No, Gateway didn't buy Cobalt -- they just sold relabeled Cobalt equipment for a time. Sun bought Cobalt quite a while ago, although there wasn't a whole lot of fanfare around the purchase itself.
And, frankly, the Sun Linux distribution isn't really a distribution at all. According to Sun, it's only available (now) on the LX50, which is pretty much par for the course for the old Cobalt equipment. What _is_ new about the LX50 is that it's built to be a general purpose server platform as opposed to a Web server (Cobalt's mainstay product lines -- the RaQ series).
I know there are a bunch of people out there that think that Sun has a motive for releasing their own distribution. I've even seen a few people claim that they might be doing this to stratify the Linux marketplace. I don't buy it. Cobalt has been shipping their units with a "distribution" called Cobalt Linux for years -- and from what I know of the LX50, it's pretty much the same thing.
Interestingly, back then, Cobalt Linux was basically a hybrid Redhat distribution. I expect that Sun Linux is exactly the same thing -- the next permutation of Cobalt Linux.
As it is, I've stopped being all that interested in the Cobalt platforms after they have wholesale _stopped_ putting out Security fixes for their older platforms. I have a poor, dejected Cobalt Qube that I don't even use anymore because the software on it is so full of holes that it would be suicide to use it as a server. The last updates to the unit were posted by Sun in 2000.
That being said, now that Sun has its nameplate on the front of the unit and its moniker on the issue.net, maybe they'll be more proficient about updates.
I would caution you to stay away from it if you aren't certain beyond a reasonable doubt that it'll work with your equipment (e.g., borrow one and test it with your access points).
Here's the deal with the WiFi Springboard module: It contains what is essentially a Cisco wireless card. The card is a fairly nice one (from what I've read), but it communicates to the Prism as if it were a PPP serial device. In other words, it emulates a modem under PalmOS. That means that you're limited to the highest speed the device can do -- and trust me, that's relatively slow.
It's not all bad... Two nice features of the unit: 1) It has a built-in Lithium Ion battery that powers the module itself when the wireless is active (in other words, it doesn't drain the Prism's internal battery). 2) It has a bit of built in flash memory that you can use to load utility programs on -- I have the Blazer browser loaded into its memory so it's immediately available to be launched when I pop the unit it the slot.
Beyond that, be aware that I've had problems getting the unit to work with most Linksys access points with WEP enabled. Trust me, folks, I know I got the settings right because I have other cards configured the same way that work fine. I've also had some oddities with D-Link access points.
I've even tried to use an ad-hoc mode with the unit and a Linux-based wireless router. I can get a wireless link up (WEP included) and get a DHCP address, but after a while, the card starts ARPing for the router and _ignoring_ the replies that it gets. That pretty much wraps up your browsing, trust me -- when you can't see the router on your segment.
Your choice of web browsers is pretty paltry, too. There's Handspring's Blazer, but it forces you to browse through the Blazer proxies that compress everything for you (and log everywhere you go, most likely). Then there's EudoraWeb, which doesn't support images. And a few more that are mainly forgettable, including a "screen scraping" browser client.
On the other hand, when I was using the unit with VNC via the Cisco wireless APs in the office, the thing worked like a champ (albeit a slow one). And when I used it with Lucent APs, it worked very well, too.
The combination of other problems and the fact that it seems that Intel (the owner of the product line now) doesn't intend to release any updates to the unit have made me move to a Sharp Zaurus and a Dlink CF 802.11b card.:>
You're missing the point, as if there was one being stressed that was worth refuting.
The processor is an offload processor. This doesn't mean a lot to the average user, but to a business user, it's gangbusters.
The "point" is that the NIC is essentially like putting a small server box in front of each of your real servers at a much lower cost. It's also platform independent: With a Linux implementation on the card, you could get a Linux firewall protecting every Linux, OS/X, or Windows server that you own. And those servers wouldn't expend any CPU just processing packets in order to reject them.
Put it this way: If you ran a business that made money on CPU cycles dedicated to a particular application, you'd want that processor dedicated full-time to the task at hand. You'd take great leaps to turn off all non-essential services, tweak the bus speeds, optimize block sizes on the filesystems, nice the process to the max, rob Peter and pay Paul -- just to get the extra 5%. In business, time is money, regardless of whether it's personnel or CPU. That's why an offload NIC is so damned attractive -- because some of us work in companies that care about the bottom line as opposed to dicking with ways to make our 1st-person shooter faster.
