HD100 --> HD200 --> ??? How soon?

[stanger89]

Originally Posted by garberfc

Hey Folks,

Thanks for all the replies. Some great lively discussions...

Why would you want a BD reader as part of an extender?

Primary reason being so you don't need a Blu-ray player next to your extender.

Why not on the PC itself so that the media can be accessed from all the extenders?

Because I'd say 99.9% of the time when somebody wants to watch a movie disc, they want to stick it in the machine sitting next to the TV they want to watch it on. They don't want to go down the hall, or to another floor, or equipment room to put the disc in a different machine. They you have to go hunt down the unnamed disc in the imported library.

People accept putting it in the server because it's possible, the only way, and means they don't need a separate player, but I think most would much rather just be able put a disc in the extender and hit "Play DVD" from the menu.

I really doubt anybody cares about being able to access a single movie disc on multiple extenders (now a pile of ripped discs is a different story there).

Originally Posted by reggie14

I pretty much agree with everything Stanger said. Some kind of Sage server-in-a-box could really make Sage accessible to a much wider audience. And, I think it would be worth quite a bit more than $300 if hard drives and tuners are included in the box (perhaps roughly $300 if you had to purchase hard drives and tuners separately, which could make sense too).

Without drives/tuners it's really no different than the HD200 (except for the processor power), so I wouldn't expect that to need to be any more.

I kind of wonder how successful it would be though, given that dealing with and setting up IR blasters is probably something novice users wouldn't do.

That's no different than a Tivo or Moxi DVR. Additionally (this is just blue skying) but a self-contained Sage DVR box would probably be much, much easier to get CableCard certified than their PC app, so a SageTV DVR box could be Sage's "in" to get CableCard support.

So, for the audience that would benefit from a Sage server-in-a-box most, you'd essentially be limited to QAM and ATSC tuning. At times I thought that market was too niche, but I'm not entirely sure. There did/does seem to be some excitement surrounding the $250 DTVPal DVR, which was essentially a more limited version of what we're envisioning. I would think a full-functioning Sage box would be worth quite a bit more.

But you've got to be careful too and not go too high. That said, if it's not more expensive than Tivo or Moxi, while retaining Sage's flexibility I think it would have a great shot at competing with them.

I think you could even imagine some really cool things too. The server boxes could have wireless-N routers in them. They could sell HD200-like clients that could just stream wirelessly off the server.

Frankly I think wireless is just asking for a support/bad PR nightmare. But maybe MOCA.

Or, you could add another server box which could work as a network encoder. Though, the ability to record more than 2 channels at once is really just useful if you can record cable/satellite, so you'd want analog inputs and IR blasters.

I think you're looking for too much "in the box". All they need is a baseline box with ATSC/QAM tuner and HDD to get started. But let that box support the HD HomeRun, USB UIRT, Hauppauge HD PVR, and maybe a couple other USB/network tuners and you've got a system with capabilities that would rival a PC based setup but with the better "out of the box" experience.

At this time standard def is fine, which you could imagine going in the server boxes themselves. Maybe some day you could imagine putting HD-PVR-like internals in a server box.

Frankly I don't think SD is fine for a DVR product anymore. Not with essentially every major network offering an HD feed, and more and more providers offering them all. I also don't think Sage would want to go down the "all in the box" path like most DVR makers. Sage could set themselves apart by support external, user-added tuners. Like supporting some PC tuners. Then people could configure the overal system to their needs.

I don't really think this is going to happen. But, I think it has the potential for Sage to either make lots of money, or lose lots of money (as there would be a sizable initial investment required).

Not really, they've got all the driver logic for supporting the HD PVR and HDHR on linux already, all they'd really need is a bit more powerful box. And heck a lot of us would probably buy that box even as just an extender for the boosted performance.

Of course I'd still like to see a "high end" extender with something like an HQV, ABT, Gennum video processor in it, and a faster processor for better menu response.

[briands]

Count me as one hoping for a BD player/Sage Client combo machine.

[reggie14]

Originally Posted by Taddeusz

To really be effective in the market they would need to have CableCard support just like the Tivo. Otherwise it would be quite a limited device.

To be option for everyone they'd need cable card. But, in theory, Sage could go after 2, relatively distinct groups. Their current niche, which wants control and customization, and the people out there that don't need, or can't get, cable. The excitement over the DTVPal DVR suggests there are quite a few people out there that want a DVR but don't want to pay for cable.

Originally Posted by stanger89

Without drives/tuners it's really no different than the HD200 (except for the processor power), so I wouldn't expect that to need to be any more.

But, adding support for those things would cost money. Certainly there would be hardware costs. SageTV server isn't all that light. I see it pegging the processor in my server while it does things like EPG updates, and you wouldn't want to interrupt playback or menu browsing when those go on. And you want additional USB ports, maybe even eSATA support. But, probably more significantly is the software cost. The software in a special purpose Sage server would quite a bit more complicated than that in an extender. And, I'm not sure how much work it would be to port Sage server over to an embedded device.

