52 years of being an electronics nerd? Wow. I'm impressed. Keep it up!:)
Hmmm.. just looking at the pile of bloated caps from todays mess..
Lelon's again..
I too have found Rubycons to be of good quality. A large part of my salvaged store of reasonable condition caps are Rubycons, I don't think i've found out a dead one yet.
I've also found G-Luxon and Sanyo to be good - going on previous history anyway. I guess any company can have a bad batch.
Teapo's been better than average, but have seen enough of them die. There are just too many damn brands out there.
I'm supprised you found as small as 220uf's playing up.. But then again i've been too lazy to ESR them/pull the little b*ggers out. I guess that's where experience comes in 8)
Still, it's nice knowing other people are finding simillar results.. An open database of this kind of information (and maybe board repair in general) would be useful. Any such thing exist yet?
Or maybe I should get cracking/creative and start developing one.. 8)
I'm no electronics guru, but I've recapped many (hundreds of) bloaty unstable boards in the last few years - I hate seeing them thrown out for no cause. I now collect them from computer stores.
It's amazing how little work is required to do this once your on a roll, and does wonders for ones soldering skills.
While I've found nothing is better than using brand new caps, I've found boards that die from other causes (eg, idiot techs with slipping screwdrivers) are a good source of the right size salvageable caps. Also I have found PC power supplies have high quality caps in them, although bloating is also a common fault in them too (causes a lot of voltage variance - they are worth recapping)
(Meter them first of course:)
Also, I have found that replacing 6.3volts with 10volt caps of the same farad rating tend to last alot longer, especially when you see all of one spec bloating..
I always replace the caps in groups, for instance, I'll replace all of the 1000uf 6.3v caps even if only a couple are bloated. I use the same brand/specs to replace the groups.
Whether or not it's economically viable may be a different story, but I've had a high success rate using salvaged (metered) parts. I've not had to do a board twice.
I've noticed it's really only caps 1000uf and over that go, 1500uf 6.3v and 1000uf 6.3v seems to be the most common caps to fail. I've replaced all sorts of brands, so I personally feel it's crummy manufacturers fault choosing the wrong spec/quality caps to save 13 cents per board.
Temperature has a big play in it too though, i've noticed quite a few caps near regulators and heatsinks go. I think slightly higher voltage or quality makes sense near those areas.
I guess the only thing my method could of really accomplished is perhaps make the inexperienced and impatient cracker bored and go after something else, which seemed to be the point of a lot of code protection schemes. If someone is patient and determined enough to break protection code, it can and will be done..
I did try to get the 386 debug registers to trap access to the IRQ table, but I couldn't do that and play with softice at the same time, and plus didn't work under windows dosbox for some dumb reason. And what would stop the debug registers being reset?
my knowledge of "encryption" back then consisted of bit-rotates & exclusive or, so no matter what I did it would of sucked.
Still, it was fun trying to step through the code, and being 15/16 at the time, I thought it was a neat idea.:}
I did something kinda (vaugely:) similar years ago using self modifying code back in my young teenage DOS/Doze assembler programming days. It didn't really do anything to stop the code being modified or screwed up, but it did a rather good job at obscuring and hiding things.
Basically there was 3 layers of "encryption" on code..
The first stage was for protection of functions/code blocks, which started with a call to the end of the block, which then called the decrypter function, used the return addresses on the stack as start and length, decrypted the block in between, ran the code, and reencrypted on return of the block.
Second stage used x86 instruction trace interrupt (int 1) to decrypt the next instruction, and encrypt the previously executed one. I used a single byte instruction (int3 i recall) to turn on and off this tracing interrupt.
Then finally, the entire code and data was encrypted, along with a software generated decryptor.
It was a little bit of a beeyatch putting the binaries together.
My original plan was to protect those software registration engines that were so easy to trace through and break with the likes of Softice.
Looking back, it probably would of made something rather fun of the viral sort, but I am thankful I wasn't into excessive dissent as a teen hacker.:)
I propose some alternative, but completely retarded solutions..
Switch on all of the switches Enter room with lights.. unscrew (with gloves) the bulbs that you don't want on - thus making the switches redundant and giving you controlling any individual light bulb.. Hooray:)
Or alternatively, open up one of the switches, and wire in a big fat resistor or dimmer circuit.. Switch one of the other ones off, and the other on. One should be off, one should be on, and the other one should be dimmer..
