(though/me wonders... why the hell not ps? Guess it doesn't have all those neat little bracketed thingies in it that say "tech!" to the average politician)
Just a thought here - have you tried using Xen or VMWare and building a 32-bit Linux as an instance?
Otherwise, yeah... I agree perfectly w/ the closed binary crap. IF they intend to do that so hard, they really ought to at least provide a 32-bit SCSI driver for the thing, even if it is 64-bit built.
'k - I just blew off a mod point somewheres else in this thread, but I think this may help you out (at least on the 'how-to-load-linux-on-the-thing' side):
Awhile back at a previous employer, I had 'inherited' some old Dell 6500-series boxes that RHEL AS3 and AS4 simply would not load on... for the same reason: LSI chips suck mud, and neither the new megaraid or the old megariad.ko files would work. This might be the case with your newer stuff as well (dunno for certain ab't the new chipsets - but what I do know is, history tells me that Dell boxes with LSI RAID chipsets remind me a whole helluva lot like my ex-wife when she got hormonal - you never really knew what you were going to get when it showed up).
The workaround is a bit complicated, but it works nicely: Install RHEL AS 2.1. When finished, stick in AS 4u(anything, really), choosing the upgrade option. You'll end up with RHEL AS 4 in every aspect but one... you get to keep the (working) 2.1 megaraid driver. (of course you could just extract the module from the AS 2.1 distro pkg and then call it as a custom driver I suppose, but I know from experience that the 2.1 to 4u* path works). It won't work with AS 3, but it will with 4.
HTH a little on the Linux side. If you're wanting to use Windows and you're having problems, well... sucks to be you:)
Okay - I understand that sometimes a kid can get more than just a little involved with computers (Hell, I remember a period of time as an ADULT when I spent nearly every waking non-work hour online playing Quake or Counter-Strike), but like anyone, they get past it, they move on.
If it starts to be an actual detriment (not eating, not sleeping, etc), okay - I can see the need for intervention. Still, this one makes me queasy a bit.
Why? Well, what about the requirements to be declared "addicted"? Isn't there a danger that safeguards could be tossed, and it would eventually boil down to just someone else's subjective opinion? Hell of a way to be got rid of in a hurry by a disgruntled low-level gov't worker, a pissed-off friend, etc. Anywhere else on the planet okay - I could understand that there would be a due process. But in a country which still prosecutes (and I quote) "hooliganism" (which can mean whatever they want it to mean), and lock dissidents up for years on end? Sounds like just an updated and modernized excuse to shut up anyone who makes the gov't feel uncomfortable.
I broke down and bought my first Mac a few years back...to replace the Windows box. This had more to do with using 3D/CG manipulation programs sans Windows than with leaving Linux (I still have Linux doing just fine @ home in a box next to the Mac).
Just depends on what I'm wanting to do which decides which KVM button I hit - if it's work, music-playing, or a non-OSX game via Cedega or native Linux, I go Linux. If it's 3D/CG stuff I want to go play with of a game that runs better on the Mac, I hit the Mac button. Sometimes I'll go back and forth (e.g. play w/ Linux while the Mac is otherwise occupied with rendering animation frames since OSX' sound card and Vue d'Esprit 5 apparently hate each other... no idea why offhand).
To really muck it all up for you, I still use AC3D and GIMP (for minor postwork on stills) on Linux.:)
Seriously... I distinctly recall having to suffer through Adobe FrameMaker 7 back in 2005 on a Win2k machine (specifically because Adobe got all uppity about refusing to make an OSX version, and the company I worked for didn't have the manual in anything but.fm format from the contractor who started the thing but failed, so...) I can sum that beast up in one phrase: "Crash-O-Rama". I couldn't get a single chapter done w/o at least restarting the thing twice. And someone tell me why in the unholy hell does it take THREE HOURS to convert an.fm file into a 150-page.pdf on a (at that time) high-end P4 3GHz box w/ 2GB of DDR RAM and nothing else running on the box!? I could've done the thing in MS Word and got it compiled faster through a 3rd-party plugin...
For the few things I need a graphics editing program for these days, GIMP does the job for me perfectly. Evince in Linux (and freakin' Preview on the Mac) kicks the dogsnot out of Acrobat Reader for speed, stability, reliability...