The article indicates that the loaded software is Secure Computing's stuff, presumably Gauntlet since it's their only firewall product. I would imagine that the console is the Gauntlet console.
As for the guy above who remarked about how silly it was to require these things to be configured by a central console, he obviously hasn't been the firewall management staff at a large company. A central console is the _only_ way to fly if you have a large number of firewall policy engines to manage. Otherwise, the flagpoles in front of most buildings would be draped with suicidal firewall admins wanting to end it all.:>
(Besides, it's not like there isn't a central console for iptables/ipchains that works pretty well -- a firewall need not be a standalone unit with a custom policy all its own to be secure. Sometimes, it's more secure to provide an administrator with an easy way to avoid screwups.)
Skipping the fact that these are clearly _server_ class NICs, they aren't as expensive as you'd imagine out there in closeout land.
Pricewatch has a vendor selling a (no doubt earlier version) 3CR990 for $59. That's a bit more than your CompUSA card, but a respectable price for a brand-name card -- especially one with an embedded ARM processor specifically for offload processing.
The article indicates that the NIC in question is the 3CR990, which, up until this point, has been the "encryption offload" high-performance NIC. The firewall simply replaces the onboard encryption "soft"-ware with something that handles packets a little differently. I find it fascinating that the NIC is simply "reloaded" with appropriate software that can directly alter its core function. It would be really intruiging to figure out just how this is done on the card.
What is especially interesting is what is loaded: Secure Computing's Gauntlet firewall product (yes, it is originally derived from the old TIS stuff, but has been commercially, er... hydrogenized:) ). This would seem to indicate that the card can support applications that weren't written for it, e.g., it can use software whose platform has been retargeted in compilation (well, at least it implies that).
I wonder what other derived applications could be loaded into that space? Hmmm... the mind wanders...
You thought I was going to mention a Beowulf cluster, didn't you? Shame on you. No cookie for you.
Don't make me go there. ATI's website is hell (dont know how bad it is currently, but historically it's been a huge mess) so it's near impossible to find out which driver you need. Not to mention Detonator is the bomb.
Hmm. Last I checked, the "Find a Driver" link on the front page went right to a selection screen for card and OS. Maybe that's a different ATI.
The poster that you're complaining about is actually right. In terms of the "all-in-one" Video Input-capable cards, ATI has always had the best set of utilities and hardware for people who didn't care about getting a bit higher in Quake's frame rate. Hydravision, ATI's multiple monitor-support software, is still head and shoulders better than any other video card manufacturer's setup. And ATI's "multimedia" applications are tightly integrated and work well. nVidia's "Personal Cinema" is quite a bit clunkier and not integrated with the other media "bits" as well. I know -- I use both.
Where ATI has always fallen down is the quality and efficiency of their drivers. They don't release performance fixes well or often enough, although they've made some good strides to get better. Now that ATI sells chipsets to other manufacturers (following nVidia's lead), we might see them start beating on the capabilities of their drivers soon enough.
Case in point: On paper, the Radeon 8500/128 has some features that could give it a definative edge on the Ti4. Unbound by drivers, it could very well have higher performance than most of the nVidia chipsets -- it already pushes the envelope set by the Ti3 very well. It has a highly efficient way of managing memory bandwidth -- of which it has more of than the nVidia card... It has an incredible shading engine that promises nearly double the performance of anything on the nVidia card... Its GPU, the PTIII, is theoretically capable of a higher fillrate at 32-bit than the nVidia card.
But, of course, it all comes down to how well the software interfaces with the hardware. The drivers need work. Maybe ATI will get it together, and maybe it won't.
It'll be fun to watch. I, frankly, can't wait until there's some good competition among video chipset vendors. I was getting bored after 3dfx tanked.
Compiler: Stackguard!
on
.NETly News
·
· Score: 5, Informative
Look here for additional details on the compiler buffer overflow.
It's not actually a _compiler_ overflow.
Instead, it's a subversion of the "buffer overflow protection" that's built-in to the compiler. The most startling piece of this technical review is that the Microsoft "Overflow Protection" in the compiler appears to be a port of StackGuard. The reviewers point out that an examination of the binary output reveals that the compiled code is nearly identical to the StackGuard output.
The "bridging code" is standard Linux bridging. I think what you're referring to is the "Host AP" or "Infrastructure" mode of the wireless card inside the AP.
