There's a fundamental problem with something like IRC or a non corporate owned decentralized IM system. When the 15 year old script kiddies flood an IRC server, effectively shutting it down, the person running the server generally doesn't have the resources to do anything about it - they bitch, they moan, they might even call the police, but the police only want to hear how much money you've lost, not that some service you were providing out of the goodness of your heart was no longer working. It would be the same as if you were a bookstore owner that had fifty free books on a table with a sign that said "Free: take one." Do you honestly expect the cops to do anything if one person came by and took them all?
Besides, how can you have granny rely on IRC-like networks to communicate with you if the whole thing is volunteer based and could collapse at any moment due to lack of funding, apathy, or whatnot? And if "everyone" can't rely on it, what good is it?
The decentralized IM system would work if, say, every ISP set up their own server and you could only send thru it if you had a username/password on their system already. Then tracking "spammers" would be easy (call/E-mail the ISP of the misbehaver) and if a particular server goes down it's between that company and its customers, and that's all.
The problem is that many people, for a variety of reasons (geography being one) can't change ISPs, and many ISPs (mine included) did nothing in response to my complaints (because they knew I wasn't going to move).
Or in many cases the spammers are paying the ISPs far more per month than the $19.99 dial up guy who's complaining about spam.
Who do you think they're going to bend over backwards to serve?
Oh sure, I vaguely remember the interview he said it in. Something to the effect of "everyone will see Matrix whether or not Jet Li is in it, so I might as well contribute to my own projects. What would be better, seeing a movie that you were already going to see that has me in it, and that's it, or seeing a movie you were already going to see PLUS a movie with me in it?"
I mean, he wasn't arrogant about it or anything, but matter of fact. Let's face it, although it would have been cool to see Li as Seraph, would people who weren't going to see the movie suddenly wanted to see it because Li was in it? Li realizes that and figures that if a movie's going to be a draw because he's in it it may as well be a project that isn't already highly anticipated.
The only problem I have with the visible patrol car thing is that many times the true idiots who were already under the speed limit will slow down even more. And don't get me started about how many times I've seen a cop doing 52 in the right lane of a 55 zone and there's a cluster of cars behind him too terrified to pass...
But otherwise, I'm down with visible cops keeping everyone honest rather than a hidden cop nailing someone and then everyone else screaming past the pulled over cars doing 15+ over.
Won't be much of a wait, considering it came out the same day the DVD did - I own a copy. Probably one of the last, if not THE last, major movie pressed in that format.
1.4 G's? you get more G's than that accelerating from a red light in a poorly-running car, seriously.
Bzzt. The 2003 Chevy Corvette Z06, one of the fastest production cars currently available, can do 0-to-60 MPH in 3.9 seconds, and that's _only_ 1 G lateral acceleration.
I would agree with you in the "losing it" department, but for me the slide started happening way earlier - although I can't decide if he was "slipping" in Difference Engine (because he was collaborating with Sterling) or Virtual Light. But by Idoru, I had a classic case of "who the fuck cares" by midway thru the book. I haven't touched ATP or Pattern Recognition yet, and if I do I'll probably check them out from the library first - I have better ways to spend my book budget than on authors who seem to be on a downward spiral.
The problem is, the Mozilla people are faw more worried about picking a name that a corporation can come after them for than stepping on other free projects' toes. What IF, after doing all the months of research to find out that Firebird doesn't violate any trademarks, they would have asked the Firebird database team and that team would have said "no"? Would you expect Mozilla to just chuck all that research and come up with another name that also isn't trademark infringing but this time might piss off some guy somewhere who named his MP3 player "Malibu" five years ago? At some point you gotta stop and use common sense. No one is going to confuse these two products. (And yes, you do the research FIRST because not violating _actual_ trademarks are the thing you're trying to avoid here, not _implied_ trademarks.)
Cool. I bet know the kind of store you speak of - Best Buy/Circuit City presentation (lots of gear and speakers in a few rooms), but higher end brands, right? If so, the salesman probably spent less than five minutes dealing with you, if at all, and it's no "loss" for them just because you're a savvy price shopper. My statement about screwing hard working dealers still stands though.
Knowing which brands ban mail order and which don't, I only hope that you are sticking it to a retailer that is truly offering you no service beyond simply selling you the product (like, letting you listen to the equipment in-store for pretty much as long as you'd like or even letting you borrow it over the weekend to try for yourself.) Because dealers that provide that higher level of service don't deserve to be screwed by price shoppers that happily take advantage of the above services so they can confidently pay less at a mail order retailer.
The problem is, most of the brands that are restricted from mail-order aren't the ones that you'd find at your average Best Buy or Circuit City (or Crutchfield's) anyway, which means chances are you're dealing with a small high-end audio/video store, who would only be able to discount product if you were either a well-established customer or were buying a complete system. There's not as much markup in the A/V world as say, clothing. Most lines are MSRPd - max sale price - at 25%-35% over what the dealer gets them for in quantity 1 - which is how most high-end dealers order product.