Post-2001 or so, Adobe has almost always equalled bloat in my personal experiences with it (across three platforms - Windows, Linux, and OSX), so I can really, really sympathize. Take it as you may.
"In any war there is always loss on both sides. The winner is simply that which has lost the least. The best way to win a war is to never have to fight it to begin with."
...so $VICTIM simply pays the Danegeld and suffer with grace?
This outlines a huge difference between OSS and MSFT, I suspect, or is it? Let's see the options: Linux fights, MSFT defers? I somehow don't think this to be in character for MSFT. Usually they fight, but prefer to do so in as many underhanded ways as they can muster (see also Baystar/SCO...)
IMHO, if there was actually any substance to Ballmer's comments, we would've seen something that would have a proxy going after RedHat, IBM, what-have-you. The Novell deal is something that strikes me as weird, but Novell publicly rebutted any such assertions of IP theft, so that can't be it.
So unless it's begun somewhere, Ballmer is apparently blowing hot air. I won't be too suprised if another shoe were to drop, but I'm not holding my breath waiting for it ATM...
That's true... I remember recently cleaning out my old bookmarks at home by confirming each of 'em as I ran through the list. I hit edonkey2000.com's old webpage and saw the content replaced by a "don't steal music" diatribe and a small javascript that flashed your IP address at you, saying that it's now been tracked. Me, I laughed at it and shuffled on (they'd have bupkis for evidence on my part - even w/ full and unfettered access to every data storage device I own). OTOH, I can see Joe Average getting a solid shock from seeing something like that.
To answer your question, they could easily glean and somewhat filter visitor IP addresses technically... Webalizer (among lots of other server-side tools) can harvest and report as many IP addys as can visit said website - there's a whole industry that once (and prolly still does) revolve around getting unique IP's and counting each as a pair of eyeballs for marketing statistics.
But, there's a whole lot of monkey wrenches in the way of them really using the list of IP addys effectively: they'd have to weed out search engine robots, obvious open proxies, other personal and commercial web crawlers of various types, results from the Tor Anonymizing Network, and similar somesuch places. Then, they'd have to weed out IP addresses that came from countries where they couldn't litigate, where it would be cost-prohibitive to litigate (e.g. Tonga), etc. But otherwise the initial intake of data would be very easy to do. While not perfect by any stretch, the results would make for an initial somewhat useable list.
The question I would have for you in return is, would it be worth their while to try?
The resulting filtered list of visitor IP's would still have to be tracked, one by one, and at least compared to IP lists they'd gleaned from their forays onto the P2P networks if they wanted at least half a chance of avoiding any goofs or snares. The results of that comparison would be somewhat useful to them, though on the legal angle I don't know if the comparison would imply guilt enough to be useful in a courtroom (which is why I do systems administration stuff, and not that semantic arm-wrestling legal stuff that you do for a living:) )
They'd still have to couple those results with evidence of infringement, I believe... and all three would have to tie into the same IP address, something I sincerely doubt would even tie to the same user considering DHCP, NATs/Proxies, Router Configs, and the fact that unless they get timestamps of all three to be match roughly within the same minute of absolute time. Then they'd have to prove that all three data points actually point to the same human being or group of human beings in a very small home network at best. Of course, all of that flies out the window if the IP in question is a proxy, or at the public-facing end of a router with lots and lots of users passing through it, etc.
They'd likely IMHO be better off just chasing P2P files and the IP's associated with that as single data points... because unless they use the site visitor IP as supporting evidence, and under very, very certain circumstances, what would be the point of throwing all the manpower and storage space at it? It'd be fairly easy to tear apart if any one of a number of caveats happen to exist, and while you're the legal type and not I, would a failure to link the two to a specific person or act not destroy the plaintiff's case, or at least damage it greatly? If true, that'd be a whole lot of unnecessary risk to take, no?