So basically, it seems like a server box without drives and tuners ought to be at least $300, and I think probably more.

That's no different than a Tivo or Moxi DVR. Additionally (this is just blue skying) but a self-contained Sage DVR box would probably be much, much easier to get CableCard certified than their PC app, so a SageTV DVR box could be Sage's "in" to get CableCard support.

I'm not sure cable card support would be very likely without compromising a lot of features that make Sage more compelling than other options. And, I'm not sure Sage would really be able to get the resources to first design a system capable of certification, and then pay to get it tested and certified. I'm not sure what cable card certification costs, but I'm familiar with costs involved with certifying cryptographic devices and some other more complicated systems, and it's not cheap.

But you've got to be careful too and not go too high. That said, if it's not more expensive than Tivo or Moxi, while retaining Sage's flexibility I think it would have a great shot at competing with them.

Because I think cable card is out of the question, I don't think you can really expect Sage to ever seriously compete with TiVo. Once you get away from the HTPC enthusiast market, I think Sage can really only hope to compete with things like the DTV Plus DVR. A smaller market, yes, but one that's still significantly bigger than Sage's current market. And, I think that makes the cost issue even more difficult. TiVo and Moxi boxes are moderately expensive. But, things like the DTV Plus DVR are pretty cheap. It would be basically impossible for Sage to offer something in the $250 range that would include all hardware and software necessary to record 2 channels at once. So, the question is, will people that are cheap enough to avoid cable/satellite TV be willing to spend twice as much on something like Sage? I'm not sure.

Frankly I think wireless is just asking for a support/bad PR nightmare. But maybe MOCA.

Perhaps, but wireless is probably required if you ever want to get Sage out to the plug-and-play demographic. I'm not even sure coax networking would be a viable alternative. At least in my experience, 802.11n works pretty well, but that could easily change as more people get 802.11n devices. Wider adoption of 802.11n could make it difficult to get consistent streaming in high-density housing. But, I think reliable wireless streaming is perfectly possible with improved buffering techniques. Apparently the new beta (and upcoming firmware) will improve buffering on extenders. Even in relatively crowded environments, 802.11n should have enough available bandwidth if you can get big enough buffers on the clients. Recording over wireless could be more problematic, but the buffering idea still applies.

I think you're looking for too much "in the box".

I'm inclined to agree with you on specifics, but maybe not with your wider view. I think non-enthusiasts would want nearly everything "in the box". And, I think enthusiasts would be likely to build their own servers so they can do things like comskip and placeshifting. So, the question is what to do about people in-between. You're probably right to say things like HD capture devices should be out of the box. But, I think that means you effectively cede a large segment of the population to TiVo.

Partially, I brought up the issue of HD tuners in the server box as a way of arguing against the usefulness of network encoders. Network encoders are mainly only useful when you want more of the tuners that are in the server box.

Not really, they've got all the driver logic for supporting the HD PVR and HDHR on linux already, all they'd really need is a bit more powerful box.

Maybe we have vastly different ideas for the intended market for a Sage server box. My thought is that it is attractive to people that wouldn't feel comfortable setting up a Sage box (as opposed to wouldn't want to set one up). So, that means that things would need to work together mostly automatically, and things would need to be very reliable. I don't really think a Sage port to a embedded linux device similar to the HD200 would be very painless, but even if it were I think there would still be a long way to go. To just use one example, look at the current amount of work required to set up multiple IR blasters on a system. You can't even use the Hauppauge blaster, and the the multi-zone setup on the USB-UIRT certainly isn't trivial. The HD-PVR is apparently getting there on reliability, but based on the number of people doing workarounds (setting a fixed output resolution, using analog audio, tape over IR receptors, etc.) it doesn't seem to be there yet.

[stanger89]

Originally Posted by reggie14

To be option for everyone they'd need cable card. But, in theory, Sage could go after 2, relatively distinct groups. Their current niche, which wants control and customization, and the people out there that don't need, or can't get, cable. The excitement over the DTVPal DVR suggests there are quite a few people out there that want a DVR but don't want to pay for cable.

A lot of the customizations don't have any impact on CableCard, the only big things that would probably be affected are comskip and transcoding (which aren't insignificant).

But, adding support for those things would cost money. Certainly there would be hardware costs. SageTV server isn't all that light. I see it pegging the processor in my server while it does things like EPG updates, and you wouldn't want to interrupt playback or menu browsing when those go on.

There's no reason an EPG update or other background task need interrupt playback. The SOC is doing all the decoding work and the background tasks can just be run at a lower priority.

And you want additional USB ports, maybe even eSATA support. But, probably more significantly is the software cost. The software in a special purpose Sage server would quite a bit more complicated than that in an extender. And, I'm not sure how much work it would be to port Sage server over to an embedded device.

There shouldn't be much software work at all. Sage already supports a number of devices under linux. The extenders don't run on a significantly different OS. No doubt there'd be work there though.

So basically, it seems like a server box without drives and tuners ought to be at least $300, and I think probably more.