If the card is based on the Prism chipset, there's already a Linux driver that will operate the card in Host AP mode by implementing some of the functions normally present in the access point's Infrastructure mode in user space. Other things are handled by the card.
For example, the card itself will handle time-dependent functions like beacon-sending. And it will handle hardware-dependent functions like WEP encrypt-decrypt (optionally). But the association table and all the rest of the functions are part of user space.
Incidentally, you don't need the Host AP driver for your Wireless card to operate a bridged network. Oddly enough, the bridge code in the Linux kernel functions just fine with a wireless card in Peer-to-Peer mode -- it is, after all, just another interface to the kernel itself. What you won't get in peer to peer is sophisticated association handling -- that means that the signal strength meter in your wireless tools won't work exactly right in most commercial packages.
I'm sad to see that eMag is going to be ceasing production, but there are plenty of other sources out there for 9-track tape. Plenty of systems out there still use it (some of them in great quantity). One of the biggest consumers of 9-track is the hospital industry -- there's a good chance a portion, if not all, of your medical history is still shuttled around on 9-track. The place where I work now has a division that does nothing but take 9-track from state health insurance programs and hospitals and produce billing runs from them.
That being said, if we're going to rid ourselves of 9-track for good, there's plenty of excellent fun to be had with it.
One of the best uses for it is to use it to prevent someone from getting into something. To wit: get a friend to help you wrap the contents of a couple of reels around someone's car. Just pass the reel back and forth underneath the car and gradually work it backwards from about where the side mirrors are located. About 3 1000' reels is enough to completely cover the doors. Do it TIGHTLY, almost to the point where the tape breaks. Once you've got a good seal (you'll know you do when you release the ends of the tape and it doesn't move at all), you're done. Damn near impossible to remove easily, and even though the door handles will be accessible, it will take the friggin' Jaws of Life to open the doors. That tape is stronger than it looks.
Another use for it is Christmas decoration. Pack away a couple of reels and use it like tinsel on your tree next year. Don't use it sparingly -- drape it on. It makes a lovely silvery-black tree.
A friend of mine and I used to take a few reels up to the top of a very large hill and "race the reels." You've got to have a really LONG runour on the hill for this. All you have to do for this one is drop the reel on the ground, stand it upright, and pull as hard as you can on the loose tape end. Once the reel starts rolling down the hill, keep pulling steadily but back off a bit in speed. You'll find that the reel will speed up quite a bit as it unspools. In fact, they can get _deadly_ fast! Doing this trick with metal reels once caused one of them to imbed itself about an inch in a cinderblock wall at the end of the hill.
Just my contribution to the end-of-life celebrations.:)
As a Tivo owner (and modifier), I can say that it's exciting that there will be an improved Tivo hitting the stores soon. I'm a little disappointed in a few things, though:
The announcement doesn't indicate what, if any, connectivity options they intend to use by default for the Tivo2. USB ports are great and everything, but if it still requires a modem line to get guide data and uses the USB network adapter for its "extended services"... Yuck.
And about the extended services: Why do I suspect that it'll be an extra charge for those?
I'm also a bit put out that Tivo isn't doing anything to announce improvements in the following areas:
- Show Send (a la ReplayTV)
- Archival of recorded shows to media or PC
- Show scheduling via Web page
- Management of recorded items via Web
- "Self-upgrade" capability via removable media
These, given the platform it's based on, would be simple to achieve. In fact, some of the same things are out there now that others outside of Tivo have created! Why not rely on the experiences of the power users, and be a truly hip company by adopting and supporting some of their work? Isn't that how the Open Source model is supposed to work -- the Adoption of What Works?
Or maybe I'm just mad that I spent all that time modifying my Tivo to add the above features and wish that I had the obviously superior Tivo2 hardware at the time.
A VS tape reader can compensate. But not on my budget. The SCSI reader I use is capable of fixed speeds only -- although it's somewhat easy to change read speed while the unit is in operation. The trouble is, the speeds are settable in blocks: 1800 bpi, 2000 bpi, etc. and if you fall in the middle... well...
It largely depends on the tape unit that recorded the tape as to how well you can recover from a stretched tape. Even those with encoded signal can get so far out of whack that you have to resort to more difficult means.
With Tivo's stock price in the toilet and analysts wondering about Tivo's "business plan and future," it only makes sense that they would try to bring something to the table as far as "on-demand video" goes.
