While this may be a reasonable conclusion to draw, I would sugest that the data to date on the efficacy of integrating DOCSIS devices with Set-Top-Boxes, or simply slapping a PVR/DVR onto a STB does not imply that this will be an easy thing to do.
There are many, many, issues, some of which are details, some of which are huge. Most of these are not apparent until you have been involved in the Cable industry for a while.
I think it's great that Tivo has finally started moving in this direction. The question will be whether or not the companies already in this space will be able to deliver products acceptable to the cable companies and end-users before Tivo gets in the space with it's brand and established user base.
The next two years will be interesting ones to watch in the Home-Theatre PC/Digital convergence/All in one Set Top Box space.
At least, consumers finally will win by having real competition for advanced services.
Back in college (about '93 or so) I got frustrated enough by this to write a little X app to build multiple cut-buffers and rotate through them. Something like 10-20 lines of C.
Bound it to a twm menu, and ta-da.
Highlight, click, highlight, delete, click, paste.
I'm sure someone has written something far more usefull in the same genre by now..
I have a very hard time believing that a Rice EE has trouble finding better work than that. Are they willing to leave swamp land?
What CS courses did he/she take? Whats the GPA?
Seriously, if they have decent grades, and are willing to move to the Northwest I might be able to point them in the right direction.
Windows update is already massively load balanced across multiple server farms. They use both a DNS based load balancer (F5 3DNS) and local area load balancers (F5 BIGIP). The server farms are in a number of locations.
Early, and not so early in the implementation of this, a number of people were concerned that Microsoft was attacking them because the 3-DNS's would create probes from each datacenter to the end-users system. I'm not sure if that is still being used.
I have no knowledge of how they Akamized so quickly since I haven't been involved with this project in years. However, it should be pointed out that the BIG-IP's make Akamizing content a very simple matter.
I'm not shilling for F5- I no longer own any of the stock, haven't been an employee for years, and I'm now just a reasonably-satisfied customer.
While I am happy that you like the Sonicwall and Watchguard firewalls, and seem to think PIXen are nice, I would have to disagree. Having used all of these products, and several others as well as home-grown firewalls, I do not hold any of these three in high regard.
Perhaps more relevant however, is that Watchguard firewalls are Intel and Linux based, Pixen (at least the older ones) are Intel based (they were originally made by Network Translations. I have heard from reliable sources that Sonicwalls are VxWorks based.
Now perhaps, these devices meet the definition of being built from the ground up to do a specific task- having worked for several equipment vendors who have sold appliances as well as custom-hardware based network devices, I have a more jaded view of what this really means in practice.
One of the few security devices that is primarily custom-hardware based is the netscreen product line. I have no idea if they use VxWorks or another embedded OS, but would be somewhat surprised if they run a fully home-grown software package.
So, in general, I see value in your argument about specific device for specific jobs- if only from an auditability and management perspective.
However, most of the specific-job devices in fact run on essentially commodity hardware and are based in large part on general purpose software- no matter how much people in marketing want you to believe otherwise.
Slashdot Mirror
While this may be a reasonable conclusion to draw, I would sugest that the data to date on the efficacy of integrating DOCSIS devices with Set-Top-Boxes, or simply slapping a PVR/DVR onto a STB does not imply that this will be an easy thing to do. There are many, many, issues, some of which are details, some of which are huge. Most of these are not apparent until you have been involved in the Cable industry for a while. I think it's great that Tivo has finally started moving in this direction. The question will be whether or not the companies already in this space will be able to deliver products acceptable to the cable companies and end-users before Tivo gets in the space with it's brand and established user base. The next two years will be interesting ones to watch in the Home-Theatre PC/Digital convergence/All in one Set Top Box space. At least, consumers finally will win by having real competition for advanced services.
Back in college (about '93 or so) I got frustrated enough by this to write a little X app to build multiple cut-buffers and rotate through them. Something like 10-20 lines of C. Bound it to a twm menu, and ta-da. Highlight, click, highlight, delete, click, paste. I'm sure someone has written something far more usefull in the same genre by now..
I have a very hard time believing that a Rice EE has trouble finding better work than that. Are they willing to leave swamp land? What CS courses did he/she take? Whats the GPA? Seriously, if they have decent grades, and are willing to move to the Northwest I might be able to point them in the right direction.
Yep. They do. BTW, I Picked up an Ultra-2 for something like $20 a while back. Not hard to do on E-bay.
Windows update is already massively load balanced across multiple server farms. They use both a DNS based load balancer (F5 3DNS) and local area load balancers (F5 BIGIP). The server farms are in a number of locations. Early, and not so early in the implementation of this, a number of people were concerned that Microsoft was attacking them because the 3-DNS's would create probes from each datacenter to the end-users system. I'm not sure if that is still being used. I have no knowledge of how they Akamized so quickly since I haven't been involved with this project in years. However, it should be pointed out that the BIG-IP's make Akamizing content a very simple matter. I'm not shilling for F5- I no longer own any of the stock, haven't been an employee for years, and I'm now just a reasonably-satisfied customer.
While I am happy that you like the Sonicwall and Watchguard firewalls, and seem to think PIXen are nice, I would have to disagree. Having used all of these products, and several others as well as home-grown firewalls, I do not hold any of these three in high regard.
Perhaps more relevant however, is that Watchguard firewalls are Intel and Linux based, Pixen (at least the older ones) are Intel based (they were originally made by Network Translations. I have heard from reliable sources that Sonicwalls are VxWorks based.
Now perhaps, these devices meet the definition of being built from the ground up to do a specific task- having worked for several equipment vendors who have sold appliances as well as custom-hardware based network devices, I have a more jaded view of what this really means in practice.
One of the few security devices that is primarily custom-hardware based is the netscreen product line. I have no idea if they use VxWorks or another embedded OS, but would be somewhat surprised if they run a fully home-grown software package.
So, in general, I see value in your argument about specific device for specific jobs- if only from an auditability and management perspective.
However, most of the specific-job devices in fact run on essentially commodity hardware and are based in large part on general purpose software- no matter how much people in marketing want you to believe otherwise.