If this becomes the case, then you could probably successfully lobby that CD-R's themselves aren't actually blank - they contain ATIP data that tells the drive stuff like the manufacturer, the disc type, the maximum write speed (I dont know if maximum write speed is in the CD-R/RW ATIP spec, but it is in all the DVD recording specs), etc.
Knoppix could already do this with normal CD-R's.. You can burn more data onto a multisession disc that isn't full. Have you actually even used a copy of knoppix that was on a pressed CD anyway? I don't think there have ever been any produced (but there probably have -- distributed with a magazine or for a tradeshow or something)
It would probably be just as cheap to publish the software onto a CD-R directly than it would be to publish it on a CD-R/CD-ROM hybrid disc anyway.
I have thought about this idea before...
on
Presenting The CDR-ROM
·
· Score: 5, Insightful
I have thought about this idea before, but it's a really half-baked idea. Here's why:
Manufacutring cost:
It's more expensive to produce a disc like this than it is to manufacture a blank CDR or a pressed CD-ROM. For any application where this type of thing would be required, it could easily be written during the manufacutring process onto a regular CDR. I would imagine that there is roughly an equivalent cost to producing one of these hybrid discs versus publishing on CD-R to begin with. In addition to this, think about all that would be involved with retooling a plant to be able to manufacture discs with a different ratio of CD-ROM to CD-R, not to mention the additional burdens it's going to place on testing equipment that will have to be able to verify both the pressed disc and the blank section.
Reduced capacity:
The CDR-ROM disk will have a reduced capacity compared to a normal pressed CD-ROM or a CD-R because no data will be able to be written at the point at which the two disc types meet. The reduction in capacity would be small, but prevalent.
Drive compatibility problems:
Secondly, the huge installed base of CDROM drives out there will not have good compatibility with this kind of a disc. Most drive firmware treats CD-R and CD-ROM media differently to achieve optimum read performance with different kinds of media. When you put this hybrid thing in your 50x cdrom you got three years ago, it's going to spin up to maximum thinking it's a pressed disc then read error all over the fucking place when it hits the CDR section. The onl thing to do is to fake the cdrom into thinking that the disc is a CD-R in its entirety, but then you don't get any of the advantages of having a pressed disc anyway, such as increased read speed without new drives that cater to this special format.
Software compatibility problems:
Due to the way ISO9660 works, the table of contents (including the TOC for the data on the pressed section) will likely have to be re-burned by any software that writes to the CD-R section of the disc. Thus, a faulty burn would render the entire disk unreadable by most systems.
The only good application I can think of for this is for a console game system where you have the luxury of ensuring a uniform set of hardware capabilities between users, and the ability to break standards to accomplish this weird hybrid design stuff. A game could keep save data on the disc or extra game data or something while protecting the game data itself. The media, though, should be CDRW and not CD-R. For those of you who remember, think about the dreamcast's data format -- use some kind of DVD format for the "outer ring" of game data, and use CDRW for the inner ring of PC/CDROM compatibility. You could pop your Xbox2 game into your PC to download new levels or whatever. Unfortunately, internal storage, and fast network connections inside of future (and some present) game consoles would render this idea pretty pointless also.
They could package the program up and sell it off the shelf as a porn filter! If it can accurately determine 'Hitability', it could save millions of masturbators from potential deflation upon running across a photograph of some ghastly beast.
Hell, they'd probably make enough money that they'd stop caring about music piracy!
~GoRK
Re:Trading disk latency for network latency
on
RAMdisk RAID?
·
· Score: 4, Informative
I'd bet money that he doesn't even have a 64-bit PCI slot in his beefy video editing client. Even marginal IDE RAID cards with 4+ 7200RPM drives can saturate that with no problem. Get a nice 64bit SATA RAID controller and pick up aobut 8 of those new WD SATA 10KRPM drives due out this month and you'll have a local solution that will easily max out a 64bit PCI bus.
Re:This is probably the problem.
on
RAMdisk RAID?
·
· Score: 1
Yeah like buying oodles of RAM and GigE network gear (good switches for what he wants cost a fuckton) is going to be any less expensive than some good SCSI drives and.
What might be a good compromise is the new WD SerialATA 10KRPM drives due to ship later this month. On a good RAID controller they should give you a bit of a speed boost for the $$, and the individual channels burst at 1.5Gb/s which is a bit faster than the GigE interconnect on a ramdisk would give you anyway.
