I/O layer refactoring
Lubomir I. Ivanov
neolit123 at gmail.com
Thu Mar 16 13:42:29 PDT 2017
On 16 March 2017 at 19:57, Dirk Hohndel <dirk at hohndel.org> wrote:
>
> I'm not sure if everyone who is likely to be interested in this is on the
> libdivecomputer mailing list...
> Linus, Anton, Lubomir, Robert... if you have comments on the below, please
> respond over on
> that mailing list (not hear, let's help keep this discussion in one place).
>
i'm guilty of not looking much at the libDC tree other than building
it and at the usage in Subsurface, so i'm going to keep a low profile
on this one.
i understand the change, the need for the change and i think that what
Jeff suggests seems reasonable in terms of future benefits.
(except that (void *) suggestion, if one breaks the ABI might as well
provide the correct type information for the content for the stream).
...but Subsurface already has a Bluetooth layer and i'm not sure that
Subsurface has much interest in that change as one of the triggers for
the change is exactly Bluetooth. if Subsurface would not benefit much
from this change then Subsurface can roll with the old API if it also
supports the latest added DCs, unless the latest added DC support only
works with the new API, in which case Subsurface needs to move to the
new API to support new DCs.
/random tinkerings
lubomir
--
P.S. yet to see a dive computer IRL, such mythical devices.
More information about the subsurface
mailing list