies I previously liked because they started engaging in these practices.
Spamming is essentially stealing. I don't know why the hell I'd want to support someone whose first action was to take from me without giving anything in return.
These new processors actually do consume heat as they operate, turning it into valuable CPU cycles. These processors require the use of a whole new CPU packaging technology that pumps heat into, rather than out of, the CPU core. Initial tests in laptop configurations have proven uncomfortable to use, due to the fact that the laptop begins to condense water out of the air, and eventually frost over as it runs.
Sweet! So a dual processor system with two of these and an aluminum bar connecting them could do away with fans entirely?
Starting out with a great product, and slowly trimming features and service, and cutting out support for the fringe customers who risk needing unique attention or service, is pretty standard these days.
Service doesn't factor into the equation the way it used to -- People buy pretty much exclusively on price or market presence.
To satisfy either, you've got to spend every available dollar on advertising until your competitors are gone, or you need to operate with little profit (or at a loss) to start with.
Getting a foothold in big business today is pretty much a starving contest. <overgeneralization type=slight>Quality products and customer service only work when you're pitching to old folks and the rich.</overgeneralization>
Some vendors are offering free images of DR-DOS on their sites for those who need to do a flash.
This is all very grand, except that floppy drives aren't guaranteed to exist in new systems. If memory serves, the 2002 computer spec put together by Intel and Microsoft actually recommends against their being there.
Ideally, there should be enough space for two copies of the BIOS to coexist, with the unused bank writable by a Windows/Linux device driver. If not, perhaps we'll start to see bootable miniature Linux ISOs capable of flashing the board. It would be practical, cost nothing for the vendors, and after all -- every computer deserves the opportunity to run Linux for a little while at least.
And if that's not enough Scott Adams for you...
on
God's Debris
·
· Score: 3, Informative
No sense in getting a DVD Drive and decoder board now, when the DVD-RW is only a few months away. I also wanted to be able to burn and rip CDs fairly fast, so I went with the Yamaha CRW2100EZ. It's a very nice, very fast drive, but has a major problem for the quiet machine: it's loud.
Worse - these drives vibrate insanely.
I don't trust a Yamaha 2100-series drive in any case that's got a hard drive too. I bought both of mine on spec, and while they're great drives, they can practically walk a loaded tower across the floor. I moved mine into external boxes with cushioned feet.
The main value of VM today is getting rid of dead code at run-time. A basic problem with shared libraries is that you load in the whole library, needed or not, when you need any function from it. This wastes memory, but after a while, the VM system will notice the unused pages and quietly release them. On a larger scale, the same problem is seen with dormant applications, a problem which has gotten totally out of hand in the Windows world, where far too much unwanted stuff launches at startup. VM ejects them from memory. That's what VM is really used for today.
So if you're actually page-faulting, VM is hurting, not helping.
That one-liner made no sense on the tail of the previous paragraph.
Apps are getting larger. A fat chunk of most modern GUI software is dead code, error checking code, or code that only gets used during startup and teardown.
Apart from application code, handling huge datasets also benefits from VM, leaving the most often used parts in memory. Do you really want 2G of memory just to lay out a multi-layer magazine-resolution image in gimp?
That said, if you really want your wish - just set VM page sizes to be larger than your largest application and have a look-see at how performance is affected.
The problems Windows users usually encounter (it's generally not so plug-and-play as you describe) are from specialized IDE drivers. Resetting your drivers to stock Windows IDE before swapping boards usually fixes that.
With Linux, you shouldn't have to do a thing, assuming you didn't have a crazy drive controller that was remapping the hard drive in a strange way. (Rare, and unheard of on anything post-486, afaik.)
At worst, you may have to make a boot disk and run lilo or equivalent on the new machine. Making a boot disk just takes a few seconds, and it may save you a headache - do that before you yank things apart.
If the thing does topple, it's more likely to be on starting X than anything else. Be sure you know how to disable kdm/gdm/xdm long enough to play with your configuration and you'll be okay.
The question isn't whether developers can work in a locked down environment. The question is whether they will work. Not being trusted by your employer well enough to use your tools is enough to demotivate a lot of employees.
The correct solution is for your IT department to invest a little time in training problem users well enough for them to stop being problems. Until you've done this, the user isn't the one causing the problem.
Once you've trained users, if any individuals persist in causing problems - stubbornly doing things you've taught them not to do - then that's an issue for management to tackle.
It's not cheap, but it's flexible. Chicago has ISPFH - a play on BOFH.
ISPFH is a non-profit, and the more it grows, the cheaper it gets. They offer colocation, DSL and ISDN access. (Possibly dialup too?)
Basically, you buy in as a member of the co-op, then pay at cost. Excess funds are returned to the members. You can do as you please, so long as it doesn't harm anyone else. (Full details in their AUP.)
ISPFH's bandwidth isn't oversold (except 2x on DSL), and I've heard nothing but good things about them. Me, I'm still doing the speakeasy DSL thing - cheap, fast. But if covad ever goes belly up, ISPFH's my next step.