Or, to set all that aside... perhaps they could use the site logs as a starting point for investigation, but not use it as part of any presented evidence? Pretty much like a sales-critter would use a bought list of leads as a starting point for cold-calling... Say, grab the IP, filter it (as above), then troll the P2P nets for it within a given time frame, to minimize accidental misidentification due to DHCP (still wouldn't eliminate the other factors, but at least one would be reduced a
1) those of us who have sufficient freedom to surf occasionally from work would make things even more interesting (e.g. the logs would show any of my http requests to be from some IP addy from $HUGE_FREAKIN_CORPORATION). I doubt they'd really dig too awful deep from there once the WHOIS showed something along those lines (or not - either way, it'd make things more problematic). I suspect as a result that they'd have to at least filter those out... Lord knows that the aforementioned Corp. would have my testicles on a platter if I were ever dumb enough to pull a stunt like P2P from their nets, but with as many employees as my little org has, well... there's bound to be some idiots in the crowd. (Then again, how would either end handle that, sans evidence of any actual piracy)?
2) guys like you - legal folk who are interested professionally and otherwise in the text and composition of such a creature.
3) anonymous proxies, open proxies, spoofed IP addys, reading it from compromised computers... way too many ways for serious and clueful pirate types to actually see what the site is all about w/o fear of getting caught. Worse, if they used it for that purpose, they'd be right back to suing grandmothers and etc. thanks to open Wifi AP's, compromised Windows boxes, etc. etc. As a corollary, I believe that if I were the pirating type and were so inclined, I'd much rather have P2P clients humming away on a pack of zombies than on my own bandwidth and IP... which handily redirects the lawsuits towards whoever has the compromised box.
IMHO, I suspect that all this site will really accomplish is to make for a big, fat DDoS target for every jerkweed out there who wants to gain a little 'hacktivist' cred w/ his buddies.
We have enough problems with idiots who leave all their backscatter-inducing defaults on @ their mail servers - coupled with the common joe-jobs, it would quickly turn the Internet into a gelatinous mass choked with bounces.
"If some DSL customer suddenly starts opening hundreds of outgoing SMTP connections, that would be a pretty reliable sign that his machine is pwned. Just block or throttle port 25, and send the customer an email telling him to fix his computer, and keep it blocked until he does - or he contacts abuse@ with a legitimate explanation."
...locking down port 25 outbound from the client would cure most of the bots out there (though not all - some jackass could set up a couple of open relays to listen on port {something-else} to then send the spam along from places where port 25 is wide open outbound. Then again, it ups the bar a bit, which isn't a bad thing either)...
Keep the port open for business commercial clients using T-1 or bigger (or who can at least demonstrate that they have an IT department), and (please!) allow it to be opened upon request by the customer w/o extra charge if he/she can demonstrably articulate on the phone that "yes, I'm setting up my own MTA here for (testing stuff / personal use / etc)".
'course, an ISP requiring clients to use IMAP w/ SSL would really rock, but I'm just dreaming by now...
I like the idea of gathering and using statistics on traffic patterns, but what they're looking for in many cases can be too easily defeated (e.g. "Junk messages are small"... now we get to watch MTA's spend more time trying to sort spam messages packed to the gills w/ random ASCII, necessitating a look through the message body all over again).
OTOH, As part of a larger array of spam-fighting tools, okay - there's bits in there I actually like and which can be used as part of other solutions, if not used in the way suggested. As someone who runs a couple of MTA's on top of everything else I do around here, I always like to find new and interesting ways of stopping spam.
N.B., all that I ask is this: Please make it useful w/o sucking down resources or requisitioning another server. I detest external RBL's - please don't suggest anything that may have an overly-subjective and/or an overly-dependant basis like that. If it isn't RFC-compliant (yes, Verizon, I'm talking to YOU when I say that!), I won't go near it.
Satisfy those, and yes, I'm interested, as would lots of other SMTP-monkeys out here.
Could be worse... all this time I kept thinking that Apple somehow added Detroit Lockers and a new set of leaf springs to its Macs in response to Moab...
(Guess that's what I get for living in Utah and owning a Jeep...)
"Which makes me wonder, what are they using now? Linux? Solaris?"
As a former Arkansan (Fayetteville, ab't 15 minutes south of Bentonville/Wal-Mart HQ) I remember a buddy of mine who worked there, and IIRC he mentioned AS/400's... a whole farm of them.
Mind you, this was 1998/1999, but it makes sense that they would use 'em for that time frame. No idea what they're using now, though.
"Don't have numbers to back it up, but most things I read say that the Secondary MX is *more* likely to be targeted by spammers on the belief that fewer filters will be in place to prevent spam."
