[Psi-Devel] PSI and file transfer

[Peter Saint-Andre]

Kevin Smith wrote:
> On Sun, Mar 9, 2008 at 10:20 AM, Ihar Mahaniok <mahaniok at gmail.com> wrote:
>
>> so, Google Talk does something different from JFT? but what Google Talk does
>> is probably also documented somewhere? And by "not supported" you mean it's
>> not implemented yet, but probably implementable, as receiving works already.
> 
> Google Talk does something different from JFT, yes. It isn't fully
> documented, as far as I know.
> If the XSF defines a UDP-based transport like GTalk uses, then we'll
> probably make best efforts to implement it.

JFT is (or will be) a general approach to the problem of negotiating
parameters for a file transfer (including the ability for the recipient
to counter-offer). Which particular transport methods you advertise are
up to you. We hope that JFT will result in file transfer that "just
works", in part because we will specify that in-band bytestreams (IBB)
is the mandatory-to-implement fallback for all clients (see XEP-0047).

Right now, the Google Talk client uses Jingle to negotiate file transfer
but also uses an HTTP-based method for the data transport, typically
over their special "Pseudo-TCP-over-UDP" technology (which AFAIK is not
yet well documented). I don't think that the Google Talk client will
fall back to IBB right now, but it may do so in the future if the XSF
recommends IBB as the mandatory-to-implement fallback method.

Peter

-- 
Peter Saint-Andre
https://stpeter.im/

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 7338 bytes
Desc: S/MIME Cryptographic Signature
Url : http://lists.affinix.com/pipermail/psi-devel-affinix.com/attachments/20080309/923c5e94/attachment-0001.bin