Paul Johnson wrote: >> The card I am using in this ThinkPad T43 laptop at the moment is >> an Orinoco Classic Silver PCMCIA card. >> Actually, I have it working at this instant under NetworkManager (NM). >> Typically, it didn't start when I re-booted under NM. >> Then when I booted again it did. > > Most wireless drivers have an option to turn on verbose debug > information. I've done it with ipw3945. I suggest that would be your > next step. Yes, thanks, I did wonder about that, and I will look into it. But I did enquire on the orinoco mailing list, and nobody mentioned such an option. > I think you are mistaken in saying that "/sbin/iwlist scan" is not > useful. It tells you if your computer can see any wireless routers, > and it gives their names and channels. Sometimes getting > system-config-network to work properly is as simple as setting the > channel, name, and wep key correctly. I should probably have said "not useful in my case". I know the ESSIDs and channels in my case. (I actually have two APs on my small system - or rather one AP and a PCI WiFi device on my desktop which I can use as an AP.) I usually set CHANNEL=Auto , but I've never found the CHANNEL setting had any effect. > Similarly, "/sbin/ifconfig" shows you if the wireless device has > obtained an IP number. It is important to know that. It shows if the > card has a driver loaded. Again, my driver (orinoco_cs) is always loaded. My problem arises in bringing the interface up. Once it is up it works fine - at least that is usually true ... > I'm a little puzzled that you say you have 2 identical laptops with > orinoco_cs and one does associate and the other does not. If you have > exactly the same hardware in the same place at the same time, it seems > to me that they will behave in the same way. IF they don't, it makes > me suspect a pcmcia card or port is damaged. Yes, it is puzzling. I don't think anything can be damaged, as (a) it works fine under Windows, and (b) once it starts working it continues indefinitely, until I turn the laptop off. One possibility that has occurred to me is that there may be some kind of timeout under Linux. I think WiFi is slightly more likely to work if I boot on mains power. > Nevertheless, on the one that sometimes joins and does not, I'd say > you are close to getting at the source of the trouble, but relying on > NetworkManager won't help you figure out what it is. NM is just one of the half-dozen resources I try. If it works who am I to complain? But I would like to understand what is going on. > Instead of rebooting, I usually do this > > /sbin/service NetworkManager stop > /sbin/service NetworkManagerDispatcher stop > > /sbin/rmmod iwl3945 > /sbin/rmmod mac80211 > > (use /sbin/lsmod to see modules you have loaded, for me these are the > ones) > > Then if you try "/sbin/ifup eth1" again, it should trigger a reload of > the modules, and you can durn dmesg to try to track the association > process. Thanks for the suggestion, I'll try that next time. > I used to have an orinoco_cs silver card and it did work pretty well. > But, to be honest, all of the laptops we have now have the builtin > wireless. I guess you choose your poison. Actually, I bought this laptop (ThinkPad T43) on eBay, assuming it had in-built WiFi, as all the T43's I had seen did have this. But I found there is a rare version that has some kind of modem instead. In fact, I have bought - also on eBay - a mini-PCI WiFi module, but installing this seems to involve a major operation dis-assembling the laptop. > Out of curiosity, I went and read some emails in the orinoco_cs user's > list. This one seems to indicate problems with that driver and newer > kernels > http://sourceforge.net/mailarchive/forum.php?thread_name=47A8BD2A.5040608%40research.telcordia.com&forum_name=orinoco-users Thanks for that. It does sound depressingly like my experience. I didn't have this problem with earlier kernels - I mean many kernels ago, maybe around 2.6.16 or so. > The part that still has me stumped is that two similarly configured > systems with orinoco_cs behave differently. I wouldn't have expected > that one. Yes, it is the un-repeatability of the problem that causes me the most frustration. But you have given me two or three things to try, next time WiFi fails on this laptop. (Maybe it will never fail again ...) Thanks very much for that. -- Timothy Murphy e-mail (<80k only): tim /at/ birdsnest.maths.tcd.ie tel: +353-86-2336090, +353-1-2842366 s-mail: School of Mathematics, Trinity College, Dublin 2, Ireland
