Read this first :) Re: [patch 0/6] staging: speakup: Introduce TTY-based comms

[Okash Khawaja]

Hi,

Thanks for the quick feedback and this helpful summary :)

On Sun, Feb 26, 2017 at 3:38 AM, Samuel Thibault
<samuel.thibault at ens-lyon.org> wrote:
> So, just a last mail, which could actually be read first :)
>
> First, I notice that it took you 12 minutes to send the patch series.
> This shows that you are not using quilt to do this :) Really, take the
> time to look up how "quilt mail" works, and notably its --send option,
> it will save you an awful lot of time (notably since I'm asking to
> split the series is rather small patches, which is preferred thing when
> pushing to Linux)

Quilt has been very helpful in preparing the patches. Sending was
tedious without it but today got quilt mail --send working which is
great.

>
> I also gave a lot of information, most of which are mostly notes for
> later.  I'd say that the priority is this:
>
> - work on the comments I made on the actual code (patch splitting,
>   migrating ltlk, not migrating dectlk, moving spk_synth_immediate,
>   exporting spk_stop_serial_interrupt, taking the ser parameter into
>   account, renaming spk_ttyio_immediate to spk_ttyio_synth_immediate,
>   and putting serial_out in a separate spk_io_ops structure)
>
> - maybe post a last round of patches here, so I make a last review
>   before we push to linux-kernel
>
> - then we push to linux-kernel and get flames :)
>
> - in the meanwhile, you can work on adding a "char *dev" synth parameter
>   which would replace the "ser" parameter, and writing a function which
>   translates it into an MKDEV(major,minor) pair (first implement it for
>   "ttyS*", then for "ttyUSB*", etc.)

Sounds good. Should this be done before pushing to linux-kernel?

>
> - hopefully at that point we can get that series commited mainstream,
>   even if we only migrate the trivial drivers.
>
> - then we can work on the other drivers as my mail series suggests.
>
> Samuel