What else does it offer besides the click and play?
Nothing, really. Every time I read about Robertson I'm reminded of Homer running CompuGlobalHyperMegaNet, staring blankly at the Comic Book Guy and asking "Can I have some money now?"
Click-and-play is a good idea, anything that simplifies the installation of packages on linux machines is a plus. End users need an easier (easy as in clicking the setup icon, apt-get command lines are not easy to the casual user) way to get stuff on the box.
But, end users don't like the idea of subscriptions. I don't. If you take politics/dogma out of the picture, I'd rather pay 200 upfront for XP Pro, and get free updates for the life of the product, than get a yearly/monthly/weekly bill.
Meh, who cares.
WTF Unit of currency is an MM?
on
OpenIPO and Lindows
·
· Score: 1, Interesting
Are there any known issues/workarounds/patches needed to get OpenSSH 3.8p1 to work with Linux-PAM-77 on a 2.6.5 kernel?
PAM is working correctly, login telnet ftpd sudo. SSH just rejects all passwords, and looking at the logs it doesnt appear to be working through pam at all.
ldd confirms it was built with lib_pam.so.0, and "UsePAM yes" is in the config.
So what's up?
(Googling for OpenSSH and/or PAM results in 9 billion security warnings so thats a dead end)
ISPs wanting to sell broadband to folks w/o a PC? Does this market really exist?
Yes, and this market is growing quickly.
I can think of about a half dozen personal friends with broadbands, but with no PCs.... Think XBox Live, PS2 Online.
Last weekend I helped a friend network his XBox, PS2 and 20 dollar email station. (Network as in we bought a cheap home router and some cords).
We'll see a lot more PC-less applications of the internet. A set top device that streams video over the 'net, 'net aware radios and boomboxes, VOIP boxes, etc, etc..
The problem affects mainly huge peering sessions between big routers, the kind that last for days. You can essentially trick the routers into dropping the peering sessions, leading to route flapping and other hassles.
Big backbone providers don't generally use home-grown linux routers.
It has no real bearing on some home/office router running linux made out of an old 486.
There is absolutely NO documentation on installing and configuring PAM on a linux install that didn't come prebuilt to support it. (Slackware, LFS, etc)
None. There are a couple "Linux-PAM HOWTO"s that say "installing PAM is beyond the scope of this document". What the fuck, bitches? Installing PAM is beyond the scope of a PAM Howto?
I went to the local bookstore, since I'm perfectly willing to pay money for this information. The only book close enough (like all O'Reilly books) was just a compilation of HOWTOs and FAQs. As useless as your friendly LUGs message board, for the low low price of 29.99.
No wonder everyone hates linux.
Like I'm going to throw away a server with 3 years of uptime because no nerd alive knows how to configure PAM.
If I install PAM, recompile just OpenSSH --with-pam (verified with ldd/usr/sbin/sshd), can I use sshd to test PAM, or do I have to rebuild shit like passwd and login or what?
I cant even find anywhere to tell me where pam_warn.so or pam_unix.so debug logs to. I cant find out whether I should use pam_pwdb or pam_unix to auth against the plain old/etc/passwd file.
No information AT ALL. Nowhere.
WHAT THE FUCK! All I want is single-signon through LDAP. LDAP works like a charm, PAM wont. Motherfuckers. I hate linux.
He was sounding too much like a sensible person who sees things for what they are, and wasn't putting enough effort into blindly hating MSFT because they're a big company.
He's just protecting his precious karma.
Frankly I think the Windows API, MFC, COM, DCOM and.Net have all been decent products from Redmond. They've all had their place, all have their quirks. MFC can make for some ugly code, but some elegant solutions too. ActiveX/COM/DCOM frankly look better on paper, there are a ton of implementation headaches, but they work and I've used them effectively.
DirectX is *the* language of PC games. There would be little, if any, PC gaming scene without it. It's evolved a lot over the years, it's gotten a lot cleaner, more tightly coupled to the HAL.. Do you ever miss the days of installing a game, and having to pick your videocard out of a list of 20 or so, then pick your soundcard, then manually configure the port/irq for your gamepad.. I sure dont.