And many brands, rather than policing stuff like B-stock (cosmetically damaged equipment that the retailer got a screamingly good deal on) being sold as "discounted A-stock" by mail order retailers - and then the customer, when said B-stock arrives, acts pissed at the _manufacturer_ for the damaged goods rather than the unscrupulous retailer - figure it's just easier all around to say "no mail order sales" and have customers personally examine every piece of equipment that they buy, and easier to then reject warranties from products purchased outside regular channels.
If you are truly buying something that is HALF what a dealer wanted to charge you, you are most likely getting a B-stock unit, and don't even think about bitching when it fails earlier than it's supposed to, or has a big gouge in it, or whatever. And don't be surprised if the manufacturer then fails to honor the regular warranty on it. It left their factory as B-stock (lesser warranty at a way cheaper price), and it's not their fault or problem if whoever you bought it from misrepresented it.
Re:Now LinkSys is going to suck as much as Cisco
on
Cisco to Acquire Linksys
·
· Score: 2, Interesting
On the low end, I completely agree. But is anything with a Cisco badge truly priced such that it can be considered "low-end"?
On the high end, I emphatically disagree. Talk to someone who's administering more than one of the platforms you mentioned (3COM, Nortel, Lucent, and Cisco). Ask them which hardware is the most reliable, flexible, configurable (no Windows-only Java programs needed), has the best web site support, and in general, has rarely if ever let them down in a pinch? Their answer will most likely be Cisco.
And also, all you have to do to get routine software upgrades for Cisco products is register for their site, for FREE, whereas 3COM/Lucent/Nortel want you to annually pay for maintenance contracts, and if the problem/bug you're experiencing isn't fixed within that year, well, then, please buy another yearly contract, repeat ad infinitum. 3COM in particular has a history of deliberately screwing customers that had been with them since the beginning, such as promising an eventual fix for a UDP latency bug (a big deal among ISPs at the time because it affected Quake players, for example), but only for "current contract customers" and then about 18 months later, dropping that product (the NetServer) and replacing it with a newer one (the HiPER ARC) that wasn't backwards compatible, rendering all the users that were waiting for the promised fix out in the cold, and thousands of dollars poorer. Many of the people burned by this little stunt switched away from 3COM after this, but others stayed and now couldn't PAY people to take the old NetServer stuff off of their hands - it is utterly useless. Lucent is better at least, because you can still get old Livingston and Ascend firmware updates for free, you just need to pay if you want software for anything more modern. And even though some of their products have been dead for some time (long live the Livingston Portmaster!) at least the legacy stuff is useful in limited capacity. A PM3
is still a great choice if you want to get a little POP going in an area where v.92 isn't much of an issue due to phone line quality anyway.
The tonality is right, but it seems a little on the slow side, tempo-wise. Or maybe because I had a 800XL versus a 400/800 that the sound played a little faster? Anyone care to comment?
Want to do it for free? Here you go. (Note: site seems to be down, no idea why, but I've been there before and the Internet Wayback Machine has copies of the old pages.) GCom and TPS Systems both have products that might do what you need on Linux. There's also the IBM Communications Server but I don't think it's what you're looking for.
Which is probably precisely what Canopy/SCO hopes happens, because otherwise Canopy has a very big, very dead albatross hanging around its neck. Seriously, is SCO actually making money from new product sales anymore?
Yeah, no kidding. I switched more than one client away from UnixWare to Linux back in 1997 and 1998, well before IBM was seriously involved, AND got all the legacy applications running under the SysV emulation, even back then. Some of those systems are still running (they never really needed to be bleeding edge), and their upgrade path has been made an order of magnitude easier (and cheaper) by switching.
SCO was irrelevant five years ago. They figure that the possibility of getting any money from IBM beats killing off what little credibility they might have, because they're sinking anyway.
Not quite true. "Interpreted" in the sense you're using it is not the same as "run time compiled" or "two stage interpreted".
When a Perl program, for example, is run, the code is _completely_ compiled into a bytecode before a single line is even run, and then the bytecode is interpreted (in the above sense) as the program is run. That's why gross syntax errors in Perl or Java are spotted before the program even runs, but logic errors that (say) result in a division by zero error might not crop up until the code is actually encountered. It's also why, even if you get Perl to spit out bytecode, you still NEED the Perl executable to RUN the bytecode - the Perl executable contains the interpreter for the bytecode. It would be another step entirely to convert that bytecode into a completely stand-alone executable - see the "Perl to C" and similar projects that are kicking around the Net.
