12 messages in org.freebsd.freebsd-mobileRe: Aviator 2.4| From | Sent On | Attachments |
|---|---|---|
| Scott Hazen Mueller | Nov 22, 2000 10:13 am | |
| Duncan Barclay | Nov 24, 2000 3:56 pm |  .txt, .txt |
| Scott Hazen Mueller | Nov 25, 2000 10:42 pm | |
| Duncan Barclay | Nov 26, 2000 4:40 am | |
| Scott Hazen Mueller | Nov 27, 2000 11:46 pm | |
| Duncan Barclay | Nov 28, 2000 12:07 am | |
| Warner Losh | Nov 28, 2000 12:24 am | |
| Peter Radcliffe | Nov 28, 2000 10:05 am | |
| Scott Hazen Mueller | Nov 28, 2000 12:42 pm | |
| Warner Losh | Nov 28, 2000 12:51 pm | |
| Richard Johnson | Nov 29, 2000 9:21 pm | |
| Peter Radcliffe | Nov 30, 2000 9:34 am |
| Subject: | Re: Aviator 2.4 | |
|---|---|---|
| From: | Duncan Barclay (dm...@ragnet.demon.co.uk) | |
| Date: | Nov 26, 2000 4:40:22 am | |
| List: | org.freebsd.freebsd-mobile | |
On 26-Nov-00 Scott Hazen Mueller wrote:
Certainly, this is how I tested on 3.x, 4.1 and -current!
I suspected as much.
I note that your raycontrol dump shows that the hopset is 0x03 and not 0x1d. Did you change it? I'm not sure that the hopset code selection actually works well on rev4 cards (conflicting information in docs etc.)
Is the hopset 'HOP_SEQ'? It seems to change itself in between power-up and configure, see below. I did observe some odd behavior when issuing the command 'raycontrol -i ray0 -f 1' repeatedly - the HOP_SEQ value seemed to go through random bytes ranging from 0x02 to 0x4-something.
Yup, the hopset is HOP_SEQ and the country code stuff. I haven't really tested the country code selection. I never change it from the defaults.
Hmm, this means that the cards aren't seeing each other at all. The seperated BSSIDs and INITED flag, mean that the cards have set up seperate networks.
OK, that confirms what I was thinking.
Things to try - plug one card in and wait a couple of minutes before plugging in the next
Tried that, to no avail.
check the hopsets
I've issued the -f 1 command against both cards. As I've noted, HOP_SEQ is all over the map, don't know if that's good or bad.
I don't think that the W98 will change anything on the cards in a permanant fashion.
I may try it as a last resort...
Try moving them away from the monitor - the scan coils may be interfering with the radio.
Tried that as well as I can, given the available space in my office. The 'base' station is about 4 feet above the nearest monitor. I did my last set of tests with the remote about 5 feet away from the nearest monitor. Of course, if the radios are sensitive to noise within these ranges, I can hardly use them anyway... :-(
Other info...
www.webgear.com is now offering a flash upgrade from plain 2.4 (& rev 4) to 2.4 Pro (& rev 5) firmware. Visit the site and find the Win2K upgrade - it includes most of a flash floppy in the kit, just add one bootable DOS floppy and run. I've applied this upgrade to my cards, no additional luck. But, as I understand it, rev 5 is worth having anyway.
The rev 5 is an upgrade to true 802.11 compliant carsd. I haven't tried it partly becuase it changes some of the software in the driver and I haven't finsihed coding it up! But you haven't yet got to that stage yet - it only affects things when the cards are seeing each other.
I ran a still more careful test tonight. I started with both systems (zork and zorba) powered down, with the ray cards inserted. I powered zork (the base station) up first. The raycontrol output, before issuing any networking commands, was
[snipped]
These diffs look about "right", but I can't check it all parameters as I don't have rev 5 cards.
Have you tried the cards under Windows to make sure that they work?
Suggestions more than welcome.
\scott
Duncan
---
________________________________________________________________________ Duncan Barclay | God smiles upon the little children, dm...@ragnet.demon.co.uk | the alcoholics, and the permanently stoned.
________________________________________________________________________
To Unsubscribe: send mail to majo...@FreeBSD.org with "unsubscribe freebsd-mobile" in the body of the message