So the question remains: Why should I waste tens of hours on writing and debugging something in C to do what can be coded in a few minutes with perl, and that will probably only be used once or twice?
If it will only be used a handful of times then, by all means, code it in perl in 5 minutes and be done with it. If it's something that the user will use over and over again, day in, day out (the windowing system, for example) then you should spend your time making it efficient:)
Languages like Perl have their place. I code in Python some of the time for small one-off bits and pieces and scripts that aren't performance critical (our build scripts spend more time waiting for the compilers to finish on a 16 processor machine, so the overhead of Python is acceptable). I certainly wouldn't write the whole compiler in Python though!
I'd like to point out another related issue. Inefficiency adherents such as Java lovers always talk about how hardware improvements make the inefficiency go away.
Not a dig at the OP rather than an observation:
Faster hardware will not make the inefficiency go away - it's still the same amount inefficient. If something wastes 30% of its time wanking on, it doesn't matter how fast you make it go, it's still wasting 30% of its time wanking on.
Faster hardware just makes minor inefficiencies less noticible, so programmers add more minor inefficiencies and apply the same "but faster hardware will make it ok" attitude instead of fixing the real problem!
You forgot to mention that keeping an 80x25 column text display updated only required moving around a maximum of 4000 bytes at a time (80x25 = 2000 bytes for the monochrome text buffer + another 80x25 colour buffer for 16 background and 16 foreground colours per text block), and that a sub 10MHz CPU would certainly struggle to animate that at 30 frames per second or higher.
I grew up in the days if Amiga - and they certainly didn't have an 80x25 text console... so your analogy is fundamentally flawed. All of my favourite Amiga software was lightweight, efficient and responsive (except for the rey tracing engines I used, but that did _actually_ have to do some serious calculation). In fact, my Amiga ticked along on a 800x600 screen quite nicely. Your analogy is also flawed becase not all of the screen is updated at any one time; only the parts that have changed. Oh, and 4000 bytes x 30 FPS is 120kb;
My Commodore 64 has a 700kHz processor in it and it can certainly animate the full screen at 25 frames per second albeit at a slightly reduced resolution. My Atari 2600 has a CPU of about the same speed and it was capable of keeping up with 25 frames per second (again, lower resolution) and still running the game engines just fine. Your argument doesn't hold water pal!
My 1920x1200 colour display requires 55296000 bytes (1920x1200*24), which is 13,824 times as much as that 80x25 text display.
Now despite my 2GHz CPU only being 200 times faster than the hypothetical 10Mhz CPU, it doesn't struggle at all - not only can it animate the whole screen at 60FPS or more, but it can also calculate positions for thousands of objects at the same time.
What exactly do you do on your 1900x1200 colour display? First, 1900x1200x24 bits is actually 1900x1200x3 bytes (6,840,000 bytes). That is a far cry from your piss-ant 55,296,000 bytes (55MB).
Given your eagerness to quote numbers that are practically meaningless to the point and blatantly inaccurate, and your "calculation of positions of thousands of objects" I suggest you are playing games on Windows.
Again, your numbers are flawed, because my old 80x25 text mode display is still drawn from individual pixels. Those pixels still have to be updated individually by the CPU (in the 10MHz days it was often the main CPU that drew every little dot). 80x25 is drawn on a 640x480 in 16 glorious colours. Now, by your argument, 640x480x2 (your 2 image argument from above) = 614,400 bytes. Therefore, your 1900x1200 screen only requires 12x as many bytes to move about but I really dont' care about those numbers because most of the work is done in the GPU now; not the main CPU.
There are certainly some areas in which the software doesn't seem to have sped up in proportion to the processor speed - but my guess is thats mostly because they ceased being CPU bound years ago, not because the CPU is flat out wasting cycles trying to do the job.
That doesn't mean it's not the programmer's fault - but if it is, it's because they decided to block the interface while they waited for a DNS query (or similar bad design decision), not because they used a pointer offset instead of an array dereference, or vice versa.
Software bloat is because programmers can get lazy when the CPU gets fast (the old "oh there'll be better CPUs by the time we release it" excuse). Looking at the power consumption figures for a modern Pentium4 CPU and figuring out that it wastes billions and billions of cycles a day doing work that it should never have needed to do is scary. If you average it out over all the PCs running in the world the amount of energy turned into heat because of sloppy programming alone would be enormous.
What about all the wasted hours waiting for the computer to do something because of some sloppy programmer being willing to waste a few million CPU cycles here and there? It doesn't seem like much at the micro level, but think about it across all the processes that run on your machine of a day and the few
for single cycles on a machine that probably bleeds off a million cycles every time you raise a window
Computers have become more powerful and programmers have become more lazy. It's not strictly true because instead of focussing a lot of time writing efficient code programmers are now focussing a lot of time writing a lot of code to fill bigger machines. That million cycles is wasted doing crap and probably half of doesn't need to be done anyway...
I can still remember the days when machines ran in the sub 10MHz range (yes, 10MHz is 400x slower than today's 4GHz). Software was generally responsive, functional and minimal. Adding a zillion features and eye candy was not considered necessary. Programs were easy to use and intuitive, and did I mention functional and minimal? In the days where "nobody will ever need more than 640k" software was designed and optimised to be small and chew up few cycles.
Now, with RAM and Gigahertz available for next to nix software has just bloated out. It's nice to see a programmer thinking about efficiency/size even if it is purely academic. We should be encouraging that; I know I'd like my applications to work faster and carry less crap than they do currently.
