Ok, maybe they have a vicious alien-superbrain-designed hellofa plan here, but really... would you bet against the thesis that SCO is just full of shit? It fscking doesn't make sense what they are doing, if you ignore that make-a-buzz->make-the-stock angle.
Damn yeah, I swear I've never read that page. I added the metrics aspect as an afterthought (to rule out responses which point to the obvious simplification as a flaw).
The QuakeNet Website is down due to issues out of our control, including the L request page, we're sorry for any inconvenience caused, but rest assured we're working on it....
But then I remembered (yes, I know quakenet, but forgot what it was...).
One of googles problems is finding the relevant link for search terms, this will help that.
You can think this even further (or speculate even wilder) Webpages are connecteced to each other via links, i.e. you get some kind of directed graph. Now you can partition this graph, by trying to find whole subgraphs which don't get links from the outside (but may link to pages outside "their" subgraph) - this subgraph then probably would be a link farm, because no "respected" page would point people into a link farm.
I think the dataming need to do the above is impossible to do.
But now, if you have "seed links" (from IRC) which point into the "good" set of pages, you probably just need to follow this direction when crawling and know with a very high probability that you are always visiting "good" pages, because, as mentioned, no respected page will contain a link to a link farm. You could also take into account how far away (w.r.t. to followed links) your crawled pages are from the "seed" link.
Now you have translated the task of giant data mining into just crawling web pages, something which should be no problem for google;).
The advantage of IRC, though, compared to the Web, is that it is more reliable - in a very weak sense, but nonetheless.
Think of the google page rank algorithm, it is in great danger to be made useless by link farms.
That is because google has problems seperating link farms from "real" pages which link to each other and by that, provide each other some trust (pagerank).
With well populated irc channels, googles bots can have a higher trust that these channels are not artifial, like the link farms are.
Although you are right, the information to be found there is crap in most cases, I could imagine google harvesting known good help channels (linux-help, etc) for urls which are posted in conversations ("#bla-expert shouts: If you want to know more about bla, goto www.bla-project.org/documentation") , in order to better qualify web pages.
Of course, if they can even reach parity with the middle high-end chips (which may be low high-end chips by the time XGI finally ships) and price things at or below the competion, we could have something interesting on our hands.
This and open sourced drivers which are managed good enough to gather the talent out there. 50% of all linux users shifting their buying decisions to XGI cards for their next buy just because of this drivers could help them a lot at the beginning. Maybe they are lucky, and are at the right time for some big linux based home pc edition.
At least, if there are no contractual or other obstacles, an opensourced _graphics_ driver is an asset.
. Now imagine if Microsoft bought out Novell/Suse and swallowed the IP, disbanded Suse and reallocated the brains. There would go another large Linux Desktop player.
That's the beauty of linux. Every SuSE employee who wouldn't want to get his brain reallocated could go out with a bunch of others and "fork" the SuSE distribution. The only stepping stone would be YaST, and this should be not too hard to recreate. After that, old SuSE clients knew where to go.
And, IIRC, she has personal ties to some of the SCO managers (wasn't she in the same company as McBride, one time). I read that somewhere on the net (she herself acknowledged it), maybe someone has a link, maybe there's something in groklaws citation database.
It would the dumbest thing IBM could do. SCO is just minor problem compared to the future IBM might face if they bought Novell. Nobody can own Linux, and there's a (yet to be destined) number of companies which will be able to live from the pure linux market. If IBM would try to monopolize this market, there would instantly be a space for newcomers. So I think they're better of in continuing what they are doing now, that is holding close ties to companies and that way cover everything they need to deliver for all their (potential) clients needs.
I also haven't understood the "matrix-in-a-matrix" explanation/theory at all. I'd think of Zion as a "honeypot", i.e. the matrix is just simulating to its inhabitants they broke out, while they were still inside. Since the matrix does _totally_ control the complete sensory input of the humans, it's no problem to simulate an "outside" within the same software. Or one could think of the zionites as virus-like agents/applets/modules/whateveryoumightcallit inside the matrix system, and zion is the "quarantine" folder of Norton Antivirus 2350 Enterprise Edition.