...and this is why my secondary MTA server is an exact synchronised replica of the first in hardware, binaries, and only the individual machine ID info differs for the scripts and configs. It may not get used as much, but I'm not much of a firm believer in making sure that one half of the boat is sound while the other half is left full of holes.
As a plus, if something falls down and goes 'splat' with smtp1, I have plenty of time to troubleshoot and fix the problem on it while smtp2 handles the pile. It's kind of nice to be able to fix a server w/o some PHB breathing down your neck because his mail is all backed-up.
Why should consumers be forced to waste resources that they paid for?
Seriously - while most users never come near the limits of what their computer can do, I have spent a ton of time waiting for 3d renders to finish thanks to a maxed-out CPU. Since any real enforceable DRM requires a bit of 'assistance' from hardware, that's just that many more CPU cycles (or GPU cycles, or...? depending) wasted on DRM that I could be putting to good use.
I buy computers on a price/performance measure - how much performance per dollar can I get is my metric. Why should I be forced to accept a lowered ratio because someone else decided that I (or any given user) could, in their eyes, potentially be a dirty little copyright pirate?
"Other than one particular issue with a RAID card, it's done a good job of setting up hardware."
Let me guess - a Dell 6300-series server w/ an old LSI card? (at least that's the one prob I'd found with RHEL (yes, RHEL), but even there there was a workaround - install AS 2.1 then upgrade to 4).
I think when most folks whine ab't hardware support and Fedora, it usually concerns the wireless Centrino thingy.
I tried that back in late 2001 by asking Dell if they could send it with Windows 2000 instead of XP - they wound up wanting to charge an extra $100 for that option, as Win2k was no longer (by that time) a supported OS option for the Inspiron 8100 laptops
Slashdot Mirror
(/me ducks and runzlakhell...)
(though /me wonders... why the hell not ps? Guess it doesn't have all those neat little bracketed thingies in it that say "tech!" to the average politician)
Otherwise, yeah... I agree perfectly w/ the closed binary crap. IF they intend to do that so hard, they really ought to at least provide a 32-bit SCSI driver for the thing, even if it is 64-bit built.
Awhile back at a previous employer, I had 'inherited' some old Dell 6500-series boxes that RHEL AS3 and AS4 simply would not load on... for the same reason: LSI chips suck mud, and neither the new megaraid or the old megariad .ko files would work. This might be the case with your newer stuff as well (dunno for certain ab't the new chipsets - but what I do know is, history tells me that Dell boxes with LSI RAID chipsets remind me a whole helluva lot like my ex-wife when she got hormonal - you never really knew what you were going to get when it showed up).
The workaround is a bit complicated, but it works nicely: Install RHEL AS 2.1. When finished, stick in AS 4u(anything, really), choosing the upgrade option. You'll end up with RHEL AS 4 in every aspect but one... you get to keep the (working) 2.1 megaraid driver. (of course you could just extract the module from the AS 2.1 distro pkg and then call it as a custom driver I suppose, but I know from experience that the 2.1 to 4u* path works). It won't work with AS 3, but it will with 4.
HTH a little on the Linux side. If you're wanting to use Windows and you're having problems, well... sucks to be you :)
If it starts to be an actual detriment (not eating, not sleeping, etc), okay - I can see the need for intervention. Still, this one makes me queasy a bit.
Why? Well, what about the requirements to be declared "addicted"? Isn't there a danger that safeguards could be tossed, and it would eventually boil down to just someone else's subjective opinion? Hell of a way to be got rid of in a hurry by a disgruntled low-level gov't worker, a pissed-off friend, etc. Anywhere else on the planet okay - I could understand that there would be a due process. But in a country which still prosecutes (and I quote) "hooliganism" (which can mean whatever they want it to mean), and lock dissidents up for years on end? Sounds like just an updated and modernized excuse to shut up anyone who makes the gov't feel uncomfortable.
I broke down and bought my first Mac a few years back...to replace the Windows box. This had more to do with using 3D/CG manipulation programs sans Windows than with leaving Linux (I still have Linux doing just fine @ home in a box next to the Mac).