Just because words sound the same doesn't mean you can switch them.
Kind of like you can't just switch your Muzak player to play regular CDs. Just because it's possible to switch it doesn't mean that you're allowed to. Just get some really really good closed-back headphones (www.sennheizer.com) and a ($PREFERED_RELIGIOUS_ICON help all us old farts) portable CD player!
One of the things that always strikes me about a lot of USB hard disk enclosures is how tiny they are and how hot disks run when in them. Try and get one with a small fan in the back. It will move a bit of air and help your disk to last longer. If you can't get one with a fan and portability isn't an issue, try and get a 5.25" one and a couple of mounting brackets for a 3.5" disk. The extra room will mean the disk runs slightly cooler. The 5.25" ones tend to have removable front panels too, so you can stick in a grille with a filter instead of a blank panel to improve ventilation.
If you're using it with your Mac Mini, I'd suggest that having another brick to plug into the wall isn't as big of an issue than if you were using it with your Powerbook or something. Bus powered enclusures won't power a lot of bigger disks - the USB spec doesn't provide for more than about 2.5W on the whole bus, and you lose a bit in hubs and controllers as well so there's not much left to power the disk. FireWire can provide a bit more power, and I've seen bus-powered FireWire enclosures that work quite well (if you have the larger, powered, FireWire socket on the Mac Mini rather than the mini, non-powered, one).
Make sure the enclosure is USB2 capable, and some come with FireWire as well. The dual support ones (in my experience) are more reliable and better built. FireWire is reportedly a bit faster than USB2 for sustained transfer rates, but I have never been able to demonstrate that.
Don't buy a bay with a disk in it. You pay a fortune for them compared to buying a good bay and a disk separately. Seagate and Maxtor both have them. Sure, they work and are good for people who can't use a screwdriver but you pay a premium for some guy in Taiwan to use his screwdriver instead.
Re:Easy counter measures, not worth killing whales
on
Sonic Torpedo Defense
·
· Score: 1
If so, what is to stop someone from figuring out the sound wavess used for this, and configuring the torpedo's sonar system to use an inverse sound to cancel the effect?
Apart from the fact that to cancel out a wave you need a wave of exactly the same intensity and 180 deg out of phase.... so you'd have to measure the incoming sound, process it, invert it and output it at the same (or high enough) intensity.
I don't think a pissy little torpedo is going to be able to carry around enough sound transducers to make a disaster area concert look like the parade of pink faries...
Well, I never said you didn't have a point there. Just that it's not the programmers who are to blame for it.
I'll take your point there; I won't (and I don't believe I originally did) blame the programmers directly.
The management are to blame for thinking in pure dollar terms. This is yet another thing that rings true of the fundamentalist capatilism that America is trying to spread to the world (fuck this democracy shit, they just don't want communism becase then the people at the top of the US wouldn't be rich or "superior" anymore).
Game companies (like the music and movie industries) are spending more and more time inventing creative ways to deter, but never stop, people from copying their software. If they invested all of that effort into innovation, simplified the games back to basics to cut on development costs and made their games cheaper people might start buying them again. I don't pay for games anymore because they're all the same as games I've already paid for with the exception of a different map or a different actor doing the voiceover or some crap like that.
It's just like the RIAA and MPAA; they make crap, overpriced shit based around the "how to make infinite profits off of morons" formula and then wonder why people don't pay for the junk they're selling. Of course, then they claim that one song is worth $15,000 when they take you to court for not paying.
Enough is enough people. Stop buying games. Boycott the industry (and while we're at it, stop paying for music and movies) to stop these assholes treating us like we're stupid and effectively stealing from us!!!!!!
So the short story is that the publisher tries to minimize their risks. That tends to mean making more of whatever sold well last year.
This just further proves my point - they release the same old crap over and over again.
As I said, do they really think I am that stupid? You occasionally see some real gems of games but most is just shit. The days when innovation and good ideas ruled the (most) industries are long gone. The dot bomb era is the reason for it becasue every moron with a computer wanted on it with a me-too mentality.
I can't write software without getting sued for patent violation. I can't play games because they're the same old boring shit over and over again. I can't look at pr0n because I've seen it all.
and I have noticed that the more realistic games look the more generic they look
I have noticed that the more realistic games look, the more the same as other crap games they tend to become. Game programmers must think we're really really stupid. They're repackaging the same old shit week after week and adding "better" graphics (where better is subjective).
Their push for "better" graphics means that those of us who cannot afford to or don't want to upgrade our PCs to the latest and greatest (WTF? I just want to look at pr0n, download my email and compile a few small applications; my current PC is already overkill for that) can't play the games in all their glory anyways.
I think new laws should be passed to force game makers to either:
* Innovate and make something new
or
* Fuck off and die
Rather than wasting our time with remakes and rehashes of old, tired themes with nothing better than "photorealism" to add.
WindRiver still sell their ScopeTools (MemScope, ProfileScope, StethoScope) that rely on a proprietary backend being loaded onto the Linux box. I, personally, don't think the tools are worth the price of the plastic in the CDs they come on. Others here seem to like them.
You also get WindRiver Workbench (an Eclipse plugin). It's cute but there are better editors and debuggers out there if you must pay for them. I prefer Emacs for editing and good old-fashioned Makefiles for building. I much prefer DDD (data display debugger) which is based on GDB for debugging.