I doubt that, considering how much more capability/hardware the HD200 had vs the HD100, I don't see a PVR version requiring a higher price due to the CPU or software work (sans tuners/drive). Like I said I think Sage could put together a box with tuner and drive for $300.

I'm not sure cable card support would be very likely without compromising a lot of features that make Sage more compelling than other options.

Well the client-server model should still be allowed (MS does it with extenders), UI customizations should work, only comskip and transcoding, which mess with the recordings seem problematic.

And, I'm not sure Sage would really be able to get the resources to first design a system capable of certification, and then pay to get it tested and certified. I'm not sure what cable card certification costs, but I'm familiar with costs involved with certifying cryptographic devices and some other more complicated systems, and it's not cheap.

Well Sigma probably has all the crypto stuff built in, Sage would just need to leverage that. Definitely not trivial, and I'd be rather surprised to see a CableCard supporting SageTV DVR regardless, I definitely don't think it's likely.

Because I think cable card is out of the question, I don't think you can really expect Sage to ever seriously compete with TiVo. Once you get away from the HTPC enthusiast market, I think Sage can really only hope to compete with things like the DTV Plus DVR. A smaller market, yes, but one that's still significantly bigger than Sage's current market. And, I think that makes the cost issue even more difficult. TiVo and Moxi boxes are moderately expensive. But, things like the DTV Plus DVR are pretty cheap. It would be basically impossible for Sage to offer something in the $250 range that would include all hardware and software necessary to record 2 channels at once. So, the question is, will people that are cheap enough to avoid cable/satellite TV be willing to spend twice as much on something like Sage? I'm not sure.

I don't think Sage has a hope of competing with the DTVPal, nor would they want to. That's only Cheap because Echostar gets the development "free" from their Dish DVRs. If I were Sage I wouldn't go after the low end. I'd be aiming at the Tivo/Moxi area where Sage can best them on PC and whole-house integration.

Further, while Sage wouldn't appeal to the whole Tivo market, I think there's a reasonably large market that sort of bridges the HTPC/CE market. People able but gudgingly willing to build an HTPC/media server to accomplish what they want, but would be very happy with an "appliance" that gives them things like customization and clients/extenders.

Perhaps, but wireless is probably required if you ever want to get Sage out to the plug-and-play demographic.

That's the problem though, you want plug and play, but Wireless is far from it.

I'm not even sure coax networking would be a viable alternative.

Why's that, MOCA seems to be working quite well for people. Giving them close to 100Mbps network performance but without having to run wires.

At least in my experience, 802.11n works pretty well, but that could easily change as more people get 802.11n devices. Wider adoption of 802.11n could make it difficult to get consistent streaming in high-density housing. But, I think reliable wireless streaming is perfectly possible with improved buffering techniques. Apparently the new beta (and upcoming firmware) will improve buffering on extenders. Even in relatively crowded environments, 802.11n should have enough available bandwidth if you can get big enough buffers on the clients. Recording over wireless could be more problematic, but the buffering idea still applies.

The problem with wireless is that it's so hard to predict, and impossible to guarantee if it will work well or not. And when you start getting into the appliance market, having to fanagle possitioning to get good, consistent performance just won't fly. Better IMO to just skip wireless and deal with the complaints about lacking it than to add it and deal with far more complaints of "novices" being unable to make the system work overwireless, and the associated bad PR.

I'm inclined to agree with you on specifics, but maybe not with your wider view. I think non-enthusiasts would want nearly everything "in the box". And, I think enthusiasts would be likely to build their own servers so they can do things like comskip and placeshifting. So, the question is what to do about people in-between.

Here's the thing, you need some sort of basic bundled capability, eg a hybrid tuner with ATSC/QAM/Cable/AV inputs (your standard maybe HVR-2250 capability) to get you out the door. But there's this market for "more". Moxi and Tivo are great unless you want 3-4 tuners and 2+ rooms in your system.

As noted there's hugely more people in the "don't want to mess with a PC" market than the HTPC market, but still quite a lot of people want more than what Tivo/Moxi provide (single room DVR). So Sage, if they went the modular route, like allowing use of an HDHR for additional tuners, and HD PVRs for recording HD cable/sat, combined with it's client/server/extender capabilities could make a nice place for itself between the Tivo and the HTPC markets.

You're probably right to say things like HD capture devices should be out of the box. But, I think that means you effectively cede a large segment of the population to TiVo.

It's important to realize that no standalone box (Tivo/Moxi) can record HD component, period. So a standalone box with that capability could be pretty huge.

Maybe we have vastly different ideas for the intended market for a Sage server box. My thought is that it is attractive to people that wouldn't feel comfortable setting up a Sage box (as opposed to wouldn't want to set one up).

Perhaps, but I'd focus it the other way, people who would rather not set up a PC, even though they could (they'd fell comfortable).

So, that means that things would need to work together mostly automatically, and things would need to be very reliable.

Oh, absolutely, no disagreement on that. A "box" must act like an appliance.

I don't really think a Sage port to a embedded linux device similar to the HD200 would be very painless, but even if it were I think there would still be a long way to go.