Ah well... This ruling is pretty stupid. The first thing any EU who installs windows is going to do is ask "Why cant I watch this video or play this song?". He's then going to go download WMP and install it, because despite the slashdot party line, it's one of the best (free) media players for the platform. I like the simplicity of ZoomPlayer myself, but WMP has a lot going for it that people like - burning directly to CDs, visualizaions, good format support, etc..
I really fail to see how this whole EU case benefits anyone, except perhaps some cousin of the judge who works at Real. It sure looks like a fat waste of time, effort and money.
I've had nothing but good luck with Kingston's ValueRAM series. I've put them in a couple dozen boxes at the office, and have 2x512s at home in my box.
They're the only sticks I've tried at home that work properly in dual channel mode at 400mhz, including some "premium" samsung-on-samsung sticks I bought at a trade show. Which makes sense since they're one of the only modules specifically listed by the manufacturer. (Gigabyte SINXP mobo to be specific)
Which is my problem with the RAM industry, all this Corsair and OCZ stuff. Matching modules to motherboards is a pain in the ass voodoo ritual. You should be able to just go and buy PC2700 (or better) RAM for a PC2700 capable mobo, but it doesnt work that way. It's a complicated rite of CAS RAS FAS SHMAS timings and jibber jabber, and you ultimately have to conclude that "The abit P3234pX Revision C doesnt work with Corsair XMEGASUPERMEM200000000+".
It just pisses me off. JDEC writes specs, don't they? So if the module is within spec, and the mobo is within spec, should this issue not exist?
Meh, oh well. My computer works, and thats all that matters.
There are two groups of people excited about AMD64...
The first are folks running datacenters that will benefit from >4G machines and 64 bit OS/CPU combos. These people are relatively few, and won't have a problem paying a small premium (whats a couple hundred in RAM on a 10 grand machine?).
The second are kids who have to own a PC hand built from the latest stuff to come out, with the highest numbers on the box. These types aren't usually dissuaded by cost, in fact, the higher the price the more appealing the product. I mean, if RAM is 50% more expensive today than a month ago, it must be 50% better!
The real move to 64 bit will be slow, since for practically all uses there's not much to be gained. People will upgrade when their old computers crap out.
There are plenty of printers and industrial applications that rely on old EDO or FPDRAM's. That old stuff is worth good bucks to the right guy at the right time. I made a couple hundred bucks unloading some 128 megs of EDOs at a swap meet to some guy who wanted to upgrade a bunch of printers at his shop.
Toasters are made using a wholly different process, to much weaker tolerances. There's no uptight timing or voltage requirements for a toaster.
Setup cost for a toaster factory would be minimal compared to chip fabs, and there are no doubt more toaster factories out there. When one toaster factory burns down (as some big chip fab did not to long ago, IIRC) its more easily replaced, and doubtless has minimal impact on the worlds toaster resources.
Plus the demand for toasters is pretty constant. People buy toasters when their old one breaks. They dont rush out to buy a 5% faster "upgraded" toaster just because it's there.
With computer tech there'll be a big rush to a tech, it'll get cheap as it reigns supreme, then get pricier as the industry moves away from it. It happened to EDO, FPDRAM, SDRAM, and now DDR as makers want to move quickly into the more lucrative DDR2 market. You can see the same trend with CPUs and other chipsets.
about a big fire/explosion/typhoon or some kinda shit at one of the big plants in asia where dimm chips are made, and it predicted fallout like this.
Apparently there are only a handful of factories in the world equipped to produce the chips, and only a handful of manufacturers doing so (samsung, winbond, etc.. but most are rebranded)
Yeah, and that's also the reason every MSFT story with a file link (patch or whatever) inevitably has some slashbot do a netcraft lookup on download.microsoft.com, then proudly announce how he discovered that microsoft (akamai) uses linux.
I was going to ask the same thing. VRML was going to replace the web with cool virtual environments, there was once a whole lot of buzz around it.
I would like to know what's lacking in VRML. A lack of foresight (didnt plan ahead for programmable pixel shaders, funtional textures, etc)?
And if it's that sort of problem, how can this new format not fall into the same traps, since the authors likely don't have magical crystal balls that tell them what types of information GPUs of the future will want to store.
What is an "ordinary" robot? It's not like I can go to the local robot dealer and look at base model "ordinary" robots vs the sports package or "pleasure model" AWESOM-O 4000.