I recently bought an Apple Powerbook G4. One day the DVD drive stoped working and I called apple tech support after they concluded that it was indeed a hardware issue. (Took 5 minutes) They told me it was still under warranty and that they will send me a box to ship it in.
That was Monday. Tuesday Airborne Express shows up with a box. I signed for it. Opened it up and inside was foam packaging for my TiPB and a Return Airborne Express slip. So I put the PB in it and called Airborne Express. They were back an hour later and picked up the box and shipped it to apple. (This is all at apples Expense) So. I send it out Tuesday. On Thursday my Powerbook is back in my hands and in Primo condition.... How cool is that?
Of all companies - Commodore was actually great about this. Commodore 64 warranty service was quick, and the support included with my Amiga 3000 went just as you described - I received a box the very next day, and had my machine back a few days later.
The treatment when my monitor (C=1950) went bad about a week after the warranty went out was amazing - I called, and had a brand new monitor at my door the very next day with instructions on how to return the bad monitor in the same box. The process never cost me a dime.
Max Payne is cool - a brilliant, beautiful game, and a lot of fun - the first time you play through.
Unfortunately it's saddled with such a linear storyline that once you're through it, it's not much fun anymore. This one's a definite renter if anything is; you can easily play it out in a weekend.
You should see the OSDN Navbar atop the page now. I don't like it any more then many of you, so if you log in, there is an option to disable it. (Click the 'X', or look in Preferences:Misc)
I don't see a Preferences:Misc - I've only got the old category options. Am I missing something here?
Slashdot Mirror
The real use of learning assembly is to validate your compiler output. It's rare to want to code much from the ground up in assembly anymore, save hardware-banging code, but knowing assembly still makes all the difference.
Yes, most compilers will perform 1,001 optimizations, but the final bit of optimization you can eck out demands that you profile, examine the hot points, and replace them with assembly or restructure your C with a good idea of the code you really want it to generate.
You'll still find things that the compiler engineers haven't thought of, or were lazy about implementing, or operations which implement unneccessary logic/precision.
For example, sometimes it's best to hoist an expensive calculation -into- a loop because it fits where the processor otherwise stalls, but most compilers take the opposite approach. It's tough to force a C compiler to do this, because most want to hoist code -out- of loops when possible, assuming it's cheapest.
Dependency on non-native signedness, float format or type sizes can mean that shifts, transcendental math, type conversion, etc is being done in software when a single opcode is what you expected.
Understanding the code will tell you if something unusual is happening, such as non-native sized bools or less than optimal variables being turned into register variables.
Some operations are simply not available to the C-only programmer. Subsequent lookups to sin/cos operations can be combined into a single opcode, but I've yet to see it done by any compiler. It's only available in assembly.
The PowerPC offers a rough inverse/division approximation which is faster than a real divide, and is good enough for operations only requiring low precision or an approximation... also completely inaccessible to C code.
Knowing a variable is guaranteed to be within a certain range or of a fixed set of values at a certain point can let you get away with all kinds of murderous assumptions which it's impossible to express in C.
That said, I haven't seen a single PPC assembler reference that was half as good as just looking at the code. Look at code, look at a lot of code, and past that, just look at system implementors' documentation. x86 through Pentium III aside, most current assembler books are just fluff, are wrong about half the pipelining, omit a million useful optimzations and don't cover the real story at all. It's really gotten to the point where the only real way to learn is by doing and doing and doing.
Just kill your extensions. If a site needs fancy songs and zooming images to keep your interest, what it's trying to tell you probably isn't very interesting.
If you really can't bear to live without Flash/Java/etc from time to time, set up a second browser, and use the clean one as your default.
Actually, if memory serves, it was the other way around. He married her, then let her spec out what she wanted and see it developed.
Anyone else and it probably wouldn't have seen the light of day.
I'll meet you half way. I still block out banner ads from companies that request cookies with the ad. Unfortunately, that's most of them these days. I'll allow unobtrusive ads supporting a site I like, except when they engage in secretive tracking of my browsing behaviors.
An advertiser keeping tabs on me without first asking me if that's okay steps way over the line of trust.
I've been there. I've seen the complaining and bitching, but the fact remains that the people who repeatedly go back and get familiar enough with the system to participate in polls, etc are largely DSL enthusiasts.
It would be too costly for Google and friends to index a site which demanded a penny for each page read.
Call me a pessimist, but my belief is that businesses are incapable of handling this kind of thing responsibly. The moment we go to penny-per-page, we'll start to see things artificially segmented across a dozen pages, and all kind of fluff and noise between the front page and any useful pages.
Make it a penny/nickel/dime a day for access to a whole domain, depending on the quantity and nature of the content within, and I might be interested.
People who go to a site to talk about how exciting DSL is come to the conclusion that DSL is better than dialup?