I'm thinking you'd be surprised. Sage already has a standalone mode, that has does things like the server, like manage the library, and running all the UI logic. And they've already got drivers for quite a number of tuners ported to Linux. So they're a good way there.

To just use one example, look at the current amount of work required to set up multiple IR blasters on a system. You can't even use the Hauppauge blaster, and the the multi-zone setup on the USB-UIRT certainly isn't trivial.

Ok, you're getting into more stuff than I was talking about, I was talking about getting the capability there to support things, while you're looking more into the logisics of making it really easy to set up. I'll defer to the linux users here for how some of those issues compare on the Linux side.

The HD-PVR is apparently getting there on reliability, but based on the number of people doing workarounds (setting a fixed output resolution, using analog audio, tape over IR receptors, etc.) it doesn't seem to be there yet.

No doubt, the HD PVR isn't bulletproof yet.

[Djc208]

Originally Posted by garberfc

Why would you want a BD reader as part of an extender? Why not on the PC itself so that the media can be accessed from all the extenders?

Personally I don't really need or want to keep my movies on the Server. When I want to watch a movie I only do it in my living room on the "good" equipment. Keeping the movies on the server at most saves me from having to put in the movie.

But to keep them on the server I'd have to upgrade to the HD200 since the 100 doesn't have the horsepower for all BD disks, plus adding the BD drive and ripping software to one of my PCs and the extra storage to keep them all. And I still can't utilize all of the BD features and extras via HD200.

I see it less as a extender with BD capability and more as a BD player that can run Sage. Which many of the new players are probably capable of. I will eventually be buying a BD player for my home theater, this would just enable me to reduce the number of devices, wiring, and setup headaches vs. just adding another device.

Plus Sage already showed a prototype of sorts, so someone liked the concept.

[reggie14]

It's almost certainly the case the Sage isn't going to make a box like this, but it's sort of fun to think about what could be. I do think there's a decent-sized potential market out there for an easy to use, prebuilt Sage box.

Originally Posted by stanger89

A lot of the customizations don't have any impact on CableCard, the only big things that would probably be affected are comskip and transcoding (which aren't insignificant).

I think cable card is more of a problem for the "control" side, not the customization side. Placeshifting would almost certainly be out of the question, for instance. But, certainly it would impact some things on the customization side. Comskip, as you mention. Some of the things that could have problems wouldn't be found on Sage-box systems, but cable card inclusion would likely mean changes to the Sage core. So, that could cause issues on PC-based systems and PC-oriented customizations (the video tools plugin comes to mind).

There shouldn't be much software work at all. Sage already supports a number of devices under linux. The extenders don't run on a significantly different OS. No doubt there'd be work there though.

While I have a CS background, this is well outside my area of expertise. But, I'm under the impression that having drivers for x86-based systems doesn't mean you're not going to have issues when you port it to a MIPS architecture. And, on the issue of porting, my understanding is that your headaches don't really go away when you're dealing with java apps. But, as you said, maybe the HD200 already implements a good portion of the server code, between the standalone mode and client/extender. Though, it seems like it's the code responsible for handling tuners that would be the real problem here, and that's not in the HD200.

Well the client-server model should still be allowed (MS does it with extenders), UI customizations should work, only comskip and transcoding, which mess with the recordings seem problematic.

I'm not sure client-server isn't a problem. Yes, Microsoft did it, but 1) they have more resources for development, and 2) they have more weight to throw around. The extender-side isn't the issue- they're specialized devices so you basically know how they will behave. The bigger concern is what happens on the PC side. There Microsoft has a greater ability to protect things like keys. Perhaps you could imagine the Sage-box and PC-based Sage products not interoperating, but I don't think that's a viable option.

Mainly, I just think getting a client-server architecture past whatever testing is required for cable card certification would add cost in various places.

Why's that, MOCA seems to be working quite well for people. Giving them close to 100Mbps network performance but without having to run wires.

I don't know exactly. Partly because I think people want wireless (right or wrong) because they think it would be easier. Partly because it actually does have compelling features, like being able to put TVs places that you otherwise couldn't very easily.

Better IMO to just skip wireless and deal with the complaints about lacking it than to add it and deal with far more complaints of "novices" being unable to make the system work overwireless, and the associated bad PR.

Again, it really seems like wireless-N streaming ought to work well with large enough buffers. No, probably not perfectly, but probably well enough. I kind of wonder how much of the anti-wifi tone around here is due to the wMVP, which was just a disaster, presumably because it's wifi chip had horrible performance compared to other 802.11G devices.

In any case, how much bad PR do you see from the big review sites on the MCE wireless extenders compared that to what the reviews say about the lack of wireless on the HD100/200? I think wireless ends up being a very important issue for novices out there.

If I were Sage I wouldn't go after the low end. I'd be aiming at the Tivo/Moxi area where Sage can best them on PC and whole-house integration.

