Our solution will replace the entire PB2K computer with brand new parts. The question is would people rather purchase the complete solution from us (motherboard, ram, ps, etc) or only the hardware/software we make? If we sold just our parts we would specify a list of compatible parts but it would require that the user has more than average techniocal expertise. If we sell only the parts our support would be somewhat limited due to possibility of user error.

Why offer only one type of package? You're going to get users on both sides of the fence. It also depends on cost. If the full meal package is $500, I'd rather build my own given that I am on the highly technical side. If the full meal is $200, though, I may just pay to get it done rather than spend limited free time building it myself.

I love the idea of putting it together myself and gaining the experience. That is why the call it a hobby

Good input, poll will be updated.

Honestly we would like to offer just the parts we make and let people build it themselves. But we think that the number of "computer experts" is small compared to novice users.

I think you should sell complete drop-in boxes, diy kits, and parts/software.

I've got a few extra pc/104 industrial cpus (wide temp range) with flash drives that would be ideal for this but I'd like to buy software and any other misc. parts from you guys. But if that's not possible a kit would be the next best thing for me.

I would prefer to build my own and just purchase the special parts from you.

If you provide recommended hardware and software specifications, those not wanting to build their own could purchase a pc from numerous places. Just provide detailed instructions, with photos, on installing your parts. This would help keep you out of the server business. For those wanting the total plug and play solution, just offer a premium priced package.

For those who've said they'd prefer to build their own, I'm curious your thoughts about not getting support. In other words, if your not able to get things working on your own, how will you feel if we say we cannot suppor or help you?

Obviously if some goofy combination of bios drivers, os drivers, flakey hardware, os bugs, etc.. prevents our solution from working, we'd really not be able to offer support unless we charged for it, which I don't think the 3 of us were planning or wishing to get involved with.

Does this change anyone's opinion?

Of course from our perspective in regards to offering support, the only way we can guarantee something will work is to build and test it ourselves. So for us, it's our prefered choice, but we recognize that there may be high demand for the DIY crowd.

It depends on what OS you're running. If it's proprietary and requires an expensive support contract on the buyers part then it would matter. If it's a public domain product with a support forum somewhere then it wouldn't matter to me at all. But it would be useful to the DIY crowd to know exact configurations that are known to work.

sellenoff wrote:For those who've said they'd prefer to build their own, I'm curious your thoughts about not getting support. In other words, if your not able to get things working on your own, how will you feel if we say we cannot suppor or help you?

It would probably be best to follow the model that other similar "official homebrew" providers use. They usually offer a specific (often small) set of models that they have tested and know work with their software. Sometimes they even sell those units but not always. The thought is that if you use hardware on that list, and do it properly, it will work. The developers will then make a reasonable effort to help if you have trouble. If the user chooses for whatever reason to use hardware not on the supported list they are on their own. Make this policy clear up front, put the hardware choice onus on the user, and you should be all set.

ChadTower wrote:

sellenoff wrote:For those who've said they'd prefer to build their own, I'm curious your thoughts about not getting support. In other words, if your not able to get things working on your own, how will you feel if we say we cannot suppor or help you?

It would probably be best to follow the model that other similar "official homebrew" providers use. They usually offer a specific (often small) set of models that they have tested and know work with their software. Sometimes they even sell those units but not always. The thought is that if you use hardware on that list, and do it properly, it will work. The developers will then make a reasonable effort to help if you have trouble. If the user chooses for whatever reason to use hardware not on the supported list they are on their own. Make this policy clear up front, put the hardware choice onus on the user, and you should be all set.

Steve and I talked about this today and it's probably the way we'll go. I have already setup a forum here so users can post problems and we can post solutions and faq's.

Chuck wrote:Steve and I talked about this today and it's probably the way we'll go. I have already setup a forum here so users can post problems and we can post solutions and faq's.

Yes I think Chads idea is the best. If I buy the product, I would probably first test with parts I have laying around, and if they dont work follow the list of supported hardware.
When excluding the pc-hardware shipping, customstax and tax for international orders is also reduced.

I like buying just the parts - I've built a MythTV box, etc. A benefit to doing this is that it should encourage you to make PB2k8 work on a wider range hardware. I know you can't support every setup, but if you supply everything there is a danger of creating a stove-pipe solution ... which is of course the dangerous situation we find ourselves in today!

I think this is the single trickiest thing we will have to deal with. I am about to start testing on all sorts of machines/cpus/chipset/graphics cards. In the end we want this to work with as many parts as possible. Realistically, we are only three guys and can only support so much ourselves. We really want everyone's experience with this solution to be top notch. The trick is that in our testing so far pc components vary so much that it's a nightmare to try and support everything. Try calling Microsoft for support with their os... and they're microsoft

We have pretty much decided (anything can change btw) we will have to sell complete boxes the we will guarantee to work. I think we will also specify components that have been tested so people who like to bargain hunt and assemble the machine themselves will be able to do so. But there will be minimum spec that must be met. The minimum specs, so far, are very reasonable and are on the lower end of today's current machines.

We want to be as flexible as we can but we also want to be able to support people. The work we are doing right now will really determine that outcome. As I'm typing this I'm setting up a high speed/inexpensive usb drive for testing.

The issue is going to be that if the whole setup is overly dependent on a particular computer/motherboard and/or specific graphics card, etc. as being the only ones that work, then I'd potentially just be searching for a different antiquated motherboard/graphics card in 5 years (assuming a failure occurred). I hear the RFM motherboards are what fail the most frequently.

I don't think you are being real clear how much of an expert someone needs to be when you talk about the kits.

My personal guess is that you'll be dealing with a mix of people but if you want to market to the larger group, you'll want to avoid having people do a bunch of soldering. If you just have to connect something to a standard connector (even if it's a slot on the motherboard) then I think you're ok with most folks.

On the software side, I think it best, if you are able, to have a straight forward user friendly GUI installation and not require programming or intimate knowledge of obscure/complicated systems settings or you'll turn yourselves into a software helpline.

Man, I don't see any reason this would have to involve soldering. About the most extreme thing hardware wise I could see this needing would be maybe some crimping - but a harness adapter isn't hard to have made or make by hand on a low volume product.

I also wouldn't worry about needing specific models of hardware. Most off the shelf PC hardware is pretty widely compatible and backwards compatible. You will have a minimum baseline requirement, sure, but there would be no reason to use specific proprietary functionality from any specific vendor that I can think of when speccing this out in my head.