| Subject | : | MMPORT and USB4REXX are now downloadable |
| Author | : | kiewitz |
| Date | : | 27 Aug, 2004 on 14:03 |
| Well, I finally released my MMPORT including Casio WMP-1 driver. I hope someone implements some more drivers, so this could get a little bigger and support more things like the RIO portables. My Casip WMP-1 driver is using my USB4REXX DLL that is also included including source-code. Have fun with it! btw. the file is on hobbes or on my personal site. Look for MMPORTv1.ZIP Regards |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | Sebadoh sys3175@optonline.net |
| Date | : | 27 Aug, 2004 on 17:49 |
| Would be nice if you included a link to the file on hobbes or your personal websites URL. Off topic do you know anything about connecting an Ipod 40gb via USB 2.0 to OS/2? Will it be detectected as a mass storage device or would it be possible for me to make a plugin for it? |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | BigWarpGuy |
| Date | : | 27 Aug, 2004 on 19:44 |
| http://hobbes.nmsu.edu/pub/incoming/mmportv1.zip Here is the temporary file location till it is in a more permenent location. --- BigWarpGuy - - - - - - - - - - - - - - - - - - OS/2 Warp-ed/eComStation-ed to the very end. - - - - - - - - - - - - - - - - - |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | BigWarpGuy |
| Date | : | 27 Aug, 2004 on 19:46 |
| http://hobbes.nmsu.edu/pub/incoming/mmportv1.txt Here is the text file at a temporary location. --- BigWarpGuy - - - - - - - - - - - - - - - - - - OS/2 Warp-ed/eComStation-ed to the very end. - - - - - - - - - - - - - - - - - |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | kiewitz |
| Date | : | 30 Aug, 2004 on 09:32 |
| Oops. I forgot to include XtraRexx support library. Will do so now... |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | magog |
| Date | : | 30 Aug, 2004 on 20:33 |
| Well a little bit off-topic, but maybe someone is interested in this.... I'm going to buy the FT245BM chip in a few days. I hope I can get that one working. That chip will transform incoming signals on the usb side and delivers parallel port signals on the other side. I want use it to control my graphic LCDs on a mainboard that has no parallel port. I germany endusers can get the chip for less than 7 EUR from Segor electronics (http://www.segor.de/). For more details: http://www.ftdichip.com/DS245B.htm
*** Java Movie Database - http://www.jmdb.de *** |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | DavidG |
| Date | : | 06 Sep, 2004 on 21:58 |
| I have tested Martin's driver on two machines. On the laptop I have, it is faster finding the USB devices than the IBM driver and works great. However, on my desktop computer which has a PCI 5 port USB 2.0 Belkins control card (this card normally works great with the IBM USB drivers), the driver fails to initialize any of the ports. Now this card works fine under Windows ME and eComStation using the IBM driver. I wonder if the driver could be made to check for USB devices using both the Standard and the Microsoft way. I wonder if the Microsoft driver does this since the USB concontroller works fine under Windows. I tested the USBD.SYS on two different eCS partitions on my Desktop computer and on both partitions, the USB controller fails to initialize (nothing lights up and no devices work). The only reason I mention this is that it would be nice to have a common driver that worked on just about every USB device. Using both methods might do this if it is possible. Also, for your information, the USBD.SYS also works great with Chris Wohlgemuth's CWUSBMSD.ADD driver on my laptop. |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | kiewitz |
| Date | : | 08 Sep, 2004 on 10:24 |
| It's not possible to use both ways, because the devices that fail on the IBM method actually "crash" internally b/c of firmware bugs. This means they will refuse any more requests inbetween the enumeration. That's why they fail. I guess you are using a later USB-basedriver on your desktop computer and that's why USBD fails. IBM implemented some new features or changed something internally. The latest base drivers are incompatible with my own USBD and cause a hang at startup (not because of my code changes!). Probably your driver combination doesn't work at all as well, but at least boots your system. If we get IBM to implement the method, everything *will* definitely work. |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | Cris |