In general I would agree, if I thought Sage really had any hope for getting cable card support. Since they don't, I think they're out of the high-end market, except for enthusiast part of that market. I think only the enthusiasts are likely to put up with HD-PVR-like devices. So, I think you're mainly looking at a Sage box competing against ATSC/QAM DVRs, of which the DTV Plus DVR is really the only thing. You seem to think that Sage could, in theory, keep prices moderately close to the DTV Plus DVR, so I don't see the problem.

People able but gudgingly willing to build an HTPC/media server to accomplish what they want, but would be very happy with an "appliance" that gives them things like customization and clients/extenders.

I just don't see a Sage box appealing that much to those people. It sounds like what you basically have in mind isn't that much easier than just getting a WHS box, adding a tuner and setting it up yourself. You don't seem to be setting that much of a higher standard for making setup and configuration easier, and that's much harder than simply installing Sage on an existing PC.

As noted there's hugely more people in the "don't want to mess with a PC" market than the HTPC market, but still quite a lot of people want more than what Tivo/Moxi provide (single room DVR).

It's less that they "don't want to mess with a PC" and more that they want it to "just work". A modular approach would be great, if things would "just work" after you plug them in. I think I can basically imagine that working with the HDHomeRun (though, I think you probably want something that would plug directly in to the box, as I think wired networking is a problematic requirement) and USB HD tuners. But, I think it's hard to imagine the HD-PVR ever living up to that. (though, maybe that doesn't matter if you cede the non-enthusiast cable TV user market to TiVos and cable STBs).

It's important to realize that no standalone box (Tivo/Moxi) can record HD component, period. So a standalone box with that capability could be pretty huge.

Why? I'd much rather use a cable card than mess around with component capture and IR blasters. Yes, there can be some limitations on what you can record, but those aren't meaningful to very many people.

[stanger89]

Originally Posted by reggie14

It's almost certainly the case the Sage isn't going to make a box like this, but it's sort of fun to think about what could be. I do think there's a decent-sized potential market out there for an easy to use, prebuilt Sage box.

Indeed.

I think cable card is more of a problem for the "control" side, not the customization side. Placeshifting would almost certainly be out of the question, for instance.

That's an interesting questions though, because Dish has Slingloaded DVRs for both Dish and Cable.

But, certainly it would impact some things on the customization side. Comskip, as you mention.

Even that is somewhat interesting, because it's just the detection that's tricky

Some of the things that could have problems wouldn't be found on Sage-box systems, but cable card inclusion would likely mean changes to the Sage core. So, that could cause issues on PC-based systems and PC-oriented customizations (the video tools plugin comes to mind).

I don't know, the playback and UI elements are pretty separate. They'd need to do a bit of work on the playback side for sure, but from a customization perspective I don't think much would have to change. The UI is quite well isolated from the actual playback engine through the Sage API.

While I have a CS background, this is well outside my area of expertise. But, I'm under the impression that having drivers for x86-based systems doesn't mean you're not going to have issues when you port it to a MIPS architecture.

True, but it should be much closer than x86 Windows -> MIPS Linux.

And, on the issue of porting, my understanding is that your headaches don't really go away when you're dealing with java apps. But, as you said, maybe the HD200 already implements a good portion of the server code, between the standalone mode and client/extender. Though, it seems like it's the code responsible for handling tuners that would be the real problem here, and that's not in the HD200.

Well think of it this way, there are several layers necessary, there's the scheduling logic error, which if done in Java, should transport pretty much without issue since it doesn't interact with the OS or any hardware. Next is the tuner managment layer, here you're going to have to do work (I assume) on an OS to OS basis. This would interact with the OS/Drivers but not directly with the hardware, so I'd expect this to come relatively cleanly from their Linux version. Finally is the drivers themselves, which are, well, the wildcard. But these boxes still use "standard" interfaces like USB and ethernet, so my guess is that the drivers need not be much different for those. Now if they tried to integrate an ATSC/hybrid tuner there might be some more work, but even that would probably use some varient of PCI so probably not a huge issue.

And I could be talking out my you know what here. I'm not an OS/Driver programmer. Not even a "programmer", just someone who works with Software Engineers all day

I'm not sure client-server isn't a problem. Yes, Microsoft did it, but 1) they have more resources for development, and 2) they have more weight to throw around. The extender-side isn't the issue- they're specialized devices so you basically know how they will behave. The bigger concern is what happens on the PC side. There Microsoft has a greater ability to protect things like keys.

Well FWIW MS doesn't do PC clients at all so...

Perhaps you could imagine the Sage-box and PC-based Sage products not interoperating, but I don't think that's a viable option.

Agreed, they'd have to integrate support for whatever content protections they create into their PC apps. Quite likely Client wouldn't work because of how it operates, but Placeshifter might (like the Dish's Sling stuff)

Mainly, I just think getting a client-server architecture past whatever testing is required for cable card certification would add cost in various places.

Of course, but likely Sage wouldn't have to get a whole content protection system certified, just their implementation of existing one(s), which I would think would be cheaper.

But as we agreed at the top, this probably isn't going to happen.