Just depends on what I'm wanting to do which decides which KVM button I hit - if it's work, music-playing, or a non-OSX game via Cedega or native Linux, I go Linux. If it's 3D/CG stuff I want to go play with of a game that runs better on the Mac, I hit the Mac button. Sometimes I'll go back and forth (e.g. play w/ Linux while the Mac is otherwise occupied with rendering animation frames since OSX' sound card and Vue d'Esprit 5 apparently hate each other... no idea why offhand).
To really muck it all up for you, I still use AC3D and GIMP (for minor postwork on stills) on Linux. :)
For the few things I need a graphics editing program for these days, GIMP does the job for me perfectly. Evince in Linux (and freakin' Preview on the Mac) kicks the dogsnot out of Acrobat Reader for speed, stability, reliability...
Post-2001 or so, Adobe has almost always equalled bloat in my personal experiences with it (across three platforms - Windows, Linux, and OSX), so I can really, really sympathize. Take it as you may.
This outlines a huge difference between OSS and MSFT, I suspect, or is it? Let's see the options: Linux fights, MSFT defers? I somehow don't think this to be in character for MSFT. Usually they fight, but prefer to do so in as many underhanded ways as they can muster (see also Baystar/SCO...)
IMHO, if there was actually any substance to Ballmer's comments, we would've seen something that would have a proxy going after RedHat, IBM, what-have-you. The Novell deal is something that strikes me as weird, but Novell publicly rebutted any such assertions of IP theft, so that can't be it.
So unless it's begun somewhere, Ballmer is apparently blowing hot air. I won't be too suprised if another shoe were to drop, but I'm not holding my breath waiting for it ATM...
To answer your question, they could easily glean and somewhat filter visitor IP addresses technically... Webalizer (among lots of other server-side tools) can harvest and report as many IP addys as can visit said website - there's a whole industry that once (and prolly still does) revolve around getting unique IP's and counting each as a pair of eyeballs for marketing statistics.
But, there's a whole lot of monkey wrenches in the way of them really using the list of IP addys effectively: they'd have to weed out search engine robots, obvious open proxies, other personal and commercial web crawlers of various types, results from the Tor Anonymizing Network, and similar somesuch places. Then, they'd have to weed out IP addresses that came from countries where they couldn't litigate, where it would be cost-prohibitive to litigate (e.g. Tonga), etc. But otherwise the initial intake of data would be very easy to do. While not perfect by any stretch, the results would make for an initial somewhat useable list.
The question I would have for you in return is, would it be worth their while to try?
The resulting filtered list of visitor IP's would still have to be tracked, one by one, and at least compared to IP lists they'd gleaned from their forays onto the P2P networks if they wanted at least half a chance of avoiding any goofs or snares. The results of that comparison would be somewhat useful to them, though on the legal angle I don't know if the comparison would imply guilt enough to be useful in a courtroom (which is why I do systems administration stuff, and not that semantic arm-wrestling legal stuff that you do for a living :) )
They'd still have to couple those results with evidence of infringement, I believe... and all three would have to tie into the same IP address, something I sincerely doubt would even tie to the same user considering DHCP, NATs/Proxies, Router Configs, and the fact that unless they get timestamps of all three to be match roughly within the same minute of absolute time. Then they'd have to prove that all three data points actually point to the same human being or group of human beings in a very small home network at best. Of course, all of that flies out the window if the IP in question is a proxy, or at the public-facing end of a router with lots and lots of users passing through it, etc.
They'd likely IMHO be better off just chasing P2P files and the IP's associated with that as single data points... because unless they use the site visitor IP as supporting evidence, and under very, very certain circumstances, what would be the point of throwing all the manpower and storage space at it? It'd be fairly easy to tear apart if any one of a number of caveats happen to exist, and while you're the legal type and not I, would a failure to link the two to a specific person or act not destroy the plaintiff's case, or at least damage it greatly? If true, that'd be a whole lot of unnecessary risk to take, no?