Now, BASIC, that's an interpreted language (not counting VB and such...I'm talking good old fashioned, IBM ROM BASIC or earlier cousins.) Each line is checked for validity and run, one at a time, as each is encountered. Try typing in a ten line "Hello World" style program in a legacy BASIC sometime, with one of the lines having a syntactical error (like "gosub" to a non-existant line number). You won't usually hear a peep out of the program until it encounters that error halfway into the program, at which point it aborts and prints the error, losing all of your work. Shell languages would count as "interpreted" as well. Each line isn't executed until it is encountered.
And don't use "assembled" and "compiled" synonymously either. One "compiles" one or more C (for example) subroutines into object (.o) files, which are then "assembled" into the final program.
----snip---- Question 22: Will model rocket engines like those sold by Estes require ATF permits to buy?
Answer: No. The ATF plans to exempt all solid rocket motors with less than 62.5 grams of propellant. ----snip----
The PROBLEM is that UPS, for now, is not shipping the engines, because apparently it's too much of a grey area for them; they'd rather just ban them than read the law and _realize_ that these engines don't apply. Fine. There aren't other shipping companies in the US? UPS isn't even the major international shipper, methinks. That honor would be either FedEx or DHL. Regardless, these engines can still be had.
So OK, as a result of the new law, after May 31, 2004, you can't teach Timmy how to build his own APCP (Ammonium Perchlorate Composite Propellant) rocket motor without having an ATF permit (or more specifically, you COULD teach him how to build one, but you'd need the permit to launch it anywhere but in your backyard.) This does suck, I agree. But I can still go out to a farm field and cook off a paper tube rocket with a B6-4 engine and show anyone who cares to watch how everything works, just like I've been doing on occasion since I was seven.
Walk into your local hobby shop (the people bitching about all this DO actually buy these engines, right?) and tell him that you'll support him with your business, and if he needs to bump the prices of engines up a bit to cover having them shipped to him by FedEx or whatever until all this blows over, you'll support him. If his stock levels are high, buy a few there. If they're not, order a few and prepay. They're not that expensive, guys, and last time I checked they last for years.
Yes, the law needs to be changed so any sort of the rocket engines that _were_ OK to build/possess _remain_ OK to build/possess. This means that many of you, instead of being apathetic toward the political process, spend time writing, calling, or E-mailing your congressman or senator and make a good case for why these engines should be exempt. And don't do the "rockets are safe because I could do even more damage with an airplane!" because that'll just get RC airplanes banned. Keep it in the family, as it were. Use anecdotes without hyperbole, about how there are x number of launches per year and how there are next to no accidents, maimings, or death as a result. (Don't have "x" right now, but I'm going to try to look it up.) Maybe be even more nebulous, about how playing with model rockets when you were a kid inspired you to go to work for an aerospace company. I don't care, just use something that they'll read and possibly remember, rather than a negative attack that'll just get round-filed.
They _are_ the same. Westwood Associates designed games for SSI (_Hillsfar_, _Dragonstrike_, and _Eye of the Beholder_ 1 & 2) and Infocom (the above mentioned _Crescent Hawk's Inception_) way back in the mid to late 80s. Around '92 or '93, they became part of Virgin, and became known as Westwood Studios at or around that time (my memory of the exact chronology is a little fuzzy.)
OK, I'm sure others will go into the "more computers doesn't equal more bandwidth consumption" argument, so I won't.
What I'm guessing fewer people will think about is IP wastage. Remember, IPs are delegated by various numbering authorities (in North America, it's ARIN, otherwise check your local listings) and they're pretty sticky about waste - to the point of not giving you more space if you're not efficiently using what you have.
Oh, and if you're giving out/29s or bigger (8 IPs or more) you're supposed to SWIP the blocks to ARIN or whomever for record keeping. SWIPing a block involves keeping track of who you've given it to, and (within reason) how they're using the block. (Running a server, workstation, etc.) If you think this doesn't apply to you, think again. Once you use up the IPs you currently have because you're burning thru them four to eight times faster then you would be if you just let people NAT to begin with, and you need to get more space, you'll most likely need to justify the new space you're asking for - either to your upstream ISP (because THEY might have to turn around and justify it upward to their ISP) or to ARIN or their ilk directly. This means that you'll need to know the percentage of use of all the subnets/29 or longer that you've given out, because if you're not using them "efficiently" enough (upwards of 75% utilization) you might not get more. And believe me, re-doing one's IP infrastructure can be a massive pain in the ass.
And it's not about "running out of IPs", before all the IPv6 people stand up and start waving their hands. It's about keeping the routing tables small enough to manage. We're nowhere near running out of IPv4 space, but the global routing tables are growing at an ungodly rate. It takes as much memory in a router to point 255 IPs (a/24 or "class C") at a given point as for a/20 (4096 IPs or 16 Class Cs) because an address is always xxx.xxx.xxx.xxx/nnn.nnn.nnn.nnn.