Please define "ordinary robot". Most of the robots I see in cartoons or movies are quite extraordinary. Thanks in advance, bitches.
Why aren't tech authors into "free as in beer?"
on
Samba 3 By Example
·
· Score: 2, Insightful
I've been struggling to get my samba PDC (and by extension every windows box on my network) and linux to authenticate against a single source, an LDAP server.
Of course, this means learning not only what LDAP is , but how to configure and test it, etc.. OpenLDAP wasnt the toughest nut to crack, but it's configuration files are out there in wackyland. This is as far as I've gotten.
Then getting samba and other services to auth against it. Of course, to use pam_ldap.so I need to have linux boxes that use PAM, and getting that running on my mutant once-slackware-but-now-fubar installs is no easy task..
Anyways, to say the documentation on such things is sparse would be an understatement. What's to be found is completely obtuse and hard to follow.
It would seem that this book would help. And if this were work-related I could get it and write off the expense. But this is just hobbyist messing-around stuff, and by this time next week I'll be messing with something different.
I just dont have the funds to spend 200 bucks on literature for my time-wasting project du jour. Google's alright most of the time, but often I just see 9 billion users group postings of the same error I'm getting (with no replies containing solutions).
Re:adds stability to Win9x/ME workgroups
on
Samba 3 By Example
·
· Score: 1
WinME (95,98,XP Home) cant join a domain, so leaving it running as a PDC with no shares would be pretty pointless.
For home use, I like samba just to turn what would otherwise be junk into fileservers. I have a samba PDC machine that also hosts a MSDFS root share. Basically its a ghetto SAN, with everything online easily browsable from one mapped drive.
The kids can find the games and cartoons, the wife can find her music and pictures, applications are all stored.
Roaming profiles and remote home directories make a format and reinstall of a machine I just cant be bothered to troubleshoot and fix virtually painless.
I've been fighting an uphill battle trying to get both samba and linux to authenticate against an ldap server. In typical linux fashion, I dont know which piece of the puzzle is missing or broken (PAM and libs, openldap, openssl, samba).
I also threw in the towel trying to get it to share my deskjet properly. It'll print, but the margins are always screwed up, ie; the top 0.5" is always trimmed on paper. And again, there's too much to troubleshoot. The formatting problem could be at the windows client, samba, cups, ghostscript, etc, etc.. There are a million filters between the users keyboard and LPT1 on my samba box.
Anyways, WinME cant join domains stupid. It cant even send out an encrypted password, so don't be having ME/9x boxes on real (important businessy) windows domains.
Will google cache pictures, which apparently is all this site consisted of?
I still say let slashdot articles include a couple photos. Let the guy do a little write up, submit it along with a picture of his little drive cannon, or whatever it is, put a link at the end for people who want to see more.. Then everyone wins, except people trying to generate ad revenue from banners - the other half of slashdot submissions.
Build a catapult or air cannon that fires hard drives?
Modify a hard drive to fire some sort of projectile?
Make a scale replica of a WWII era Howitzer using old hard drives?
Who knows?
It annoys me cuz this is the kind of geeky shit I actually like reading about and discussing. Who cares about the latest round of RIAA threats or MSFT hiring some goober? This could well be a neat lil project dude has going, but I'll never know.
When people submit something like this, why can't they submit a descriptive little write up, and why can't slashdot hold a jpeg or two in the story text?
And why can't editors just flat out refuse non-descript submissions like this, which consist of nothing more than a hyperlink to some guys little home server?
I mean, it's not hard to scope out a webserver and come to the conclusion that it's hosted on junk that isn't up to the task.
Slashdot Mirror
What else does it offer besides the click and play?
Nothing, really. Every time I read about Robertson I'm reminded of Homer running CompuGlobalHyperMegaNet, staring blankly at the Comic Book Guy and asking "Can I have some money now?"
Click-and-play is a good idea, anything that simplifies the installation of packages on linux machines is a plus. End users need an easier (easy as in clicking the setup icon, apt-get command lines are not easy to the casual user) way to get stuff on the box.
But, end users don't like the idea of subscriptions. I don't. If you take politics/dogma out of the picture, I'd rather pay 200 upfront for XP Pro, and get free updates for the life of the product, than get a yearly/monthly/weekly bill.
