Yes, well, you have experienced exactly why I quit caring if speakup got into the official linux kernel or not. I know a lot of folks hoped that once speakup did manage to get in there might be help from the kernel community to fix problems with the system but I never believed there would be help and I still don't. I believe that unless Samuel or someone else that has good kernel savvy decide to attack the problems and straighten them out or find other solutions, it is just a short period of time before speakup is back out of the kernel again. That will leave another problem though which is speakup's entire git repo structure has been gutted to accomodate a community that doesn't care and who's going to piece things back together afterword? If you wish to be helpful and I personally am glad of it, my suggestion is to forget about trying to make the serial synth modules work correctly inside the kernel and write user space drivers/programs to read the output of the soft synth device and feed hardware synths from user land. That should make it fairly easy to handle them properly either rs232C or usb. Then the serial modules could be gutted and concentrate on fixing the problems speakup has that are show stoppers like cut-and-past bug/goto position bug and smp problem with screwing up the order of output. That's my .0000002 bitcoins worth. Kirk On Wed, 7 Mar 2012, John Heim wrote: > From: "Samuel Thibault" samuel.thibault at ens-lyon.org >> Note: it's not a proper fix, as it won't prevent you from screwing your >> serial port if you concurrently access it through /dev/ttyS0 while >> speakup is running, but it's probably better than nothing. > > Right. I had intended to mention that. I said in another message that I'd > talk about the controversy I stirred up on the linux-kernel list last > weekend. This is a related issue. So here goes... > > I signed up to the email list for kernel developers and last weekend I asked > for help with this bug. The avice I gout was not exactly helpful. To tell you > the truth, I found the attitude of the people on that list shockingly > snobbish. They seemed more interested in criticizing the speakup developers > than in helping me with the bug. In fact, I got no useful advice what so ever > in actually fixing the bug. The patch I posted is the result of my own > efforts to figure out what was going on and just make it so it would work. > Screw correctness. > > But there was a debate on the list about the "right" way to fix speakup. It > looked to me as if the developers offering advice didn't even understand what > is required from speakup. Never the less, that didn't stop them from opining > on how it should be rewritten. I was trying to tactfully suggest that the > speakup developers must have had good reasons for what they did but > fortunately, Samuel was there to correct many of their misconceptions. To > tell you the truth though, I thought some of them should have been smacked. > But, I guess you can't do that. > > So now I'm not entirely sure how to proceed. I think I will have to post my > patch to the linux-kernel list to see if I can get it incorporated into the > linux kernel code. Admittedly, its a bad fix. All it really does is to cut > out some code that didn't work anyway. I don't know what the point of having > speakup at all is if it doesn't work. The patch doesn't really make things > worse than they were before unless you consider not working at all better > than working incorrectly. > > Then there is the bigger question of fixing speakup so the linux kernel > developers will approve of it and so that it can support USB synths. I would > like to start working on that but I don't really know how to get started. I > think its going to take some talking to the linux developers. I'm sure > there's an answer there somewhere. If you can get video as soon as the kernel > loads, you ought to be able to get speech via a hardware synth. The linux > kernel sends characters to your video card immediately... Why can't it do the > same to your hardware synth? An even better example is serial consoles. The > linux kernel can be made to use a serial console immediately upon boot. So > the magic is there. Its just a matter of figuring it out. > > So I think that what will have to be done is to make a plan that everyone can > agree on. I am reluctant for obvious reasons to volunteer anyone else for > that job. I'm willing to give it a try unless someone objects. > > _______________________________________________ > Speakup mailing list > Speakup at braille.uwo.ca > http://speech.braille.uwo.ca/mailman/listinfo/speakup > -- Kirk Reiser The Computer Braille Facility e-mail: kirk at braille.uwo.ca University of Western Ontario phone: (519) 661-3061
