Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - darrenmason

Pages: 1 ... 32 33 [34] 35 36
Users / Re: Multiple Cores
« on: August 21, 2007, 05:44:02 am »
I don't think that you can really have two cores. You would need to have only only DCERouter running, which is the main part of the core.
However, a MediaDirector can act as the interface for security devices and other things. This may do what you require. Just make sure the software devices that you require reside under the MD rather than the core in the device tree.


Users / Re: Still a little confused about diff of Core and MDs
« on: July 24, 2007, 02:09:12 am »
So I put my TV capture card in my Core? or MD? or both?
All of the above works. MythTV has two components, a frontend and a backend. Each MD (including Hybrids) will launch a frontend when you press the TV button. Multiple backends can work together in Master/Slave configurations. Therefore in any machine that has a TV Capture Card it will run a backend (linuxMCE runs a backend on all MDs and core by default). The backends will work together so that they appear as one and all frontends can use any of the capture cards. My suggestion would be to put them all in the core as you are guarenteed that this will be running.

Do the IR devices to turn on my TV and Reciever connect to my MD? or Core? (This one seems simple but would like to make sure)
Usually the MD or a dedicated device like a GC100. You basically want it near them because IR must be line of sight.

Does the MD need a HDD at all if its PXE booting?
No, but you can has a HDD with another OS installed if you want and sort of sual boot them.

So if I want to add new devices to my MD(such as Bluetooth), I plug it into the MD I want, but configure it on the Core? Cause the MD is Volitile, right?
Sort of, you are plugging them into MDs and configuring them through the Pluto interface or the admin web application (which is hosted on the core).
btw: The whole file system for the diskless MD can be found on the CORE and this can be useful for checking logs etc without turning the MD on.

In the case of IR devices, if I am suppose to have IR devices connected to the MD(See above question #2) do I need multiple IR Recievers for each entertaiment device. For example, say I have a TV, VCR, and a Reciever. Each can be controlled by their remotes, which work via IR, so I need 1 IR Reciever for each device? Cause isn't IR "line of site" comm.  Furthermore do the IR entertaiment devices (TV, VCR, Reciever) IR sensor need to be pointed directly at the IR device going to the MD?
I think you may have confused the terminology somewhat. You need an IR sender to control your IR devices. The devices themselves will have receivers built in (if they can be controlled by remote). You should only need one sender but it may have multiple IR LEDS which can be positioned to control multiple devices.


I can understand that these questions are elemantrary, but its more of a reasurance than anything else.

Thank you a bunch for your replies.


Hope that helps

Developers / Re: Using XBox as an MD
« on: July 05, 2007, 01:35:19 am »
It is a fair statement, and this is what I had tried as well.

The main problem, at least as far as Pluto was concerned - assume it still applies for LMCE, is that there is code/scripts that forces checks between the CORE and all MD's to ensure they are using the same Kernel version.

My solution to this was to recompile a matching version with the XBox patches. I still think this should probably work. The only reason I didn't finish it was that I ran out of disk space on the XBox and lost enthusiasm to play around with it.

The other problem is the OS dependencies that exist. LMCE seems to be dependant upon specific distributions and their releases, so I guess you would need to get Kubuntu 7.04 running on the XBox. I don't think that anyone has done this.

I thought there was more to the Kernel patch than just the fatx support, but I havn't looked at it for a while.

I think it is desirable to determine what is actually required for just an MD installation and work  on just installing the MD components on a running OS. This then would open up a variety of machines to being MDs including XBoxs. Once you have it running as an installed machine then it should just be a packaging exercise to turn that into a network boot.

All that said, I was sortr of hoping that if the XBox network boot would work and it had a compatible kernel then the initial install that is done when adding an MD might just work, hence I am very interested to see what Nick can get working.


Feature requests & roadmap / Re: Is LinuxMCE what I want?
« on: July 04, 2007, 01:33:41 am »
I know the first post is now quite old, but in answer to the subject question, I think the answer is no. LinuxMCE does not seem to be a good fit for your requirements.

Maybe try geexbox which has apparently just had a new release.


