[t029248]

perhaps a strange idea but isn’t there a way to make a hardware clone of the xbox?
hardware which understands the software instructions, but being more powerful!
i always understood the xbox is a pc, so why not going that way.

with the new an powerful hardware,   usb 2.0 and more memory the pvr could be realized. it could be seprate or all integrated...

if mod chips can be engineered, then why not a complete clone?

[Hullebulle]

perhaps a strange idea but isn’t there a way to make a hardware clone of the xbox?
hardware which understands the software instructions, but being more powerful!
i always understood the xbox is a pc, so why not going that way.

with the new an powerful hardware,   usb 2.0 and more memory the pvr could be realized. it could be seprate or all integrated...

if mod chips can be engineered, then why not a complete clone?

erhmm... may i ask why?
if you wanna an app similar to xbmc but on a pc try mediaportal.

[yngvebn]

perhaps a strange idea but isn’t there a way to make a hardware clone of the xbox?
hardware which understands the software instructions, but being more powerful!
i always understood the xbox is a pc, so why not going that way.

with the new an powerful hardware,   usb 2.0 and more memory the pvr could be realized. it could be seprate or all integrated...

if mod chips can be engineered, then why not a complete clone?

i guess paying $50-$100 on a engineered modchip, and getting the best program on earth, xbmc, for free is better than spending a year constructing an xbox-clone. "xbox 2" is beeing released next year anyway (maybe).
also, everyone who wanted this had to either buy the clone from someone who knew what they were doing, or know what they were doing themselves.

alot of hazzle for basically nothing. i love xlash's idea, and i can't wait to see the result!

[Xlash]

that's what gamester17 point to my attention. xbmc is not a native operating system like xbox linux and no usb drivers are availaible. progress hasn't been made to take use of them.

as a results, i'll stick to the more open and interesting plan, the antenna2 lan project. i'm working to know more of all the pieces, and i will be in discussion with friendtech this week, thanks to gamester17.

stay tuned

[t029248]

i only tried to give some ideas!
of course i agree the antenna2 lan project is the main thing, but hardware is crucial so i thought about using more powerful xbox hardware, or itx form factor pc hardware. if i’m not mistaken this antenna2 lan project will need separate “custom” hardware and software. of course it would be more convenient to use already engineered hardware. (is a dreambox usb tv tuner add-on not an option, since it’s using linux, and there is a big community supporting it? ) (or a xbox modchip with usb2.0 support could do the job, if someone could create a driver for the tuner of course!

well i’m definitely going to stay tuned…..to stay where this is going….

[Xlash]

i understand your idea but i do not have the money and time to buy a few xboxes, to test and install a usb 2.0 that would be useless in that case. by doing this, we will minimize our marketing (if any is ever done) because it will apply to only a small group of people (xbox + usb2.0 + i want a dvr).

in the other hand, that other project which may look easier offer a product for pc owner, xbox owner, mac owner, and will be compatible for next generation. moreover, it will support a wireless lan broadcast (via the router) so one would not have to worry about where is xbox is in the house.

finally, from a technical point of view, the xbox isn't design to accomodate a tv-tuner hardware integrated. i haven't studied enough the xbox motherboard, but that would require more study that i would do. for the usb2.0 on the mod, note that it will still require an interface to the xbox! it's like adding a firewire port to the mod. it's useless till it's connected to the bus of device, understood as a firewire port, and as a processor to process is data.

in conclusion, thanks for your idea, it's by thinking of everything that we will figure the best product out, but i think you could forget about that usb. xbmc cannot access already integrated usb 1.1, has already a good stream capability, and i think that we could use that over reinventing the wheel.

c ya!

[Gamester17]

this antenna2lan project will need separate “custom” hardware and software. of course it would be more convenient to use already engineered hardware.

your right that it would be more convenient and easy to build on existing hardware (a tv=>network box that is, not a replacement xbox),
however there is a big downside to it and that is cost! and cost is very important if planing to mass market it (which drives cost down more).
sure it would be relativly simply to use ex. via embedded platform with a mini-disto of linux on it, but that would be much more expensive.
(one would simple get one with built-in tv-turner or pci adapter, put the os and software on compact flash and add a small psu in a box)
even though it wouldn't be a smart idea to go that way for a final product it could be useful when building a concept prototype box(?)

a xbox modchip with usb2.0 support could do the job, if someone could create a driver for the tuner of course!

haha, funny :lol:  that's impossible (best you can achieve via a modchip is parallell/spi interface, forget video on those :p there's no pci slots),
and besides even if it was possible you did not read my last posts; the problem with xdk/xbox is not speed but the usb-drivers and interface,
(if it's too hard to code support for a usb-tv-turner on xbox own usb-ports, how hard would it be to also add usb 2.0 ports on top of that)

ps! please, eveyone, move/keep the disucssion about usb tv-turners to its existing thread (link) in suggestions (unless your a guru coder)

