On my side, I have queued a patch for adding generic send & receive commands to libticalcs, so as to provide an API close to TI_NN_Write and TI_NN_Read, but I haven't published it yet.

Good job you two

Also, Extended, why not both ?
(well, eventually).

And this made me think that some files (and not just a few) are not .tns directly, but are zipped (or rar[er] - pun intended, whoooo - anyway, rar would be a pain on-calc, zip, probably less.)
So, should I add another "type" field to the api response, or will that just be left to the client to parse from the file url ? (it would be just one line in php, probably same or a bit more in C...)


Edit : Levak / ExtendeD : I reordered the doc (see Levak's post ^ )

The only FOSS software for extracting rar that supports rarv3 is the Unarchiver.

EDIT: It's written in Objective C, though, so I don't know how easy it would be to port it to the Nspire.

No, not unar, not unrar.
The FOSS version of unrar supports only rarv2, but unar supports a variety of archive formats, including rarv3.

BTW, the official version of unrar is proprietary.