Or just forget the damn switches and bring a torch or have fun in the dark.. less likely to get burnt/electrocuted that way.
I reckon it looks more like a muddy rock with a couple of barnicles/whatever they are called from the beach..
Well, at least thats what I think they look like.. I can't remember last time I left the comforting glow of my CRT and stood in the searing heat of that bright yellow thing in the sky that makes it difficult to sleep during non-coding periods.
Mmmmm.. Yummy.. but got to eat them naked in the bath tub.:)
But then i guess nerdy people like us would then need a waterproof keyboard..
What i've found is even worse is 2yo bubs eating them, and then touching everything in and out of my (sidecover-less) computer(s) with her evil sticky child hands..
I too have spilt my share of fizzy drinks, pulled many a keyboard to a million bits (well, 108 or so) and washed the membrane in the sink/under the garden hose, dry em off and then put those stupid keys back in qwerpy order.. Usually works too. 8)
Re:Back to the old days? Maybe it WAS a good idea.
on
Blu-Ray vs. HD-DVD
·
· Score: 0
I think you are on to something:-)
When you think about it, floppy discs had protective cases, so do zip discs, harddrives are concealed in armour to protect their sensitive platters.
They are talking about discs coming that hold almost as much as a basic harddrive today coming in the starkers. Are they nuts?
How much does a little bit of protective padding and a well designed plastic molding the size of a current cd case cost? Not much I bet..
I think people are just scared of using something that looks the size of a 5 1/4" disc.
I wish they would scrap the cruddy platter idea and bring on cheap solidstate.:)
Hmm.. Integration is definitly a key to making things more energy efficient. Couple that with killing off some legacy stuff, and I see a good future in low power consumption for computing in general.
While my notebook doesn't have serial ports, parallel ports, floppy drives, etc, I know the (cheap sis) chipset still has some wasted circuitry inside (while although probably doesn't consume much power in it's unused state, the space could be reused to bring things closer together). (And how long until they get rid of 16bit mode in the cpu.. yuck.)
In the meantime, maybe it's time to regain some of that currently wasted heat back into electricity. Help run the fans off CPU heat?
And maybe a plugin solar panel to charge the battery?
Hmm. I've always wondered if large multisession count discs have to load all the tracks in sequence to reconstruct the directory, or if each track has the entire directory structure replicated?
If it doesn't replicate, and providing the disc doesn't get scratched in the meantime, might get a bit slow after a while..:-}
Not that your worth a reply, but why not... You are a clueless moron.
It's assholes like you that drag this industry down.
I was just expressing my opinion. You are obviously one of those idiots who thinks they are something when they are not. And you are too chicken to post with your own name...
Please commit suicide and cleanse the earth of yourself you rude pathetic arrogant disgusting filthy PIG!
Sorry, half detecting sarcasm. but my sensors faulty at the moment.:-}
Back then, people coded because they loved to, but today, a lot more people just code because they can.
C64 was 1mhz (well, 2mhz, but the gfx chip stole half of it), and had kickass games for the time. Creatures 2 and Mayhem in Monsterland were the best I can recall off the top of my head.
Plus a thousand hideously overpriced software licenses, product keys, and too much stinking paperwork.:-}
Has microsoft even considered that companies who need huge computing power are using linux/etc.. cos they can spend more money on hardware resources?
IMHO, Moores law is utter bullocks, they are well and truly capable of building a lot bigger and better equipment now.
It appears to me to be more of a marketing standard to determine how long they can draw out their lines in order to milk the maximum profit from unwitting consumers.
Sure some of this milk money has gone towards research, which in turn has lead to smaller transistor size and other improvements in technology, but doesn't stop the fact they are milking the market.
Still, if software people put the same love and devotion into the chips we have now, as they did back when toy computers roamed the earth, maybe us poor consumers would see some real value.. Instead of hardware people designing stupid chips to run the extra bloated software all the more efficiently.
The easier a CPU's is to program, more idiots will try to program...
Sync the software to hardware clock at boot time, and update the software clock in a low level interrupt. With a few subtle little variances worked in, and balances out a few seconds later.
I thought most OS's did most of that already (apart from the subtle variance bit)?
As far as NAT goes, why not b0rk the timestamps a bit on the packets? I'm no networking guru, but someone will come up with something soon enough to stop this.
Sure it's handy to have as evidence, but what can be used to solve crimes can also be used to carry out crimes.