Tivo is hamstrung in that it has -- for most consumers who don't specifically modify the device -- only a 56K modem to get video into the device. Tivo's got to come up with something else, and darned if RealNetworks doesn't already have ready-made code that can run on Linux. What else would they choose? Microsoft's Media format?
Seems to me that Tivo needs to take a page from SonicBlue's playbook and start making broadband-capable Tivos ASAP. You might as well forget about asking me to download.rm files and display them on my Sony WEGA TV, blocks and all. Give me something a little better by sticking with MPEG and upping the connection speed. I'd pay for it.
The article isn't kidding about the difficulty of finding a reader for your typical nine-track tape these days. I spent lots of bucks on a SCSI nine-track a few years ago for archiving system and application software on nine-track from old computer systems. And although the purchase helped, there are still occassions when I have to fire up some very old Big Iron to read one tape or another.
An interesting thing about these tapes: They stretch over time and can sometimes become unreadable because of that. There are times when, to extract the information on the tape, I would put a number of them in my freezer for an hour or so, then try again. Nine times out of ten that would actually work.
Another note about the article: I can still remember discussions with others who had modems about 1200 baud being just "too fast". The reasoning was that the average person couldn't read much faster than 300 baud.:)
A lot of cities have tried to do this, most of them by setting up interconnections in the local Metro area. For example, in Nashville, Tennessee, the city government actually funded an initiative to interconnect all businesses in the city at a site they co-owned with BellSouth. The goal was to centralize and provide a faster "lane" of access to local citizenry for area businesses.
It worked like this: a business with an existing Internet connection would be provided another one at a very low cost by the city. The city also provided interconnections to various local ISPs (Sprint, BBN, local ISPs). Then when the user dialed in to the local POPs in each location, each local business would be only 2 or 3 hops away. Interesting idea.
It failed. Turns out, local city businesses with the desire to operate large Internet business could care less about the access speeds of the communities in which they were located. They dispised the extra complexity the initiative forced upon them, even though part of their "backup line" costs were underwritten. And most of them felt that the government wasn't up to the task anyway. ("It took 4 months to repair the sidewalk in front of my house, and you want me to peer with you? Hah!")
The second part of the plan was to offer Internet access. But that plan was shot down by a multitude of ISPs that didn't want to compete with a city government that was intent on taxing them anyway. They went to court, and now the whole effort is a distant memory.
So although it sounds nice, having a city government -- many of whom collect sales taxes from ISPs -- competing against their tax base. Well, it never tends to work out to benefit the citizens in any meaningful way.
Preface: The way things are posted here is really confusing to me, even though I've been a user for years. Seems to me like I submitted this one a long time ago. Anyway...
These units have the capability to send shows from one ReplayTV unit to another. There's not a whole lot of detail given about this functionality, but I wonder whether it can be fooled into thinking your PC is a ReplayTV unit. I slobber uncontrollably when I think about a DVR that would let me archive shows to my file server.
I've been a Tivo owner for almost a couple of years now, and in that time I've modified mine with extra disk space, a web interface, an ethernet port, and a shell prompt on a serial port.:> And there is some work going on right now to play raw video streams from the unit streamed over the network (Andrew Tridgell of Samba fame is the main culprit there), but something like this -- and the stand that SonicBlue is taking on this issue -- makes me sorta want a ReplayTV 4000.
For those interested, there's very little information on the "Send Show" functionality listed on the ReplayTV web site, but I am curious how a user with multiple ReplayTV units and a broadband hardware firewall would allow people to send video to them. I assume it's a TCP session and let-'er-rip, but the site is annoyingly lacking on details. I'd love to know.
I don't know about upgrading from 7.0 to 7.1 _through_ RHN, but you can upgrade from any sort of media (CD, ftp, etc.). Once the upgrade is complete, just run up2date to update the package list with RHN and it'll move your system profile to the new version.
Volution is, no doubt, a lovely piece of work. The screenshots make one drool.
But it doesn't support either RedHat 7.0 or 7.1, and only supports certain other distributions if you replace certain components in them with Caldera Volution "Compatible" packages. For example, to manage a Mandrake 7.1 box from Volution, you have to replace the Mandrake version of apache with one that comes with the Volution install package.