I don't know exactly. Partly because I think people want wireless (right or wrong) because they think it would be easier. Partly because it actually does have compelling features, like being able to put TVs places that you otherwise couldn't very easily.

Oh I completely understand the appeal of wireless and why people want it. I just think until n is "proven" it's not a good thing to stick in your product. A good compromise would be to put a mini-PCI or Expresscard slot in and offer a wireless "upgrade" option. But even then you're risking the bad PR when people's environements aren't well suited to wireless.

Again, it really seems like wireless-N streaming ought to work well with large enough buffers. No, probably not perfectly, but probably well enough.

See, I don't think "not perfectly, but probably well enough" is good enough for a standalone out of the box product. But maybe I just think to much like James Dyson, that things should just work properly. Probably why I've got a URC MX-880 instead of a Harmony for my HT system

I kind of wonder how much of the anti-wifi tone around here is due to the wMVP, which was just a disaster, presumably because it's wifi chip had horrible performance compared to other 802.11G devices.

That's definitely part of it I'm sure. But just looking for n devices for my home, it's quite discouraging just how bad they are. They claim "300Mbps", but from the real tests I can find they seem like they really only do about 80-90Mbps, and that's not without interruptions, and if I read stuff right, that's also without encryption which further kills bandwidth.

In any case, how much bad PR do you see from the big review sites on the MCE wireless extenders compared that to what the reviews say about the lack of wireless on the HD100/200? I think wireless ends up being a very important issue for novices out there.

Frankly I've never read a review of and MCE extender.

In general I would agree, if I thought Sage really had any hope for getting cable card support. Since they don't, I think they're out of the high-end market, except for enthusiast part of that market. I think only the enthusiasts are likely to put up with HD-PVR-like devices. So, I think you're mainly looking at a Sage box competing against ATSC/QAM DVRs, of which the DTV Plus DVR is really the only thing. You seem to think that Sage could, in theory, keep prices moderately close to the DTV Plus DVR, so I don't see the problem.



I just don't see a Sage box appealing that much to those people. It sounds like what you basically have in mind isn't that much easier than just getting a WHS box, adding a tuner and setting it up yourself. You don't seem to be setting that much of a higher standard for making setup and configuration easier, and that's much harder than simply installing Sage on an existing PC.

I was just thinking something similar. What I'd really like to see, is Sage get together with Netgear/Infrant and get SageTV running on a ReadyNAS Pro and/or NVX, those things have more than enough horsepower for a Sage Server, they're small, relatively quiet, have huge storage capacity, dual-gig-E. Seems overall like a perfect match. Heck my ReadyNAS X6 has a PCI slot (unused) in it

It's less that they "don't want to mess with a PC" and more that they want it to "just work". A modular approach would be great, if things would "just work" after you plug them in. I think I can basically imagine that working with the HDHomeRun (though, I think you probably want something that would plug directly in to the box, as I think wired networking is a problematic requirement) and USB HD tuners. But, I think it's hard to imagine the HD-PVR ever living up to that. (though, maybe that doesn't matter if you cede the non-enthusiast cable TV user market to TiVos and cable STBs).

I think you're in trouble if you try to take on the cable company's DVR.

Why? I'd much rather use a cable card than mess around with component capture and IR blasters. Yes, there can be some limitations on what you can record, but those aren't meaningful to very many people.

Well one big one is CableCard doesn't work with Dish or DirecTV, the two carriers with the most HD by far in most of the US.

[reggie14]

Originally Posted by stanger89

But just looking for n devices for my home, it's quite discouraging just how bad they are. They claim "300Mbps", but from the real tests I can find they seem like they really only do about 80-90Mbps, and that's not without interruptions, and if I read stuff right, that's also without encryption which further kills bandwidth.

I wasn't going to respond, since I think we basically see each others' points, but I wanted to respond to this, since it's very much related to my field. Because of the way the 802.11 protocols work, you simply can't expect more than about 100mbps in ideal conditions for 802.11n. The 300mbps number is mostly bogus, but I understand where they get it. If it weren't for the protocol, the hardware basically could handle 300mbps. But, since you need to design the protocol for shared access in an environment where you can't sense collisions, you need to use a protocol which slows things down. While the manufacturers do seem to be advertising the theoretical max less than they did during the 802.11g days, they still make ridiculous statements about speed. I can't think of any way to justify statements like a particular 802.11n router is 15 times faster than 802.11g routers. So, I agree there's a problem there, but it's a marketing problem, not a technical one.

Anyways, encryption is what I really wanted to address. There's really no reason not to enable AES encryption on new routers. AES encryption is almost always faster than TKIP encryption, for a paradoxical reason. AES is a much, much slower encryption algorithm than TKIP (in its defense, it is also much more secure). So much slower than AES encryption operations are typically performed on separate crypto chips (at least on things like routers). TKIP is fast enough that it can just use the host CPU. It's my understanding that you typically can't get more than 60mbps of performance when using TKIP encryption because of the strain on the host CPU. I'm not saying there's a zero performance hit when turning on AES encryption, but it should be pretty minimal compared to the early ways of wifi and WEP encryption.