Their "system Viewer" (aka WindView) is nothing more than the free (speech/beer) Linux Trace Tool (LTT) from http://www.opersys.com/LTT/.
WindRiver also have a very restrictive license that tries to claim you're not allowed to redistribute the source code to any of the GPL tools. I haven't read it in full but that is the general gist of it.
Management types seem to love them because they have "support". Support is fine, when support can help. I don't see how paying approx $50k US/year for "support" works when I have solved all of the problems we find with the Linux distributions we have used. Speaking of problems, the only ones with real problems were the expensive "commercial" ones.
MontaVista are probably the most trimmed down and GPL friendly of the embedded Linux crowds. They give out their distro on CD and because It's GPL'd they can't stop you using it. Their supplied "tools" are based on OSS tools like GDB. We weren't allowed to even see the tools before we paid for them. They claim that there are screenshots available; like that really helps decide if it's good!
BlueCat won't work unless your development host is RedShit9. They use an encrypted ISO image that they mount via loopback to install. Of course, they were too dumb to use a real encryption and, instead, used XOR so the password is visible just by looking into the encrypted file with Less. You can then xor the file agains the password to get an unencrypted ISO that you can mount and install. Their distro looks like it was based on Scratchbox (see below). Their "tools" are also based on OSS tools. They let us see the tool at a demo by the sales guy, but they wouldn't let us actually get hands-on to see what the tool could really do.
BuildRoot is GPL and free (speech and beer). It comes from the uClibc group and makes a uClibc-based distro. It works beautifully and has a wonderful mailing list/support. It is a piece of cake to add packages to it and get it to build them for you. I added some very complex pieces of software with almost zero effort apart from making a makefile to do various parts of the build (download the tarball, uncompress it, untar it, configure it, make it, install it, clean it). All you have to do is put in the right commands to do each of those steps under the right targets. You can have your package appear in their make menuconfig menu and it has a cool dependency arrangement so it's easy to remove packages without breaking anything.
Scratchbox is similar, but it attempts to create an entire hidden build environment inside your host that you chroot into to use. You can build/install everything inside of it without really worrying about cross compile issues (you're in a chroot so the compiler/headers/libs are defaulted to cross compile correctly). It's also free as in beer. Scratchbox can build anything, including Linux, BSD, glibc, uclibc on i386, arm, ppc, etc. It doesn't care as long as you can build a toolchain to put inside of it before you start building packages.
There are also umpteen other tiny Linux builds to do different things available and they are all free. There are routers, firewalls, mail servers, single disk console-based things, two disk X-things with web browsers, etc.
These companies make a profit by relying on the ignorance of management-types that want "support" and think that paying for something is good business sense because it must be good. This is the problem you get when you let accountants run businesses instead of engineers! If you want support, pay for it, otherwise get buildroot or Scratchbox and have quality.
Not quite the worst Slashdot story ever, but this is right up there with Zonk's "performing cunnilingus on a hardwood floor."
You're just jealous 'cos nobody has ever let you perform cunnilingus on them; let-alone on a hardwood floor;)
Comic Book Guy: "Worst Slashdot Article, Ever".
Windows and WEP just works. If you can't make it work you're a tool. Windows and WPA doesn't just work (if you use AES encryption you have to reconfigure your access point to use TKIP, then connect to it, then reconfigure the accesspoint back to AES to get it).
So, this is the way forward; copy a known flakey product that hides all of the important features you would actually need but makes up for it with 10^27 useless ones that you really won't ever need.
Gmail is innovative and tries to solve the age-old problem of how to categorise mail effectively. It is a solution that meets the needs of the countless millions that decided to sign up and use it daily. It is not the solution for everyone.
This can be seen in the way there exists many different mail clients for the desktop. They all look and feel slightly different. Some go out on a limb and try to innovate while others follow the mail client paradigm.
Choice is good, and saying one thing is better than another purely on the basis of the user interface is purely subjective.
Take, for example, two things; Thing1 and Thing2 (aka Second Thing, but prefers to be called Superthing or somesuch). Thing1 and Thing2 make widgets. Thing1 has a handle to pull and Superthing has a simple pushbutton. Thing1 is painted red and Superthing is painted yellow.
Both Thing1 and Thing2 make exactly the same widget at about the same rate, all things considered. It is not possible to tell the difference between a Thing1 and Thing2 widget.
Some fucktard pensioner (who's spent too much of his retirement playing the one armed bandits) decided to write a review on Thing1 and Superthing. Fucktard tells the world that Thing1 is better because it follows the "standard" widget-arm interface and Second Thing sucks because users don't understand that buttons are used to make widgets too. Fucktard also says that he doesn't like the colour yellow because it detracts from the widget making quality.
Fucktard is wrong because some users prefer the button paradigm - it's easier for them becuse they have bad joints or poor muscle strength and are unable to operate the lever/arm machine. Some users prefer red, some yellow. They are now free to choose the combination of arm/button/red/yellow that suits them best. Widgets will still get made regardless.
Both Thing1 and Thing2 do not suck. The both perform the widget-making task equally well, and the existance of both methods of widget making allows a more varied array of users to make widgets instead of just Fucktard and his old slots skills.
This was probably a little bit of a stupid example, but I think it proves my point. Fucktards are spending too much time worrying about what everyone else is using (because it is different and scary more often than not) and not enough time worrying about real important things...