Alan Sugar was my hero, I loved the Amstrad CPC664 and CP/M 2.2 |CPM - plus it was the only way to format the stupid 3 inch discs out of the box.
Had a brilliant editor wordstar, and logo, plus other cool stuff I've well and truly forgotten. Spent more time in their than in Basic or playing games.
Was only a kid, but used to code stuff in masm, found the different register names on z80 and 8080 a bit confusing. z80 made more sense.
On the Microbee cp/m, I discovered turbo pascal. That was fun.
Commodore64 had a z80 expansion cartridge that came with CP/M. was a bit boring cos the C64 had sooo much better potential with it's own CPU and Turbo Assembler 5.:-)
I'm only 26, but I certainly had a misspent childhood and a lot of good memories thanks to CP/M. Thankyou too all the guys alive and dead who made those things happen.
I am by no means a music professional, it's only a side hobby of mine, but I play a fair bit with midi stuff, and multitrack too, but only using everyday hardware.
With the Jack Audio Connection Kit things start to get pretty awesome, I'm getting close to zero latency (5-10ms) (with 2.4.26 low latency patches) with qsynth/fluidsynth. Something I never achieved on windows.. (And i'm not talking about fancy expensive hardware solutions)
Rosegarden? I admit it's not perfect, but a great foundation. There's quite a few multitrack recorders around too. Audicity's not too bad.
There are starting to be some very useful tools for music. There's no Reason yet, but the foundation for that kind of power is definitly there. The rest of the work is just eye candy and pretty front ends..
And what can a well trained professional do in photoshop that a well trained professional can't do in the Gimp? not all that much... Gimp does have a slightly gimpy name though.:-}
You sad mortals who do not control your subconcious...;-)
And whats with this outdated coffee cup technology? haven't you heard of intraveinous caffeine injections?
I was running a P3 600, on an 810 board for a while running with 3 seagate 120gb drives for my brother who lives in a flat under my house. it was definitely not quick, but stayed up. The performance was pretty crap compared to the same hardware running a mirror, but a large part of that was the crappy controller on board.
I upped the box to an athlon 1200 on a sis mainboard and things improved quiet a bit. (both boxes had 256mb sdram) The raid and network performance doubled, and tons of delays disappeared.
After complaints (from the windows user) that it still wasn't utilising the bandwidth well enough, I reluctantly used one of my main PCs (athlon 2400, via kt266, 1Gb) and also tried an extremely cheap Sil680 controller (used just for the disk interface). The individual drive speeds went over 50mb/s. The raid5 was still only running at about 35mb/s. And of course no faster over the network - just a waste of a machine.:/
After not being able to stand the fan noises of the 2400, I split the drives off and put them in my brothers windows box as a mirror, seeing he's the only one who needed the reduntant space (he has some bizarre psychological need for 'warez' and mp3s) They failed within a day with loads of data recovery fun. He's since warranty swapped two of the for seagate sata's, and reinstalls once a month, and i've stopped caring about the plight of his O/S or warez..:-)
These seagate drives were less than 5 months old, and were getting loads of seek & ecc errors from the start. The were all running with enough cooling (it's summer here 9 months of the year), and the temperatures were monitored constantly. Just crap I guess.
The moral of the story: use decent long lifetime drives from the start, of different batches, don't throw more money on hardware than you have to (a cheap faster ide controller does wonders), and use a decent O/S that is stable and lets you optimized to hell (ie gentoo:-)
I kinda agree with you.. If the ability to lift a copy for 'free' is taken away from the masses, the natural migration path for those who dont want to pay will end up being towards free software.
I'd love to help out with writing some of this antipiracy stuff, but i'm a self righteous bastard who wouldn't run windows if I was paid to, besides, I don't wanna have to download a stinking compiler to do it.:-)
I wonder how well microsoft will survive the reality check, when their user base has a huge chunk taken out of it, because they wont give freebe's anymore to the 'pirate scum'?
They won't be taking in that much money from those few scared pirate fools rushing out to buy a legit copy.
Sure makes me glad my O/S is free, works fast and stable, comes with both C compiler and source code, and every app i'll ever need.. and on top of that, I get a clean conscience in the bargain.