Developers / Re: Using XBox as an MD
« on: June 29, 2007, 03:38:04 am »

I don't know about emulators, but are you flashing to your actual bios or to a modchip?
My modchip has switchable banks so that you can flash to either and then switch between them. Therefore if you stuff one up it doesn't matter too much. I thought most modchips had this.

Sounds like you are making some progress.
You will definately need a new kernel though, I think linuxMCE is using an unmodified 2.6.20 kernel from other posts.
The XBOX patches only seem to go up to 2.6.16 - not sure if this will create problems.

Anyway, keep us informed.


Developers / Re: Using XBox as an MD
« on: June 22, 2007, 02:02:35 am »

Have you built a patched kernel for it yet?



I can't offer you much help, although I can see no reason why it shouldn't work if you can somehow tell it a different device ID and core IP.
The core will generate an Orbiter interface for a particular device ID that will only work with that Core, so if you do that on each core then as long as you can tell it the Core IP (Which you can) and the device ID (which is the bit I can't remember) then there should be no reason why it shouldn't work.

Is this what you have tried?

On a side note, how have you set up your two cores? Different subnets using different hubs/switches? Is one of them still acting as a gateway? Is DHCP still running on both?

I assume you are running both as an evaluation to see which to proceed with, is this correct?

The web orbiter is definately an option, but it is a pain to use with the way it refreshes.

On another side note, the Windows XP orbiter works under Wine on Ubuntu 7.04. Although, it does seem a bit silly to be running a windows emulated version of what is a predominately linux product. But if anyone else has managed to just run a linux Orbiter then I would be keen to hear, I had no luck. (Note, the machine does not have linuxMCE installed, this is my everyday desktop at home)


Developers / Re: Using XBox as an MD
« on: June 14, 2007, 01:36:41 am »
One question - to build stuff for the XBox, do I have to use the XDK? As a side effect, would that mean the binaries could not be included in the LMCE install as they're illegal (at least, that seems to be what XBMC believe since they will not distribute binaries)?

No, you won't be using the XDK. The cromwell bios was reverse engineered without any microsoft code and then you are just installing a linux distribution on the machine.
I might have play with it on the weekend again.

Developers / Re: Using XBox as an MD
« on: June 14, 2007, 01:31:40 am »
the sources of LinuxMCE are released under GPL. and the "pluto" kernel is not a special one. it's just the kernel from Kubuntu. so, you can apply some patches and build a custom kernel which will be used by linuxmce to boot or to install on an Xbox MD.
Where are the latest sources? When I tried to find them I could only find the original sources from the 1.0 fork than ran on Ubuntu 6.10. A few people have asked this on other threads as well.

wrt. the kernel, I have only done the above with Pluto (ie: not linuxMCE) and it certainly had a custom built kernel which I was pretty sure had some updates. Since linuxMCE does not have the same dependency on controlling the OS it may be easier to get this working.

Developers / Re: Using XBox as an MD
« on: June 13, 2007, 02:14:53 am »
I have not done much more since that thread unfortunately.

I was having no luck getting a kernel built that I could get to work using the PXE software that comes with cromwell.
Each time I tried something new I ran into a problem. Shifted my kernel compile to the actual XBox (Was doing it on the core) and didn't have the disk space, tried mounting remote drives and then got some permission problems. Sort of lost enthusiasm/motivation at this point.
Was also a bit put off as I was not getting good performance from PCs with similar specs (but more memory) to the XBox.
I think the tasks needed to do still are;
* Build a kernel using a mixture of the Pluto kernel source and XBox patches.
* Get the XBox to run this kernel, either using PXE and manually updating the CORE or by just installing it on the disk.
* Setup swap files etc on local XBox disks (even with network boot). I think this will definately be necessary.
* Then get all thej bits to work - this should not be too hard as most of the XBox linux projects have already attacked this (eg . lirc for remote and controllers, nvidia drivers, USB hubs).

Big problem with doing it with linuxMCE is that as far as I know they are not releasing the source to their releases. This is really worrying and certainly hinders anyone else from doing any modification that can be merged back into the project. This is the one reason why I still have most of my main components running Pluto (rather than linuxMCE) and I just have a single machine that I have started to play round with linuxMCE on.

There are heaps of websites telling you how to get cromwell onto the XBox, which if you just want to run linux on it - then that is all that you need.