Apple's XServe solution, while it has a FibreChannel interconnect, is an 7200 RPM ATA RAID solution inside the box, so that's going to be a per-client speed hit, and for what he's doing could be accomplished at the same performance level for much cheaper with an internal local ATA RAID controller (which is what he's already doing) - probably individually cheaper than a decent FibreChannel host adaptor!
All in all, I agree with the grandparent post that 15K SCSI drives with a good 64-bit RAID controller with oodles of ram on a nice motherboard will probably SMOKE anything you could get with some kind of bizzare frankenstein Network-raid-ramdisk, though the frankenstein network-raid-ramdisk would be a fun hack just to pull for the hell of it.
You could write an app that installs with ActiveX, but provide a java version for fallback. This has the disadvantage of requiring two applications. You might also look into Citrix MetaFrame for Windows or Solars if you would like to keep the app serverside (written using anything you awnt). Java and ActiveX and standalone viewers are written for ICA, so you have pretty good flexibility there.
What would be useful is writing an algorithm for some media player, xine for instance, that would detect and filter out any bug in the corner. You'd lose some (or a lot) of the color detail, but you could get rid of it.
Not that having a higher samplerate would not benefit them, but keep in mind that their HDTV implementation is getting them a datastream that's just as good as they can pull off-air. The color information is encoded in the MPEG data. The poor colors you see on their screenshots are JPEG artifacts. Pull up one of those 4MB.PNG's and you'll see what I mean.
Also, as a followup, it's very interesting to note that Texas does not make this kind of assumption when ordering merchandise via mail-order. They do not collect sales tax on an interstate mail order, but they do on an interstate e-commerce transaction. It's totally fucking bogus.
Texas does; however, have a law that says they get to collect state sales tax on EVERY e-commerce transaction that's conducted on a server within its borders. Texas is also exempt from the current federal Internet sales tax moratorium. And, that lack of state income tax, well, you really pay for it with an 8.25% state sales tax (higher in some localities). Can you imagine making a resident of another state pay a sales tax rate that is double or triple what he'd pay in his own state?
This is a *MAJOR* sticking point that is ignored by a lot of small companies that have webservers here and could easily get bit. The big companies HQ'd here -- Dell, for instance, do not keep servers within the state borders for this reason.
Yeah you know with all of those aluminum bullets shooting at that stupid moon over the last few billion years, it's bound to have created something!!!!!
On a budget is fine, but $1500 is not exactly an "ultra-expensive" telescope for an amateur.. In fact, it's barely entry-level for a very serious hobbiest.
Anyway, I don't know that much about specific models, but at least this guy can get what he's after for the price he wants - a scope powerful enough to check out Saturn that can do (limited) tracking, probably with a PC doing the hard work, and an eyepiece adaptor to hook to a webcam or something, and he's got a pretty good setup.
Since the "several hops" are actually just a tunnel, BGP would still run fine, but the other alternatives such as OSPF as I mentioned would be 1) simpler to set up and 2) more "in tune" with the actual underlying routes and 3) way easier to load balance.. if you're running two pipes, you might as well have them both carrying traffic if they can.
You're going to have to do your own redundant routing in between you and a network that is properly multi-homed with BGP out to the larger internet to make this work like I think you are really wanting it to work.
First, find an upstream ISP that is multi-homed to your satisfaction. Buy some IP's from them and put in a router or two for redundancy.
Next, build two or more tunnels to the ISP over different circuts or providers and run your own small BGP network on private IP's between the router at your multihomed isp and the routers on either end of your connection. Assign the IP's that belong on the multihomed network locally and let your own routers run BGP (or OSPF or whatever else you want to use instead like load balancing) between your LAN and the multihomed network.
It's hackish. It will be fairly expensive. It will also, however, work, let you keep your servers on-site, and give you greater control over redundancy and failover than you'd get with two upstream providers allowing you to use BGP anyway.
In the end, it might work out to be cheaper to do this in the long run since you wont have to pay any upstream ISP for letting you do BGP. You'll just have to pay for colocation somewhere, which could be a lot cheaper.
They won't do it because the digital signature would undoubtadely be applied to the bootloader. A bootloader could in theory load an appropriate second stage loader to boot anything you wanted - windows, linux, copied game off of the hard drive, whatever.
I suppose they could have the signature verify both the bootloader and the kernel, but then... you're stuck with one kernel.
Erm, no not really. Konfabulator is not that integrated. It's more like Active Desktop without a lot of the flexibility Active Desktop has. For instance, Konfabulator widgets cannot register file handlers or generally be loaded/unloaded on demand by the OS itself such as the OpenDoc modules could. Konfabulator widgets also cannot automatically update themselves with new code as it becomes available (unless of course it's implemented internally in the widget), unlike Active Desktop widgets which can.