[Gamester17]

on second thought, if you do not plan to make the hardware part a commerial-product that people buy but rather a open/public-project where people has to buy the parts themself and build the box, then the modular design of a via embedded platform (or similar) might be a best way to go, (even though it will be much more expensive to for end-users to build, it will be very easy to put together and find/buy parts for it).

exampels of possible via embedded platform solutions for this project:
- micro form-factor: nano-itx epia n mainboard, mini-disto linux-os on compact flash memory card with ide to compactflash adapter,
  tv-turner can either be a custom mini-pci adapter* or a standard usb tv-adapter (preferable with built-in hardware mpeg-2 encoder).
  use an internal mini-itx psu or a dc-dc adapter/convertor with a suitable external ac-dc power-supply. then put in all in a small box.
- medium to small form-factor: mini-itx epia series-mainboard**, mini-disto linux on compactflash memory in a built-in compactflash slot,
  (**there's many models in via mini-itx series to choose from, even a few with built-in compactflash and/or cardbus/pcmcia slots)
  tv-turner can either be a custom/retail pci adapter* or a standard usb tv-adapter (preferable with built-in hardware mpeg-2 encoder).
  use an internal mini-itx psu or a dc-dc adapter/convertor with a suitable external ac-dc power-supply. then put in all in a small box.

* = custom mini-pci adapter, or custom (low-profile) pci adapter, an own designed adapter the project create/build (and sell) yourself.
     this way you get control over what exacly the adapter feature, (you can even make different models) (make them fw upgradable)

since all via epia mainboard series are all native x86 cpu platforms which is ideal to run linux and existing applications on, same goes for hardware and their device drivers. you could quitte easily code a specicial mini/micro-distro of linux-os os it and make it boot really fast by using a hybernation image. using a small compact flash memory module for storage of the os is a cheap and noise-less option. an cardbus (pcmcia) slot would be ideal for an optional wireless-network-adapter. one downside with the via epia mainboard series is that none have a built-in tv-turner/encoder so you would have to add that either via a mini-pci adapter if select the nano-itx platform (don't think there are any existing mini-pci adapters on the market so a custom built adapter would be the only option then), or a existing usb adapter (whcih is not ideal since more expnsive and would have to be plugged in externaly), or best a pci adapter if choose to use a mini-itx platform, that pci adapter could either be a custom built one or an existing retail one, with a pci adapter you also have an option of selecting a half-hight/low-profile one or standard size plus you can either install it direct in the via epia mainboards pci slot or you can get a "v" bracket/covertor to install the card horisontaly so the adapter lies paralell with the mainboard so can fit it all in a smaller box. of course choosing a modular designed x86 platform also has other percs such as the already mentioned options of having/selling different models, one model could have one pci adapter for analog-tv/radio and a other model have a pci adapter for digital-tv/radio, a third with both, a forth with a pci card with multiple tv-turners so can record more than one channel at once, and a fifth be in a bigger box with harddrive slot(s) so could be used as a nas (network attatched storage) box with samba and/or local-storage for the pvr or even function as a vcr-server. ...hell, it could even work as a standalone tunnel-box with xlink kai or xbconnect for xbox system-link online gaming if the pc-side if it was ported to it.

i will stop ranting for this time, just rememeber that even though existing via platform is easy to use and flexible, it's much more expensive.

ps! if any case, no matter if you plan to use existing hardware or built your own from scratch if you plan to make the project open source and platform independand then ask other relative osd developers if they like to help/join, such as devs from mplayer, xine, xmms, vlc, mythtv, etc.

[Xlash]

that via type embedded motherboard is way too expensive for the project. moreover it is built for mobile multimedia and has much more power that what is really needed. i don't think people are ready to spend 200$ to buy the adapter. personnaly i won't aim at that for 2 reason : 1-the price, 2- the useless power.

i'm planning to conceive a small embedded system without every components of a regular pc. it could end with a boot flashrom, a cpu, some memory and cache memory, a data bus for i/o and a small adapter for lan broadcast. i didn't think about the power supply needed yet to convert regular voltage to cmos compatible. with that solution, it's gonna be cheap, small, noiseless

on the other hand, most of the processing job will be left on the xbox cpu (for recording and buffer size). i have still to check about real-time mpeg encoding. it may need too much processing power for that small unit.