Sorry Mr Gates, you aint gonna see a single cent from me.. unless I find Micro$ucks has crumbled and your destitute begging on the street corner.. in which case I might let you work for a couple of bucks an hour bootstrapping gentoo boxes for me...;-)
I agree that it is damn handy having everything at your fingertips... But I was kinda refering to a fair bit into the future of linux. (well, hopefully not too far away)
With a restructure and perhaps the benifits of using a little C++ in the driver/module interface we could make a decent API available.
Something like a standardized driver source code repository would be awesome. And a little front end app to fetch and build... (emerge anyone?;-) Distributions can play a key role here if they dont sh*tfight with each other over and actually adpot something.
This seems to be a major problem with Linux - the forces that be are a little too slow to standardize on anything.. (apart from the apparent "rejection of new ideas" standard;-) IMHO the linux kernel is stagnating a little bit in this area..
Still, theres not much better to use at the moment, so I will stick to what we got... At home, I have 10 gentoo boxen for development, and 1 mac running OS-X. It'll be running gentoo too the second my wife turns her head...;-)
Yeah, but if it can be built as a module, it dont really need to be in the mainline...:-/ You need kernels source to build such modules, but not the other way around..
I agree the platform dependant stuff is a mess, but if they split it, it could cause a few rifts in the kernel..? Last thing we need is 1024 seperate branches...
Patches are fine if you run a stock kernel... (me speaking as a lazy gentoo-dev-sources user:)
If anything split out of it, I do believe it should be available through the same distribution channel though as the mainline kernel, just _optional_.
Sure you can - if you like doing all the dirty work yourself... (not that anythings wrong with dirty work:-)
Although sometimes it is good to throw in a little abstraction. Being able to visualise some things on a different level really does help a programmer wrap their head around difficult issues that can't easily be solved going the hard yards..
For example, learning forth as a kid did wonders for me being able to get a very solid grasp on recursion. True, it's not something you use everyday, but it's a different way to think, and some problems can only be solved (efficiently) that way.
(My point is...) Having a few different programming models at your disposal is not such a bad thing. Yes, along with C++ comes a little bloat, but it also includes C in the bargain. It'd be nice having a little choice for those who want it.
My language of choice these days is a hybrid of C and C++ (well, more lazy C++) - sure it means I have to link in an extra library, but I can break out with a different model when I see a real need for it.
It is a little different at the kernel level, but their really isn't much stopping that extra bloat being included when a module is loaded.
Slashdot Mirror
52 years of being an electronics nerd? Wow. I'm impressed. Keep it up! :)
Hmmm.. just looking at the pile of bloated caps from todays mess..
Lelon's again..
I too have found Rubycons to be of good quality. A large part of my salvaged store of reasonable condition caps are Rubycons, I don't think i've found out a dead one yet.
I've also found G-Luxon and Sanyo to be good - going on previous history anyway. I guess any company can have a bad batch.
Teapo's been better than average, but have seen enough of them die. There are just too many damn brands out there.
I'm supprised you found as small as 220uf's playing up.. But then again i've been too lazy to ESR them/pull the little b*ggers out. I guess that's where experience comes in 8)
Still, it's nice knowing other people are finding simillar results.. An open database of this kind of information (and maybe board repair in general) would be useful. Any such thing exist yet?
Or maybe I should get cracking/creative and start developing one.. 8)
I'm no electronics guru, but I've recapped many (hundreds of) bloaty unstable boards in the last few years - I hate seeing them thrown out for no cause. I now collect them from computer stores.
:)
It's amazing how little work is required to do this once your on a roll, and does wonders for ones soldering skills.
While I've found nothing is better than using brand new caps, I've found boards that die from other causes (eg, idiot techs with slipping screwdrivers) are a good source of the right size salvageable caps. Also I have found PC power supplies have high quality caps in them, although bloating is also a common fault in them too (causes a lot of voltage variance - they are worth recapping)
(Meter them first of course
Also, I have found that replacing 6.3volts with 10volt caps of the same farad rating tend to last alot longer, especially when you see all of one spec bloating..
I always replace the caps in groups, for instance, I'll replace all of the 1000uf 6.3v caps even if only a couple are bloated. I use the same brand/specs to replace the groups.
Whether or not it's economically viable may be a different story, but I've had a high success rate using salvaged (metered) parts. I've not had to do a board twice.
I've noticed it's really only caps 1000uf and over that go, 1500uf 6.3v and 1000uf 6.3v seems to be the most common caps to fail. I've replaced all sorts of brands, so I personally feel it's crummy manufacturers fault choosing the wrong spec/quality caps to save 13 cents per board.