To say nothing of the extra "administration" charges that ARIN and whatnot charge the customers for IP space. The more IPs you burn, the more IPs your upstream is burning, and eventually, someone somewhere will go from paying $2500 a year to $5000 a year or to $10000 a year so ARIN can "administer" their IP block. Want to talk about unreasonable and preventable costs? There you go.
Back in the day, we used to have a different IP address assigned to every virtual web domain we hosted. Once it was appropriate to move all those domains to one IP address (IE 3 and Netscape 2 becoming common browsers), we did so. If one still gives out IPs like candy to web hosting customers, ARIN may or may not consider that "efficient use of space" depending on the circumstances.
In a nutshell, don't make people use more IP space than they really need to. The customers who are savvy enough to NAT aren't going to change their bandwidth habits if they have one versus several computers, and the business customers who actually want to run multiple servers will happily pony up for the extra IP space. But charging me extra because I'm reading Slashdot while my girlfriend is checking her E-mail on another computer just causes ill will - it would be like the gas stations charging my parents for owning four cars instead of how much fuel they actually consumed. If you're worried about bandwidth charges, charge for bandwidth. Period.
And, of course, that was when distributed.net was really the only game in town when it came to the whole "turn idle CPUs into something productive" thing. The prize money was probably a good incentive too.:) Now of course, we have SETI@Home, the various protein folding projects, all stuff that many people would argue is a "better" use of time. Plus, and feel free to correct me if I'm wrong, wasn't the original point of the RC5 projects to show how weak limited-length keys were?
Until Symantec makes PCAnywhere for Linux, I'll have to run Windows at work and on my work computer at home. Dual booting is out of the question - imagine explaning to my boss that I had to close my six SSH windows, my web browser, and my E-mail program so I can reboot to Windows to use PCA to fix a two minute problem, then switch back to Linux to continue doing my work. And running PCAnywhere under VMWare for Linux is...problematic. Certainly not reliable enough (last time I tried it) to rely on it working when I'm going to need it. Yes, I know about VNC, et al. I don't need something _like_ PCAnywhere. I need _PCAnywhere_. I don't need it to be free. I just need it to exist.
A quote from a long ago "What life would be like if operatings systems drove your car":
"Taligent/Pink - You walk to the store with Ricardo Montalban, who tells you how wonderful it will be when he can fly you there in his Learjet."
The long and short of it is, it's not innovative until people can _use it_. Linux was at 0.12 in February 1992 and at 1.0 in March 1994, and by then Slackware 1.0 (pretty much the first complete Linux distribution) had been out for almost a year. Yes, all based on GNU tools, blah blah woof woof, but nevertheless it was _useable_. The ISP I once worked for was running all Linux servers by early 1996. Not bad for an OS/kernel that was less than four years old.
Don't pull the "But other OSs are monolithic kernels, and ours is a microkernel!" rhetoric. That's like those old Apple "its better because it's RISC!" ads that had to get pulled after a few short months because of the backlash. Most users don't care what's going on that far under the hood of their server, they just want it to meet their computing needs. I think the Hurd team is finding out the hard way that if you take long enough when developing something, paradigms have shifted and what your on the verge of delivering is irrelevant if not completely obsolete. Then, you start over, and then act surprised in two more years when people _still_ don't care. Telling us in 1998 (or even 2000) that "Hurd is here, but there's a 2GB partition limit for now" would not have been nearly as big of a deal, because it would gotten the OS out there and have people pounding on it, and other parts of the OS would get updated and improved while the 2GB limit was being fixed. Telling us in late 2002 that "Hurd still isn't here because we've finally realized that 2GB partition limits are not appropriate" is winning the award for obvious statement.
Don't tell me how cool it's going to be. _Show me how cool it actually is_. Then I will lavish you with praise, hire young nubile virgin women to lick your toes, and proclaim you the next great innovator.
Oh, and by the way, HERE is the (IMHO) far more meaningful and potentially important section of the article (paragraphs condensed by me):
"The FSF is also modifying the GNU General Public License (GPL), though the fundamental principles will remain unchanged, according to Stallman. 'We have looked at, for example, adding a clause that explicitly states that you give a patent license when you redistribute the software,' Stallman added. FSF also plans to incorporate into the GNU GPL a section covering use of software on a computer network. This new section is likely to be based on a similar section in the Affero GPL adopted by San Francisco-based Affero Inc. The Affero GPL requires anyone modifying a software program to give immediate access by HTTP (Hypertext Transfer Protocol) to the complete source code of the modified software to other users interacting with the software on the network, if the original program had a provision for this kind of access."
I would expect _that_ to cause far more discussion on Slashdot than "This just in - GNU/Hurd still isn't out".