...like the summary execution of all staff involved with the inspection and enforcement of parking fines; parking inspectors are the lowest of human beings. They are too dumb to rise to the top by skill, but want to sit at the top and enforce power over people. They are even lower than telephone sanitizers and management consultants.
If you've been using it for 3 years, and it's directly (or obviously) derivable from your business name then it doesn't need to be registered as a trademark to get the same protections under the law.
A trademark can be either registered with the PTO or it can be a commonly used word, phrase or graphic that people commonly associate with your business. Your domain name, being in use for over 3 years, would be commonly associated with your business by people who know of or use it.
Send the new owner a harsh letter demanding some untold number of millions of $$$ for illegal use of your trademark.
You'll win if you get to court. If it was sold off within less than 8 hours and the new owner is only displaying ads on it, it sounds like they were just waiting for it to expire, or did actually pre-buy it while you were still using it...
I've heard of several issues with Gentoo kernels on Multi-processor systems.
I just remembered the exact problems I'd heard of, and it wasn't Gentoo-specific (but it only appeared on Gentoo for some strange reason)... It was a CPU freq scaling thing with AMD64 CPUS. Apparantly the latest driver is broken and when the frequency scales down in one CPU the kernel detects a loss of sync and panics, instead of realising that the CPU frequency is scaling and compensating for it.
It's fixed by disabling CPU frequency scaling. Apparantly AMD are working on a PowerNow patch for it, but that is just hear-say AFIK.
I've heard of several issues with Gentoo kernels on Multi-processor systems.
I run a 2.6.12 on my desktop, and we are playing with 2.6.12 in a high-load embedded system. They both seem to work well enough here. I've only ever seen kernel panics when my network switch fails (damn dicky power connector; been meaning to replace) and the NFS-mounted root on the embedded box goes away.
I recently upgraded my laptop to 2.6.13 and it brought all manner of problems (wireless didn't work anymore. Sound problems that were fixed in 2.6.12 reappeared, etc). I think most of my problems are with the IPW2200 driver modules I have loaded, so I just rolled back to 2.6.12 where it all works well.
Stick with 2.6.12 for now if you're scared of problems. I can safely say that it is pretty damned reliable.
n answer to your question, I was going by the fact that this copyright notice....
True, but any GPL'd work includes a copyright notice from the author as well. Having a copyright does not mean you are not allowed to change and redistribute; and the GNU/FSF is about freedom:)
Can text that is part of court filings be copyrighted?
Having the text listed as an exhibit does not change the fact that the FSF holds the copyright on the GPL. There are various parts of SCO code that were called as exhibits that are not available publicly through the court. SCO pulled some sort of trade secret garbage to avoid having to disclose them.
That was the whole reason that the SCO/Linux thing became such a big affair; SCO were making bold claims that Linux stole their code, but they wouldn't produce the original code and changeset logs to back up their claims.
Gold. Platinum. Rhodium. Diamonds and other "precious" stones. etc., etc., etc..
But everyone wants to use fossil fuels in their cars and boats. They are used in the making of plastics, the refining of all of the precious metals above (for power, heat, etc), so the world grinds to a slow painful stop without them.
All of those other things - gold, platinum, etc can be recycled from whatever you turn them into. Once you burn the fossil fuels, they are gone forver.
You cannot modify the text of the GPL document. The GPL license document text is a copyright work, with a copyright notice on it. The GPL license text is Copyright by the FSF.
Where did you come up with this?
You are allowed to use all of the GPL terms without the preamble, see here.
But many create cars, sell them to reckless teenagers and alcoholics.
Those cars are used to kill people (often inadvertantly, often through inexperience, speed, etc, and sometimes maliciously), yet car manufacturers haven't been brought into the courts and tried for their crimes...
Many create knives. Some knives are designed to be good at killing things; hunting knives. When was the last time a knife maker was in court because his product was used to kill a person?
It's kind of akin to DVD John; he did nothing more than make some software (in a country other than USA). The MPAA and others decided it was a good idea to try and punish him for breaking a USA law, on the grounds that his software would be used in a way that wasn't legal in the US... and he wasn't even fucking well in the USA. What he did was perfectly legal where he came from!
Worm writers should be prosecuted. Their software is designed with the sole purpose of seeking out vulnerable computers and attacking them. This "loveware" or whatever it was was less malicious than that, and required that one party make a conscious decision to install it on someone's computer. The guy who made it can't be responsible for the actions of the person paying him.
I hate spyware. I hate spam. This guy should be hung on the grounds that he created spyware and spam, but blaming him for the actions of others that were using his software is ludicrous.
Grrrr!!! One standard for one thing, one for another.
No s'pose about it. If Microsoft did it you know what the data charges would be like... unaffordable to all but those who earn a larger salary than the GNP a small island nation. Not to mention that they'd try to listen to every single byte transmitted and be the GAC's that they are about it.
If Google do it, at least their "don't be evil" policy would make me feel a little better about using it.
Slashdot Mirror
So the question remains: Why should I waste tens of hours on writing and debugging something in C to do what can be coded in a few minutes with perl, and that will probably only be used once or twice?
If it will only be used a handful of times then, by all means, code it in perl in 5 minutes and be done with it. If it's something that the user will use over and over again, day in, day out (the windowing system, for example) then you should spend your time making it efficient :)
Languages like Perl have their place. I code in Python some of the time for small one-off bits and pieces and scripts that aren't performance critical (our build scripts spend more time waiting for the compilers to finish on a 16 processor machine, so the overhead of Python is acceptable). I certainly wouldn't write the whole compiler in Python though!
