Hi! I have really BIG problems running any application, which needs to comm over a serial port: - My ham TCVR - My GPS - My multimeter - A lot of my mobiles - A lot of devices used in my work (special telco equipment, comms varying from simple AT-style commands up to complex proprietary binary protocols). In all the cases there are very similar problems - either the program doesn't detect the device properly, or the device doesn't respond, or in some cases the device does, what the program wants, but it then can't read it back. It looks that some archaic wine versions are better, for example I have to use wine-0.9.40 to communicate with a device in my work (simple text-based command-reply communication), any wine newer than say 0.9.50 doesn't work, the communication times out. Because I'm a telco engineer and serial comm is my daily work, and there are only windows versions of many tools I have to use, proper and mature serial port support is essential for me. So, what can I do to help fixing these problems ? Should I open a bug for every such a program ? Or just open one bug and state all the programs there ? I think that bugzilla contains a lot of bugs related to serial port; is it good to add a new ones ? I'm afraid I can't bisect a problem between 0.9.40 and current wine, but I can make a serial trace for both of them and look at the difference. Would it help to find at least this one particular problem ? Or what more can I do to make the serial port working ? I can code in C in the Linux environment, but I know NOTHING about windows... With regards, Pavel Troller
Pavel Troller wrote:
So, what can I do to help fixing these problems ? Should I open a bug for
To solve most/all problems related to hardware one needs to have such hardware. It's nearly impossible to "simulate" the behavior while reading logs.
every such a program ? Or just open one bug and state all the programs there ? I think that bugzilla contains a lot of bugs related to serial port; is it good to add a new ones ?
No, creating duplicates never helps. You are correct there are already pliantly of bugs for serial communications.
I'm afraid I can't bisect a problem between 0.9.40 and current wine, but I can make a serial trace for both of them and look at the difference. Would it help to find at least this one particular problem ? Or what more can I do to make the serial port working ? I can code in C in the Linux environment, but
There are few bugs identified and are set as target for 1.0.0. You can look at those bugs. Few even list the commit that caused the regression. While those patches should have been no-ops.
So what you can do is to figure out what really changed and what part of the change broke the serial port functionality (in relation to each bug/patch). For this part you will need your hardware. Hopefully this will be enough to come up with a fix.
Vitaliy.
"Pavel" == Pavel Troller patrol@sinus.cz writes:
Pavel> Hi! I have really BIG problems running any application, which Pavel> needs to comm over a serial port: - My ham TCVR - My GPS - My Pavel> multimeter - A lot of my mobiles - A lot of devices used in my Pavel> work (special telco equipment, comms varying from simple AT-style Pavel> commands up to complex proprietary binary protocols). In all the Pavel> cases there are very similar problems - either the program Pavel> doesn't detect the device properly, or the device doesn't Pavel> respond, or in some cases the device does, what the program Pavel> wants, but it then can't read it back. It looks that some Pavel> archaic wine versions are better, for example I have to use Pavel> wine-0.9.40 to communicate with a device in my work (simple Pavel> text-based command-reply communication), any wine newer than say Pavel> 0.9.50 doesn't work, the communication times out. Because I'm a Pavel> telco engineer and serial comm is my daily work, and there are Pavel> only windows versions of many tools I have to use, proper and Pavel> mature serial port support is essential for me. So, what can I Pavel> do to help fixing these problems ? Should I open a bug for every Pavel> such a program ? Or just open one bug and state all the programs Pavel> there ? I think that bugzilla contains a lot of bugs related to Pavel> serial port; is it good to add a new ones ? I'm afraid I can't Pavel> bisect a problem between 0.9.40 and current wine, but I can make Pavel> a serial trace for both of them and look at the difference. Would Pavel> it help to find at least this one particular problem ? Or what Pavel> more can I do to make the serial port working ? I can code in C Pavel> in the Linux environment, but I know NOTHING about windows... Pavel> With regards, Pavel Troller
Pavel,
perhaps compile wine yourself and instrument the serial code with debugging output. Try to see what's going wrong. Look at the codes where this happens and read the old changelogs, where somebody fiddled witrh the code. Ask about what people thought (or smoked) when implementing or offer a better implementation.
There is a big problem with those special devices, as only few if any wine developpers have it...
Bye
Hi Uwe,
Pavel,
perhaps compile wine yourself and instrument the serial code with debugging output. Try to see what's going wrong. Look at the codes where this happens and read the old changelogs, where somebody fiddled witrh the code. Ask about what people thought (or smoked) when implementing or offer a better implementation.
OK, will do. I'm always compiling wine from source, so I will try to dig into the driver and search for the bugs there. One of the cases is I think simple, the device uses normal ASCII protocol and I can talk to it using kermit or minicom, but the windows program intended to control it simply stucks on the first command and never gets a reply. I think there is a good place to start.
There is a big problem with those special devices, as only few if any wine developpers have it...
Yes, I know. I hope that there is just a limited number of bugs, but they affect many such devices, so it looks that the driver is totally broken.
With regards, Pavel Troller
-
Pavel Troller -
Uwe Bonnes -
Vitaliy Margolen