How is the parent insightful?!? "If Belkin's routers shipped with this "feature" disabled, who in their right mind would turn it on?" The feature is parental controls...
Wrong, this is the feature they want to sell, your parent was talking about the "feature" of getting an advertisment for this product (i.e. feature). This they have just the nerve to a feature ("due to popular demand"), and nobody in his right might would opt in to watch this silly ad. Ergo, they shoved the customers something unwanted down the throat.
Whenever I see "is this the beginning of the end" I know the submitter is full of it. Ahem, whatever this submitter might be full of, I wish some guys at netscape (navigator), IBM(OS/2), Corel(a whole lot of products), Novell (Netware)..., would have also been full of it. Yes, not everything MS does will end as a monopoly, but if I were Macromedia, I were concerned, very concerned. I think this will result in Macromedia being much more forthcoming in interoperability issues and general openess.
Take a virtual +1 insightful from me. Yes, dammit, everytime I think of some new move MS makes, I forget the inertia of their users. I always have it in my mind when I ask me "why the hell are all these people still using outlook (express) and IE", but this is a typical case of your best ally being your worst enemy at the same time.
I do lament the new breed of techies, who think any movie that has the wherewithall to incorporate 10 seconds of a computer screen showing a shell prompt as worthy of respect.
Two remarks: - Was it the bash? - Hey, just a shell prompt would've been lame, but hey, they showed nmap!
And yes, I have lost about 30 pounds since I started exercising. I now have a basic metabolism that seems to burn off about 500kcal more each day than it did before. I can now more or less eat moderate amounts of whatever I want without gaining weight. When I have time to really put in some exercise (which means doing something physical for an hour each day or so), it is more or less impossible to gain weight.
Yeah. That is the point. All these damn wunder diets... People, if you want to loose weight, get off your lazy ass and do some sports.. Burn more calories than you consume, that's all that is to it. No pain, no gain, it's that easy. And, btw., lifting weights and stuff like that is perfect, because it's quite energy consuming.
No. This event is actually the result of SCO painting themselves into a corner when they started the whole thing. Since day one, everybody said (rightly) that SCO themselves, by distributing linux with the GPL attached when they "knew" (in their surreal world) about the "contamination" of the kernel with their "IP", have practically cleared that IP for use in the linux kernel.
So they had to do something. They couldn't stop distributing linux (I assume their clients would be too happy), so now we see what they decided would be their best course of action.
I assume they went this route, because by stopping to distribute linux they would have opened themselves to more imminent charges by their clients, compared to the threat of lawsuits etc. by copyright holders of their linux version, which they might face now.
All in all, for me this just proofs that they (Canopy and other shareholders of SCO) are really expecting the company to sink, they just want enough time to do god-knows-what.
ty bits for SMP systems, like the ones in the test, that has SCO suing everybody and their brother) among many other things. Let's think about a place where an improved thread model might make a big difference... hmm, maybe in a web server load test? That might involve a lot of threads, what do you think?
Sorry, but you're talking out of your ass. He tested Apache 1.3.28, and this apache doesn't use threads at all. Go read up about it. This is the reason why apache was aways so much faster on linux than on solaris or windows (talking UP machines here), because linux' processes are cheaper. And that is the reason why the apache people heavily changed their code to allow for threading. Again, inform yourself about apache 2.
Btw. ironically the reason mainly told for making this change to apache was windows performance, seems solaris isn't important enough (ok, maybe on solaris a commercial http server like netscape is prevalent anyway),
Frex, I can reliably freeze my Mandrake box into an inert lump, just by running the Netscape that was part of the MDK default install. In an ideal world, MDK should kill NS and be none the worse for wear.
You ignored what he wrote, all that blathering doesn't change the fact that you shouldn't be able to panic the OS from userspace. It's just pure coincidence that the panic occured correlated with resource contention, ab bug like that could also happen in otherwise complete uncritical situations, because it shows, where the OS designers made a mistake and allowed something unpredictable to happen
A seemingly hanging OS, caused by resource contention cannot be completely prevented, because you'll always be able to construct situations where the heuristics will cause the OS slowing or killing the wrong process. That is/was the reason for the opposition of some linux kernel developers against the OOM killer.