I'd like to point out another related issue. Inefficiency adherents such as Java lovers always talk about how hardware improvements make the inefficiency go away.
Not a dig at the OP rather than an observation:
Faster hardware will not make the inefficiency go away - it's still the same amount inefficient. If something wastes 30% of its time wanking on, it doesn't matter how fast you make it go, it's still wasting 30% of its time wanking on.
Faster hardware just makes minor inefficiencies less noticible, so programmers add more minor inefficiencies and apply the same "but faster hardware will make it ok" attitude instead of fixing the real problem!
You forgot to mention that keeping an 80x25 column text display updated only required moving around a maximum of 4000 bytes at a time (80x25 = 2000 bytes for the monochrome text buffer + another 80x25 colour buffer for 16 background and 16 foreground colours per text block), and that a sub 10MHz CPU would certainly struggle to animate that at 30 frames per second or higher.
I grew up in the days if Amiga - and they certainly didn't have an 80x25 text console... so your analogy is fundamentally flawed. All of my favourite Amiga software was lightweight, efficient and responsive (except for the rey tracing engines I used, but that did _actually_ have to do some serious calculation). In fact, my Amiga ticked along on a 800x600 screen quite nicely. Your analogy is also flawed becase not all of the screen is updated at any one time; only the parts that have changed. Oh, and 4000 bytes x 30 FPS is 120kb;
My Commodore 64 has a 700kHz processor in it and it can certainly animate the full screen at 25 frames per second albeit at a slightly reduced resolution. My Atari 2600 has a CPU of about the same speed and it was capable of keeping up with 25 frames per second (again, lower resolution) and still running the game engines just fine. Your argument doesn't hold water pal!
My 1920x1200 colour display requires 55296000 bytes (1920x1200*24), which is 13,824 times as much as that 80x25 text display. Now despite my 2GHz CPU only being 200 times faster than the hypothetical 10Mhz CPU, it doesn't struggle at all - not only can it animate the whole screen at 60FPS or more, but it can also calculate positions for thousands of objects at the same time.
What exactly do you do on your 1900x1200 colour display? First, 1900x1200x24 bits is actually 1900x1200x3 bytes (6,840,000 bytes). That is a far cry from your piss-ant 55,296,000 bytes (55MB).
Given your eagerness to quote numbers that are practically meaningless to the point and blatantly inaccurate, and your "calculation of positions of thousands of objects" I suggest you are playing games on Windows.
Again, your numbers are flawed, because my old 80x25 text mode display is still drawn from individual pixels. Those pixels still have to be updated individually by the CPU (in the 10MHz days it was often the main CPU that drew every little dot). 80x25 is drawn on a 640x480 in 16 glorious colours. Now, by your argument, 640x480x2 (your 2 image argument from above) = 614,400 bytes. Therefore, your 1900x1200 screen only requires 12x as many bytes to move about but I really dont' care about those numbers because most of the work is done in the GPU now; not the main CPU.
There are certainly some areas in which the software doesn't seem to have sped up in proportion to the processor speed - but my guess is thats mostly because they ceased being CPU bound years ago, not because the CPU is flat out wasting cycles trying to do the job. That doesn't mean it's not the programmer's fault - but if it is, it's because they decided to block the interface while they waited for a DNS query (or similar bad design decision), not because they used a pointer offset instead of an array dereference, or vice versa.
Software bloat is because programmers can get lazy when the CPU gets fast (the old "oh there'll be better CPUs by the time we release it" excuse). Looking at the power consumption figures for a modern Pentium4 CPU and figuring out that it wastes billions and billions of cycles a day doing work that it should never have needed to do is scary. If you average it out over all the PCs running in the world the amount of energy turned into heat because of sloppy programming alone would be enormous.
What about all the wasted hours waiting for the computer to do something because of some sloppy programmer being willing to waste a few million CPU cycles here and there? It doesn't seem like much at the micro level, but think about it across all the processes that run on your machine of a day and the few
then don't wank around optimizing
Dude, best use of the word wank. Ever!
for single cycles on a machine that probably bleeds off a million cycles every time you raise a window
Computers have become more powerful and programmers have become more lazy. It's not strictly true because instead of focussing a lot of time writing efficient code programmers are now focussing a lot of time writing a lot of code to fill bigger machines. That million cycles is wasted doing crap and probably half of doesn't need to be done anyway...
I can still remember the days when machines ran in the sub 10MHz range (yes, 10MHz is 400x slower than today's 4GHz). Software was generally responsive, functional and minimal. Adding a zillion features and eye candy was not considered necessary. Programs were easy to use and intuitive, and did I mention functional and minimal? In the days where "nobody will ever need more than 640k" software was designed and optimised to be small and chew up few cycles.
Now, with RAM and Gigahertz available for next to nix software has just bloated out. It's nice to see a programmer thinking about efficiency/size even if it is purely academic. We should be encouraging that; I know I'd like my applications to work faster and carry less crap than they do currently.
*grumbles* It's too late in the day for this..
I meant to say http://www.sennheiserusa.com/
I pressed the Z accidently before, and also should have checked the URL before I hit submit...
Just because words sound the same doesn't mean you can switch them.