If you get it and are still keen I would be willing to work together to try and get something working. I don't think that there is any reason why it won't work. I think it is just a matter of persistence. Most of the big problems have already been solved by others. Like I said I am a bit unsure about the performance because it looks like Orbiter and some of the other pieces seem to be very memory hungry, but it is ceratinly worth trying.

If not, then running XBMC on it works well and you can certainly see your media from the core and play it etc. You can also use it as a Myth Frontend.


I have not done this but thinking about it I can't see why it is not possible. You may need to play with some scripts though.
You may not need two XSessions, 2 screens on the same XServer should be OK.
If it will let you add two Onscreen orbiters to one MD you should get two device numbers. starts the orbiter and could probably be modified to take an optional screen number parameter I would think.
Then add the parameter to the DeviceTemplate for OnscreenOrbiter.
It should still launch all the media on the main screen.

Let us know if you have a go at this as I have thought about the same sort of setup using a cheap ebay touch screen but never done anything about it.


Users / Re: LinuxMCE focus for the future?
« on: June 05, 2007, 01:51:51 am »
Totally agree with what you are saying. Do the guys that did the code fork for linuxMCE read this? Care to comment?

Feature requests & roadmap / Re: PowerPc architecture support
« on: June 04, 2007, 01:53:13 am »
You would have to build it yourself and I would suspect that you would struggle to get all the dependencies satisfied. But you could try :)

Users / Re: LinuxMCE focus for the future?
« on: June 04, 2007, 01:44:52 am »

I guess the answer will somewhat be demand driven and the demand seems to certainly be focussed on the media components, at least judging from the forums.
If the project maintains the same code base then I assume support for everything else will certainly remain and the more people that are interested in it, the more likely that people will log issues, developers will code for it and users will test.
I also would like to see it evolve with all its features, but I must admit looking at what I use now if the media components don't work then noone would be using the rest of the system. But in media components I would also include the control of existing non-IP components such as TVs, amplifiers etc.


Users / Re: Is Linux MCE better than Pluto?
« on: May 31, 2007, 02:54:02 am »
What do other Pluto Users think?

I am also using Pluto and have been for a while now. I have downloaded everything to do a linuxMCE install and will probably setup a standalone subnet to try it on but I am still a bit hestitant.

I have a few opinions which I would like to hear feedback on;

* I have always considered Pluto an OS distribution in its own right. Sure, it is built on top of a faily standard Debian build but it really is designed and built to be in total control of the machine. I think this works well for the core. I do think it is desirable and certainly possible within the Pluto design to have pieces of Pluto software running all around the network on various devices. This is already done to some success with Orbiter software but could/should extend to any device software. linuxMCE seems to be aiming at having all the devices running on standard OS desktop platform. I think there is a mixture of the two required.

* I think linuxMCE is named inaccurately and is confusing a lot of newcomers. A media centre is highly desirable and is certainly attracting people to the project but the Pluto software, in my opinion, is primarily about integration between devices. Where devices are quite loosely defined as anything electronic. This includes software packages like mythtv, xine, asterisk etc. I think there is limited value in all the overhead of pluto for a standalone machine that is only for media viewing.

* Pluto has a great design (although I would like to see a bit more documentation in this space) but, in my opinion, has a poor implementation. Now I don't want to offend the Pluto developers that are possibly/probably reading this, but after looking at the code I would think that they would agree that the implementation has a high amount of coupling and dependency  between objects and has dirtied the code with specific workarounds for certain situations and hardware. I think that this has now caused the project to have an increasing number of 'hard to fix' software bugs causing an unpredictable project schedule and release plan. I believe they have encouraged the linuxMCE fork as hopefully a way to get some of this fixed by the wider community whilst still retaining the control of what they actually use and release.

But instead of complaining about everything I am willing to help out. I am just trying to decide which of the two projects to devote the time to.
To the other past Pluto users that have tried linuxMCE -
once you were up and running again was there much difference?
is it still conceivable to have a big dedicated core machine that sits under the house with all the disks?
has anyone built and run from the lastest SVN code and does this work OK?
Any other experiences worth sharing?


Pages: 1 ... 32 33 [34] 35 36