Slashdot Mirror
There's a fundamental problem with something like IRC or a non corporate owned decentralized IM system. When the 15 year old script kiddies flood an IRC server, effectively shutting it down, the person running the server generally doesn't have the resources to do anything about it - they bitch, they moan, they might even call the police, but the police only want to hear how much money you've lost, not that some service you were providing out of the goodness of your heart was no longer working. It would be the same as if you were a bookstore owner that had fifty free books on a table with a sign that said "Free: take one." Do you honestly expect the cops to do anything if one person came by and took them all?
Besides, how can you have granny rely on IRC-like networks to communicate with you if the whole thing is volunteer based and could collapse at any moment due to lack of funding, apathy, or whatnot? And if "everyone" can't rely on it, what good is it?
The decentralized IM system would work if, say, every ISP set up their own server and you could only send thru it if you had a username/password on their system already. Then tracking "spammers" would be easy (call/E-mail the ISP of the misbehaver) and if a particular server goes down it's between that company and its customers, and that's all.
I liken the whole "death spiral of more and more time" trap that you talk about above to the old saying:
"It's like playing pinball. The reward for doing it once is the opportunity to do it again."
The problem is that many people, for a variety of reasons (geography being one) can't change ISPs, and many ISPs (mine included) did nothing in response to my complaints (because they knew I wasn't going to move).
Or in many cases the spammers are paying the ISPs far more per month than the $19.99 dial up guy who's complaining about spam.
Who do you think they're going to bend over backwards to serve?
Oh sure, I vaguely remember the interview he said it in. Something to the effect of "everyone will see Matrix whether or not Jet Li is in it, so I might as well contribute to my own projects. What would be better, seeing a movie that you were already going to see that has me in it, and that's it, or seeing a movie you were already going to see PLUS a movie with me in it?"
I mean, he wasn't arrogant about it or anything, but matter of fact. Let's face it, although it would have been cool to see Li as Seraph, would people who weren't going to see the movie suddenly wanted to see it because Li was in it? Li realizes that and figures that if a movie's going to be a draw because he's in it it may as well be a project that isn't already highly anticipated.
The only problem I have with the visible patrol car thing is that many times the true idiots who were already under the speed limit will slow down even more. And don't get me started about how many times I've seen a cop doing 52 in the right lane of a 55 zone and there's a cluster of cars behind him too terrified to pass...
But otherwise, I'm down with visible cops keeping everyone honest rather than a hidden cop nailing someone and then everyone else screaming past the pulled over cars doing 15+ over.
One can be correct without having any credibility. As they say, even a broken clock is right twice a day.
Won't be much of a wait, considering it came out the same day the DVD did - I own a copy. Probably one of the last, if not THE last, major movie pressed in that format.
Yes indeed, I was talking about "lateral" acceleration - generally found by the skidpad test - when what we're worried about is "linear" acceleration.
:)
Feh. Looks like the cold medicine isn't as non-drowsy as I thought.
1.4 G's? you get more G's than that accelerating from a red light in a poorly-running car, seriously.
Bzzt. The 2003 Chevy Corvette Z06, one of the fastest production cars currently available, can do 0-to-60 MPH in 3.9 seconds, and that's _only_ 1 G lateral acceleration.
I would agree with you in the "losing it" department, but for me the slide started happening way earlier - although I can't decide if he was "slipping" in Difference Engine (because he was collaborating with Sterling) or Virtual Light. But by Idoru, I had a classic case of "who the fuck cares" by midway thru the book. I haven't touched ATP or Pattern Recognition yet, and if I do I'll probably check them out from the library first - I have better ways to spend my book budget than on authors who seem to be on a downward spiral.
The problem is, the Mozilla people are faw more worried about picking a name that a corporation can come after them for than stepping on other free projects' toes. What IF, after doing all the months of research to find out that Firebird doesn't violate any trademarks, they would have asked the Firebird database team and that team would have said "no"? Would you expect Mozilla to just chuck all that research and come up with another name that also isn't trademark infringing but this time might piss off some guy somewhere who named his MP3 player "Malibu" five years ago? At some point you gotta stop and use common sense. No one is going to confuse these two products.
(And yes, you do the research FIRST because not violating _actual_ trademarks are the thing you're trying to avoid here, not _implied_ trademarks.)
Cool. I bet know the kind of store you speak of - Best Buy/Circuit City presentation (lots of gear and speakers in a few rooms), but higher end brands, right? If so, the salesman probably spent less than five minutes dealing with you, if at all, and it's no "loss" for them just because you're a savvy price shopper.
My statement about screwing hard working dealers still stands though.
Knowing which brands ban mail order and which don't, I only hope that you are sticking it to a retailer that is truly offering you no service beyond simply selling you the product (like, letting you listen to the equipment in-store for pretty much as long as you'd like or even letting you borrow it over the weekend to try for yourself.) Because dealers that provide that higher level of service don't deserve to be screwed by price shoppers that happily take advantage of the above services so they can confidently pay less at a mail order retailer.
