You have to keep in mind that the more nodes you add in a wireless LAN (or any other shared medium LAN), the more contention you get. For doing a cluster, the practical limit would likely be in the neighborhood of 10 machines... Depending of course on what you mean by "cluster".
As for a cable/dsl router system, you should check out karlbridge software ( http://www.karlbridge.com/ ). Their software is what is running under the hood of most of the "Access Point" class machines out there.
The Apple Airport Basestation, which retails for
about US$300 or so, contains a Lucent WaveLan IEEE silver card (64bit encryption), an ethernet port, and a 56K modem. I have one that I have upgraded to the Lucent WaveLan IEEE gold card (128bit encryption) and it works wonderfully. I had to use some karlbridge windows-based config software to configure it, but it works!
Beware of one flaw in the Lucent cards if you choose to try to build your own bridge with it. The Lucent cards will *not* send packets with a different MAC address than their own. Hence, bridging does not work with Lucent cards. In order to get bridging to work with a Lucent card, specialized firmware (which is not freely available) has to be downloaded into the card before it is put into bridging mode.
I realize I should have included all the relevant data in the initial question.
What I have done:
I used 'xev' to see if any events were generated for the extra keys. Only the MickeySoft keys generated any events. From my (admittedly limited) knowledge of the X event system, if the keyboard handler isn't generating a key pressed event, then nothing I do with xmodmap or xkeycaps is going to help. I first have to convince the X server's keyboard code to generate some kind of events for key presses of those keys.
Once that part is done, I'm fairly confident that I can deal with xkeycaps/xmodmap to make the keys actually useful (and yes, I intend to try to create the stuff for xkeycaps so others can use it later).
I suppose I'm a bit confused as to why X has to have it's own code to do hardware scancode conversion to software events, when (at least under linux) this seems like a kernel thing to do. I have my linux kernel being very happy about the extra keys, but X seems to just quietly ignore them. I fear I'll have to delve into the XKB extensions to get it to work.
Slashdot Mirror
You have to keep in mind that the more nodes you add in a wireless LAN (or any other shared medium LAN), the more contention you get. For doing a cluster, the practical limit would likely be in the neighborhood of 10 machines... Depending of course on what you mean by "cluster".
As for a cable/dsl router system, you should check out karlbridge software ( http://www.karlbridge.com/ ). Their software is what is running under the hood of most of the "Access Point" class machines out there.
The Apple Airport Basestation, which retails for
about US$300 or so, contains a Lucent WaveLan IEEE silver card (64bit encryption), an ethernet port, and a 56K modem. I have one that I have upgraded to the Lucent WaveLan IEEE gold card (128bit encryption) and it works wonderfully. I had to use some karlbridge windows-based config software to configure it, but it works!
Beware of one flaw in the Lucent cards if you choose to try to build your own bridge with it. The Lucent cards will *not* send packets with a different MAC address than their own. Hence, bridging does not work with Lucent cards. In order to get bridging to work with a Lucent card, specialized firmware (which is not freely available) has to be downloaded into the card before it is put into bridging mode.
What I have done:
I used 'xev' to see if any events were generated for the extra keys. Only the MickeySoft keys generated any events. From my (admittedly limited) knowledge of the X event system, if the keyboard handler isn't generating a key pressed event, then nothing I do with xmodmap or xkeycaps is going to help. I first have to convince the X server's keyboard code to generate some kind of events for key presses of those keys.
Once that part is done, I'm fairly confident that I can deal with xkeycaps/xmodmap to make the keys actually useful (and yes, I intend to try to create the stuff for xkeycaps so others can use it later).
I suppose I'm a bit confused as to why X has to have it's own code to do hardware scancode conversion to software events, when (at least under linux) this seems like a kernel thing to do. I have my linux kernel being very happy about the extra keys, but X seems to just quietly ignore them. I fear I'll have to delve into the XKB extensions to get it to work.
Thanks for all the comments so far folx.