Volution is also a proprietary package. At least the client pieces of RHN are available in source code form. Not that I'm above using proprietary software for management -- we've spent enough on Tivoli to fund the armies of several small countries -- but the idea of sticking to something I can vet is nice.
It's also almost useless for individual users who manage a number of machines for personal use only. There's a 60-day trial version available, but after that, you've got to buy it -- and it isn't cheap.
Then again, I suppose it's cheaper than RHN's $19.95/mo. per system over the long haul. But I don't expect Caldera would sell Volution to capital-cost poor individuals on a payment plan.:>
Slashdot Mirror
As far as I know, there's no MIPS distribution of either. Sorry.
No, Gateway didn't buy Cobalt -- they just sold relabeled Cobalt equipment for a time. Sun bought Cobalt quite a while ago, although there wasn't a whole lot of fanfare around the purchase itself.
And, frankly, the Sun Linux distribution isn't really a distribution at all. According to Sun, it's only available (now) on the LX50, which is pretty much par for the course for the old Cobalt equipment. What _is_ new about the LX50 is that it's built to be a general purpose server platform as opposed to a Web server (Cobalt's mainstay product lines -- the RaQ series).
I know there are a bunch of people out there that think that Sun has a motive for releasing their own distribution. I've even seen a few people claim that they might be doing this to stratify the Linux marketplace. I don't buy it. Cobalt has been shipping their units with a "distribution" called Cobalt Linux for years -- and from what I know of the LX50, it's pretty much the same thing.
Interestingly, back then, Cobalt Linux was basically a hybrid Redhat distribution. I expect that Sun Linux is exactly the same thing -- the next permutation of Cobalt Linux.
As it is, I've stopped being all that interested in the Cobalt platforms after they have wholesale _stopped_ putting out Security fixes for their older platforms. I have a poor, dejected Cobalt Qube that I don't even use anymore because the software on it is so full of holes that it would be suicide to use it as a server. The last updates to the unit were posted by Sun in 2000.
That being said, now that Sun has its nameplate on the front of the unit and its moniker on the issue.net, maybe they'll be more proficient about updates.
Er, anyone notice that "osama.bin.l@den.net" is supposed to be attending the Mozilla party in Pakistan ("at Tariq's place")?
I would caution you to stay away from it if you aren't certain beyond a reasonable doubt that it'll work with your equipment (e.g., borrow one and test it with your access points).
:>
Here's the deal with the WiFi Springboard module: It contains what is essentially a Cisco wireless card. The card is a fairly nice one (from what I've read), but it communicates to the Prism as if it were a PPP serial device. In other words, it emulates a modem under PalmOS. That means that you're limited to the highest speed the device can do -- and trust me, that's relatively slow.
It's not all bad... Two nice features of the unit:
1) It has a built-in Lithium Ion battery that powers the module itself when the wireless is active (in other words, it doesn't drain the Prism's internal battery).
2) It has a bit of built in flash memory that you can use to load utility programs on -- I have the Blazer browser loaded into its memory so it's immediately available to be launched when I pop the unit it the slot.
Beyond that, be aware that I've had problems getting the unit to work with most Linksys access points with WEP enabled. Trust me, folks, I know I got the settings right because I have other cards configured the same way that work fine. I've also had some oddities with D-Link access points.
I've even tried to use an ad-hoc mode with the unit and a Linux-based wireless router. I can get a wireless link up (WEP included) and get a DHCP address, but after a while, the card starts ARPing for the router and _ignoring_ the replies that it gets. That pretty much wraps up your browsing, trust me -- when you can't see the router on your segment.
Your choice of web browsers is pretty paltry, too. There's Handspring's Blazer, but it forces you to browse through the Blazer proxies that compress everything for you (and log everywhere you go, most likely). Then there's EudoraWeb, which doesn't support images. And a few more that are mainly forgettable, including a "screen scraping" browser client.
On the other hand, when I was using the unit with VNC via the Cisco wireless APs in the office, the thing worked like a champ (albeit a slow one). And when I used it with Lucent APs, it worked very well, too.
The combination of other problems and the fact that it seems that Intel (the owner of the product line now) doesn't intend to release any updates to the unit have made me move to a Sharp Zaurus and a Dlink CF 802.11b card.
Natch. But that was posted after hours. ;>
The processor is an offload processor. This doesn't mean a lot to the average user, but to a business user, it's gangbusters.