Show the linux developers a userspace program which is able to panic the kernel and they'll acknowledge this as a bug.
Yup, can you say, "Duh!" I can't believe I still hear this bullshit about strict typing all the time like it's some kind of fuckin' magic bullet that makes everything maintainable.
Especially if the people touting it are in reality talking about static typing (and not strong, or strict, as you call it).
Wrong! At least for the inherent proposition ("full speed ahead") that running something like SETI@home is the same as letting the box sit idle.
Try this at home, if you have a temperatur sensor on your CPU (even on something like an Athlon).
- Download cpuburn (interesting proggy btw., I have never had my CPU _that_ stressed) - run it - watch your CPU get hot
hotter CPU = more power consumption, that's for sure.
Btw. I that might be the real issue with things like SETI at home. If you run this stuff for excessive periods, you _will_ pay for it. Someone should do the math, comparing reduced life expecation for a proc (and price to buy _the_same_ proc after it dies prematurely) against cost for the additional power consumption.
Yeah, neuroklinik pointed that out, too, I didn't know that this pdf existed and speculated purely by the numbers they paid.
Just one note, they didn't say it was point-to-point performance;) - actually, the whole architecute is "fat tree", so no point-to-point.
So we can I guess a 96 port Mellanox switch costs around 50000$ + around 1000 $ per card (someone may look that up, but the numbers surely aren't that off). Since the 5.5E+6$ seem to include that money, we are at around 3.5E+6$ for 1100 dual G5s. Makes around 3000$ per dual G5, good deal, I'd say.
Anyway, who really believes that the quoted costs are really comparable is nuts.
Slashdot Mirror
Ok, maybe they have a vicious alien-superbrain-designed hellofa plan here, but really ... would you bet against the thesis that SCO is just full of shit?
It fscking doesn't make sense what they are doing, if you ignore that make-a-buzz->make-the-stock angle.
IBM has a vested interest in making any investment firm that might have actually helped SCO think better of doing the same thing again.
So true. I said from the beginning I wait for the moment IBM goes against Canopy. I think we might find out soon.
Damn yeah, I swear I've never read that page. I added the metrics aspect as an afterthought (to rule out responses which point to the obvious simplification as a flaw).
One of googles problems is finding the relevant link for search terms, this will help that.
;).
You can think this even further (or speculate even wilder)
Webpages are connecteced to each other via links, i.e. you get some kind of directed graph.
Now you can partition this graph, by trying to find whole subgraphs which don't get links from the outside (but may link to pages outside "their" subgraph) - this subgraph then probably would be a link farm, because no "respected" page would point people into a link farm.
I think the dataming need to do the above is impossible to do.
But now, if you have "seed links" (from IRC) which point into the "good" set of pages, you probably just need to follow this direction when crawling and know with a very high probability that you are always visiting "good" pages, because, as mentioned, no respected page will contain a link to a link farm. You could also take into account how far away (w.r.t. to followed links) your crawled pages are from the "seed" link.
Now you have translated the task of giant data mining into just crawling web pages, something which should be no problem for google
The advantage of IRC, though, compared to the Web, is that it is more reliable - in a very weak sense, but nonetheless.
Think of the google page rank algorithm, it is in great danger to be made useless by link farms.
That is because google has problems seperating link farms from "real" pages which link to each other and by that, provide each other some trust (pagerank).
With well populated irc channels, googles bots can have a higher trust that these channels are not artifial, like the link farms are.
Although you are right, the information to be found there is crap in most cases, I could imagine google harvesting known good help channels (linux-help, etc) for urls which are posted in conversations ("#bla-expert shouts: If you want to know more about bla, goto www.bla-project.org/documentation") , in order to better qualify web pages.
Of course, if they can even reach parity with the middle high-end chips (which may be low high-end chips by the time XGI finally ships) and price things at or below the competion, we could have something interesting on our hands.
This and open sourced drivers which are managed good enough to gather the talent out there.