The moral of the story is to turn on WPA2-AES encryption, unless you have some device that you need to use that doesn't support it. And don't even bother with MAC filtering. I really don't understand why people enable it. It's completely trivial to get past, compared to WEP/WPA/WPA2 cracking where even the weak schemes take some amount of work to get past.

What I'd really like to see, is Sage get together with Netgear/Infrant and get SageTV running on a ReadyNAS Pro and/or NVX, those things have more than enough horsepower for a Sage Server, they're small, relatively quiet, have huge storage capacity, dual-gig-E. Seems overall like a perfect match. Heck my ReadyNAS X6 has a PCI slot (unused) in it

Those have always seemed ridiculously priced to me. It seems like you could imagine Atom-based systems costing far less. Maybe there is some good reason for the prices, but it seems like it's mostly just that they are relatively high-end (for SOHO-level devices), niche products able to command a high profit margin. It seems like there ought to be better alternatives if the idea is to make Sage accessible to a larger population of consumers.

I think you're in trouble if you try to take on the cable company's DVR.

It depends on what you mean. TiVo basically does that, with some degree of success. Cable company DVRs will, for the foreseeable future, vastly outnumber other options, but I think there is a viable (smaller) market for alternatives.

Well one big one is CableCard doesn't work with Dish or DirecTV, the two carriers with the most HD by far in most of the US.

True. Though, I just can't imagine ordinary people turning to component capture. Doesn't DirecTV have a TiVo box now anyway? But Dish Network's DVR software is downright awful. I'd probably be willing to go to extreme measures to avoid using that (though, switching to DirecTiVo would be the obvious choice).

[stanger89]

Originally Posted by reggie14

I wasn't going to respond, since I think we basically see each others' points, but I wanted to respond to this, since it's very much related to my field. Because of the way the 802.11 protocols work, you simply can't expect more than about 100mbps in ideal conditions for 802.11n. The 300mbps number is mostly bogus, but I understand where they get it. If it weren't for the protocol, the hardware basically could handle 300mbps. But, since you need to design the protocol for shared access in an environment where you can't sense collisions, you need to use a protocol which slows things down. While the manufacturers do seem to be advertising the theoretical max less than they did during the 802.11g days, they still make ridiculous statements about speed. I can't think of any way to justify statements like a particular 802.11n router is 15 times faster than 802.11g routers. So, I agree there's a problem there, but it's a marketing problem, not a technical one.

Oh, understood on the limits of wireless tech. This is part of the reason why I'm so hard on it for use in the home with streaming high bandwidth content. It was mainly just the observation that I'd finally actually started looking an "n" devices (I bought my laptop with N just in case but since it's just my laptop never really looked for an AP). Given all the "talk" about how much better N is, I was kind of expecting something more.

Anyways, encryption is what I really wanted to address. There's really no reason not to enable AES encryption on new routers. AES encryption is almost always faster than TKIP encryption, for a paradoxical reason. AES is a much, much slower encryption algorithm than TKIP (in its defense, it is also much more secure). So much slower than AES encryption operations are typically performed on separate crypto chips (at least on things like routers). TKIP is fast enough that it can just use the host CPU. It's my understanding that you typically can't get more than 60mbps of performance when using TKIP encryption because of the strain on the host CPU. I'm not saying there's a zero performance hit when turning on AES encryption, but it should be pretty minimal compared to the early ways of wifi and WEP encryption.

Interesting, honestly I hadn't read that much. I'd just been researching WAPs over on smallnetbuilder (which "seems" to have good reviews/tests), and saw a few comments about enabling encryption causing a ~50% throughput hit. I wish they actually had tests for that though.

The moral of the story is to turn on WPA2-AES encryption, unless you have some device that you need to use that doesn't support it. And don't even bother with MAC filtering. I really don't understand why people enable it. It's completely trivial to get past, compared to WEP/WPA/WPA2 cracking where even the weak schemes take some amount of work to get past.




Those have always seemed ridiculously priced to me. It seems like you could imagine Atom-based systems costing far less. Maybe there is some good reason for the prices, but it seems like it's mostly just that they are relatively high-end (for SOHO-level devices), niche products able to command a high profit margin. It seems like there ought to be better alternatives if the idea is to make Sage accessible to a larger population of consumers.

Rediculous is, as always, a matter of opinion. They command the price though because they stand pretty much alone in terms of features and performance (there's not a lot out there that supports 4-6 drives with redundancy and can easilly saturate a gig-E lan).

Of course we've got some "high end" type people here. I've just always thought Sage on a ReadyNAS would be a perfect match. Sage, Squeezecenter, webserver, and whatever else they've got, I wouldn't need my PC server anymore.

It depends on what you mean. TiVo basically does that, with some degree of success. Cable company DVRs will, for the foreseeable future, vastly outnumber other options, but I think there is a viable (smaller) market for alternatives.