And to the guy who said that this was more "Integrated" than Active Desktop, it's really not. If you put a local (d)html page on your drive in the "Local" internet "zone" (sorry bout the lame m$ terminology), scripts on the page can be given access to local resources they need to do anything these Konfabulator widgets could do. You can script active desktop widgets in VBScript, JavaScript, Perl, Python (Using ActiveState PerlScript/PythonScript), and using these languages access enough of the underlying API that you could easily put an entire little application in there (think PerlScript + Win32::GUI app in an Active Desktop widget)
Don't get me wrong; I hate Active Desktop and Windows and I think the entire idea of little desktop apps will be forever relegated to 'toys' and other useless crap, but I think Active Desktop is a superior implementation over Konfabulator because of automatic updating and scripting language flexibility. At least it has a better name and doesn't sound like it belongs in KDE.
Interestingly enough, webcams are the only thing that adorned my Active Desktop space for the longest amount of time. After a while the resources running Active Desktop weren't worth it.
Take a look at this crap to see a rather extreme example of Active Desktop.
Woo. Active Desktop comes to the Mac about four years after everyone decided it was pretty much useless. Oh well, I guess some people will really like to fool with it and say how great and superior it is to active desktop even though it is the same exact fucking thing. I made lots of little Active Desktop widgets at one point that are now lost to the annals of time I suppose. It was fun while it lasted.
In 1993 - 1994, they used to let secondary schools and in some cases even elementary schools and middle schools register.edu domains. If they open it up to vocational institutions, they also should open it up to the legitimate primary and secondary schools they took it away from before.
Even better, they ought to move the k12 subdomain to.edu and create a lot more subdomains for.edu to handle this kind of stuff.. Training institutions and other vocational schools can get a domain in.voc.edu after a certain amount of time or whatnot, other schools can register in elem.edu or somesuch..
Slashdot Mirror
Yeah - that's it... The secret handshake!
Why would you need it when the headphones automatically spring off of your head when the levels get too high?
More info can be found on the manufacturer's site here:
http://www.optical-disc.com/CDR_ROM.htm
~GoRK
If this becomes the case, then you could probably successfully lobby that CD-R's themselves aren't actually blank - they contain ATIP data that tells the drive stuff like the manufacturer, the disc type, the maximum write speed (I dont know if maximum write speed is in the CD-R/RW ATIP spec, but it is in all the DVD recording specs), etc.
~GoRK
Knoppix could already do this with normal CD-R's.. You can burn more data onto a multisession disc that isn't full. Have you actually even used a copy of knoppix that was on a pressed CD anyway? I don't think there have ever been any produced (but there probably have -- distributed with a magazine or for a tradeshow or something)
It would probably be just as cheap to publish the software onto a CD-R directly than it would be to publish it on a CD-R/CD-ROM hybrid disc anyway.
I have thought about this idea before, but it's a really half-baked idea. Here's why:
Manufacutring cost:
It's more expensive to produce a disc like this than it is to manufacture a blank CDR or a pressed CD-ROM. For any application where this type of thing would be required, it could easily be written during the manufacutring process onto a regular CDR. I would imagine that there is roughly an equivalent cost to producing one of these hybrid discs versus publishing on CD-R to begin with. In addition to this, think about all that would be involved with retooling a plant to be able to manufacture discs with a different ratio of CD-ROM to CD-R, not to mention the additional burdens it's going to place on testing equipment that will have to be able to verify both the pressed disc and the blank section.
Reduced capacity:
The CDR-ROM disk will have a reduced capacity compared to a normal pressed CD-ROM or a CD-R because no data will be able to be written at the point at which the two disc types meet. The reduction in capacity would be small, but prevalent.
Drive compatibility problems:
Secondly, the huge installed base of CDROM drives out there will not have good compatibility with this kind of a disc. Most drive firmware treats CD-R and CD-ROM media differently to achieve optimum read performance with different kinds of media. When you put this hybrid thing in your 50x cdrom you got three years ago, it's going to spin up to maximum thinking it's a pressed disc then read error all over the fucking place when it hits the CDR section. The onl thing to do is to fake the cdrom into thinking that the disc is a CD-R in its entirety, but then you don't get any of the advantages of having a pressed disc anyway, such as increased read speed without new drives that cater to this special format.