50% of all linux users shifting their buying decisions to XGI cards for their next buy just because of this drivers could help them a lot at the beginning. Maybe they are lucky, and are at the right time for some big linux based home pc edition.
At least, if there are no contractual or other obstacles, an opensourced _graphics_ driver is an asset.
. Now imagine if Microsoft bought out Novell/Suse and swallowed the IP, disbanded Suse and reallocated the brains. There would go another large Linux Desktop player.
That's the beauty of linux. Every SuSE employee who wouldn't want to get his brain reallocated could go out with a bunch of others and "fork" the SuSE distribution. The only stepping stone would be YaST, and this should be not too hard to recreate.
After that, old SuSE clients knew where to go.
And, IIRC, she has personal ties to some of the SCO managers (wasn't she in the same company as McBride, one time). I read that somewhere on the net (she herself acknowledged it), maybe someone has a link, maybe there's something in groklaws citation database.
It would the dumbest thing IBM could do. SCO is just minor problem compared to the future IBM might face if they bought Novell.
Nobody can own Linux, and there's a (yet to be destined) number of companies which will be able to live from the pure linux market. If IBM would try to monopolize this market, there would instantly be a space for newcomers. So I think they're better of in continuing what they are doing now, that is holding close ties to companies and that way cover everything they need to deliver for all their (potential) clients needs.
I also haven't understood the "matrix-in-a-matrix" explanation/theory at all.
I'd think of Zion as a "honeypot", i.e. the matrix is just simulating to its inhabitants they broke out, while they were still inside. Since the matrix does _totally_ control the complete sensory input of the humans, it's no problem to simulate an "outside" within the same software. Or one could think of the zionites as virus-like agents/applets/modules/whateveryoumightcallit inside the matrix system, and zion is the "quarantine" folder of Norton Antivirus 2350 Enterprise Edition.
How is the parent insightful?!? "If Belkin's routers shipped with this "feature" disabled, who in their right mind would turn it on?" The feature is parental controls ...
Wrong, this is the feature they want to sell, your parent was talking about the "feature" of getting an advertisment for this product (i.e. feature). This they have just the nerve to a feature ("due to popular demand"), and nobody in his right might would opt in to watch this silly ad. Ergo, they shoved the customers something unwanted down the throat.
Whenever I see "is this the beginning of the end" I know the submitter is full of it. ..., would have also been full of it.
Ahem, whatever this submitter might be full of, I wish some guys at netscape (navigator), IBM(OS/2), Corel(a whole lot of products), Novell (Netware)
Yes, not everything MS does will end as a monopoly, but if I were Macromedia, I were concerned, very concerned. I think this will result in Macromedia being much more forthcoming in interoperability issues and general openess.
Take a virtual +1 insightful from me. Yes, dammit, everytime I think of some new move MS makes, I forget the inertia of their users.
I always have it in my mind when I ask me "why the hell are all these people still using outlook (express) and IE", but this is a typical case of your best ally being your worst enemy at the same time.
I do lament the new breed of techies, who think any movie that has the wherewithall to incorporate 10 seconds of a computer screen showing a shell prompt as worthy of respect.
Two remarks:
- Was it the bash?
- Hey, just a shell prompt would've been lame, but hey, they showed nmap!
- They should learn from the past, and invest in the desktop. That's where they'll sell this to potential customers, as and end to end solution.
Yeah! Anyone wants to take a bet that Novell might buy Trolltech?
And yes, I have lost about 30 pounds since I started exercising. I now have a basic metabolism that seems to burn off about 500kcal more each day than it did before. I can now more or less eat moderate amounts of whatever I want without gaining weight. When I have time to really put in some exercise (which means doing something physical for an hour each day or so), it is more or less impossible to gain weight.
...
Yeah. That is the point. All these damn wunder diets
People, if you want to loose weight, get off your lazy ass and do some sports.. Burn more calories than you consume, that's all that is to it. No pain, no gain, it's that easy.
And, btw., lifting weights and stuff like that is perfect, because it's quite energy consuming.