True. Though, I just can't imagine ordinary people turning to component capture. Doesn't DirecTV have a TiVo box now anyway? But Dish Network's DVR software is downright awful. I'd probably be willing to go to extreme measures to avoid using that (though, switching to DirecTiVo would be the obvious choice).

Supposedly Dish's DVR software is "wonderful" lately, though I haven't tried it (it is tempting to enable the DVR features on my 211 just to see how they work.

[reggie14]

Interesting, honestly I hadn't read that much. I'd just been researching WAPs over on smallnetbuilder (which "seems" to have good reviews/tests), and saw a few comments about enabling encryption causing a ~50% throughput hit. I wish they actually had tests for that though.

Take the review of the WRT320N, for example. They found a 60% hit when enabling WEP/WPA-TKIP and a 12% hit when enabling WPA2-AES. The actual performance penalties vary from device to device, but the TKIP penalty is pretty much always much higher than that for WPA2-AES, since the AES operations are done in hardware.

Originally Posted by stanger89

They command the price though because they stand pretty much alone in terms of features and performance (there's not a lot out there that supports 4-6 drives with redundancy and can easilly saturate a gig-E lan).

Can they? Maybe I've been interpreting reviews wrong, but I've never been terribly impressed with the performance results out of those things. For instance, the SmallNetBuilder review of the ReadyNAS Pro says that, for very large files, write speeds top out at 70mbps (RAID0) and 40mbps (RAID5). That doesn't seem all that impressive to me.

I could be completely wrong, but it seems like the actual hardware costs involved in things like the ReadyNAS are pretty cheap. Is there some reason you couldn't take an Atom board with 4 SATA ports (I think that's what Atom boards top out at), have it run something like NASLite (or, perhaps, WHS) and get something basically similar for a substantially lower price?

Supposedly Dish's DVR software is "wonderful" lately, though I haven't tried it (it is tempting to enable the DVR features on my 211 just to see how they work.

Who said that? My mom has a Dish DVR (though, no matter how many times I tell her, she doesn't realize the pause, fast forward and rewind buttons actually do anything), and I find the interface atrocious. Somehow she keeps adding various shows to Dish's equivalent of favorites. It took me quite a while to figure out how to remove those favorites so the DVR wouldn't keep changing the channel on her to record those shows.

I basically think a DVR should be simple enough to use that I can perform relatively common tasks without going to a manual. If that's not possible, the interface should at least be intuitive enough that after doing it once it's basically trivial to remember how to do it again. I don't think the Dish DVR software meets either of those standards.

[stanger89]

Originally Posted by reggie14

Take the review of the WRT320N, for example. They found a 60% hit when enabling WEP/WPA-TKIP and a 12% hit when enabling WPA2-AES. The actual performance penalties vary from device to device, but the TKIP penalty is pretty much always much higher than that for WPA2-AES, since the AES operations are done in hardware.

Ah, interesting....

Can they? Maybe I've been interpreting reviews wrong, but I've never been terribly impressed with the performance results out of those things. For instance, the SmallNetBuilder review of the ReadyNAS Pro says that, for very large files, write speeds top out at 70mbps (RAID0) and 40mbps (RAID5). That doesn't seem all that impressive to me.

That's MB/sec, not Mbps, they're getting 82MB/sec, (650Mbps). And other people who have tested it in the default X-RAID (IIRC) config with more drives were getting ~100MB/sec.

I could be completely wrong, but it seems like the actual hardware costs involved in things like the ReadyNAS are pretty cheap. Is there some reason you couldn't take an Atom board with 4 SATA ports (I think that's what Atom boards top out at), have it run something like NASLite (or, perhaps, WHS) and get something basically similar for a substantially lower price?

It's not the hardware you're paying the premium for with a ReadyNAS, it's the software/support. Sure you can build your own cheaper, but then you've got to manage the redundancy yourself (buy a RAID card, or something, WHS has a much higher space penalty for redundancy). It's the overall package. Try to find and off the shelf NAS that can do 6 drives, 100MB/sec throughput, offers lots of plugins, and you can just plug in and go.

Who said that? My mom has a Dish DVR (though, no matter how many times I tell her, she doesn't realize the pause, fast forward and rewind buttons actually do anything), and I find the interface atrocious. Somehow she keeps adding various shows to Dish's equivalent of favorites. It took me quite a while to figure out how to remove those favorites so the DVR wouldn't keep changing the channel on her to record those shows.

Just the impression I had, maybe they've improved since your moms, or maybe my impression was just wrong. It's not something I researched since I use Sage and aren't looking to change.

[Djc208]

Why? I'd much rather use a cable card than mess around with component capture and IR blasters. Yes, there can be some limitations on what you can record, but those aren't meaningful to very many people.

I haven't found cable cards much more reliable than an HD-PVR from the cable box. I consitently have minor issues with the CC in my TV not properly tuning scrambled channels. I usually have to re-sync it after a power outage or I get garbage, and support is abysmal. The cable companies don't realy like CC since they make less money on rentals, which I think is one reason why they've been slow to adopt a new revision that allows interactive menus and PPV/on-demand features. Most new TVs don't even offer CC slots any more.