Temperature has a big play in it too though, i've noticed quite a few caps near regulators and heatsinks go. I think slightly higher voltage or quality makes sense near those areas.
I guess the only thing my method could of really accomplished is perhaps make the inexperienced and impatient cracker bored and go after something else, which seemed to be the point of a lot of code protection schemes. If someone is patient and determined enough to break protection code, it can and will be done..
:}
I did try to get the 386 debug registers to trap access to the IRQ table, but I couldn't do that and play with softice at the same time, and plus didn't work under windows dosbox for some dumb reason. And what would stop the debug registers being reset?
my knowledge of "encryption" back then consisted of bit-rotates & exclusive or, so no matter what I did it would of sucked.
Still, it was fun trying to step through the code, and being 15/16 at the time, I thought it was a neat idea.
I did something kinda (vaugely :) similar years ago using self modifying code back in my young teenage DOS/Doze assembler programming days. It didn't really do anything to stop the code being modified or screwed up, but it did a rather good job at obscuring and hiding things.
:)
Basically there was 3 layers of "encryption" on code..
The first stage was for protection of functions/code blocks, which started with a call to the end of the block, which then called the decrypter function, used the return addresses on the stack as start and length, decrypted the block in between, ran the code, and reencrypted on return of the block.
Second stage used x86 instruction trace interrupt (int 1) to decrypt the next instruction, and encrypt the previously executed one. I used a single byte instruction (int3 i recall) to turn on and off this tracing interrupt.
Then finally, the entire code and data was encrypted, along with a software generated decryptor.
It was a little bit of a beeyatch putting the binaries together.
My original plan was to protect those software registration engines that were so easy to trace through and break with the likes of Softice.
Looking back, it probably would of made something rather fun of the viral sort, but I am thankful I wasn't into excessive dissent as a teen hacker.
I propose some alternative, but completely retarded solutions..
:)
Switch on all of the switches
Enter room with lights..
unscrew (with gloves) the bulbs that you don't want on - thus making the switches redundant and giving you controlling any individual light bulb.. Hooray
Or alternatively, open up one of the switches, and wire in a big fat resistor or dimmer circuit.. Switch one of the other ones off, and the other on. One should be off, one should be on, and the other one should be dimmer..
Or just forget the damn switches and bring a torch or have fun in the dark.. less likely to get burnt/electrocuted that way.
I'll go back in my hole now...
I reckon it looks more like a muddy rock with a couple of barnicles/whatever they are called from the beach..
Well, at least thats what I think they look like.. I can't remember last time I left the comforting glow of my CRT and stood in the searing heat of that bright yellow thing in the sky that makes it difficult to sleep during non-coding periods.
Mmmmm.. Yummy.. but got to eat them naked in the bath tub. :)
But then i guess nerdy people like us would then need a waterproof keyboard..
What i've found is even worse is 2yo bubs eating them, and then touching everything in and out of my (sidecover-less) computer(s) with her evil sticky child hands..
I too have spilt my share of fizzy drinks, pulled many a keyboard to a million bits (well, 108 or so) and washed the membrane in the sink/under the garden hose, dry em off and then put those stupid keys back in qwerpy order.. Usually works too. 8)
I think you are on to something :-)
:)
When you think about it, floppy discs had protective cases, so do zip discs, harddrives are concealed in armour to protect their sensitive platters.
They are talking about discs coming that hold almost as much as a basic harddrive today coming in the starkers. Are they nuts?
How much does a little bit of protective padding and a well designed plastic molding the size of a current cd case cost? Not much I bet..
I think people are just scared of using something that looks the size of a 5 1/4" disc.
I wish they would scrap the cruddy platter idea and bring on cheap solidstate.
Hmm.. Integration is definitly a key to making things more energy efficient. Couple that with killing off some legacy stuff, and I see a good future in low power consumption for computing in general.
:-(
While my notebook doesn't have serial ports, parallel ports, floppy drives, etc, I know the (cheap sis) chipset still has some wasted circuitry inside (while although probably doesn't consume much power in it's unused state, the space could be reused to bring things closer together). (And how long until they get rid of 16bit mode in the cpu.. yuck.)
In the meantime, maybe it's time to regain some of that currently wasted heat back into electricity. Help run the fans off CPU heat?
And maybe a plugin solar panel to charge the battery?
But I guess it kind of hampers portability..
Thats my 2 cents..
Now if only I could convince my stinking ISP that downloading linux ISO's is not illegal :-)
Hmm. I've always wondered if large multisession count discs have to load all the tracks in sequence to reconstruct the directory, or if each track has the entire directory structure replicated?
:-}
If it doesn't replicate, and providing the disc doesn't get scratched in the meantime, might get a bit slow after a while..
Damn. now they know.. Now how am I going to smuggle US military secrets? ;-)
Not that your worth a reply, but why not... You are a clueless moron.
It's assholes like you that drag this industry down.
I was just expressing my opinion. You are obviously one of those idiots who thinks they are something when they are not. And you are too chicken to post with your own name...
Please commit suicide and cleanse the earth of yourself you rude pathetic arrogant disgusting filthy PIG!
Sorry, half detecting sarcasm. but my sensors faulty at the moment. :-}
Back then, people coded because they loved to, but today, a lot more people just code because they can.
C64 was 1mhz (well, 2mhz, but the gfx chip stole half of it), and had kickass games for the time. Creatures 2 and Mayhem in Monsterland were the best I can recall off the top of my head.
Plus a thousand hideously overpriced software licenses, product keys, and too much stinking paperwork. :-}
Has microsoft even considered that companies who need huge computing power are using linux/etc.. cos they can spend more money on hardware resources?
IMHO, Moores law is utter bullocks, they are well and truly capable of building a lot bigger and better equipment now.
It appears to me to be more of a marketing standard to determine how long they can draw out their lines in order to milk the maximum profit from unwitting consumers.
Sure some of this milk money has gone towards research, which in turn has lead to smaller transistor size and other improvements in technology, but doesn't stop the fact they are milking the market.
Still, if software people put the same love and devotion into the chips we have now, as they did back when toy computers roamed the earth, maybe us poor consumers would see some real value.. Instead of hardware people designing stupid chips to run the extra bloated software all the more efficiently.
The easier a CPU's is to program, more idiots will try to program...
Sync the software to hardware clock at boot time, and update the software clock in a low level interrupt. With a few subtle little variances worked in, and balances out a few seconds later.
I thought most OS's did most of that already (apart from the subtle variance bit)?
As far as NAT goes, why not b0rk the timestamps a bit on the packets? I'm no networking guru, but someone will come up with something soon enough to stop this.
Sure it's handy to have as evidence, but what can be used to solve crimes can also be used to carry out crimes.
Alan Sugar was my hero, I loved the Amstrad CPC664 and CP/M 2.2 |CPM - plus it was the only way to format the stupid 3 inch discs out of the box.
:-)
Had a brilliant editor wordstar, and logo, plus other cool stuff I've well and truly forgotten. Spent more time in their than in Basic or playing games.
Was only a kid, but used to code stuff in masm, found the different register names on z80 and 8080 a bit confusing. z80 made more sense.
On the Microbee cp/m, I discovered turbo pascal. That was fun.
Commodore64 had a z80 expansion cartridge that came with CP/M. was a bit boring cos the C64 had sooo much better potential with it's own CPU and Turbo Assembler 5.
I'm only 26, but I certainly had a misspent childhood and a lot of good memories thanks to CP/M. Thankyou too all the guys alive and dead who made those things happen.
I am by no means a music professional, it's only a side hobby of mine, but I play a fair bit with midi stuff, and multitrack too, but only using everyday hardware.
:-}
With the Jack Audio Connection Kit things start to get pretty awesome, I'm getting close to zero latency (5-10ms) (with 2.4.26 low latency patches) with qsynth/fluidsynth. Something I never achieved on windows.. (And i'm not talking about fancy expensive hardware solutions)
Rosegarden? I admit it's not perfect, but a great foundation.
There's quite a few multitrack recorders around too. Audicity's not too bad.
There are starting to be some very useful tools for music. There's no Reason yet, but the foundation for that kind of power is definitly there. The rest of the work is just eye candy and pretty front ends..
And what can a well trained professional do in photoshop that a well trained professional can't do in the Gimp? not all that much...
Gimp does have a slightly gimpy name though.
You sad mortals who do not control your subconcious... ;-)
And whats with this outdated coffee cup technology? haven't you heard of intraveinous caffeine injections?