Software compatibility problems:
Due to the way ISO9660 works, the table of contents (including the TOC for the data on the pressed section) will likely have to be re-burned by any software that writes to the CD-R section of the disc. Thus, a faulty burn would render the entire disk unreadable by most systems.
The only good application I can think of for this is for a console game system where you have the luxury of ensuring a uniform set of hardware capabilities between users, and the ability to break standards to accomplish this weird hybrid design stuff. A game could keep save data on the disc or extra game data or something while protecting the game data itself. The media, though, should be CDRW and not CD-R. For those of you who remember, think about the dreamcast's data format -- use some kind of DVD format for the "outer ring" of game data, and use CDRW for the inner ring of PC/CDROM compatibility. You could pop your Xbox2 game into your PC to download new levels or whatever. Unfortunately, internal storage, and fast network connections inside of future (and some present) game consoles would render this idea pretty pointless also.
~GoRK
Idea --
They could package the program up and sell it off the shelf as a porn filter! If it can accurately determine 'Hitability', it could save millions of masturbators from potential deflation upon running across a photograph of some ghastly beast.
Hell, they'd probably make enough money that they'd stop caring about music piracy!
~GoRK
I'd bet money that he doesn't even have a 64-bit PCI slot in his beefy video editing client. Even marginal IDE RAID cards with 4+ 7200RPM drives can saturate that with no problem. Get a nice 64bit SATA RAID controller and pick up aobut 8 of those new WD SATA 10KRPM drives due out this month and you'll have a local solution that will easily max out a 64bit PCI bus.
Yeah like buying oodles of RAM and GigE network gear (good switches for what he wants cost a fuckton) is going to be any less expensive than some good SCSI drives and .
What might be a good compromise is the new WD SerialATA 10KRPM drives due to ship later this month. On a good RAID controller they should give you a bit of a speed boost for the $$, and the individual channels burst at 1.5Gb/s which is a bit faster than the GigE interconnect on a ramdisk would give you anyway.
Apple's XServe solution, while it has a FibreChannel interconnect, is an 7200 RPM ATA RAID solution inside the box, so that's going to be a per-client speed hit, and for what he's doing could be accomplished at the same performance level for much cheaper with an internal local ATA RAID controller (which is what he's already doing) - probably individually cheaper than a decent FibreChannel host adaptor!
[Source: Apple XServe RAID Technology Overview ]
All in all, I agree with the grandparent post that 15K SCSI drives with a good 64-bit RAID controller with oodles of ram on a nice motherboard will probably SMOKE anything you could get with some kind of bizzare frankenstein Network-raid-ramdisk, though the frankenstein network-raid-ramdisk would be a fun hack just to pull for the hell of it.
You could write an app that installs with ActiveX, but provide a java version for fallback. This has the disadvantage of requiring two applications. You might also look into Citrix MetaFrame for Windows or Solars if you would like to keep the app serverside (written using anything you awnt). Java and ActiveX and standalone viewers are written for ICA, so you have pretty good flexibility there.
~GoRK
What would be useful is writing an algorithm for some media player, xine for instance, that would detect and filter out any bug in the corner. You'd lose some (or a lot) of the color detail, but you could get rid of it.
Then implement it in a TV!
Not that having a higher samplerate would not benefit them, but keep in mind that their HDTV implementation is getting them a datastream that's just as good as they can pull off-air. The color information is encoded in the MPEG data. The poor colors you see on their screenshots are JPEG artifacts. Pull up one of those 4MB .PNG's and you'll see what I mean.
A little correction -- he hooked it to (non-rechargable) Lithium cells which completely kick the ass of LiIon rechargables.
Also, as a followup, it's very interesting to note that Texas does not make this kind of assumption when ordering merchandise via mail-order. They do not collect sales tax on an interstate mail order, but they do on an interstate e-commerce transaction. It's totally fucking bogus.
Texas does; however, have a law that says they get to collect state sales tax on EVERY e-commerce transaction that's conducted on a server within its borders. Texas is also exempt from the current federal Internet sales tax moratorium. And, that lack of state income tax, well, you really pay for it with an 8.25% state sales tax (higher in some localities). Can you imagine making a resident of another state pay a sales tax rate that is double or triple what he'd pay in his own state?
This is a *MAJOR* sticking point that is ignored by a lot of small companies that have webservers here and could easily get bit. The big companies HQ'd here -- Dell, for instance, do not keep servers within the state borders for this reason.