The problem is, most of the brands that are restricted from mail-order aren't the ones that you'd find at your average Best Buy or Circuit City (or Crutchfield's) anyway, which means chances are you're dealing with a small high-end audio/video store, who would only be able to discount product if you were either a well-established customer or were buying a complete system. There's not as much markup in the A/V world as say, clothing. Most lines are MSRPd - max sale price - at 25%-35% over what the dealer gets them for in quantity 1 - which is how most high-end dealers order product.
And many brands, rather than policing stuff like B-stock (cosmetically damaged equipment that the retailer got a screamingly good deal on) being sold as "discounted A-stock" by mail order retailers - and then the customer, when said B-stock arrives, acts pissed at the _manufacturer_ for the damaged goods rather than the unscrupulous retailer - figure it's just easier all around to say "no mail order sales" and have customers personally examine every piece of equipment that they buy, and easier to then reject warranties from products purchased outside regular channels.
If you are truly buying something that is HALF what a dealer wanted to charge you, you are most likely getting a B-stock unit, and don't even think about bitching when it fails earlier than it's supposed to, or has a big gouge in it, or whatever. And don't be surprised if the manufacturer then fails to honor the regular warranty on it. It left their factory as B-stock (lesser warranty at a way cheaper price), and it's not their fault or problem if whoever you bought it from misrepresented it.
On the low end, I completely agree. But is anything with a Cisco badge truly priced such that it can be considered "low-end"?
On the high end, I emphatically disagree. Talk to someone who's administering more than one of the platforms you mentioned (3COM, Nortel, Lucent, and Cisco). Ask them which hardware is the most reliable, flexible, configurable (no Windows-only Java programs needed), has the best web site support, and in general, has rarely if ever let them down in a pinch? Their answer will most likely be Cisco.
And also, all you have to do to get routine software upgrades for Cisco products is register for their site, for FREE, whereas 3COM/Lucent/Nortel want you to annually pay for maintenance contracts, and if the problem/bug you're experiencing isn't fixed within that year, well, then, please buy another yearly contract, repeat ad infinitum. 3COM in particular has a history of deliberately screwing customers that had been with them since the beginning, such as promising an eventual fix for a UDP latency bug (a big deal among ISPs at the time because it affected Quake players, for example), but only for "current contract customers" and then about 18 months later, dropping that product (the NetServer) and replacing it with a newer one (the HiPER ARC) that wasn't backwards compatible, rendering all the users that were waiting for the promised fix out in the cold, and thousands of dollars poorer. Many of the people burned by this little stunt switched away from 3COM after this, but others stayed and now couldn't PAY people to take the old NetServer stuff off of their hands - it is utterly useless.
Lucent is better at least, because you can still get old Livingston and Ascend firmware updates for free, you just need to pay if you want software for anything more modern. And even though some of their products have been dead for some time (long live the Livingston Portmaster!) at least the legacy stuff is useful in limited capacity. A PM3
is still a great choice if you want to get a little POP going in an area where v.92 isn't much of an issue due to phone line quality anyway.
The tonality is right, but it seems a little on the slow side, tempo-wise. Or maybe because I had a 800XL versus a 400/800 that the sound played a little faster? Anyone care to comment?
Want to do it for free? Here you go. (Note: site seems to be down, no idea why, but I've been there before and the Internet Wayback Machine has copies of the old pages.)
GCom and TPS Systems both have products that might do what you need on Linux.
There's also the IBM Communications Server but I don't think it's what you're looking for.
Which is probably precisely what Canopy/SCO hopes happens, because otherwise Canopy has a very big, very dead albatross hanging around its neck. Seriously, is SCO actually making money from new product sales anymore?
Yeah, no kidding. I switched more than one client away from UnixWare to Linux back in 1997 and 1998, well before IBM was seriously involved, AND got all the legacy applications running under the SysV emulation, even back then. Some of those systems are still running (they never really needed to be bleeding edge), and their upgrade path has been made an order of magnitude easier (and cheaper) by switching.
SCO was irrelevant five years ago. They figure that the possibility of getting any money from IBM beats killing off what little credibility they might have, because they're sinking anyway.
Not quite true. "Interpreted" in the sense you're using it is not the same as "run time compiled" or "two stage interpreted".
When a Perl program, for example, is run, the code is _completely_ compiled into a bytecode before a single line is even run, and then the bytecode is interpreted (in the above sense) as the program is run. That's why gross syntax errors in Perl or Java are spotted before the program even runs, but logic errors that (say) result in a division by zero error might not crop up until the code is actually encountered.
It's also why, even if you get Perl to spit out bytecode, you still NEED the Perl executable to RUN the bytecode - the Perl executable contains the interpreter for the bytecode. It would be another step entirely to convert that bytecode into a completely stand-alone executable - see the "Perl to C" and similar projects that are kicking around the Net.