Kind of like you can't just switch your Muzak player to play regular CDs. Just because it's possible to switch it doesn't mean that you're allowed to. Just get some really really good closed-back headphones (www.sennheizer.com) and a ($PREFERED_RELIGIOUS_ICON help all us old farts) portable CD player!
:)
One of the things that always strikes me about a lot of USB hard disk enclosures is how tiny they are and how hot disks run when in them. Try and get one with a small fan in the back. It will move a bit of air and help your disk to last longer. If you can't get one with a fan and portability isn't an issue, try and get a 5.25" one and a couple of mounting brackets for a 3.5" disk. The extra room will mean the disk runs slightly cooler. The 5.25" ones tend to have removable front panels too, so you can stick in a grille with a filter instead of a blank panel to improve ventilation.
If you're using it with your Mac Mini, I'd suggest that having another brick to plug into the wall isn't as big of an issue than if you were using it with your Powerbook or something. Bus powered enclusures won't power a lot of bigger disks - the USB spec doesn't provide for more than about 2.5W on the whole bus, and you lose a bit in hubs and controllers as well so there's not much left to power the disk. FireWire can provide a bit more power, and I've seen bus-powered FireWire enclosures that work quite well (if you have the larger, powered, FireWire socket on the Mac Mini rather than the mini, non-powered, one).
Make sure the enclosure is USB2 capable, and some come with FireWire as well. The dual support ones (in my experience) are more reliable and better built. FireWire is reportedly a bit faster than USB2 for sustained transfer rates, but I have never been able to demonstrate that.
Don't buy a bay with a disk in it. You pay a fortune for them compared to buying a good bay and a disk separately. Seagate and Maxtor both have them. Sure, they work and are good for people who can't use a screwdriver but you pay a premium for some guy in Taiwan to use his screwdriver instead.
If so, what is to stop someone from figuring out the sound wavess used for this, and configuring the torpedo's sonar system to use an inverse sound to cancel the effect?
Apart from the fact that to cancel out a wave you need a wave of exactly the same intensity and 180 deg out of phase.... so you'd have to measure the incoming sound, process it, invert it and output it at the same (or high enough) intensity.
I don't think a pissy little torpedo is going to be able to carry around enough sound transducers to make a disaster area concert look like the parade of pink faries...
Well, I never said you didn't have a point there. Just that it's not the programmers who are to blame for it.
I'll take your point there; I won't (and I don't believe I originally did) blame the programmers directly.
The management are to blame for thinking in pure dollar terms. This is yet another thing that rings true of the fundamentalist capatilism that America is trying to spread to the world (fuck this democracy shit, they just don't want communism becase then the people at the top of the US wouldn't be rich or "superior" anymore).
Game companies (like the music and movie industries) are spending more and more time inventing creative ways to deter, but never stop, people from copying their software. If they invested all of that effort into innovation, simplified the games back to basics to cut on development costs and made their games cheaper people might start buying them again. I don't pay for games anymore because they're all the same as games I've already paid for with the exception of a different map or a different actor doing the voiceover or some crap like that.
It's just like the RIAA and MPAA; they make crap, overpriced shit based around the "how to make infinite profits off of morons" formula and then wonder why people don't pay for the junk they're selling. Of course, then they claim that one song is worth $15,000 when they take you to court for not paying.
Enough is enough people. Stop buying games. Boycott the industry (and while we're at it, stop paying for music and movies) to stop these assholes treating us like we're stupid and effectively stealing from us!!!!!!
I'm no fan of Java, but at least it gives your program a chance to recover (exceptions and stuff).
And C++ doesn't have exceptions and stuff? Java is just as insecure as C/C++. It's just insecure in different ways.
As others have said, most bugs and security flaws come from lazy programmers not doing the right thing.
I'd also like to see you run Java in a lightweight embedded DSP situation!
So the short story is that the publisher tries to minimize their risks. That tends to mean making more of whatever sold well last year.
This just further proves my point - they release the same old crap over and over again.
As I said, do they really think I am that stupid? You occasionally see some real gems of games but most is just shit. The days when innovation and good ideas ruled the (most) industries are long gone. The dot bomb era is the reason for it becasue every moron with a computer wanted on it with a me-too mentality.
I can't write software without getting sued for patent violation. I can't play games because they're the same old boring shit over and over again. I can't look at pr0n because I've seen it all.
I think I might go outside!
and I have noticed that the more realistic games look the more generic they look
I have noticed that the more realistic games look, the more the same as other crap games they tend to become. Game programmers must think we're really really stupid. They're repackaging the same old shit week after week and adding "better" graphics (where better is subjective).
Their push for "better" graphics means that those of us who cannot afford to or don't want to upgrade our PCs to the latest and greatest (WTF? I just want to look at pr0n, download my email and compile a few small applications; my current PC is already overkill for that) can't play the games in all their glory anyways.
I think new laws should be passed to force game makers to either:
* Innovate and make something new
or
* Fuck off and die
Rather than wasting our time with remakes and rehashes of old, tired themes with nothing better than "photorealism" to add.
WindRiver still sell their ScopeTools (MemScope, ProfileScope, StethoScope) that rely on a proprietary backend being loaded onto the Linux box. I, personally, don't think the tools are worth the price of the plastic in the CDs they come on. Others here seem to like them.
You also get WindRiver Workbench (an Eclipse plugin). It's cute but there are better editors and debuggers out there if you must pay for them. I prefer Emacs for editing and good old-fashioned Makefiles for building. I much prefer DDD (data display debugger) which is based on GDB for debugging.