The "point" is that the NIC is essentially like putting a small server box in front of each of your real servers at a much lower cost. It's also platform independent: With a Linux implementation on the card, you could get a Linux firewall protecting every Linux, OS/X, or Windows server that you own. And those servers wouldn't expend any CPU just processing packets in order to reject them.
Put it this way: If you ran a business that made money on CPU cycles dedicated to a particular application, you'd want that processor dedicated full-time to the task at hand. You'd take great leaps to turn off all non-essential services, tweak the bus speeds, optimize block sizes on the filesystems, nice the process to the max, rob Peter and pay Paul -- just to get the extra 5%. In business, time is money, regardless of whether it's personnel or CPU. That's why an offload NIC is so damned attractive -- because some of us work in companies that care about the bottom line as opposed to dicking with ways to make our 1st-person shooter faster.
As for the guy above who remarked about how silly it was to require these things to be configured by a central console, he obviously hasn't been the firewall management staff at a large company. A central console is the _only_ way to fly if you have a large number of firewall policy engines to manage. Otherwise, the flagpoles in front of most buildings would be draped with suicidal firewall admins wanting to end it all. :>
(Besides, it's not like there isn't a central console for iptables/ipchains that works pretty well -- a firewall need not be a standalone unit with a custom policy all its own to be secure. Sometimes, it's more secure to provide an administrator with an easy way to avoid screwups.)
Pricewatch has a vendor selling a (no doubt earlier version) 3CR990 for $59. That's a bit more than your CompUSA card, but a respectable price for a brand-name card -- especially one with an embedded ARM processor specifically for offload processing.
What is especially interesting is what is loaded: Secure Computing's Gauntlet firewall product (yes, it is originally derived from the old TIS stuff, but has been commercially, er... hydrogenized :) ). This would seem to indicate that the card can support applications that weren't written for it, e.g., it can use software whose platform has been retargeted in compilation (well, at least it implies that).
I wonder what other derived applications could be loaded into that space? Hmmm... the mind wanders...
You thought I was going to mention a Beowulf cluster, didn't you? Shame on you. No cookie for you.
Hmm. Last I checked, the "Find a Driver" link on the front page went right to a selection screen for card and OS. Maybe that's a different ATI.
The poster that you're complaining about is actually right. In terms of the "all-in-one" Video Input-capable cards, ATI has always had the best set of utilities and hardware for people who didn't care about getting a bit higher in Quake's frame rate. Hydravision, ATI's multiple monitor-support software, is still head and shoulders better than any other video card manufacturer's setup. And ATI's "multimedia" applications are tightly integrated and work well. nVidia's "Personal Cinema" is quite a bit clunkier and not integrated with the other media "bits" as well. I know -- I use both.
Where ATI has always fallen down is the quality and efficiency of their drivers. They don't release performance fixes well or often enough, although they've made some good strides to get better. Now that ATI sells chipsets to other manufacturers (following nVidia's lead), we might see them start beating on the capabilities of their drivers soon enough.
Case in point: On paper, the Radeon 8500/128 has some features that could give it a definative edge on the Ti4. Unbound by drivers, it could very well have higher performance than most of the nVidia chipsets -- it already pushes the envelope set by the Ti3 very well. It has a highly efficient way of managing memory bandwidth -- of which it has more of than the nVidia card... It has an incredible shading engine that promises nearly double the performance of anything on the nVidia card... Its GPU, the PTIII, is theoretically capable of a higher fillrate at 32-bit than the nVidia card.
But, of course, it all comes down to how well the software interfaces with the hardware. The drivers need work. Maybe ATI will get it together, and maybe it won't.
It'll be fun to watch. I, frankly, can't wait until there's some good competition among video chipset vendors. I was getting bored after 3dfx tanked.
It's not actually a _compiler_ overflow.
Instead, it's a subversion of the "buffer overflow protection" that's built-in to the compiler. The most startling piece of this technical review is that the Microsoft "Overflow Protection" in the compiler appears to be a port of StackGuard. The reviewers point out that an examination of the binary output reveals that the compiled code is nearly identical to the StackGuard output.
The "bridging code" is standard Linux bridging. I think what you're referring to is the "Host AP" or "Infrastructure" mode of the wireless card inside the AP.
If the card is based on the Prism chipset, there's already a Linux driver that will operate the card in Host AP mode by implementing some of the functions normally present in the access point's Infrastructure mode in user space. Other things are handled by the card.
