<div dir="ltr"><br><div class="gmail_extra"><br><br><div class="gmail_quote">On Mon, Mar 24, 2014 at 8:40 AM, Dan Williams <span dir="ltr"><<a href="mailto:dcbw@redhat.com" target="_blank">dcbw@redhat.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="">On Mon, 2014-03-24 at 16:18 +0100, Aleksander Morgado wrote:<br>
> On Mon, Mar 24, 2014 at 12:25 PM, Arnaud Desmier <<a href="mailto:adesmier@sequans.com">adesmier@sequans.com</a>> wrote:<br>
> > We get into the same troubles on our side but it's for UUID, not CID. you<br>
> > can try attached patch to allow you to register custom UUID and so get<br>
> > access to your own CID.<br>
> ><br>
> > * use mbim_register_custom_service API to register a new UUID<br>
> > * use mbim_unregister_custom_service API unregister it<br>
><br>
><br>
> I believe we should make this logic part of libmbim. What do others think?<br>
> That patch would need some cleanup to follow coding style and other<br>
> minor things, though.<br>
<br>
</div>Yeah, agreed.</blockquote><div><br></div><div>For a library API perspective, it's not a bad idea to add hooks in libmbim, which would allow an application to implement custom device services using libmbim.</div>
<div><br></div><div>The extensibility in MBIM is unavoidable for practical reasons but does come at a cost, IMHO. Let's hope subsequent MBIM specifications would standardize more common services, so that we don't end up having many versions of the same service, like we did to AT.</div>
</div></div></div>