Their "system Viewer" (aka WindView) is nothing more than the free (speech/beer) Linux Trace Tool (LTT) from http://www.opersys.com/LTT/.
WindRiver also have a very restrictive license that tries to claim you're not allowed to redistribute the source code to any of the GPL tools. I haven't read it in full but that is the general gist of it.
Management types seem to love them because they have "support". Support is fine, when support can help. I don't see how paying approx $50k US/year for "support" works when I have solved all of the problems we find with the Linux distributions we have used. Speaking of problems, the only ones with real problems were the expensive "commercial" ones.
MontaVista are probably the most trimmed down and GPL friendly of the embedded Linux crowds. They give out their distro on CD and because It's GPL'd they can't stop you using it. Their supplied "tools" are based on OSS tools like GDB. We weren't allowed to even see the tools before we paid for them. They claim that there are screenshots available; like that really helps decide if it's good!
BlueCat won't work unless your development host is RedShit9. They use an encrypted ISO image that they mount via loopback to install. Of course, they were too dumb to use a real encryption and, instead, used XOR so the password is visible just by looking into the encrypted file with Less. You can then xor the file agains the password to get an unencrypted ISO that you can mount and install. Their distro looks like it was based on Scratchbox (see below). Their "tools" are also based on OSS tools. They let us see the tool at a demo by the sales guy, but they wouldn't let us actually get hands-on to see what the tool could really do.
BuildRoot is GPL and free (speech and beer). It comes from the uClibc group and makes a uClibc-based distro. It works beautifully and has a wonderful mailing list/support. It is a piece of cake to add packages to it and get it to build them for you. I added some very complex pieces of software with almost zero effort apart from making a makefile to do various parts of the build (download the tarball, uncompress it, untar it, configure it, make it, install it, clean it). All you have to do is put in the right commands to do each of those steps under the right targets. You can have your package appear in their make menuconfig menu and it has a cool dependency arrangement so it's easy to remove packages without breaking anything.
Scratchbox is similar, but it attempts to create an entire hidden build environment inside your host that you chroot into to use. You can build/install everything inside of it without really worrying about cross compile issues (you're in a chroot so the compiler/headers/libs are defaulted to cross compile correctly). It's also free as in beer. Scratchbox can build anything, including Linux, BSD, glibc, uclibc on i386, arm, ppc, etc. It doesn't care as long as you can build a toolchain to put inside of it before you start building packages.
There are also umpteen other tiny Linux builds to do different things available and they are all free. There are routers, firewalls, mail servers, single disk console-based things, two disk X-things with web browsers, etc.
These companies make a profit by relying on the ignorance of management-types that want "support" and think that paying for something is good business sense because it must be good. This is the problem you get when you let accountants run businesses instead of engineers! If you want support, pay for it, otherwise get buildroot or Scratchbox and have quality.
Not quite the worst Slashdot story ever, but this is right up there with Zonk's "performing cunnilingus on a hardwood floor."
You're just jealous 'cos nobody has ever let you perform cunnilingus on them; let-alone on a hardwood floor ;)
Comic Book Guy: "Worst Slashdot Article, Ever".
Windows and WEP just works. If you can't make it work you're a tool. Windows and WPA doesn't just work (if you use AES encryption you have to reconfigure your access point to use TKIP, then connect to it, then reconfigure the accesspoint back to AES to get it).
Yay! Finally some sense in the court system!!!
The new yahoo mail looks like outlook
So, this is the way forward; copy a known flakey product that hides all of the important features you would actually need but makes up for it with 10^27 useless ones that you really won't ever need.
Gmail is innovative and tries to solve the age-old problem of how to categorise mail effectively. It is a solution that meets the needs of the countless millions that decided to sign up and use it daily. It is not the solution for everyone.
This can be seen in the way there exists many different mail clients for the desktop. They all look and feel slightly different. Some go out on a limb and try to innovate while others follow the mail client paradigm.
Choice is good, and saying one thing is better than another purely on the basis of the user interface is purely subjective.
Take, for example, two things; Thing1 and Thing2 (aka Second Thing, but prefers to be called Superthing or somesuch). Thing1 and Thing2 make widgets. Thing1 has a handle to pull and Superthing has a simple pushbutton. Thing1 is painted red and Superthing is painted yellow.
Both Thing1 and Thing2 make exactly the same widget at about the same rate, all things considered. It is not possible to tell the difference between a Thing1 and Thing2 widget.
Some fucktard pensioner (who's spent too much of his retirement playing the one armed bandits) decided to write a review on Thing1 and Superthing. Fucktard tells the world that Thing1 is better because it follows the "standard" widget-arm interface and Second Thing sucks because users don't understand that buttons are used to make widgets too. Fucktard also says that he doesn't like the colour yellow because it detracts from the widget making quality.
Fucktard is wrong because some users prefer the button paradigm - it's easier for them becuse they have bad joints or poor muscle strength and are unable to operate the lever/arm machine. Some users prefer red, some yellow. They are now free to choose the combination of arm/button/red/yellow that suits them best. Widgets will still get made regardless.
Both Thing1 and Thing2 do not suck. The both perform the widget-making task equally well, and the existance of both methods of widget making allows a more varied array of users to make widgets instead of just Fucktard and his old slots skills.