No. This event is actually the result of SCO painting themselves into a corner when they started the whole thing. Since day one, everybody said (rightly) that SCO themselves, by distributing linux with the GPL attached when they "knew" (in their surreal world) about the "contamination" of the kernel with their "IP", have practically cleared that IP for use in the linux kernel.
So they had to do something. They couldn't stop distributing linux (I assume their clients would be too happy), so now we see what they decided would be their best course of action.
I assume they went this route, because by stopping to distribute linux they would have opened themselves to more imminent charges by their clients, compared to the threat of lawsuits etc. by copyright holders of their linux version, which they might face now.
All in all, for me this just proofs that they (Canopy and other shareholders of SCO) are really expecting the company to sink, they just want enough time to do god-knows-what.
ty bits for SMP systems, like the ones in the test, that has SCO suing everybody and their brother) among many other things. Let's think about a place where an improved thread model might make a big difference ... hmm, maybe in a web server load test? That might involve a lot of threads, what do you think?
Sorry, but you're talking out of your ass. He tested Apache 1.3.28, and this apache doesn't use threads at all. Go read up about it. This is the reason why apache was aways so much faster on linux than on solaris or windows (talking UP machines here), because linux' processes are cheaper. And that is the reason why the apache people heavily changed their code to allow for threading. Again, inform yourself about apache 2.
Btw. ironically the reason mainly told for making this change to apache was windows performance, seems solaris isn't important enough (ok, maybe on solaris a commercial http server like netscape is prevalent anyway),
Frex, I can reliably freeze my Mandrake box into an inert lump, just by running the Netscape that was part of the MDK default install. In an ideal world, MDK should kill NS and be none the worse for wear.
You ignored what he wrote, all that blathering doesn't change the fact that you shouldn't be able to panic the OS from userspace.
It's just pure coincidence that the panic occured correlated with resource contention, ab bug like that could also happen in otherwise complete uncritical situations, because it shows, where the OS designers made a mistake and allowed something unpredictable to happen
A seemingly hanging OS, caused by resource contention cannot be completely prevented, because you'll always be able to construct situations where the heuristics will cause the OS slowing or killing the wrong process. That is/was the reason for the opposition of some linux kernel developers against the OOM killer.
Show the linux developers a userspace program which is able to panic the kernel and they'll acknowledge this as a bug.
Yup, can you say, "Duh!" I can't believe I still hear this bullshit about strict typing all the time like it's some kind of fuckin' magic bullet that makes everything maintainable.
Especially if the people touting it are in reality talking about static typing (and not strong, or strict, as you call it).
Wrong! At least for the inherent proposition ("full speed ahead") that running something like SETI@home is the same as letting the box sit idle.
Try this at home, if you have a temperatur sensor on your CPU (even on something like an Athlon).
- Download cpuburn (interesting proggy btw., I have never had my CPU _that_ stressed)
- run it
- watch your CPU get hot
hotter CPU = more power consumption, that's for sure.
Btw. I that might be the real issue with things like SETI at home. If you run this stuff for excessive periods, you _will_ pay for it. Someone should do the math, comparing reduced life expecation for a proc (and price to buy _the_same_ proc after it dies prematurely) against cost for the additional power consumption.
Go, look up for TMS320C6701 on google. This is a DSP which gets to 1 GFLOP at 2 Watts (I think).
A DSP is just not comparable to a desktop CPU, and I didn't see anything which suggests this thing can do more.
Makes around 3000$ per dual G5, good deal, I'd say
Ahem, 4GB, I forgot to mention.
Yeah, neuroklinik pointed that out, too, I didn't know that this pdf existed and speculated purely by the numbers they paid.
;) - actually, the whole architecute is "fat tree", so no point-to-point.
Just one note, they didn't say it was point-to-point performance
So we can
I guess a 96 port Mellanox switch costs around 50000$ + around 1000 $ per card (someone may look that up, but the numbers surely aren't that off). Since the 5.5E+6$ seem to include that money, we are at around 3.5E+6$ for 1100 dual G5s.
Makes around 3000$ per dual G5, good deal, I'd say.
Anyway, who really believes that the quoted costs are really comparable is nuts.