Gee, now there's a story...
I'd kill to have a Dual 2000+ MP system...
That doesn't stop the new noisy banner ads.
This doesn't stop Shockwave, etc from making sound. The audio settings only affect the standard html and javascript-controlled sound. You'd have to remove all your plugins to stop the sounds in IE.
It's already had an effect. The kinds of places that spam are the kinds of places that don't give a fuck about anything but growing the business. By avoiding spammers, I'm increasing the probability that I'm dealing with a company that has some measure of a concern about the consumer.
The rest will sort itself out. The worse spam gets, the more likely it is to get legislated into a black hole, or the more likely it is that enough ISPs will start deploying cooperative spam circumvention devices to make it effective.
It'll get there, through law or common sense. In this day and age, the problem just has to become expensive to big business first.
I actively avoid supporting marketers who annoy me, and I urge you to do the same. If a company is advertising in a magazine I like, in a banner ad on a web page I like, or on a TV show I like, they're starting out with an even trade. A little support for my interests in exchange for a little bit of my attention.
If they spam, throw junk mail at me, pop up annoying windows all over my desktop, stick business cards or flyers on my windshield or otherwise inconvenience me, I ensure that I never ever support them. I've stopped dealing with companies I previously liked because they started engaging in these practices.
Spamming is essentially stealing. I don't know why the hell I'd want to support someone whose first action was to take from me without giving anything in return.
Service doesn't factor into the equation the way it used to -- People buy pretty much exclusively on price or market presence.
To satisfy either, you've got to spend every available dollar on advertising until your competitors are gone, or you need to operate with little profit (or at a loss) to start with.
Getting a foothold in big business today is pretty much a starving contest. <overgeneralization type=slight>Quality products and customer service only work when you're pitching to old folks and the rich.</overgeneralization>
This is all very grand, except that floppy drives aren't guaranteed to exist in new systems. If memory serves, the 2002 computer spec put together by Intel and Microsoft actually recommends against their being there.
Ideally, there should be enough space for two copies of the BIOS to coexist, with the unused bank writable by a Windows/Linux device driver. If not, perhaps we'll start to see bootable miniature Linux ISOs capable of flashing the board. It would be practical, cost nothing for the vendors, and after all -- every computer deserves the opportunity to run Linux for a little while at least.
It's bad, but bad enough to be funny, I guess. Or maybe not.
Worse - these drives vibrate insanely.
I don't trust a Yamaha 2100-series drive in any case that's got a hard drive too. I bought both of mine on spec, and while they're great drives, they can practically walk a loaded tower across the floor. I moved mine into external boxes with cushioned feet.
That one-liner made no sense on the tail of the previous paragraph.
Apps are getting larger. A fat chunk of most modern GUI software is dead code, error checking code, or code that only gets used during startup and teardown.
Apart from application code, handling huge datasets also benefits from VM, leaving the most often used parts in memory. Do you really want 2G of memory just to lay out a multi-layer magazine-resolution image in gimp?
That said, if you really want your wish - just set VM page sizes to be larger than your largest application and have a look-see at how performance is affected.
With Linux, you shouldn't have to do a thing, assuming you didn't have a crazy drive controller that was remapping the hard drive in a strange way. (Rare, and unheard of on anything post-486, afaik.)
At worst, you may have to make a boot disk and run lilo or equivalent on the new machine. Making a boot disk just takes a few seconds, and it may save you a headache - do that before you yank things apart.
If the thing does topple, it's more likely to be on starting X than anything else. Be sure you know how to disable kdm/gdm/xdm long enough to play with your configuration and you'll be okay.
The correct solution is for your IT department to invest a little time in training problem users well enough for them to stop being problems. Until you've done this, the user isn't the one causing the problem.
Once you've trained users, if any individuals persist in causing problems - stubbornly doing things you've taught them not to do - then that's an issue for management to tackle.
Basically, you buy in as a member of the co-op, then pay at cost. Excess funds are returned to the members. You can do as you please, so long as it doesn't harm anyone else. (Full details in their AUP.)
ISPFH's bandwidth isn't oversold (except 2x on DSL), and I've heard nothing but good things about them. Me, I'm still doing the speakeasy DSL thing - cheap, fast. But if covad ever goes belly up, ISPFH's my next step.
Of all companies - Commodore was actually great about this. Commodore 64 warranty service was quick, and the support included with my Amiga 3000 went just as you described - I received a box the very next day, and had my machine back a few days later.
The treatment when my monitor (C=1950) went bad about a week after the warranty went out was amazing - I called, and had a brand new monitor at my door the very next day with instructions on how to return the bad monitor in the same box. The process never cost me a dime.
I wish C= had advertised this service a bit more.
Unfortunately it's saddled with such a linear storyline that once you're through it, it's not much fun anymore. This one's a definite renter if anything is; you can easily play it out in a weekend.
I don't see a Preferences:Misc - I've only got the old category options. Am I missing something here?