This was probably a little bit of a stupid example, but I think it proves my point. Fucktards are spending too much time worrying about what everyone else is using (because it is different and scary more often than not) and not enough time worrying about real important things...
...like the summary execution of all staff involved with the inspection and enforcement of parking fines; parking inspectors are the lowest of human beings. They are too dumb to rise to the top by skill, but want to sit at the top and enforce power over people. They are even lower than telephone sanitizers and management consultants.
If you own a trademark to the name
If you've been using it for 3 years, and it's directly (or obviously) derivable from your business name then it doesn't need to be registered as a trademark to get the same protections under the law.
A trademark can be either registered with the PTO or it can be a commonly used word, phrase or graphic that people commonly associate with your business. Your domain name, being in use for over 3 years, would be commonly associated with your business by people who know of or use it.
Send the new owner a harsh letter demanding some untold number of millions of $$$ for illegal use of your trademark.
You'll win if you get to court. If it was sold off within less than 8 hours and the new owner is only displaying ads on it, it sounds like they were just waiting for it to expire, or did actually pre-buy it while you were still using it...
That is ILLEGAL! Take it to court!
What tips do you have for surviving (even thriving) in this type of environment?"
Quit - if the machine room leaks and they won't even fix that then you're never going to be able to get them to move to better computing solutions...
I've heard of several issues with Gentoo kernels on Multi-processor systems.
I just remembered the exact problems I'd heard of, and it wasn't Gentoo-specific (but it only appeared on Gentoo for some strange reason)... It was a CPU freq scaling thing with AMD64 CPUS. Apparantly the latest driver is broken and when the frequency scales down in one CPU the kernel detects a loss of sync and panics, instead of realising that the CPU frequency is scaling and compensating for it.
It's fixed by disabling CPU frequency scaling. Apparantly AMD are working on a PowerNow patch for it, but that is just hear-say AFIK.
I've heard of several issues with Gentoo kernels on Multi-processor systems.
I run a 2.6.12 on my desktop, and we are playing with 2.6.12 in a high-load embedded system. They both seem to work well enough here. I've only ever seen kernel panics when my network switch fails (damn dicky power connector; been meaning to replace) and the NFS-mounted root on the embedded box goes away.
I recently upgraded my laptop to 2.6.13 and it brought all manner of problems (wireless didn't work anymore. Sound problems that were fixed in 2.6.12 reappeared, etc). I think most of my problems are with the IPW2200 driver modules I have loaded, so I just rolled back to 2.6.12 where it all works well.
Stick with 2.6.12 for now if you're scared of problems. I can safely say that it is pretty damned reliable.
n answer to your question, I was going by the fact that this copyright notice....
True, but any GPL'd work includes a copyright notice from the author as well. Having a copyright does not mean you are not allowed to change and redistribute; and the GNU/FSF is about freedom :)
Can text that is part of court filings be copyrighted?
Having the text listed as an exhibit does not change the fact that the FSF holds the copyright on the GPL. There are various parts of SCO code that were called as exhibits that are not available publicly through the court. SCO pulled some sort of trade secret garbage to avoid having to disclose them.
That was the whole reason that the SCO/Linux thing became such a big affair; SCO were making bold claims that Linux stole their code, but they wouldn't produce the original code and changeset logs to back up their claims.
Gold. Platinum. Rhodium. Diamonds and other "precious" stones. etc., etc., etc..
But everyone wants to use fossil fuels in their cars and boats. They are used in the making of plastics, the refining of all of the precious metals above (for power, heat, etc), so the world grinds to a slow painful stop without them.
All of those other things - gold, platinum, etc can be recycled from whatever you turn them into. Once you burn the fossil fuels, they are gone forver.
You cannot modify the text of the GPL document. The GPL license document text is a copyright work, with a copyright notice on it. The GPL license text is Copyright by the FSF.
Where did you come up with this?
You are allowed to use all of the GPL terms without the preamble, see here.
But many create cars, sell them to reckless teenagers and alcoholics.
Those cars are used to kill people (often inadvertantly, often through inexperience, speed, etc, and sometimes maliciously), yet car manufacturers haven't been brought into the courts and tried for their crimes...
Many create knives. Some knives are designed to be good at killing things; hunting knives. When was the last time a knife maker was in court because his product was used to kill a person?
It's kind of akin to DVD John; he did nothing more than make some software (in a country other than USA). The MPAA and others decided it was a good idea to try and punish him for breaking a USA law, on the grounds that his software would be used in a way that wasn't legal in the US... and he wasn't even fucking well in the USA. What he did was perfectly legal where he came from!
Worm writers should be prosecuted. Their software is designed with the sole purpose of seeking out vulnerable computers and attacking them. This "loveware" or whatever it was was less malicious than that, and required that one party make a conscious decision to install it on someone's computer. The guy who made it can't be responsible for the actions of the person paying him.
I hate spyware. I hate spam. This guy should be hung on the grounds that he created spyware and spam, but blaming him for the actions of others that were using his software is ludicrous.
Grrrr!!! One standard for one thing, one for another.
I suppose better Google than Microsoft, right...?
No s'pose about it. If Microsoft did it you know what the data charges would be like... unaffordable to all but those who earn a larger salary than the GNP a small island nation. Not to mention that they'd try to listen to every single byte transmitted and be the GAC's that they are about it.
If Google do it, at least their "don't be evil" policy would make me feel a little better about using it.