| Date | : | 08 Sep, 2004 on 15:17 |
| Martin, I have the latest IBM drivers installed on my eCS 1.13 (at least the latest that are available on the ecomstation site), and they DO work with your USBD.SYS. I have noticed that the boot process is significantly longer when using your USBD.SYS, but it doesn't hang. I am now using your driver on a regular basis, because it's the ONLY one that makes my USB key work under eCS / OS/2. BTW, after I posted a news item on os2voice regarding your driver, I received a message from David Graser (IIRC) which tells me his USB key does work with the standard IBM drivers, while it DOESN'T work with your USBD.SYS driver. Thankyou! Cris |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | DavidG |
| Date | : | 08 Sep, 2004 on 16:37 |
| Hi Chris The post about the Belkin's USB adapter is mine. This post was after further experimentation on my part. It wasn't the flash key that was the problem, but the Belkins adapter. The laptop works great with IBM's laptest driver package and replacing the USBD.SYS with Martin's driver. The desktop also has the latest IBM USB driver package, does not hang, but nothing inializes using the Belkins card. Maybe Martin could add a switch to his driver to tell the driver to use the Standard way or the Microsoft way. example" BASEDEV=USBD.SYS /S /* The Standard way */ if no option is listed, then the default would be the Microsoft way. What IBM fails to realize is that Microsoft, with its 95+% of the market is the stardard, being the monopoly they are. Martin is correct that it would be nice to have the IBM driver supporting the Microsoft way, but being the realistic person I am, I don't see it happening. It would be nice if IBM updated their USB driver code in the DDK package, but that won't happen. Their too busy charging people for driver devopment. By the way Martin, thanks for all the driver developement and driver fixes you have done. I use them all when I can. David Graser |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | DavidG |
| Date | : | 08 Sep, 2004 on 17:26 |
| I have been following the Sane development and they have apparently come across several bugs which IBM has failed to fix. I have no idea which USB drivers these bugs would pertain to. If any of it is related to the USBD.SYS, Martin might be interested. From posts found on Comp.Os.os2.bugs. KS> Again I have to blame USB basic drivers for failing with KS> At a first attempt everything looked fine. But when it Try to transfer ONLY 4k (one page) of data with a buffer allocated with KS> We've discussed debug.logs too with Gerhard Jaeger (the It seems it is the misbehaviour of physical/virtuel memory... KS> With such drivers it seems a 'mission impossible' to An other idea: Commandblocks for USB have always (?) a certain 'framenumber' |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | kiewitz |
| Date | : | 09 Sep, 2004 on 13:01 |
| >I have been following the Sane development and they have >apparently come across several bugs which IBM has failed to >fix. I have no idea which USB drivers these bugs would >pertain to. If any of it is related to the USBD.SYS, Martin >might be interested. From posts found on Comp.Os.os2.bugs. Yes, the 4k limit exists but only within the base-driver. You are supposed to use USBCALLS *anyway* and at least my fixed version of USBCALLS gets around it by simply transfering at most 4k at a time. Please note that you may call it and transfer up to some gigabyte data. Fixing it won't help anyone and IBM won't fix it as well. Anyway, I would be interested in what that USB key is doing. Is it working under Windows 98? Also do you have 2 computers, where you could connect both with a serial cable and check debug messages? I could send you a debug version, that logs all sorts of traffic, so I would be able to find out what happens. |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | Cris |
| Date | : | 09 Sep, 2004 on 13:49 |
Hi Martin.
