[nanomsg] Re: nanomsg iOS
- From: Martin Sustrik <sustrik@xxxxxxxxxx>
- To: nanomsg@xxxxxxxxxxxxx
- Date: Thu, 27 Nov 2014 22:31:05 +0100
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi Bent,
As for the process, either send the patch to the mailing list or make
a GitHub pull request. Don't forget to state that the patch is
submitted under MIT license.
Please, be careful not to break the build on other platforms.
Thanks!
Martin
On 27/11/14 21:57, Bent Cardan wrote:
> Hi Martin,
>
> If any of these changes/files could be deemed acceptable for
> upstreaming, I would likely prefer to close the repo I just started
> so as to centralize the thing and make it easier for people to
> find. Therefore people could unambiguously understand what files to
> apply in their own projects without being distracted about
> something else out there to test for iOS.
>
> This minor offshoot is minimal enough to be represented reasonably
> well in a branch or maybe some nanomsg repo files with the org?
> Please inform me of the team's requirements and anything else in
> mind so that I might be able meet with approval or even just
> upstream consideration. My main question is: what is the best way
> for me to send potential upstream nanomsg.ios patches? My personal
> goal behind efforts here would be to share the simplest and most
> convenient distro for moving iOS work forward. Ideally, I just want
> quick access to embedded, pluggable nanomsg topologies across all
> the iOS stuff.
>
> - Bent
>
> On Thu, Nov 27, 2014 at 2:26 PM, Martin Sustrik
> <sustrik@xxxxxxxxxx <mailto:sustrik@xxxxxxxxxx>> wrote:
>
> Hi Bent,
>
> What's the chance of upstreaming the changes?
>
> Martin
>
> On 27/11/14 15:04, Bent Cardan wrote:
>> Hi All!
>
>> Earlier this week I was having some weird compiler errors trying
>> to do I/O multiplexing poll and select for determining when the
>> descriptor at nn_recv() endpoint is ready for I/O.
>
>> Granted, I was trying to do this on new iOS equipment like with
>> 64 bit arm processors.
>
>> I did a little more messing around with definition macros and
>> code loading/linking happening particularly in nanomsg/src/aio
>> directory, i.e., the library’s support for Apple/Darwin kqueue
>> kevents..
>
>> And now today I report having extreme high performance using
>> poll and select for nn_recv on all iOS devices, new and old
>> alike.
>
>> Here’s how i’ve setup nanomsg on iOS for the transports in beta
>> release 0.5 : https://github.com/reqshark/nanomsg.ios
>
>> Thank you and kind regards!
>
>> -Bent
>
>
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
iQEcBAEBAgAGBQJUd5gYAAoJENTpVjxCNN9Ye7oH/AgBUeW5SCukk02/evb0Mdzm
bN/XeFkwaL3N3HhgnMYnaGhlB7JlTlUVYs/bzIY1+igTB9kcmkzQeNN/yt9J/YB4
urQ1p9HMxFGWXVn2jD4wyDE6OKxr3y6Q5FGIFtv4AL5COkiYZ4wEGVnZs1eAh699
10YvH3iIJCwDqoMrVeAymqkipYsR2EXiziWNn5aDi3ZM2PEj9Ga6IZ/IOmmhVoWr
KK0/9NNWBg3nBiPDrLhSatAr/xlAXTOT1DpNKY8nqnIcibL8qZmruvIGFDJsr+Vg
t4/k5eufYfaftQcDTxwXRXDBh2ibIMFwyIvDsoQD/khjVVPOFu3M7iMy+wsvJvo=
=xQDG
-----END PGP SIGNATURE-----
Other related posts:
