I think your bang on here. Linux is just too hard, because in windows land you only have to worry about 2k, XP, and Vista, in linux land you have people with custom hacks to the kernel, not to mention the flavor of the month for installers, development environments, compilers etc. There is just too many choices to be made, and they'd rather come across as offering great support to 99% of their users, than fight to try and help the 1%, and then have the occasional screw up. I hear this a lot, but it's bullshit. There's no "Windows" out there, except perhaps for exactly windows XP SP2. (what, 80%+ of the market still?) 95/98/ME/2000/XP/2003/Vista, plus each service pack, what version of.net, bla bla bla. Are they on a home computer or using a managed profile? What parts of the registry can we touch, and if we can't touch locked parts on newer OSs, are the appropriate replacements usable on older ones? Where do we install? Where do we keep user data? What APIs require Administrator privledges. Do we require Administrator for everything? Do we laugh at the customers then who upgrade to vista?
Or, you know, linux. If you're LSB compliant you're probably set. Target Ubuntu for desktop apps, Redhat for enterprise. Everybody using something else is smart enough to figure it out (or read the support forums). You can assume they're not running 10 year old software because it's free to upgrade. You can require new (not bleeding edge, but new) packages because again, you're not asking someone to shell out hundreds of dollars just to run your crap.
Realistically, it's not that hard to write decent unix software. You know you can't run as root because nobody does. You know you can't update system files except via the installer, and you know they'll have libX11 installed and the API is pretty fixed. After that, you've got a subset of options (GTK or another widget set) and possibly some kernel glue (2.6+ only) but mostly you just use libUSB for consumer-grade hardware.
Your second paragraph sounds like someone who hasn't actually required technical support for a windows based product, ever. When your target market can't tell the difference between their display and their computer, (and all the assorted other horror stories, cupholders and blackouts), it doesn't take very skilled support to explain to them how to plug in the cable and click Start/My Computer/Widget Installer CD. This means that anyone who actually has a PROBLEM with it is told the three Rs of tech support: restart, reboot, reinstall. Yay, helpful.
"the dumbass" probably wanted people to be able to reach him in an emergency, which is why he took a cellular phone with him. Turning it off/leaving it at home rather defeats the purpose, doesn't it?
I mean, call me crazy, but that's what I do with my phone while I'm out of the country, I just refrain from making calls on it unless they're important. Then again, I guess I'm "the dumbass" for expecting slashdot readers to have crawled out of their mother's basement and actually gotten out of the country themselves.
That's what I was asking, thanks. I missed that it hadn't added any new SSE registers. Don't be so quick on the "No such thing as a SSE5 process" though - there IS such a thing as a FPU process, because of an ancient design decision from intel that had the FPU as a coprocessor. That's stuck with us right to the point of 64bit processors - and they still have to emulate it in 32bit mode.
So, where's the analysis by people who write optimized media encoders/decoders? How useful are these new instructions, or are they just toys? How well did they handle context switching? What's the CX overhead? Is there a penalty for all processes, or only when you are switching to/from a SSE5 process? Will this be safely usable under all operating systems, or will they need a patch?
These clowns? Quality secure configuration there. That's called "information leak", and is one of the first things you're supposed to do when trying to secure a site. Hell, modern webservers have all that crap turned off by default.
Anyway, your posts are full of words and hyperbole, but I'll try your test if I can actually get it for linux. I'm not going to pay for it though, and it appears to be free for.gov use only?
Sure, if you get root on a linux box, you can load a module to grab the syscall handler and rewrite the ones you want. It's not as trivial on linux - the syscall table has no hooks for being tampered with at all, and has no fixed position for binary compatability. Still, it's been done. Adore traps sys_readdir, sys_open and a few others to hide processes and directories from view. And of course, it only takes one person to write the initial code - the armies of script kiddies will happily spread it without understanding what it does.
Other options are using the *trace(2) functions to hook into all running processes and trap/modify their syscall use, or simply replace ld.so with one that silently adds libtrojan.so to every program. The trace functions can be raced against (fork/clone followed immedately by the syscall you want to use. The latter can be defeated by directly using sysenter/int80 via inline asm.
Actually, "rootkit" told me all I ever needed to know about their "security". It's nothing but a USB image aqusition device and PC-side software to handle the matching and authorization. In other words - completely useless from a security standpoint. Think DRM - plug in the USB stick, it copies the decryption software, image matcher AND THE SECRET KEY to your harddrive, then uses a rootkit to "obscure" it.
The trick here is it's cheap as shit. Doing it properly on the keychain costs money - you'll need a decent processor to handle image aquisition and processing. Why bother with that when there's a 2+ghz CPU right next door on the bus? Worse, because they sell this crap as "security devices", they undercut everyone who spends the money to do it right. And of course they lie about how it really works, throwing buzzwords like "biometrically encrypted data storage" out.
How about mere unprofessional behavior? With the number of employers doing background checks for decent jobs, and googles all-seeing eye, it's rather nice to be able to drop back to a nym rather then a name. It also helps when my real name only shows up on technical mailing lists, and I have other nyms for other activities.
Yeah, it's not illegal to talk about dragging an entire cooler of booze to the beach and throwing an all-day party where we're blitzed out of our minds, but that's not how I like to be defined on google. The way pagerank worked, cool parties and other hobbies get linked to a hell of a lot more then archived mailing list posts discussing DMA timeouts on specific PCI busses.
So, should I be required to make a report of everything I've ever done in my entire life and let people pour over it? Or can I take a shit in private sometimes?
Yes it is, and no it isn't. It's really useful for "databases" that are only storing a small amount of data, or mostly static data. While I generally use Postgresql for any serious work, I've made trivial apps out of php+sqlite, because I can send someone the entire application as a single tarball. It's mostly useful to avoid having to write your own storage routines (which will invariably be buggier then something with lots of eyeballs on it).
Um, you care about oracle's storage because it forces you to care. You can't do ANYTHING until it's properly configured. It's a serious bitch that way.
And you've seriously GOT to be trolling about postgresql. "user@host w/password on database A" is seperate from "user@host w/password2 at database B" is a mysqlism that drives me up the wall.
Great, virtualized user domains would be nice, but that's not what MySQL is giving you. It's a best-match regex that may change depending on the whims of RDNS (that you may not even have control over). Thanks, but no thanks.
I'll take "create user 'foo' with password 'bla'; grant all on database 'bar' to 'foo';" any day of the week. Both postgres and MySQL have neigh-identical backup, unless you are incapable of changing 'mysqladmin dump database' to 'pgdump database' in your scripts.
Database workload tuning is a black art on all databases. Some just hide the values in obscure non-populated-by-default registry keys and require paid support services to learn about them.
Uh, yeah. I won't be the first nor last to point this out, but if Cable was truely "half duplex" then you'd have no upstream - period. Because if you'd get up from your computer desk, walk over to your television, and turn it on, guess what? Yeah, all those channels are broadcasting 24/7.
Thicknet/Thinnet lans used a single frequency for 'data transmission', modern cablemodems have vastly different upstream/downstream frequencies, and you are dynamically assigned a 'channel' thats much faster then your cap, shared with multiple other people.
Also, try Skype. Look ma, it works, full duplex. Guess you were wrong.
Don't post on copyright subjects until you can grasp the basic concept that Copyright is NOT Trademark, and doesn't have to be enforced to maintain it.
Copyright is for a period of time, period. Trademark is forever until failed to be maintained.
And you're an idiot for telling people they HAVE to file copyright infringement suits - even worse, this isn't even copyright infringement, it's a civil issue over which of the (many) royalty schemes allofmp3.com should be paying.
except despite having spent hundreds of dollars at allofmp3, they never did transfer my account (or balance) to the new site. And they don't take any form of payment, either. They should get into linden dollars or e-gold or something goofy and obviously money-laundering like that.
I mean, hell, how the fuck do offshore casinos move cash around?
Congrats, you successfully managed to snip out the part where he pointed out that the VAT (what's being enforced in South Korea) isn't the same as Income Tax like we have in the US. It's a sales tax.
Your post is like where you take a car and drop out the engine and wonder why it doesn't work! There, an analogy!
In other words - it's nearly identical to the situation in the US.
Some branches offer great low mortgage rates - but you have to have great credit score to get it. Some branches offer horrible non-fixed APR interest-only mortgages - but anyone can get them.
See how your credit rating DOES impact what kind of rate you get?
It takes 5 to 7 years before a company gets credit. It's _TOUGH_. Even with a Dunn & Bradstreet listing, and 4 years of audited recordkeeping, we still couldn't get terms on network gear. Ended up having to run the business on a personal Amex until year 5, when magically "Oh, here, have lines of credit! Good rates!".
That's a small business, with ~150k/year gross. Much like what you're reccommending to this guy. Still, there ARE benefits to incorperating when you don't have a direct 'employer' to deal with. It's worth doing. And besides, if he's pulled in a cool mil in a decade, he can afford someone to handle the paperwork. Unless it's all gone straight up his nose...
Saving up for a house requires exceptional restraint? Like what? Getting a good job but living in a cardboard box? Hint: If you're not living in a house, you're paying rent for an apartment, not saving up.
Your second piece of advice isn't fiscally sound either - your home is at it's most expensive in the first few years you own it. Your 30 year mortgage is a constant dollar cost, while inflation keeps rising. Paying expensive money now vs. cheap money later isn't a good plan, especially at a 5 or 6% rate. Put the money into moderate growth funds, and then if something goes wrong (and it will) you have money to cover it. If you don't? When you miss a monthly payment, you go into forclosure even if you've been overpaying.
Oh yeah, and in the US, mortage interest is tax-deductable, but principal payments are not.
When talking about credit cards - if you're already in debt, yes, pay off as much as you can each month.
Any more bad financial advice you'd like to share?
Actually, there is. You have to send exactly the same content to everyone in order to recieve the deepest discounts. In fact, once you have gotten bulk-identical mail delivery names and addresses for 80% of a zipcode, you can just give them a big pile of your crap for 100% coverage and they'll stick it in everyone's mailbox.
In effect, my 41c/letter is subsidizing junk mail. Strangely, I don't seem to recieve junk-fedex deliveries. Oh, that's because they don't have such an abuse-prone rate system.
Also, since I use e-billpay and correspond via email, the only things I get in the mail are replacement credit cards and some financial statements that they won't deliver as PDF. Oh, and piles and piles and piles of flyers, credit offers, coupon booklets and other junk.
Without money as an option? 9.08 gigabits/second. For internet1 apps, quite a bit less. I can trivially saturate 10mb from my dedicated server to my cablemodem. Two 100meg-linked boxes can transfer at 100meg point-to-point (not just point-to->many). If you had dark fiber, you could do whatever signalling you could support on it.
You're mistaking a single person as gatekeeper of the 'official' source as the only model to operate DVCS, though. In fact, even the linux setup defies that. Linus maintains 'linus-brand' linux, but the -mm tree is extremely used, and lots of development goes through there for a shakeout before going to 'official'. So there's seperate gatekeepers.
There's also the possibility of running the 'official' repository on a server that multiple release managers have access to.
Really, DVCS can handle any situation as well as (or generally better) then CVCS can. Want every developer's stuff backed up? Have a central repo that pulls every developer repo as a branch, and back it up. His PC dies, he can just re-grab his branch from the repo. No merges required, and shared history is preserved. (This would require a push-pull mode in some network setups, such as the case of a telecommuting worker behind NAT - the central repo can't reach him directly for a pull)
Re:Distributed version control gaining ground in F
on
Linus on GIT and SCM
·
· Score: 1
... and the non-networked medical instruments had a functioning development suite on them? Bzzzt. You develop on a networked PC and load the software/firmware to the instruments via whatever means. If the medical instruments have a PC controlling them, you throw a goddamned nic card in your development box.
Or are technicians supposed to say "whoops, I can fix that glitch!" as they're examining someone, drop out of the exam, fire up the IDE, fix the bug, re-start the program and complete examining the patient?
Actually, no, and for the same reasons file locking is an abysmally bad idea.
If you have to make a core API change, it's going to suck. Period. But if you lock the API that's used through the entire code (think resource allocation, for an example) that means nobody can work on anything, and all existing outstanding work has to be merged and stopped before you can even start to make the change. Refactoring is difficult enough as it is, you can't make the entire project grind to a halt over it.
Slashdot Mirror
Or, you know, linux. If you're LSB compliant you're probably set. Target Ubuntu for desktop apps, Redhat for enterprise. Everybody using something else is smart enough to figure it out (or read the support forums). You can assume they're not running 10 year old software because it's free to upgrade. You can require new (not bleeding edge, but new) packages because again, you're not asking someone to shell out hundreds of dollars just to run your crap.
Realistically, it's not that hard to write decent unix software. You know you can't run as root because nobody does. You know you can't update system files except via the installer, and you know they'll have libX11 installed and the API is pretty fixed. After that, you've got a subset of options (GTK or another widget set) and possibly some kernel glue (2.6+ only) but mostly you just use libUSB for consumer-grade hardware.
Your second paragraph sounds like someone who hasn't actually required technical support for a windows based product, ever. When your target market can't tell the difference between their display and their computer, (and all the assorted other horror stories, cupholders and blackouts), it doesn't take very skilled support to explain to them how to plug in the cable and click Start/My Computer/Widget Installer CD. This means that anyone who actually has a PROBLEM with it is told the three Rs of tech support: restart, reboot, reinstall. Yay, helpful.
I'm guessing your dayjob consists of "Would you like fries with that?"
"the dumbass" probably wanted people to be able to reach him in an emergency, which is why he took a cellular phone with him. Turning it off/leaving it at home rather defeats the purpose, doesn't it?
I mean, call me crazy, but that's what I do with my phone while I'm out of the country, I just refrain from making calls on it unless they're important. Then again, I guess I'm "the dumbass" for expecting slashdot readers to have crawled out of their mother's basement and actually gotten out of the country themselves.
That's what I was asking, thanks. I missed that it hadn't added any new SSE registers. Don't be so quick on the "No such thing as a SSE5 process" though - there IS such a thing as a FPU process, because of an ancient design decision from intel that had the FPU as a coprocessor. That's stuck with us right to the point of 64bit processors - and they still have to emulate it in 32bit mode.
So, where's the analysis by people who write optimized media encoders/decoders? How useful are these new instructions, or are they just toys? How well did they handle context switching? What's the CX overhead? Is there a penalty for all processes, or only when you are switching to/from a SSE5 process? Will this be safely usable under all operating systems, or will they need a patch?
Anyway, your posts are full of words and hyperbole, but I'll try your test if I can actually get it for linux. I'm not going to pay for it though, and it appears to be free for .gov use only?
Sure, if you get root on a linux box, you can load a module to grab the syscall handler and rewrite the ones you want. It's not as trivial on linux - the syscall table has no hooks for being tampered with at all, and has no fixed position for binary compatability. Still, it's been done. Adore traps sys_readdir, sys_open and a few others to hide processes and directories from view. And of course, it only takes one person to write the initial code - the armies of script kiddies will happily spread it without understanding what it does.
Other options are using the *trace(2) functions to hook into all running processes and trap/modify their syscall use, or simply replace ld.so with one that silently adds libtrojan.so to every program. The trace functions can be raced against (fork/clone followed immedately by the syscall you want to use. The latter can be defeated by directly using sysenter/int80 via inline asm.
Actually, "rootkit" told me all I ever needed to know about their "security". It's nothing but a USB image aqusition device and PC-side software to handle the matching and authorization. In other words - completely useless from a security standpoint. Think DRM - plug in the USB stick, it copies the decryption software, image matcher AND THE SECRET KEY to your harddrive, then uses a rootkit to "obscure" it.
The trick here is it's cheap as shit. Doing it properly on the keychain costs money - you'll need a decent processor to handle image aquisition and processing. Why bother with that when there's a 2+ghz CPU right next door on the bus? Worse, because they sell this crap as "security devices", they undercut everyone who spends the money to do it right. And of course they lie about how it really works, throwing buzzwords like "biometrically encrypted data storage" out.
tl;dr: snake oil.
How about mere unprofessional behavior? With the number of employers doing background checks for decent jobs, and googles all-seeing eye, it's rather nice to be able to drop back to a nym rather then a name. It also helps when my real name only shows up on technical mailing lists, and I have other nyms for other activities.
Yeah, it's not illegal to talk about dragging an entire cooler of booze to the beach and throwing an all-day party where we're blitzed out of our minds, but that's not how I like to be defined on google. The way pagerank worked, cool parties and other hobbies get linked to a hell of a lot more then archived mailing list posts discussing DMA timeouts on specific PCI busses.
So, should I be required to make a report of everything I've ever done in my entire life and let people pour over it? Or can I take a shit in private sometimes?
That'd be a great defense for Hans Reiser.
Yes it is, and no it isn't. It's really useful for "databases" that are only storing a small amount of data, or mostly static data. While I generally use Postgresql for any serious work, I've made trivial apps out of php+sqlite, because I can send someone the entire application as a single tarball. It's mostly useful to avoid having to write your own storage routines (which will invariably be buggier then something with lots of eyeballs on it).
Um, you care about oracle's storage because it forces you to care. You can't do ANYTHING until it's properly configured. It's a serious bitch that way.
And you've seriously GOT to be trolling about postgresql. "user@host w/password on database A" is seperate from "user@host w/password2 at database B" is a mysqlism that drives me up the wall.
Great, virtualized user domains would be nice, but that's not what MySQL is giving you. It's a best-match regex that may change depending on the whims of RDNS (that you may not even have control over). Thanks, but no thanks.
I'll take "create user 'foo' with password 'bla'; grant all on database 'bar' to 'foo';" any day of the week. Both postgres and MySQL have neigh-identical backup, unless you are incapable of changing 'mysqladmin dump database' to 'pgdump database' in your scripts.
Database workload tuning is a black art on all databases. Some just hide the values in obscure non-populated-by-default registry keys and require paid support services to learn about them.
Uh, yeah. I won't be the first nor last to point this out, but if Cable was truely "half duplex" then you'd have no upstream - period. Because if you'd get up from your computer desk, walk over to your television, and turn it on, guess what? Yeah, all those channels are broadcasting 24/7.
Thicknet/Thinnet lans used a single frequency for 'data transmission', modern cablemodems have vastly different upstream/downstream frequencies, and you are dynamically assigned a 'channel' thats much faster then your cap, shared with multiple other people.
Also, try Skype. Look ma, it works, full duplex. Guess you were wrong.
Don't post on copyright subjects until you can grasp the basic concept that Copyright is NOT Trademark, and doesn't have to be enforced to maintain it.
Copyright is for a period of time, period.
Trademark is forever until failed to be maintained.
And you're an idiot for telling people they HAVE to file copyright infringement suits - even worse, this isn't even copyright infringement, it's a civil issue over which of the (many) royalty schemes allofmp3.com should be paying.
But hey, feel free to think you knew something.
except despite having spent hundreds of dollars at allofmp3, they never did transfer my account (or balance) to the new site. And they don't take any form of payment, either. They should get into linden dollars or e-gold or something goofy and obviously money-laundering like that.
I mean, hell, how the fuck do offshore casinos move cash around?
Congrats, you successfully managed to snip out the part where he pointed out that the VAT (what's being enforced in South Korea) isn't the same as Income Tax like we have in the US. It's a sales tax.
Your post is like where you take a car and drop out the engine and wonder why it doesn't work! There, an analogy!
In other words - it's nearly identical to the situation in the US.
Some branches offer great low mortgage rates - but you have to have great credit score to get it.
Some branches offer horrible non-fixed APR interest-only mortgages - but anyone can get them.
See how your credit rating DOES impact what kind of rate you get?
It takes 5 to 7 years before a company gets credit. It's _TOUGH_. Even with a Dunn & Bradstreet listing, and 4 years of audited recordkeeping, we still couldn't get terms on network gear. Ended up having to run the business on a personal Amex until year 5, when magically "Oh, here, have lines of credit! Good rates!".
That's a small business, with ~150k/year gross. Much like what you're reccommending to this guy. Still, there ARE benefits to incorperating when you don't have a direct 'employer' to deal with. It's worth doing. And besides, if he's pulled in a cool mil in a decade, he can afford someone to handle the paperwork. Unless it's all gone straight up his nose...
Saving up for a house requires exceptional restraint? Like what? Getting a good job but living in a cardboard box? Hint: If you're not living in a house, you're paying rent for an apartment, not saving up.
Your second piece of advice isn't fiscally sound either - your home is at it's most expensive in the first few years you own it. Your 30 year mortgage is a constant dollar cost, while inflation keeps rising. Paying expensive money now vs. cheap money later isn't a good plan, especially at a 5 or 6% rate. Put the money into moderate growth funds, and then if something goes wrong (and it will) you have money to cover it. If you don't? When you miss a monthly payment, you go into forclosure even if you've been overpaying.
Oh yeah, and in the US, mortage interest is tax-deductable, but principal payments are not.
When talking about credit cards - if you're already in debt, yes, pay off as much as you can each month.
Any more bad financial advice you'd like to share?
Actually, there is. You have to send exactly the same content to everyone in order to recieve the deepest discounts. In fact, once you have gotten bulk-identical mail delivery names and addresses for 80% of a zipcode, you can just give them a big pile of your crap for 100% coverage and they'll stick it in everyone's mailbox.
In effect, my 41c/letter is subsidizing junk mail. Strangely, I don't seem to recieve junk-fedex deliveries. Oh, that's because they don't have such an abuse-prone rate system.
Also, since I use e-billpay and correspond via email, the only things I get in the mail are replacement credit cards and some financial statements that they won't deliver as PDF. Oh, and piles and piles and piles of flyers, credit offers, coupon booklets and other junk.
It won't be cheap, though.
Marketers know that user+tag@domain.com = user@domain.com. It's not precicely difficult to undo.
You're mistaking a single person as gatekeeper of the 'official' source as the only model to operate DVCS, though. In fact, even the linux setup defies that. Linus maintains 'linus-brand' linux, but the -mm tree is extremely used, and lots of development goes through there for a shakeout before going to 'official'. So there's seperate gatekeepers.
There's also the possibility of running the 'official' repository on a server that multiple release managers have access to.
Really, DVCS can handle any situation as well as (or generally better) then CVCS can. Want every developer's stuff backed up? Have a central repo that pulls every developer repo as a branch, and back it up. His PC dies, he can just re-grab his branch from the repo. No merges required, and shared history is preserved. (This would require a push-pull mode in some network setups, such as the case of a telecommuting worker behind NAT - the central repo can't reach him directly for a pull)
... and the non-networked medical instruments had a functioning development suite on them? Bzzzt. You develop on a networked PC and load the software/firmware to the instruments via whatever means. If the medical instruments have a PC controlling them, you throw a goddamned nic card in your development box.
Or are technicians supposed to say "whoops, I can fix that glitch!" as they're examining someone, drop out of the exam, fire up the IDE, fix the bug, re-start the program and complete examining the patient?
Yeah, didn't think so.
Actually, no, and for the same reasons file locking is an abysmally bad idea.
If you have to make a core API change, it's going to suck. Period. But if you lock the API that's used through the entire code (think resource allocation, for an example) that means nobody can work on anything, and all existing outstanding work has to be merged and stopped before you can even start to make the change. Refactoring is difficult enough as it is, you can't make the entire project grind to a halt over it.