From: Andy R. (Aleksander) <an...@tr...> - 2006-12-27 22:53:31
|
"Hal V. Engel" <hv...@as...> wrote: > On Wednesday 27 December 2006 04:07, Aleksander Rozman - Andy wrote: > > At 27.12.2006 02:16, you wrote: > > ... snip ... > > My feeling is any company that does not want OSS projects to have access to > their protocols or that alternatively does not make interface libraries (when > these exist) available for Linux/Unix systems does not want my business and I > do not want to do business with them. So as far as I am concerned LifeScan > does not want my business and I will take my meter/test strip business else > where. By the way I have already told them that I would not be doing any > more business with them. > > I support a piece of software named LProf which is used to create ICC device > profiles for doing color critical work. I have been working with a color > meter/spectrophotometer device manufacture to try to get Linux/Unix support > for these meters so that I can use these in my software. So far I have > protocol information for some of these devices but not all of them. But they > are still talking to me about the other devices. Until I have complete > information from the company I will not support any of their devices in my > software. > > Its a matter of principle and I think OSS projects that run into problems with > > device manufacturers not making their interface information available should > make sure that their users also know that these companies are withholding > this information. That way those that want to use OSS software can avoid > buying these devices. In other words I think you should publish a list of > the manufacturers that you have contacted on you web pages and also state > which have been forthcoming with support which are withholding support. > We, developer of free software must usually take what we can get, but we know that we get just a little. I would think that for hardware that is made for public usage, should have this protocols made available... > > I got protocols only from Ascensia (Bayer) so I > > can add their meters, plus I have two of thoose at home (Contour, Dex). > > Even with protocol description is sometimes hard to implement this. > > Ascensia protocol document is very badly written, so I had to listen to > > communication with meter. > > It appears that Bayer is at least trying doing the right thing and I think I > will get one of their meters. The Contour looks like it is comparable to > the one I am using now. How do you like yours. Also how close is the > protocol code for this to being in a working condition? Days, weeks, months? > I am not sure. I plan to have this ready by end of January. I am working on my Diploma and it will be due in January so I don't know when I will have time to do this. Maybe it will be sooner, but I doubt it. When you get new version from SVN you will see that I have done some things, but data class is still not finished, so I haven't tied it to GUI yet.. And also GUI is not finished. > My experience with my software is that once you get it into a state that users > > can get it (at least mostly) working on their machines without too much > trouble that they will test things like this and give you all kinds of > feedback both good and bad. This is real helpful since then you will know > where the problems actually are. In addition having a this mature enough to > be usable tends to attract additional contributers (programmers, translators, > documentation writers...). > > Your project is currently seeing about 30 downloads a month and I suspect that > > if it were in a more mature state that you would have significantly higher > download activity. I can't stress it enough how important having active > users is to an OSS project. > I don't think that this number will go up much. Lot of poeple with diabetes don't have idea how to use computers, so until we crate proper install nobody will use it. Software still needs a lot of work and I am almost the only developer that creates new features. > > > > So what is easiest for you to do. Get original software from your meter, > > install it, install sniffer (I used free-serial-port-monitor), and let > > original program get data. In sniffer every communication will be writen > > and send this file to me, and I will see what I can do. Maybe you can do > > this by yourself. > > I don't have a data cable because I have yet to find software that will run on > > my system and all of the available free software runs on Windows only. So I > can't do this. > > There is another, apparently dead, sourceforge project named kpumpe that has a > > set of meter modules in a separate project named glucomodul on sourceforge > that includes support for the euroflash, onetouch fast take ultra and the > onetouch ultra meters (these last two appear to be the same protocol). In > CVS is a copy of a PDF from Lifescan with the Ultra RS-232 com spec. The > PDF does not have anything about not making the spec available to others. So > this appears to be a new policy on the part of LifeScan (J&J). I will attach > a copy for you. > I have this protocol descriptions already, but I can't add support for something I can't test. Let me know if you can build it. You must have ant installed and libraries in correct directory. README and BUILD ANT documents should be sufficient to build it. If you can't build it I will send you all files so you can run them with little changes (you will have to change run.sh probably). Andy |