For example, the card itself will handle time-dependent functions like beacon-sending. And it will handle hardware-dependent functions like WEP encrypt-decrypt (optionally). But the association table and all the rest of the functions are part of user space.
Incidentally, you don't need the Host AP driver for your Wireless card to operate a bridged network. Oddly enough, the bridge code in the Linux kernel functions just fine with a wireless card in Peer-to-Peer mode -- it is, after all, just another interface to the kernel itself. What you won't get in peer to peer is sophisticated association handling -- that means that the signal strength meter in your wireless tools won't work exactly right in most commercial packages.
I'm sad to see that eMag is going to be ceasing production, but there are plenty of other sources out there for 9-track tape. Plenty of systems out there still use it (some of them in great quantity). One of the biggest consumers of 9-track is the hospital industry -- there's a good chance a portion, if not all, of your medical history is still shuttled around on 9-track. The place where I work now has a division that does nothing but take 9-track from state health insurance programs and hospitals and produce billing runs from them.
:)
That being said, if we're going to rid ourselves of 9-track for good, there's plenty of excellent fun to be had with it.
One of the best uses for it is to use it to prevent someone from getting into something. To wit: get a friend to help you wrap the contents of a couple of reels around someone's car. Just pass the reel back and forth underneath the car and gradually work it backwards from about where the side mirrors are located. About 3 1000' reels is enough to completely cover the doors. Do it TIGHTLY, almost to the point where the tape breaks. Once you've got a good seal (you'll know you do when you release the ends of the tape and it doesn't move at all), you're done. Damn near impossible to remove easily, and even though the door handles will be accessible, it will take the friggin' Jaws of Life to open the doors. That tape is stronger than it looks.
Another use for it is Christmas decoration. Pack away a couple of reels and use it like tinsel on your tree next year. Don't use it sparingly -- drape it on. It makes a lovely silvery-black tree.
A friend of mine and I used to take a few reels up to the top of a very large hill and "race the reels." You've got to have a really LONG runour on the hill for this. All you have to do for this one is drop the reel on the ground, stand it upright, and pull as hard as you can on the loose tape end. Once the reel starts rolling down the hill, keep pulling steadily but back off a bit in speed. You'll find that the reel will speed up quite a bit as it unspools. In fact, they can get _deadly_ fast! Doing this trick with metal reels once caused one of them to imbed itself about an inch in a cinderblock wall at the end of the hill.
Just my contribution to the end-of-life celebrations.
As a Tivo owner (and modifier), I can say that it's exciting that there will be an improved Tivo hitting the stores soon. I'm a little disappointed in a few things, though:
The announcement doesn't indicate what, if any, connectivity options they intend to use by default for the Tivo2. USB ports are great and everything, but if it still requires a modem line to get guide data and uses the USB network adapter for its "extended services"... Yuck.
And about the extended services: Why do I suspect that it'll be an extra charge for those?
I'm also a bit put out that Tivo isn't doing anything to announce improvements in the following areas:
- Show Send (a la ReplayTV)
- Archival of recorded shows to media or PC
- Show scheduling via Web page
- Management of recorded items via Web
- "Self-upgrade" capability via removable media
These, given the platform it's based on, would be simple to achieve. In fact, some of the same things are out there now that others outside of Tivo have created! Why not rely on the experiences of the power users, and be a truly hip company by adopting and supporting some of their work? Isn't that how the Open Source model is supposed to work -- the Adoption of What Works?
Or maybe I'm just mad that I spent all that time modifying my Tivo to add the above features and wish that I had the obviously superior Tivo2 hardware at the time.
:)
I think the real reason broadband is being held up are called, uhm... "bankrupcy proceedings?"
:)
A VS tape reader can compensate. But not on my budget. The SCSI reader I use is capable of fixed speeds only -- although it's somewhat easy to change read speed while the unit is in operation. The trouble is, the speeds are settable in blocks: 1800 bpi, 2000 bpi, etc. and if you fall in the middle... well...
It largely depends on the tape unit that recorded the tape as to how well you can recover from a stretched tape. Even those with encoded signal can get so far out of whack that you have to resort to more difficult means.
With Tivo's stock price in the toilet and analysts wondering about Tivo's "business plan and future," it only makes sense that they would try to bring something to the table as far as "on-demand video" goes.
.rm files and display them on my Sony WEGA TV, blocks and all. Give me something a little better by sticking with MPEG and upping the connection speed. I'd pay for it.
