libcurl vs libsoup



Hey so, I'd just like to put out a feeler among ostree's users:
Who would strongly prefer libcurl over libsoup?

The reason I'm asking is right now, my primary consumer (rpm-ostree)
uses libcurl indirectly via librepo, so it's linked in process anyways.

The other things driving me asking are:
 - It's common for things lower than the GNOME stack to use libcurl (e.g. systemd, although it's a soft dep 
there)
 - Right now at least in Fedora: libsoup -> glib-networking -> gnutls, and libcurl -> nss.
   I'm aware ostree isn't going to solve the crypto library problem it self, but it would
  be beneficial from an Atomic Host and container perspective to drop gnutls.
- libcurl has http/2.0 work which is likely dramatically faster for what ostree does

In the end I suspect we'd likely make it a build time option or perhaps
go all the way to runtime pluggability so that desktop-y things like flatpak
could still use libsoup and benefit from the glib-networking proxy support
and such, whereas "system" things like rpm-ostreed would use libcurl.


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]