I was running a P3 600, on an 810 board for a while running with 3 seagate 120gb drives for my brother who lives in a flat under my house. it was definitely not quick, but stayed up. The performance was pretty crap compared to the same hardware running a mirror, but a large part of that was the crappy controller on board.
:/
:-)
:-)
;-)
I upped the box to an athlon 1200 on a sis mainboard and things improved quiet a bit. (both boxes had 256mb sdram) The raid and network performance doubled, and tons of delays disappeared.
After complaints (from the windows user) that it still wasn't utilising the bandwidth well enough, I reluctantly used one of my main PCs (athlon 2400, via kt266, 1Gb) and also tried an extremely cheap Sil680 controller (used just for the disk interface). The individual drive speeds went over 50mb/s. The raid5 was still only running at about 35mb/s. And of course no faster over the network - just a waste of a machine.
After not being able to stand the fan noises of the 2400, I split the drives off and put them in my brothers windows box as a mirror, seeing he's the only one who needed the reduntant space (he has some bizarre psychological need for 'warez' and mp3s) They failed within a day with loads of data recovery fun. He's since warranty swapped two of the for seagate sata's, and reinstalls once a month, and i've stopped caring about the plight of his O/S or warez..
These seagate drives were less than 5 months old, and were getting loads of seek & ecc errors from the start. The were all running with enough cooling (it's summer here 9 months of the year), and the temperatures were monitored constantly. Just crap I guess.
The moral of the story: use decent long lifetime drives from the start, of different batches, don't throw more money on hardware than you have to (a cheap faster ide controller does wonders), and use a decent O/S that is stable and lets you optimized to hell (ie gentoo
Oh yeah, and kill all windows users..
I'd love to help out with writing some of this antipiracy stuff, but i'm a self righteous bastard who wouldn't run windows if I was paid to, besides, I don't wanna have to download a stinking compiler to do it. :-)
I wonder how well microsoft will survive the reality check, when their user base has a huge chunk taken out of it, because they wont give freebe's anymore to the 'pirate scum'?
They won't be taking in that much money from those few scared pirate fools rushing out to buy a legit copy.
Sure makes me glad my O/S is free, works fast and stable, comes with both C compiler and source code, and every app i'll ever need.. and on top of that, I get a clean conscience in the bargain.
Sorry Mr Gates, you aint gonna see a single cent from me.. unless I find Micro$ucks has crumbled and your destitute begging on the street corner.. in which case I might let you work for a couple of bucks an hour bootstrapping gentoo boxes for me... ;-)
With a restructure and perhaps the benifits of using a little C++ in the driver/module interface we could make a decent API available.
Something like a standardized driver source code repository would be awesome. And a little front end app to fetch and build... (emerge anyone? ;-) Distributions can play a key role here if they dont sh*tfight with each other over and actually adpot something.
This seems to be a major problem with Linux - the forces that be are a little too slow to standardize on anything.. (apart from the apparent "rejection of new ideas" standard ;-) IMHO the linux kernel is stagnating a little bit in this area..
Still, theres not much better to use at the moment, so I will stick to what we got... At home, I have 10 gentoo boxen for development, and 1 mac running OS-X. It'll be running gentoo too the second my wife turns her head... ;-)
Debian rules too, but just not in my house. :-)
You need kernels source to build such modules, but not the other way around..
I agree the platform dependant stuff is a mess, but if they split it, it could cause a few rifts in the kernel..? Last thing we need is 1024 seperate branches...
Patches are fine if you run a stock kernel... (me speaking as a lazy gentoo-dev-sources user :)
If anything split out of it, I do believe it should be available through the same distribution channel though as the mainline kernel, just _optional_.
Although sometimes it is good to throw in a little abstraction. Being able to visualise some things on a different level really does help a programmer wrap their head around difficult issues that can't easily be solved going the hard yards..
For example, learning forth as a kid did wonders for me being able to get a very solid grasp on recursion. True, it's not something you use everyday, but it's a different way to think, and some problems can only be solved (efficiently) that way.
(My point is...) Having a few different programming models at your disposal is not such a bad thing. Yes, along with C++ comes a little bloat, but it also includes C in the bargain. It'd be nice having a little choice for those who want it.
My language of choice these days is a hybrid of C and C++ (well, more lazy C++) - sure it means I have to link in an extra library, but I can break out with a different model when I see a real need for it.
It is a little different at the kernel level, but their really isn't much stopping that extra bloat being included when a module is loaded.