Now, BASIC, that's an interpreted language (not counting VB and such...I'm talking good old fashioned, IBM ROM BASIC or earlier cousins.) Each line is checked for validity and run, one at a time, as each is encountered. Try typing in a ten line "Hello World" style program in a legacy BASIC sometime, with one of the lines having a syntactical error (like "gosub" to a non-existant line number). You won't usually hear a peep out of the program until it encounters that error halfway into the program, at which point it aborts and prints the error, losing all of your work.
Shell languages would count as "interpreted" as well. Each line isn't executed until it is encountered.
And don't use "assembled" and "compiled" synonymously either. One "compiles" one or more C (for example) subroutines into object (.o) files, which are then "assembled" into the final program.
How the Homeland Security Act Affects You!
----snip----
Question 22: Will model rocket engines like those sold by Estes require ATF permits to buy?
Answer: No. The ATF plans to exempt all solid rocket motors with less than 62.5 grams of propellant.
----snip----
The PROBLEM is that UPS, for now, is not shipping the engines, because apparently it's too much of a grey area for them; they'd rather just ban them than read the law and _realize_ that these engines don't apply. Fine. There aren't other shipping companies in the US? UPS isn't even the major international shipper, methinks. That honor would be either FedEx or DHL. Regardless, these engines can still be had.
So OK, as a result of the new law, after May 31, 2004, you can't teach Timmy how to build his own APCP (Ammonium Perchlorate Composite Propellant) rocket motor without having an ATF permit (or more specifically, you COULD teach him how to build one, but you'd need the permit to launch it anywhere but in your backyard.) This does suck, I agree. But I can still go out to a farm field and cook off a paper tube rocket with a B6-4 engine and show anyone who cares to watch how everything works, just like I've been doing on occasion since I was seven.
Walk into your local hobby shop (the people bitching about all this DO actually buy these engines, right?) and tell him that you'll support him with your business, and if he needs to bump the prices of engines up a bit to cover having them shipped to him by FedEx or whatever until all this blows over, you'll support him. If his stock levels are high, buy a few there. If they're not, order a few and prepay. They're not that expensive, guys, and last time I checked they last for years.
Yes, the law needs to be changed so any sort of the rocket engines that _were_ OK to build/possess _remain_ OK to build/possess. This means that many of you, instead of being apathetic toward the political process, spend time writing, calling, or E-mailing your congressman or senator and make a good case for why these engines should be exempt. And don't do the "rockets are safe because I could do even more damage with an airplane!" because that'll just get RC airplanes banned. Keep it in the family, as it were. Use anecdotes without hyperbole, about how there are x number of launches per year and how there are next to no accidents, maimings, or death as a result. (Don't have "x" right now, but I'm going to try to look it up.) Maybe be even more nebulous, about how playing with model rockets when you were a kid inspired you to go to work for an aerospace company. I don't care, just use something that they'll read and possibly remember, rather than a negative attack that'll just get round-filed.
Back to lurking...
They _are_ the same.
Westwood Associates designed games for SSI (_Hillsfar_, _Dragonstrike_, and _Eye of the Beholder_ 1 & 2) and Infocom (the above mentioned _Crescent Hawk's Inception_) way back in the mid to late 80s.
Around '92 or '93, they became part of Virgin, and became known as Westwood Studios at or around that time (my memory of the exact chronology is a little fuzzy.)
OK, I'm sure others will go into the "more computers doesn't equal more bandwidth consumption" argument, so I won't.
/29s or bigger (8 IPs or more) you're supposed to SWIP the blocks to ARIN or whomever for record keeping. SWIPing a block involves keeping track of who you've given it to, and (within reason) how they're using the block. (Running a server, workstation, etc.) If you think this doesn't apply to you, think again. Once you use up the IPs you currently have because you're burning thru them four to eight times faster then you would be if you just let people NAT to begin with, and you need to get more space, you'll most likely need to justify the new space you're asking for - either to your upstream ISP (because THEY might have to turn around and justify it upward to their ISP) or to ARIN or their ilk directly. This means that you'll need to know the percentage of use of all the subnets /29 or longer that you've given out, because if you're not using them "efficiently" enough (upwards of 75% utilization) you might not get more. And believe me, re-doing one's IP infrastructure can be a massive pain in the ass.
/24 or "class C") at a given point as for a /20 (4096 IPs or 16 Class Cs) because an address is always xxx.xxx.xxx.xxx/nnn.nnn.nnn.nnn.
What I'm guessing fewer people will think about is IP wastage. Remember, IPs are delegated by various numbering authorities (in North America, it's ARIN, otherwise check your local listings) and they're pretty sticky about waste - to the point of not giving you more space if you're not efficiently using what you have.