Or, to set all that aside... perhaps they could use the site logs as a starting point for investigation, but not use it as part of any presented evidence? Pretty much like a sales-critter would use a bought list of leads as a starting point for cold-calling... Say, grab the IP, filter it (as above), then troll the P2P nets for it within a given time frame, to minimize accidental misidentification due to DHCP (still wouldn't eliminate the other factors, but at least one would be reduced a
1) those of us who have sufficient freedom to surf occasionally from work would make things even more interesting (e.g. the logs would show any of my http requests to be from some IP addy from $HUGE_FREAKIN_CORPORATION). I doubt they'd really dig too awful deep from there once the WHOIS showed something along those lines (or not - either way, it'd make things more problematic). I suspect as a result that they'd have to at least filter those out... Lord knows that the aforementioned Corp. would have my testicles on a platter if I were ever dumb enough to pull a stunt like P2P from their nets, but with as many employees as my little org has, well... there's bound to be some idiots in the crowd. (Then again, how would either end handle that, sans evidence of any actual piracy)?
2) guys like you - legal folk who are interested professionally and otherwise in the text and composition of such a creature.
3) anonymous proxies, open proxies, spoofed IP addys, reading it from compromised computers... way too many ways for serious and clueful pirate types to actually see what the site is all about w/o fear of getting caught. Worse, if they used it for that purpose, they'd be right back to suing grandmothers and etc. thanks to open Wifi AP's, compromised Windows boxes, etc. etc. As a corollary, I believe that if I were the pirating type and were so inclined, I'd much rather have P2P clients humming away on a pack of zombies than on my own bandwidth and IP... which handily redirects the lawsuits towards whoever has the compromised box.
IMHO, I suspect that all this site will really accomplish is to make for a big, fat DDoS target for every jerkweed out there who wants to gain a little 'hacktivist' cred w/ his buddies.
Thx in advance,
Keep the port open for business commercial clients using T-1 or bigger (or who can at least demonstrate that they have an IT department), and (please!) allow it to be opened upon request by the customer w/o extra charge if he/she can demonstrably articulate on the phone that "yes, I'm setting up my own MTA here for (testing stuff / personal use / etc)".
'course, an ISP requiring clients to use IMAP w/ SSL would really rock, but I'm just dreaming by now...
OTOH, As part of a larger array of spam-fighting tools, okay - there's bits in there I actually like and which can be used as part of other solutions, if not used in the way suggested. As someone who runs a couple of MTA's on top of everything else I do around here, I always like to find new and interesting ways of stopping spam.
N.B., all that I ask is this: Please make it useful w/o sucking down resources or requisitioning another server. I detest external RBL's - please don't suggest anything that may have an overly-subjective and/or an overly-dependant basis like that. If it isn't RFC-compliant (yes, Verizon, I'm talking to YOU when I say that!), I won't go near it.
Satisfy those, and yes, I'm interested, as would lots of other SMTP-monkeys out here.
So IOW, if you don't fit their ideology and/or political agendae, you're not among the intelligent cities on Earth?
Not a very intelligent way to measure intelligence, is it?
(Guess that's what I get for living in Utah and owning a Jeep...)
As a former Arkansan (Fayetteville, ab't 15 minutes south of Bentonville/Wal-Mart HQ) I remember a buddy of mine who worked there, and IIRC he mentioned AS/400's... a whole farm of them.
Mind you, this was 1998/1999, but it makes sense that they would use 'em for that time frame. No idea what they're using now, though.
As a plus, if something falls down and goes 'splat' with smtp1, I have plenty of time to troubleshoot and fix the problem on it while smtp2 handles the pile. It's kind of nice to be able to fix a server w/o some PHB breathing down your neck because his mail is all backed-up.
Seriously - while most users never come near the limits of what their computer can do, I have spent a ton of time waiting for 3d renders to finish thanks to a maxed-out CPU. Since any real enforceable DRM requires a bit of 'assistance' from hardware, that's just that many more CPU cycles (or GPU cycles, or ...? depending) wasted on DRM that I could be putting to good use.
I buy computers on a price/performance measure - how much performance per dollar can I get is my metric. Why should I be forced to accept a lowered ratio because someone else decided that I (or any given user) could, in their eyes, potentially be a dirty little copyright pirate?
Let me guess - a Dell 6300-series server w/ an old LSI card? (at least that's the one prob I'd found with RHEL (yes, RHEL), but even there there was a workaround - install AS 2.1 then upgrade to 4).
I think when most folks whine ab't hardware support and Fedora, it usually concerns the wireless Centrino thingy.