Dear Greg, Thank you very much for your kind response, valuable warnings and suggestions. I see a very good road-map in your answer. On the other hand I ( probably together with Alan and somebody else interested in participating us) will appreciate all other comments and suggestions.
Best regards, Murat On 2021/04/15 20:40:32, Gregory Nutt <spudan...@gmail.com> wrote: > Before you start writing code, I think you should talk with the group > about the architecture that you would develop. > > One of the essential, unbend-able rules is that any new development must > not add new operating system interfaces that are not standard, not > documented at OpenGroup.org, or are not supported by Linux. New logic > can use, for examples, standard character driver interfaces, a BSD > socket interface, or the file system, but no made up interfaces and no > direct calls into non-standard OS functions. > > I don't know much about DALI other than having scanned some websites. > My recommendation is that you consider this as a user-space library like > apps/modbus, perhaps at apps/dali. The actual, low-level hardware > interface could be implemented, say, via a character driver known to the > apps/dali logic. The user, application interface could then be purely > of you choosing and exported via a header file at apps/include/dali/dali.h > > The dali drivers would go at nuttx/drivers/dali (probably) and the > interface (IOCTL commands and internal OS setup interfaces) might go in > nuttx/include/nuttx/drivers/dali.h. > > Does that make sense? In any case, let's get concurrence on the > interfaces before starting code development. That will save a lot of > problems down the road and will probably also engage more people, get a > good review of the design, and might recruit people help you with the job. > > Greg > > On 4/15/2021 9:43 AM, murat tologlu wrote: > > Hi Alan, > > I am glad to hear that you found my proposal as a nice feature for Nuttx to > > have. I see you have made a good intruction; let me add something: Yes, > > DALI interface standard has DALI and DALI2 versions. DALI2 version was also > > extended with a feature set named as D4i. Therefore we have to cover all. > > Pysical layer is very simple, we can use any of ST STEVAL-ILM001V1 and > > Mikroe DALI 2 Click, or we can make our own hardware interface for our > > tests, no problem. Manchester encoding is also very simple, as the and > > since the clodck frequency is very low we can implement it by software with > > register operations without using any special counter therefore we can > > easily obtain portability of our code. > > In this work what I can do is, I can get all the information required such > > as IEC62386 standard and others, I can order all the required hardware, I > > can setup the hardware and I can do necessary tests. I can also participate > > implementing these in Nuttx codebase as much as I can with your help. So, > > let's get started, cd nuttxspace/nuttx make distclean :) > > > > On 2021/04/14 14:11:09, Alan Carvalho de Assis <acas...@gmail.com> wrote: > >> Hi Murat, > >> > >> I think DALI support should be a nice feature to have! > >> > >> Well, I already search for this protocol some time ago, but I don't > >> know much about it yet. > >> > >> The protocol uses Manchester encoding, maybe driver interface should > >> be implemented using GPIO and freerunning timer. Suggestions are > >> welcome! > >> > >> For HW I think we have two options: ST STEVAL-ILM001V1 and Mikroe DALI 2 > >> Click. > >> > >> It seams there are two protocol version: DALI and DALI 2. Probably > >> those DALI dimmers on Aliexpress are pretty old DALI protocol. > >> > >> BR, > >> > >> Alan > >> > >> On 4/14/21, murat toloğlu <mtolo...@hotmail.com> wrote: > >>> I would very much like the DALI interface to be in Nuttx and I would like > >>> to > >>> learn your opinions on this issue. My knowledge and experience in Nuttx is > >>> not enough to do this work alone, but if we get a few people together, I > >>> can > >>> participate in the development work. > >>> >