[Source: Report of Texas Internet Tax Policy Working Group ]
Yeah you know with all of those aluminum bullets shooting at that stupid moon over the last few billion years, it's bound to have created something!!!!!
~GoRK
On a budget is fine, but $1500 is not exactly an "ultra-expensive" telescope for an amateur.. In fact, it's barely entry-level for a very serious hobbiest.
Anyway, I don't know that much about specific models, but at least this guy can get what he's after for the price he wants - a scope powerful enough to check out Saturn that can do (limited) tracking, probably with a PC doing the hard work, and an eyepiece adaptor to hook to a webcam or something, and he's got a pretty good setup.
~GoRK
Since the "several hops" are actually just a tunnel, BGP would still run fine, but the other alternatives such as OSPF as I mentioned would be 1) simpler to set up and 2) more "in tune" with the actual underlying routes and 3) way easier to load balance .. if you're running two pipes, you might as well have them both carrying traffic if they can.
~GoRK
You're going to have to do your own redundant routing in between you and a network that is properly multi-homed with BGP out to the larger internet to make this work like I think you are really wanting it to work.
First, find an upstream ISP that is multi-homed to your satisfaction. Buy some IP's from them and put in a router or two for redundancy.
Next, build two or more tunnels to the ISP over different circuts or providers and run your own small BGP network on private IP's between the router at your multihomed isp and the routers on either end of your connection. Assign the IP's that belong on the multihomed network locally and let your own routers run BGP (or OSPF or whatever else you want to use instead like load balancing) between your LAN and the multihomed network.
It's hackish. It will be fairly expensive. It will also, however, work, let you keep your servers on-site, and give you greater control over redundancy and failover than you'd get with two upstream providers allowing you to use BGP anyway.
In the end, it might work out to be cheaper to do this in the long run since you wont have to pay any upstream ISP for letting you do BGP. You'll just have to pay for colocation somewhere, which could be a lot cheaper.
~GoRK
They won't do it because the digital signature would undoubtadely be applied to the bootloader. A bootloader could in theory load an appropriate second stage loader to boot anything you wanted - windows, linux, copied game off of the hard drive, whatever.
... you're stuck with one kernel.
I suppose they could have the signature verify both the bootloader and the kernel, but then
~GoRK
Erm, no not really. Konfabulator is not that integrated. It's more like Active Desktop without a lot of the flexibility Active Desktop has. For instance, Konfabulator widgets cannot register file handlers or generally be loaded/unloaded on demand by the OS itself such as the OpenDoc modules could. Konfabulator widgets also cannot automatically update themselves with new code as it becomes available (unless of course it's implemented internally in the widget), unlike Active Desktop widgets which can.
And to the guy who said that this was more "Integrated" than Active Desktop, it's really not. If you put a local (d)html page on your drive in the "Local" internet "zone" (sorry bout the lame m$ terminology), scripts on the page can be given access to local resources they need to do anything these Konfabulator widgets could do. You can script active desktop widgets in VBScript, JavaScript, Perl, Python (Using ActiveState PerlScript/PythonScript), and using these languages access enough of the underlying API that you could easily put an entire little application in there (think PerlScript + Win32::GUI app in an Active Desktop widget)
Don't get me wrong; I hate Active Desktop and Windows and I think the entire idea of little desktop apps will be forever relegated to 'toys' and other useless crap, but I think Active Desktop is a superior implementation over Konfabulator because of automatic updating and scripting language flexibility. At least it has a better name and doesn't sound like it belongs in KDE.
~GoRK
Interestingly enough, webcams are the only thing that adorned my Active Desktop space for the longest amount of time. After a while the resources running Active Desktop weren't worth it.
Take a look at this crap to see a rather extreme example of Active Desktop.
Woo. Active Desktop comes to the Mac about four years after everyone decided it was pretty much useless. Oh well, I guess some people will really like to fool with it and say how great and superior it is to active desktop even though it is the same exact fucking thing. I made lots of little Active Desktop widgets at one point that are now lost to the annals of time I suppose. It was fun while it lasted.
~GoRK
You really like that heap stuff, don't you buddy?
In 1993 - 1994, they used to let secondary schools and in some cases even elementary schools and middle schools register .edu domains. If they open it up to vocational institutions, they also should open it up to the legitimate primary and secondary schools they took it away from before.
.edu and create a lot more subdomains for .edu to handle this kind of stuff.. Training institutions and other vocational schools can get a domain in .voc.edu after a certain amount of time or whatnot, other schools can register in elem.edu or somesuch..
Even better, they ought to move the k12 subdomain to