Tivo is hamstrung in that it has -- for most consumers who don't specifically modify the device -- only a 56K modem to get video into the device. Tivo's got to come up with something else, and darned if RealNetworks doesn't already have ready-made code that can run on Linux. What else would they choose? Microsoft's Media format?
Seems to me that Tivo needs to take a page from SonicBlue's playbook and start making broadband-capable Tivos ASAP. You might as well forget about asking me to download
Wired is spot-on in a couple of areas, but I'm afraid they missed the biggest "Vaporware" of the entire year: The antitrust "breakup" of Microsoft.
I don't know about anyone else, but that's been the most frustrating waiting game in the Industry for a long time now, IMHO.
It's been posted here before, but a list of "first mentions" are here. Notably absent is the first mention of Kibo... just an early post BY him. :)
An interesting thing about these tapes: They stretch over time and can sometimes become unreadable because of that. There are times when, to extract the information on the tape, I would put a number of them in my freezer for an hour or so, then try again. Nine times out of ten that would actually work.
Another note about the article: I can still remember discussions with others who had modems about 1200 baud being just "too fast". The reasoning was that the average person couldn't read much faster than 300 baud. :)
It worked like this: a business with an existing Internet connection would be provided another one at a very low cost by the city. The city also provided interconnections to various local ISPs (Sprint, BBN, local ISPs). Then when the user dialed in to the local POPs in each location, each local business would be only 2 or 3 hops away. Interesting idea.
It failed. Turns out, local city businesses with the desire to operate large Internet business could care less about the access speeds of the communities in which they were located. They dispised the extra complexity the initiative forced upon them, even though part of their "backup line" costs were underwritten. And most of them felt that the government wasn't up to the task anyway. ("It took 4 months to repair the sidewalk in front of my house, and you want me to peer with you? Hah!")
The second part of the plan was to offer Internet access. But that plan was shot down by a multitude of ISPs that didn't want to compete with a city government that was intent on taxing them anyway. They went to court, and now the whole effort is a distant memory.
So although it sounds nice, having a city government -- many of whom collect sales taxes from ISPs -- competing against their tax base. Well, it never tends to work out to benefit the citizens in any meaningful way.
Judging by its looks alone, I'm wagering that someone in Apple's marketing department MUST have suggested calling it the "iStalk"...
These units have the capability to send shows from one ReplayTV unit to another. There's not a whole lot of detail given about this functionality, but I wonder whether it can be fooled into thinking your PC is a ReplayTV unit. I slobber uncontrollably when I think about a DVR that would let me archive shows to my file server.
I've been a Tivo owner for almost a couple of years now, and in that time I've modified mine with extra disk space, a web interface, an ethernet port, and a shell prompt on a serial port. :> And there is some work going on right now to play raw video streams from the unit streamed over the network (Andrew Tridgell of Samba fame is the main culprit there), but something like this -- and the stand that SonicBlue is taking on this issue -- makes me sorta want a ReplayTV 4000.
For those interested, there's very little information on the "Send Show" functionality listed on the ReplayTV web site, but I am curious how a user with multiple ReplayTV units and a broadband hardware firewall would allow people to send video to them. I assume it's a TCP session and let-'er-rip, but the site is annoyingly lacking on details. I'd love to know.
I don't know about upgrading from 7.0 to 7.1 _through_ RHN, but you can upgrade from any sort of media (CD, ftp, etc.). Once the upgrade is complete, just run up2date to update the package list with RHN and it'll move your system profile to the new version.
But it doesn't support either RedHat 7.0 or 7.1, and only supports certain other distributions if you replace certain components in them with Caldera Volution "Compatible" packages. For example, to manage a Mandrake 7.1 box from Volution, you have to replace the Mandrake version of apache with one that comes with the Volution install package.
Volution is also a proprietary package. At least the client pieces of RHN are available in source code form. Not that I'm above using proprietary software for management -- we've spent enough on Tivoli to fund the armies of several small countries -- but the idea of sticking to something I can vet is nice.
It's also almost useless for individual users who manage a number of machines for personal use only. There's a 60-day trial version available, but after that, you've got to buy it -- and it isn't cheap.
Then again, I suppose it's cheaper than RHN's $19.95/mo. per system over the long haul. But I don't expect Caldera would sell Volution to capital-cost poor individuals on a payment plan. :>