[winterbox]

athough you were getting abit caried away gamester17 with some of the mini-itx rant  :d. i do like the idea of keeping the tv wantena  software abstract. i guess the kind of kit that is being considerd would be underpowerd to run the  mono  framework (plarform agnostic .net vm) as this we be sweet for many types of hardware and os or  would you be looking at c to squess as much horse power out of the embeded device as possible. (personal opinion comming up) i feel that the more abstract and the larger the platform suface area the more bods you could atract to the project
(i suppose c was the origonal portable language /write onece compile some places/) my comments are driven from my desier for an isolated  meda stoage device also


davidk

[Xlash]

there is no current thinking about implementing a remote storage.

you could do that easily with your pc, or your xbox with bigger hdd. (even at the same time, same channel stream on the pc). (i don't want to use a pc for that piece of hardware "tv2lan". it isn't answering the project needs.)

for the code in c, there is a lot of possibility that the software will be so simple (in the means of the calculation), that it will be easier and faster if writter in assembly (for a x86 cpu). i repeat, for now, i'm looking as the end interface hardware (xbox, or pc) would do most, if not all the processing job. i want to keep that add-on as simple as possible

[floink]

a side note: if/when you come to that state in the project. if youre planning to design your own casing of this device and not use a standard of the shelf casing, i would gladly help you out with the design of it. i'm an industrial designer with experience in working with hardware design (from a visual design/production point of view). still, of the shelf stuff tend to be cheaper, which is a big issue here...

[Gamester17]

i'm planning to conceive a small embedded system without every components of a regular pc. it could end with a boot flashrom, a cpu, some memory and cache memory, a data bus for i/o and a small adapter for lan broadcast. i didn't think about the power supply needed yet to convert regular voltage to cmos compatible. with that solution, it's gonna be cheap, small, noiseless

i too think that is the best and cheapest way to go if you plan to manufacture/market/sell them, (just remember that most end-users could not build one like that themselves so they would have to buy a preassembled box). as for powersupply for a such box i think an internal dc/dc-converter be best as then you could sell/use it with a country specific ac/dc adapter.

though i don't think you should totally dismiss a via mini-itx platform just yet (so keep it in thought in the back of your head), as by taking the via mini-itx route you could skip most if not all the hardware research and development steps, thus get a final product on the shelves for end-users to buy much sooner, (if that does well then you could design and build a cheaper but improved follow-up model from scratch, ...just look at microsoft with the first xbox and xbox2, hehe). my guess is that if you only include the bare-necessities on a via mini-itx platform you could get it down in price so much that you could sell it in shops or on the internet for under $100/£100/€100: just think about it, both friendtech and via are based in taiwan, so are many ram-memory, tv-turner and psu manufacturers, plus if get a company in taiwan or china to mass-manufacture the box then they buy every part at bulk-prices direct from the original manufacturer so they get a much better price on each part which in turn brings down the price of the complete box. here is a hardware-recipe example/idea; take the cheapest via mini-itx mainboard with built-in compactflash slot, add a 32mb compactflash card and 64mb ram-memory, also add a cheap/value pci tv-turner adapter plus a 90-degree pci slot bracket, all you then have left is the powersupply (dc-dc converter internally in the box and an external ac-dc adapter) and the box itself, i believe even the cheapest via mini-itx mainboard come with a powerfull enough cpu to do mpeg-2/ts encoding on-the-fly in software thus you would not have to pay for a tv-turner adapter with video-encoding-chip in its hardware.

i'm looking as the end interface hardware (xbox, or pc) would do most, if not all the processing job.

i have to raise an objection to that as i strongly disagree with taking that approach. it is not a good idea for many reasons; streaming uncompressed video (like raw avi) takes a lot of network-bandwith, streaming uncompressed video is also be very memory-intensive on the end-interface side (xbox/pc), plus if you're even thinking bout using the end-interface (xbox/pc) as a pvr which actually record/store the stream onto its permanent storage (harddrive) then the video must already be compressed or compressed on-the-fly before stored as uncompressed video takes up way to much harddrive space (even for short-time temporary/intermediate storage), and finally i do not believe xbox cpu is powerful enough to encode/compress video on-the-fly, ((especially not if you're also planning to watch the stream at the same time, or do time-shifting). i strongly believe that it will be required for the external-tv2lan-box to do all the encoding/compressing to either mpeg-1, mpeg-2, mpeg-ts or mpeg-4 before streaming it onto the network (of those formats i think mpeg-2, or better mpeg-ts is the best compromise), the compression could be done by the tv2lan-box on-the-fly via either a dedicated hardware chip (those exist for mpeg-2), or via software (libmpeg2 or ffmpeg) if the cpu in the tv2lan-box  is powerfull enough (libmpeg2 and ffmpeg are both ported to many platforms so the cpu does not have to be x86, it can be a arm or other risc cpu), ...again just remember to make the codec/software on the box upgradable via firmware, even if the hardware-encoder-chip method is chosen.

