Page 34 of 66 FirstFirst ... 24293031323334353637383944 ... LastLast
Results 661 to 680 of 1320
  1. #661  
    This is pretty obvious and probably tried but I don't recall the idea being posted....have folks tested with the lowest level of backlighting and any other power-saving ideas (minimal background apps running, etc.)
  2. #662  
    Quote Originally Posted by vish0402
    Shadow: Would I need to add the PM Net file while I am testing on the Treo 600 as well?
    Honestly, I have not yet verified the hex code from the 600 pm net file, if it even exists. Mainly this is focused at the 650, but I'm glad you guys are testing with the 600 as we go. Honestly this driver is designed to work with most any 5.2+ OS device it looks like. Maybe you could use filez and see if there is a pm net file, and post it here if so for someone to compare...

    From what we have seen so far, on the 650 adding the new pm net file appears to make the phone want to use the phone for connections. This is probably due to the pm net file only having 1 network stack. Whereas the built-in pm net file appears to know about a secondary stack already! (VERY ODD).

    I think this is all going to come down to the fake lib files, and if I can get the wifi drivers to reference them without the phone seeing them. As I mentioned over the weekend though, if anyone has the needed equipment to hook up a pass through sdio circuit and feed it's own power supply I'd still like to know the results... I could do it, but I'll have to order the parts from digikey or mouser. So that'll be a while. We'll see where the day brings us.
  3. #663  
    Quote Originally Posted by shadowmite
    Honestly, I have not yet verified the hex code from the 600 pm net file, if it even exists. Mainly this is focused at the 650, but I'm glad you guys are testing with the 600 as we go. Honestly this driver is designed to work with most any 5.2+ OS device it looks like. Maybe you could use filez and see if there is a pm net file, and post it here if so for someone to compare...
    In the 600 ROM, there's a Net Library with CreatorID netl and type LiBr (size: 92273 bytes), and a NetPatch Library with CreatorID netl and type libr (size: 10628)...
  4. #664  
    Hey euroclie, while I'm pretty good with assembly, I still don't entirely understand the header sections of these files (your area of expertise). Is there a difference between LiBr and libr? Is it case sensitive?

    Here is beta 3... I've changed some references to the fake files, Don't spend to much time testing this, just a pass or fail will suffice... (I'm at work, don't have my wifi card).
    Attached Files Attached Files
    Last edited by shadowmite; 12/06/2004 at 10:34 AM.
  5. #665  
    Quote Originally Posted by shadowmite
    Hey euroclie, while I'm pretty good with assembly, I still don't entirely understand the header sections of these files (your area of expertise). Is there a difference between LiBr and libr? Is it case sensitive?
    As far as I know, the CreatorID and type are handled like 4 bytes integer, so they are certainly case sensitive. That being said, I have never checked that, I'll try to find out and let you know tomorrow...
  6. #666  
    Quote Originally Posted by shadowmite
    Here is beta 3... I've changed some references to the fake files, Don't spend to much time testing this, just a pass or fail will suffice... (I'm at work, don't have my wifi card).
    Exact same packet trace, I'm getting soft-resets every few network attempts.

    Adam
  7. #667  
    Keep on rolling lads...keep it up...I can see the light......aw lord I can see the light...
    -----------------------------
    Adzo
    Treo since November 2003!

    Still got the original.....
  8. #668  
    LOL @ adzo

    Shadowmite, a lot have expressed their willingness to give donation for your efforts, whether it pays off or no...why dont u make u an account on paypal or something so they can donate?
  9.    #669  
    Quote Originally Posted by drtansa
    LOL @ adzo

    Shadowmite, a lot have expressed their willingness to give donation for your efforts, whether it pays off or no...why dont u make u an account on paypal or something so they can donate?
    Follow the thread from the beginning and you'll have your answer. There are plans in place to get the donations going. It just requires some action by Marcus of TreoCentral to set up the collection bin.
  10. #670  
    Quote Originally Posted by bmacfarland
    Follow the thread from the beginning and you'll have your answer. There are plans in place to get the donations going. It just requires some action by Marcus of TreoCentral to set up the collection bin.
    LOL, thirty some pages!

    Well, I'm running out of ideas for a easy fix here. It really looks like a problem because of the network libraries. It sure would be nice if palm would open source the wifi driver they have! (HINT HINT!) This is the last thing I can think of to try. Most likely this will go into a reset loop, but give it a try if you're bored.

    The problem is these calls aren't in 68k assembly, they are in arm assembly, and that's something I'm not at all ready to deal with. So I made a change here to the reference header of the code, if this doesn't do it, well, let's not hold our breath. We've proven that there is no power issue (well, 95% proven), and we've proven it's a EASY fix if we had source code... It's been a interesting week...

    Let me explain a little more:

    The NetServLib makes the calls to the Fake Net Libs, we've verified this as either of these can be left out and the driver becomes stable and appears to work other than not being able to connect successfully. These calls are all in ARM code (the native Intel cpu being used now). If we could somehow change the creator ID of the fake libs, they no longer conflict with the phones built in libraries. But the calls needs to be changed in the netservlib, and I can't do that. Palm is holding all the keys here unless someone has some more advice...
    Attached Files Attached Files
    Last edited by shadowmite; 12/06/2004 at 05:21 PM.
  11. #671  
    Quote Originally Posted by shadowmite
    Most likely this will go into a reset loop, but give it a try if you're bored.
    Unfortunatly, your guess was correct (reset loop as soon as you reset). I'm working on a application to see what exactly I can get the network stack to do, but without being able to attach a full debugger this is going to be only somewhat useful.