Meh, who cares.
A Euro, dollar, peso, ruble, what?
Are there any known issues/workarounds/patches needed to get OpenSSH 3.8p1 to work with Linux-PAM-77 on a 2.6.5 kernel?
PAM is working correctly, login telnet ftpd sudo. SSH just rejects all passwords, and looking at the logs it doesnt appear to be working through pam at all.
ldd confirms it was built with lib_pam.so.0, and "UsePAM yes" is in the config.
So what's up?
(Googling for OpenSSH and/or PAM results in 9 billion security warnings so thats a dead end)
And any type of google search involving "OpenSSH" and "PAM" returns a 90 page flood of security warnings.
/OpenSSH/Linux just cant cut it, security-wise, but I want to live dangerously.
Which should be enough of a clue that PAM
Login services work fine with PAM, looking up in LDAP like they should.
Can anyone point me to a patch/working version/config file voodoo?
ISPs wanting to sell broadband to folks w/o a PC? Does this market really exist?
Yes, and this market is growing quickly.
I can think of about a half dozen personal friends with broadbands, but with no PCs.... Think XBox Live, PS2 Online.
Last weekend I helped a friend network his XBox, PS2 and 20 dollar email station. (Network as in we bought a cheap home router and some cords).
We'll see a lot more PC-less applications of the internet. A set top device that streams video over the 'net, 'net aware radios and boomboxes, VOIP boxes, etc, etc..
I don't want to hear about WIFI radio. Not until someone tells me how you install and configure PAM on a pam-less linux box (slackware originally).
The problem affects mainly huge peering sessions between big routers, the kind that last for days. You can essentially trick the routers into dropping the peering sessions, leading to route flapping and other hassles.
Big backbone providers don't generally use home-grown linux routers.
It has no real bearing on some home/office router running linux made out of an old 486.
There is absolutely NO documentation on installing and configuring PAM on a linux install that didn't come prebuilt to support it. (Slackware, LFS, etc)
/usr/sbin/sshd), can I use sshd to test PAM, or do I have to rebuild shit like passwd and login or what?
/etc/passwd file.
None. There are a couple "Linux-PAM HOWTO"s that say "installing PAM is beyond the scope of this document". What the fuck, bitches? Installing PAM is beyond the scope of a PAM Howto?
I went to the local bookstore, since I'm perfectly willing to pay money for this information. The only book close enough (like all O'Reilly books) was just a compilation of HOWTOs and FAQs. As useless as your friendly LUGs message board, for the low low price of 29.99.
No wonder everyone hates linux.
Like I'm going to throw away a server with 3 years of uptime because no nerd alive knows how to configure PAM.
If I install PAM, recompile just OpenSSH --with-pam (verified with ldd
I cant even find anywhere to tell me where pam_warn.so or pam_unix.so debug logs to. I cant find out whether I should use pam_pwdb or pam_unix to auth against the plain old
No information AT ALL. Nowhere.
WHAT THE FUCK! All I want is single-signon through LDAP. LDAP works like a charm, PAM wont. Motherfuckers. I hate linux.
Thats how linux broke the internet. Goddamnit.
He was sounding too much like a sensible person who sees things for what they are, and wasn't putting enough effort into blindly hating MSFT because they're a big company.
.Net have all been decent products from Redmond. They've all had their place, all have their quirks. MFC can make for some ugly code, but some elegant solutions too. ActiveX/COM/DCOM frankly look better on paper, there are a ton of implementation headaches, but they work and I've used them effectively.
He's just protecting his precious karma.
Frankly I think the Windows API, MFC, COM, DCOM and
DirectX is *the* language of PC games. There would be little, if any, PC gaming scene without it. It's evolved a lot over the years, it's gotten a lot cleaner, more tightly coupled to the HAL.. Do you ever miss the days of installing a game, and having to pick your videocard out of a list of 20 or so, then pick your soundcard, then manually configure the port/irq for your gamepad.. I sure dont.
Ah well... This ruling is pretty stupid. The first thing any EU who installs windows is going to do is ask "Why cant I watch this video or play this song?". He's then going to go download WMP and install it, because despite the slashdot party line, it's one of the best (free) media players for the platform. I like the simplicity of ZoomPlayer myself, but WMP has a lot going for it that people like - burning directly to CDs, visualizaions, good format support, etc..
