Python bindings: we should remove mappings for http_seeds

[AllSeeingEyeTolledEweSew]

atp = lt.add_torrent_params()
atp.ti = ...
atp.http_seeds = ["http://httpseed.com"]
atp.url_seeds = ["http://urlseed.com"]
handle = session.add_torrent(atp)
handle.http_seeds()  # []
handle.url_seeds()  # ["http://urlseed.com"]

Also write_resume_data() doesn't seem to write http_seeds and read_resume_data() doesn't read it.

This doesn't appear to be a python bindings issue, the C++ code just drops the value. I know http_seeds is deprecated, but is it intended to be dropped?

[arvidn]

yes, support for http_seeds (BEP 17) have been removed from master. As far as I know, nobody uses them. support for url seeds is still there (BEP 19).

[arvidn]

mailing list announcement here.

[AllSeeingEyeTolledEweSew]

Ahhh, I see

IMO, once a feature is "vestigial" (function exists but has no effect), it can be removed from python bindings altogether. It shouldn't break any python clients until they go to call the function. At that point they'd break anyway if they depend on the effect

[AllSeeingEyeTolledEweSew]

(I'm assuming you only kept http_seeds in the C++ code to avoid link errors, which doesn't apply to python)

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[AllSeeingEyeTolledEweSew]

unstale

[AllSeeingEyeTolledEweSew]

Actually, never mind. As I look through it, I can think of a lot of ways an app might touch http_seeds for passthrough purposes, and we'd break them needlessly if we removed it.

We should still deprecate http_seeds though, but that's #5967.