    Adam
  12. #672  
    Thanks adamp, I expected as much. I do have one more idea here. I'm going to try mixing and matching with the zodiac drivers since they appear to have a different version of the libraries. We might get lucky. But it's to early to tell...
  13. spotter's Avatar
    Posts
    316 Posts
    Global Posts
    327 Global Posts
    #673  
    shadowmite's done a hell of a job so far, but I think the best idea is for someone to sit down and try to understand the entry points into the libs Palm is giving us.

    based on skimming documentation available at http://www.palmos.com/dev/support/docs/68k_books.html my gut feeling (though only a feeling, I've never programmed a palm before, but have done plenty of linux kernel hacking) is that we should be able to hook the driver they give us into the tcp stack. though I could be misunderstanding the documentation.

    someone has to figure out how the slot/expansion driver palm is giving us can be made to work w/ the Treo 650. Not an easy job, but should be doable if someone has the time to put in and learn the platform.
  14. #674  
    I have a question. The proven power thing you mentioned, ( that power isn't the issue) is that true in palm treo 600 ??

    Also my dad used to program on assembly. I will ask him to look a this thread. He hasn't programmed in a long time. Maybe he can help if he has time ? fingers crossed !
  15. #675  
    Hey, shadomite, I woke up this morning with an interesting idea: I think you changed a few bytes here and there in some files. Did you modify any file (application or library) that has an associated overlay? Because if that's the case, then there's a checksum in there that has to be updated for the overlay to work again, and if that's not done, then the file/library may fail to be loaded... This might be (part of) the source of our problems.

    Just a passing thought, but worth checking, I think.


    If you could give me a list of the files you modified, I'll see if that's the case or not.
    Last edited by euroclie; 12/06/2004 at 11:42 PM.
  16. #676  
    Nothing useful content-wise to add, but I do want to express my gratitude to the expertise of shadowmite. If something like this would work, I would be immensely greatful, and it would make the T650 a communications powerhouse, what with IR/BT/Wireless Web/WiFi.

    Keep up the excellent work.
  17. #677  
    Shadow & Euroclie: No power issues on the Treo 600 either...no blown SDIO or card. Am happy so far. Though I have a larger problem to deal with, dropped my Treo and the screen is shattered . was about to test with a Linksys B-only router but that's up in smoke. Can someone else please give it a shot with the Treo 600 (make sure you have insurance on your phone though).

    Cheers (off to ebay to buy another Treo 600)
  18. #678  
    to vish0402: sorry to hear the bad news about your Treo 600... Hope you can find one cheap on eBay. As far as hardware tests are concerned, I won't be able to make any test before next week, and that would be with a Sandisk 256MB+WiFi SD card, so it most certainly won't work anyway!
  19. #679  
    I haven't been keeping a close watch on this thread since my last post notifying you all about the availability of the Wi-Fi drivers from Tapwave. I got a SanDisk v2.1 Wi-Fi card delivered yesterday and got it up and running great on my Zodiac 1. I then decided to throw caution to the wind and try it out on my Treo 650. The drivers seemed to install just fine and much of the interface was even D-Pad-aware on the 650. When I popped in the Wi-Fi card the first time, it automatically fired up the Wi-Fi configuration screen, so it was definitely detecting the card (not sure if that indicates anything about the power concerns). Unfortunately, the bad news is that it didn't work. When I click on the "Connect" button (I had to manually set up access to my access point since I have WEP enabled) it didn't do anything. When I click "Rescan" the window that says "Looking for wireless networks..." appears and disappears on the screen so quickly that you can barely notice it. On my Zodiac, that pop-up window would stay on the screen for several seconds before it would time out and say that it couldn't find anything.

    My theory is that it has to do with palmOne hard-coding network OS calls such that it assumes that it always needs to access the internet over the cell phone carrier's network (understandably). If so, I don't see this as a quick-fix for someone trying to hack this from the outside.

    Nevertheless, shadowmite, if you have some freaky OS library files that you want me to play with in conjunction with either your hacked p1 drivers or the Tapwave drivers, let me know. I'm thinking that I may end up returning my 650 to palmOne for various reasons, so I don't mind messing around with it before then. Perhaps you could return the favor by hacking at the T5 web browser to get it working on the Zodiac. The free browser that comes with the Zodiac (based on the PalmSource Web Browser) leaves a lot to be desired. You send files for me to try (and directions) to scott AT tapland.com

    Scott
    Now THIS is the future of smartphones.
  20. worthb's Avatar
    Posts
    241 Posts
    Global Posts
    259 Global Posts
    #680  
    Quote Originally Posted by euroclie
    ..... and that would be with a Sandisk 256MB+WiFi SD card, so it most certainly won't work anyway!
    You know, I was just thinking, if there are no power problems (and there shouldn't be as p1 has said wifi support is just a matter of drivers, and the drivers appear to be 95% there) why aren't Sandisk and others who manufacture wifi cards able to demand enough info/docs from p1 to write their own drivers or finish the existing p1 drivers? Can we gather enough collective clout to pressure one of the wifi makers to push p1 for whatever is needed? We have seen the power of pledges. Maybe we should have people pledge to buy the first (non p1) wifi card with working drivers for both the t600 and t650. (or at least a survey)

Posting Permissions