Oh, and if you're giving out
And it's not about "running out of IPs", before all the IPv6 people stand up and start waving their hands. It's about keeping the routing tables small enough to manage. We're nowhere near running out of IPv4 space, but the global routing tables are growing at an ungodly rate. It takes as much memory in a router to point 255 IPs (a
To say nothing of the extra "administration" charges that ARIN and whatnot charge the customers for IP space. The more IPs you burn, the more IPs your upstream is burning, and eventually, someone somewhere will go from paying $2500 a year to $5000 a year or to $10000 a year so ARIN can "administer" their IP block. Want to talk about unreasonable and preventable costs? There you go.
Back in the day, we used to have a different IP address assigned to every virtual web domain we hosted. Once it was appropriate to move all those domains to one IP address (IE 3 and Netscape 2 becoming common browsers), we did so. If one still gives out IPs like candy to web hosting customers, ARIN may or may not consider that "efficient use of space" depending on the circumstances.
In a nutshell, don't make people use more IP space than they really need to. The customers who are savvy enough to NAT aren't going to change their bandwidth habits if they have one versus several computers, and the business customers who actually want to run multiple servers will happily pony up for the extra IP space. But charging me extra because I'm reading Slashdot while my girlfriend is checking her E-mail on another computer just causes ill will - it would be like the gas stations charging my parents for owning four cars instead of how much fuel they actually consumed. If you're worried about bandwidth charges, charge for bandwidth. Period.
And, of course, that was when distributed.net was really the only game in town when it came to the whole "turn idle CPUs into something productive" thing. The prize money was probably a good incentive too. :)
Now of course, we have SETI@Home, the various protein folding projects, all stuff that many people would argue is a "better" use of time.
Plus, and feel free to correct me if I'm wrong, wasn't the original point of the RC5 projects to show how weak limited-length keys were?
Until Symantec makes PCAnywhere for Linux, I'll have to run Windows at work and on my work computer at home. Dual booting is out of the question - imagine explaning to my boss that I had to close my six SSH windows, my web browser, and my E-mail program so I can reboot to Windows to use PCA to fix a two minute problem, then switch back to Linux to continue doing my work.
And running PCAnywhere under VMWare for Linux is...problematic. Certainly not reliable enough (last time I tried it) to rely on it working when I'm going to need it.
Yes, I know about VNC, et al. I don't need something _like_ PCAnywhere. I need _PCAnywhere_.
I don't need it to be free. I just need it to exist.
A quote from a long ago "What life would be like if operatings systems drove your car":
"Taligent/Pink - You walk to the store with Ricardo Montalban, who tells you how wonderful it will be when he can fly you there in his Learjet."
The long and short of it is, it's not innovative until people can _use it_. Linux was at 0.12 in February 1992 and at 1.0 in March 1994, and by then Slackware 1.0 (pretty much the first complete Linux distribution) had been out for almost a year. Yes, all based on GNU tools, blah blah woof woof, but nevertheless it was _useable_. The ISP I once worked for was running all Linux servers by early 1996. Not bad for an OS/kernel that was less than four years old.
Don't pull the "But other OSs are monolithic kernels, and ours is a microkernel!" rhetoric. That's like those old Apple "its better because it's RISC!" ads that had to get pulled after a few short months because of the backlash. Most users don't care what's going on that far under the hood of their server, they just want it to meet their computing needs.
I think the Hurd team is finding out the hard way that if you take long enough when developing something, paradigms have shifted and what your on the verge of delivering is irrelevant if not completely obsolete. Then, you start over, and then act surprised in two more years when people _still_ don't care. Telling us in 1998 (or even 2000) that "Hurd is here, but there's a 2GB partition limit for now" would not have been nearly as big of a deal, because it would gotten the OS out there and have people pounding on it, and other parts of the OS would get updated and improved while the 2GB limit was being fixed. Telling us in late 2002 that "Hurd still isn't here because we've finally realized that 2GB partition limits are not appropriate" is winning the award for obvious statement.
Don't tell me how cool it's going to be. _Show me how cool it actually is_. Then I will lavish you with praise, hire young nubile virgin women to lick your toes, and proclaim you the next great innovator.
Oh, and by the way, HERE is the (IMHO) far more meaningful and potentially important section of the article (paragraphs condensed by me):
"The FSF is also modifying the GNU General Public License (GPL), though the fundamental principles will remain unchanged, according to Stallman. 'We have looked at, for example, adding a clause that explicitly states that you give a patent license when you redistribute the software,' Stallman added. FSF also plans to incorporate into the GNU GPL a section covering use of software on a computer network. This new section is likely to be based on a similar section in the Affero GPL adopted by San Francisco-based Affero Inc. The Affero GPL requires anyone modifying a software program to give immediate access by HTTP (Hypertext Transfer Protocol) to the complete source code of the modified software to other users interacting with the software on the network, if the original program had a provision for this kind of access."
I would expect _that_ to cause far more discussion on Slashdot than "This just in - GNU/Hurd still isn't out".