I really fail to see how this whole EU case benefits anyone, except perhaps some cousin of the judge who works at Real. It sure looks like a fat waste of time, effort and money.
I've had nothing but good luck with Kingston's ValueRAM series. I've put them in a couple dozen boxes at the office, and have 2x512s at home in my box.
They're the only sticks I've tried at home that work properly in dual channel mode at 400mhz, including some "premium" samsung-on-samsung sticks I bought at a trade show. Which makes sense since they're one of the only modules specifically listed by the manufacturer. (Gigabyte SINXP mobo to be specific)
Which is my problem with the RAM industry, all this Corsair and OCZ stuff. Matching modules to motherboards is a pain in the ass voodoo ritual. You should be able to just go and buy PC2700 (or better) RAM for a PC2700 capable mobo, but it doesnt work that way. It's a complicated rite of CAS RAS FAS SHMAS timings and jibber jabber, and you ultimately have to conclude that "The abit P3234pX Revision C doesnt work with Corsair XMEGASUPERMEM200000000+".
It just pisses me off. JDEC writes specs, don't they? So if the module is within spec, and the mobo is within spec, should this issue not exist?
Meh, oh well. My computer works, and thats all that matters.
There are two groups of people excited about AMD64...
The first are folks running datacenters that will benefit from >4G machines and 64 bit OS/CPU combos. These people are relatively few, and won't have a problem paying a small premium (whats a couple hundred in RAM on a 10 grand machine?).
The second are kids who have to own a PC hand built from the latest stuff to come out, with the highest numbers on the box. These types aren't usually dissuaded by cost, in fact, the higher the price the more appealing the product. I mean, if RAM is 50% more expensive today than a month ago, it must be 50% better!
The real move to 64 bit will be slow, since for practically all uses there's not much to be gained. People will upgrade when their old computers crap out.
There are plenty of printers and industrial applications that rely on old EDO or FPDRAM's. That old stuff is worth good bucks to the right guy at the right time. I made a couple hundred bucks unloading some 128 megs of EDOs at a swap meet to some guy who wanted to upgrade a bunch of printers at his shop.
That the local Best Buy now lists them as "ReAM Modules"
Toasters are made using a wholly different process, to much weaker tolerances. There's no uptight timing or voltage requirements for a toaster.
Setup cost for a toaster factory would be minimal compared to chip fabs, and there are no doubt more toaster factories out there. When one toaster factory burns down (as some big chip fab did not to long ago, IIRC) its more easily replaced, and doubtless has minimal impact on the worlds toaster resources.
Plus the demand for toasters is pretty constant. People buy toasters when their old one breaks. They dont rush out to buy a 5% faster "upgraded" toaster just because it's there.
With computer tech there'll be a big rush to a tech, it'll get cheap as it reigns supreme, then get pricier as the industry moves away from it. It happened to EDO, FPDRAM, SDRAM, and now DDR as makers want to move quickly into the more lucrative DDR2 market. You can see the same trend with CPUs and other chipsets.
about a big fire/explosion/typhoon or some kinda shit at one of the big plants in asia where dimm chips are made, and it predicted fallout like this.
Apparently there are only a handful of factories in the world equipped to produce the chips, and only a handful of manufacturers doing so (samsung, winbond, etc.. but most are rebranded)
Supply and demand.. life moves on
There's always more waiting to take their places though, so it's all good. Like reading about a big pot bust at the harbour.
Yeah, and that's also the reason every MSFT story with a file link (patch or whatever) inevitably has some slashbot do a netcraft lookup on download.microsoft.com, then proudly announce how he discovered that microsoft (akamai) uses linux.
I was going to ask the same thing. VRML was going to replace the web with cool virtual environments, there was once a whole lot of buzz around it.
I would like to know what's lacking in VRML. A lack of foresight (didnt plan ahead for programmable pixel shaders, funtional textures, etc)?
And if it's that sort of problem, how can this new format not fall into the same traps, since the authors likely don't have magical crystal balls that tell them what types of information GPUs of the future will want to store.
Where do I get an ordinary robot?