oh, and yes i also think that assembly and/or c should be the programming languages for both tv2lan box and the library/api

ps! some project-name ideas (and code/market-names for the hardware-box); "tv2lan", "wantenna", "nettenna" or "lantenna", ...if they are not already taken(?) all are play with words combination of lan, network, tv and antenna)

ops, that became rather a long rant too but please take the time read it, it should give you one or two good ideas

[Xlash]

it has giving me some ideas, thanks a lot for the post i didn't find it long enoug :-p

first, i know the idea of another small pcinabox is interesting for it's power, it's future possibility and it's easy design. however, that could already be done via a pc running linux with a video-in card. the result is a pc in the way for the pvr.

however, i agree that i doubt the xbox could handle real-time video, pvr functionnalities, buffering, recording and encoding.

that lead to the point : it is gonna be crucial for the normal xbox, to receive a mpeg stream. i dunnot know yet if the dreamx could handle a good encoding by itself too. knowing the ram is already use in part by xbmc, a part for buffering, there won't be much left of 64 mb or even 128. even a top pc would like to redistribute is tv-in functionnalities to a dedicated machine to free up is work power.

they may not be a real need for a full via motherboard, because their are some chip encoder. some of those encoders are now optimize with mpeg4 compatibility.

i'll start looking at 'em, to find how much they cost, what do they need (as some look to be full independant if a video signal is coming), etc.

i want to conclude by saying that the actual tv2pc2lan is already possible via software, and won't be the direction the project will take. i know it would be easier for everyone if they can just set-up a small box, but it is not a full answer to our needs. i will still have to evaluate the final cost (individually, for someone to set up one), and i will keep it in back of my thoughts for now, as a alternative option.


==> floink :
thanks for your offer, i'll be sure to keep it in mind. it's a great way to show how everybody could set up their ressources to give a hand! however, it may end in the hands of a business to conceive the final retail box. keep watchin' ;-)

for the name, i like a lost tv2lan, it's clear, cool and attractive. (edit:and google didn't find any meaning for it! what do you think guys?

[Gamester17]

they may not be a real need for a full via motherboard, because there are some chip encoder. some of those encoders are now optimize with mpeg4 compatibility. i'll start looking at 'em, to find how much they cost, what do they need (as some look to be full independant if a video signal is coming), etc.

yes a encoder-chip in the box will be needed to encode/compress the video on-the-fly. i didn't know encoder chips for mpeg-4 existed (are you sure?,  if they do then they are probebely more expensive than the mpeg-2 chips), i only knew that there were decoder chips for mpeg-4 available, and that mpeg-2 encoder chips been available for some time, (note that mpeg-2 and mpeg-ts use same/similar compression but different headers so i believe a chips that can encode mpeg2 could probebely be made to stream mpeg-ts, and mpeg-ts is designed from streaming which mpeg-2 is not directly optimized for). you should also note the difference between "hardware-assisted encoder-chips" and full "hardware-encoder-chips", "hardware-assisted encoder-chips" as the name gives away only assist with the encoding like a co-processor and thus you would still require a main-cpu in addition to that chip. full hardware-encoder-chips does all the processing themself meaning you do not require to add a cpu in addition in the box, (these full hardware-encoder-chips are the ones that most of those new standalone dvd-recorders you can buy use). (i know "hardware-assisted encoder-chips" are available for mpeg-2 and mpeg-4, but again i haven't heard of any full hardware-encoder-chips for mpeg-4, only for mpeg-2).

i want to conclude by saying that the actual tv2pc2lan is already possible via software, and won't be the direction the project will take. i know it would be easier for everyone if they can just set-up a small box, but it is not a full answer to our needs. i will still have to evaluate the final cost (individually, for someone to set up one), and i will keep it in back of my thoughts for now, as a alternative option.

at least mentioned several possible routes to go and we now all know which ruote you prefer for this specific project

for the name, i like a lost tv2lan, it's clear, cool and attractive. (edit:and google didn't find any meaning for it!

persoanly i rather prefered lantenna however google'd for it and it's already taken, also wantenna and nettenna is taken.

btw, wonder what ever happened to the dreamix-project? (they announced their concept for a mass-market commersial pvr/dvr adapter on xbox-linux a couple of years ago), rumour was that they were gonna do it via a network-based-tv-adapter, last i heard they abandoned the xbox as a platform, (ah, found their website www.dreamix.tv and looks like they're developing their own hardware platform for a set-top-box pvr, not a standalone network-tv-turner at all, and not even a mention of the xbox).