On Fri, 12 Mar 2004, Jerry Kilpatrick wrote: > ALL NEW INFORMATION: > > I've done quite a bit of meticulous testing and found out some things that > blow parts of my previous mail out of the water.
Okay, I'll ignore your earlier mail. > 1) Rebooted my laptop, plugged in the digital camera. Mounted it, copied > files to it. Unmounted it, removed the device. Plugged it back in > mounted it and my files were there that I had added. > (Figured this would work, but you always need a control variable right?) > > 2) Rebooted my laptop, plugged in my pen drive. Mounted it, copied > files to it. Unmounted it, removed the device. Plugged it back in > mounted it and my files were there that I had added. > (Didn't expect this to work, but now that it does, it points out that > there is definitely nothing wrong with my laptop.) You need to be more careful about saying things like that. Strictly speaking, it's not the computer _or_ the device that's broken -- it's the combination of the two put together. To say that the computer is broken implies that it doesn't work _at all_, no matter what device you connect to it. Since the computer worked with the digital camera, you already knew that the computer isn't broken. That doesn't mean there's nothing wrong with it, however. It's entirely possible that the USB hardware is just slightly marginal, and as a result it works okay with robust devices but fails randomly with devices that are slightly marginal themselves. > 3) Rebooted my laptop, plugged in my watch. Mounted it, copied files to > it, got errors, the rest you know from previous emails. > (Didn't expect this to work, and I was right.) > > 3.a) Removed the watch, plugged in my pen cam and it started doing the > same thing, actually it wouldn't mount something like every other time, > and I had problem writing to it. It was as if using the Watch had caused > everything to go wonkey. > (Was doubtful this would work, and proved to be correct) > > 3.b) Removed the pen cam, put in the camera, and it worked no problem. > (It was working before so I figured it would remain working.) This certainly points to hardware problems, with the watch not working at all on your laptop and the pen cam only working sometimes. If it were a software issue then almost certainly the same result would happen each time you tried it. It's even possible that you're suffering from a bad USB cable connection. If some electrical contact is getting broken intermittently it could cause the sort of problems that you see. > 4) Rebooted my Linux computer that the watch HAD been working on, and now > it won't even mount the thing, gives me a whole new set of errors that I > can give you if it's important. It uses ohci instead of uhci. I have yet > to make this work again. > (This one shocked me. Though I don't know how I got it to work in the > first place, at least I have proven the watch is the only real device with > an issue.) Depends what the errors are, but it does sound like the watch doesn't work very well. > 5) Plugged the watch into a Windows XP machine. Came up as a drive almost > instantly, I opened the new drive it made, copied a few programs like > putty.exe into it, removed the device, then plugged it back in and all the > files are there. I did the whole right click on the drive and had it scan > for bad blocks and whatnot then figured it wouldn't hurt to do a full > format, so I wiped it, then copied my putty programs back onto it and > removed it again. > (A little surprised, yet happy, yet disgusted that it worked here, but I > now know the watch at least kind of works.) Interesting. And we've seen examples before of hardware that worked under Windows but not (or at least, not as well) under Linux. It may be a question of custom-made drivers that the manufacturer has supplied, or it might be something else. It's almost impossible to tell. > 6) Moved it to the laptop in question, mounted it. Saw all the new files > via ls. Tried to copy putty.exe off of it, and it failed with a IO error > on the command line. Unmounted. > (Now I can't read from it, but this is what you had assumed would have > been the case before so no big surprise for you.) You can't read from it? Is that just on the laptop or on the Windows machine as well? And what goes wrong when you try to read from it? Does it survive the initial probing done by the scsi core and sd drivers and then fail the READ commands, or does something go wrong at an earlier stage of the process? > 7) Moved back to the Windows XP machine, it popped open and I saw all my > programs. So you _can_ read from it after all. You've got to be more precise. > I decided to just make sure that those files weren't corrupt, > so I double clicked putty.exe and it pulled up like it should have and I > ssh'd to one of my servers. > (Well, I am having a hard time believing the 'watch' per say is the > problem other than possibly not conforming to the standards which is why > the drivers in linux aren't supporting it properly?) I couldn't say. But lots of other devices that do conform to the standards work just fine with Linux. Can you try replacing the USB cable? > So, I suppose the summary is, the watch works in windows, remotely kind of > works in uhci, and works fine during the right phase of the moon with > additional solar flares on ohci? > > I believe I saw that there's some kind of handler for devices that don't > follow the norms? Is this where you believe my solution may come from? I don't know what you might be thinking of. > I hope this sheds new light on the problem at hand. > > Thanks again for the help, > > Jerry Kilpatrick Sorry I can't suggest anything else. Alan Stern ------------------------------------------------------- This SF.Net email is sponsored by: IBM Linux Tutorials Free Linux tutorial presented by Daniel Robbins, President and CEO of GenToo technologies. Learn everything from fundamentals to system administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