What is an "ordinary" robot? It's not like I can go to the local robot dealer and look at base model "ordinary" robots vs the sports package or "pleasure model" AWESOM-O 4000.
Please define "ordinary robot". Most of the robots I see in cartoons or movies are quite extraordinary. Thanks in advance, bitches.
I've been struggling to get my samba PDC (and by extension every windows box on my network) and linux to authenticate against a single source, an LDAP server.
Of course, this means learning not only what LDAP is , but how to configure and test it, etc.. OpenLDAP wasnt the toughest nut to crack, but it's configuration files are out there in wackyland. This is as far as I've gotten.
Then getting samba and other services to auth against it. Of course, to use pam_ldap.so I need to have linux boxes that use PAM, and getting that running on my mutant once-slackware-but-now-fubar installs is no easy task..
Anyways, to say the documentation on such things is sparse would be an understatement. What's to be found is completely obtuse and hard to follow.
It would seem that this book would help. And if this were work-related I could get it and write off the expense. But this is just hobbyist messing-around stuff, and by this time next week I'll be messing with something different.
I just dont have the funds to spend 200 bucks on literature for my time-wasting project du jour. Google's alright most of the time, but often I just see 9 billion users group postings of the same error I'm getting (with no replies containing solutions).
WinME (95,98,XP Home) cant join a domain, so leaving it running as a PDC with no shares would be pretty pointless.
For home use, I like samba just to turn what would otherwise be junk into fileservers. I have a samba PDC machine that also hosts a MSDFS root share. Basically its a ghetto SAN, with everything online easily browsable from one mapped drive.
The kids can find the games and cartoons, the wife can find her music and pictures, applications are all stored.
Roaming profiles and remote home directories make a format and reinstall of a machine I just cant be bothered to troubleshoot and fix virtually painless.
I've been fighting an uphill battle trying to get both samba and linux to authenticate against an ldap server. In typical linux fashion, I dont know which piece of the puzzle is missing or broken (PAM and libs, openldap, openssl, samba).
I also threw in the towel trying to get it to share my deskjet properly. It'll print, but the margins are always screwed up, ie; the top 0.5" is always trimmed on paper. And again, there's too much to troubleshoot. The formatting problem could be at the windows client, samba, cups, ghostscript, etc, etc.. There are a million filters between the users keyboard and LPT1 on my samba box.
Anyways, WinME cant join domains stupid. It cant even send out an encrypted password, so don't be having ME/9x boxes on real (important businessy) windows domains.
They bill hours to the client. Lawyers making 7 figures bill hourly. So do people with fat gravy contracts from the gummint, or other corporations.
Executives and corporate officers, and general beurocrats need salaries since they only do an hours worth of actual work in a week.
But a high priced tech on a jobsite is no different from a high prices plumber or electrician. You'll pay for the time he's there.
Will google cache pictures, which apparently is all this site consisted of?
I still say let slashdot articles include a couple photos. Let the guy do a little write up, submit it along with a picture of his little drive cannon, or whatever it is, put a link at the end for people who want to see more.. Then everyone wins, except people trying to generate ad revenue from banners - the other half of slashdot submissions.
So what did he do?
Build a catapult or air cannon that fires hard drives?
Modify a hard drive to fire some sort of projectile?
Make a scale replica of a WWII era Howitzer using old hard drives?
Who knows?
It annoys me cuz this is the kind of geeky shit I actually like reading about and discussing. Who cares about the latest round of RIAA threats or MSFT hiring some goober? This could well be a neat lil project dude has going, but I'll never know.
When people submit something like this, why can't they submit a descriptive little write up, and why can't slashdot hold a jpeg or two in the story text?
And why can't editors just flat out refuse non-descript submissions like this, which consist of nothing more than a hyperlink to some guys little home server?
I mean, it's not hard to scope out a webserver and come to the conclusion that it's hosted on junk that isn't up to the task.
OK, you know that some guys private webspace on his dsl connected linux server is going to be slashdotted within seconds.
So why not actually put SOMETHING in the submission that describes, in some way, WHAT THE SITE IS ABOUT?
"Check out what this crazy guy did with his computer!!!11!!!!11ROFLOL!"
It's an absolute waste of everyones time. Why even put shit like that on the front page?