OS/2 Capi for Zyxel Elite and Omni terminal adapters
From this page you can download my capi 1.1 and 2.0 device drivers for OS/2 for the Zyxel Elite and Omni ISDN products.Please see Read Me for further information.
Actual version: 0.41 Beta
Read Me
These capis are freeware and are provided AS ARE. The Author makes NO WARRANTY OF ANY KIND.
For this software still is in beta state, I would be glad to get some answering signals concerning problems and suggestions regarding this program. It also would be nice to know, which software you've tested the product with. Please contact me via E-Mail: Mail@Humbert-Online.de or Fax: +49-5131-476024
When this betatest is finished, I plan to implement the capi into a device driver. So the com port will be no longer locked for other programs and it should by possible to use the capi from Win/OS2.
Installation:
This Capi requires at least firmware 2.07d / 2.02 for the Elite / Omni products. To install this Capi, simply copy it to a path pointed to by the LIBPATH variable in config.sys. For example X:\OS2\DLL (where X must be replaced by your bootdrive letter) will work fine. Some programs like Opalis or VarioX4 require the setting of path and name of the Capi on the "Einstellungen \ Schnittstelle" notebookpage. This should look like 'X:\OS2\DLL\CAPI.DLL'.
History:
Rel. 0.41 Beta July 20th, 1997
Known problems:
- With Elite Firmware Rel. 2.07e and later the throughput of both capis is bad (3400 to 4200 cps) while with Rel. 2.07d and before there is no problem (7400 to 7600 cps). I wonder what Zyxel might have changed. Perhaps someone is able to tell me.
- A 2nd problem with this Releases is, that all messages for the 2nd, 3rd, 4th etc. application will be routed to application no 1. With Rel. 2.07d there is no problem like this.
- With Rel. 2.07d the EAZ2MSN mapping maps incoming calls only correct to numbers, stored with AT&Z7, while Rel. 2.07e and 2.08 do this only for AT&Z9. (AT&ZI0 - AT&ZI7 empty)
- Opalis and VarioX4 will crash when selecting "Test Capi". (Only if one of these programs is the only or first one to run. If they are the second program, this test will work. Nevertheless they will work without any problem).
- The first application (that one which installs the Capi), must be the last one to close when you want to finish the use of the Capi.
- added Support for the USA ISDN protocol
Known problems:
- With Elite Firmware Rel. 2.07e and 2.08 the throughput of both capis is bad (3400 to 4200 cps) while with Rel. 2.07d there is no problem (7400 to 7600 cps). I wonder what Zyxel might have changed. Perhaps someone is able to tell me.
- A 2nd problem with this Releases is, that all messages for the 2nd, 3rd, 4th etc. application will be routed to application no 1. With Rel. 2.07d there is no problem like this.
- With Rel. 2.07d the EAZ2MSN mapping maps incoming calls only correct to numbers, stored with AT&Z7, while Rel. 2.07e and 2.08 do this only for AT&Z9. (AT&ZI0 - AT&ZI7 empty)
- Opalis and VarioX4 will crash when selecting "Test Capi". (Only if one of these programs is the only or first one to run. If they are the second program, this test will work. Nevertheless they will work without any problem.
- The first application (that one which installs the Capi), must be the last one to close when you want to finish the use of the Capi.
- added a 3rd thread for better keyboard Performance
- fixed a small bug referring to the B3 Data message length
- fixed the timing problem with IBM's original COM.SYS device driver
- first public beta release of CAPI 2.0
Known problems:
- Opalis and VarioX4 will crash when selecting "Test Capi". Nevertheless they will work without problem
- Input from keyboard will take some time
- Download of small files will be slow, while larger ones will reach 7400 to 7500 cps
- The first application (that one which installs the Capi), must be the last one to close when you want to finish the use of the Capi.
- There still is a timing problem with IBM's COM.SYS device driver. Sometimes it will 'hang' and you have to finish the application. SIO.SYS Rel. 1.60d from R. Gwinn works fine
- fixed a timing problem with IBM's original COM.SYS device driver (the problem mentioned above still exists)
- implemented the use of concurrent programs
- implemented the use of the Omni products. (In lack of such a device not tested. I hope it will work)
- implemented a "workaround" for a bug in Capi Snooper 1.01. This program doesn't give back the correct message number for response messages
First public beta version
[ Top of page | back to homepage ]