Who's USB key are you referring to? If it is mine, I am more than willing to help you. I have two computers connected with a network. The primari one runs eCS 1.13, the other runs Win2K, mostly for my wife's needs. See you... Cris |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | DavidG |
| Date | : | 09 Sep, 2004 on 20:29 |
| Anyway, I would be interested in what that USB key is doing. Is it working under Windows 98? Also do you have 2 computers, where you could connect both with a serial cable and check debug messages? I could send you a debug version, that logs all sorts of traffic, so I would be able to find out what happens.[/quote] Martin My USB hardware: Both machines has the very latest IBM USB drivers. Desktop system with the PCI 5 port 2.0 USB control fails to power up USB with your USBD.SYS installed. Replacing with IBM's and all my USB devices work with the exception of the generic 6-in-1 card reader. The system becomes unstable with it inserted. However, if I replace Chris's CWUSBMSD.ADD with IBM's USBMSD.ADD, the reader works fine. However, although my other USB devices are found, they cannot be read. On the laptop, everyone of my USB devices work with your driver and Chris's installed. One thing I have to do with your driver installed is rebooted with some of the devices in. The USB flash devices I could insert any time. The other devices have to be plugged in during bootup. Using the IBM driver (USBD.SYS), I can insert my USB devices at any time. All USB devices work on both Windows ME and Windows XP. I have never done debugging. Not sure if I would know how. I don't have a serial carble, but I I could buy one. I assume it would need to plug in serial port to serial port on both machines. How would one get one computer talking to another? My email address is dwgras at swbell dot net if you want to contact me further concerning debugging. I am a home user, but I will do my best to help with further driver development. David Graser |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | kiewitz |
| Date | : | 10 Sep, 2004 on 10:54 |
| I have never done debugging. Not sure if I would know how. I don't have a serial carble, but I I could buy one. I assume it would need to plug in serial port to serial port on both machines. How would one get one computer talking to another? ------ You will need a null-mode cable instead of a normal serial cable. The talking will get done by USBD.SYS and you will need to run a terminal software on the other computer. Anyway, what do you mean by "is detected but can not get read from". Do you mean the devices show up in USB-Resourcemanager? If that's the case, everything with USBD is fine. The problem is USBMSD then. btw. some more explanation about the 4k limit in the base-driver. That's because the driver DIRECTLY maps the memory into physical space and transfers that to the device. Virtual memory is physical memory mapped together. Of course, the physical memory does not need to be aligned. Because of that transfer above 4k will fail, because 4k is the page size. This means that if you transfer more than 4k (or less but you don't use page aligned memory), it will garble up the data. Still there will be 4k+ transfered, but only the first few bytes (up to 4k) will be valid data. Everything else will be fubar. This "problem" is actually the only way of doing it in a fast manner. The other solution would be to COPY OVER from virtual memory to physical aligned memory. This takes time and will slow down everything. So it's actually not a bug, but a feature. Just use my USBCALLS. This will align the memory (at least afaik) and will also transfer 4k chunks internally each time, so everything should work. |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | DavidG |
| Date | : | 11 Sep, 2004 on 00:41 |
| [quote]kiewitz (10 Sep, 2004 10:54): I have never done debugging. Not sure if I would know how. I don't have a serial carble, but I I could buy one. I assume it would need to plug in serial port to serial port on both machines. How would one get one computer talking to another? ------ You will need a null-mode cable instead of a normal serial cable. The talking will get done by USBD.SYS and you will need to run a terminal software on the other computer. When you say terminal software, are you referring to a program like ZOC?. As for the cable, I will need drive some ways to see if I can find one. No one has one in the area where I live. Next time I leave for the major city near where I live, I will check.
The drives are found (drive letters show up), but cannot be read. I agree that the problem is USBMSD. That is why I use Chris's driver. With his and your driver on my laptop, I can read everything I can attach. |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | kiewitz |
| Date | : | 12 Sep, 2004 on 19:25 |
| -quote- The drives are found (drive letters show up), but cannot be read. I agree that the problem is USBMSD. That is why I use Chris's driver. With his and your driver on my laptop, I can read everything I can attach. -quote- so everything is working fine up to my level. If any device does not get recognized, then it's me to call If you get a drive letter, run LVM and tell me what it says. Also did you try to reformat the partition? Most of the time that helps. If OS/2 assigns a drive letter, at least the device is read-able. Regards |
| Subject | : | Re:MMPORT and USB4REXX are now downloadable |
| Author | : | DavidG |
| Date | : | 13 Sep, 2004 on 05:54 |
(so everything is working fine up to my level. If any device does not get recognized, then it's me to call Chris is already trying to fix USBMSD, so I won't even think about doing so as well. My Sony digital camera isn't working as well. It has a 16MB memory stick in it and it's detected as being 96MB. Also the device does not work (LVM fails, no drive letter assigned).If you get a drive letter, run LVM and tell me what it says. Also did you try to reformat the partition? Most of the time that helps. If OS/2 assigns a drive letter, at least the device is read-able. One thing I did do on all my USB removable media was to use DFSee to create New Master Boot Records while destroying the old tables. I then used LVM to create the partitions on the removable drives. I also assigned permanent drive letters. It is easier for me to keep track of what USB device are inserted by the assigned drive letter. All my USB devices are formatted FAT32 with the exception of the camera media. All are properly recognized. I do believe creating new partition tables helped. Before I did this, most of my media was recognized improperly (listed with incorrect sizes). David Regards |