This is merely a historical archive of years 2008-2021, before the migration to mailman3.
A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/gerrit-log@lists.osmocom.org/.
Neels Hofmeyr gerrit-no-reply at lists.osmocom.orgPatch Set 4: This is breaking ABI, but not API. In the commit log, I wrote: - I decided against packing the mnc with the mnc_3_digits field into a sub-struct because it would enlarge this patch; plus, with a separate flag, current code can decide when to start heeding leading zeros. Maybe that wasn't clear enough, let me strengthen that case: If we pack mnc+flag into a sub-struct, as soon as we merge that patch to libosmocore master, all of our projects flare up red and stop building. Keeping the 'mnc' member as it were and just adding a flag alongside allows us to tackle each code base when we're ready to use the flag. It would be quite a stunt to have all of the patches ready in gerrit before merging the libosmocore one; and we're making compiling old code against new libosmocore hard, unnecessarily so. If we decide to indeed go for a new struct packing the mnc+flag or a string as it were, I'd also go the extra mile and actually "bump" the names of all the structs using it, i.e. introducing struct gprs_ra_id2, struct bssgp_bvc_ctx2, struct bssgp_paging_info2 etc. In that case we would be both API *and* ABI compatible (-- I'd have gone that way, the only reason I'm not doing it is because you said to rather bump ABI instead). Either way, I would discourage changing the current structs in ways that the current dependent code cannot compile as-is. -- To view, visit https://gerrit.osmocom.org/6659 To unsubscribe, visit https://gerrit.osmocom.org/settings Gerrit-MessageType: comment Gerrit-Change-Id: Id2240f7f518494c9df6c8bda52c0d5092f90f221 Gerrit-PatchSet: 4 Gerrit-Project: libosmocore Gerrit-Branch: master Gerrit-Owner: Neels Hofmeyr <nhofmeyr at sysmocom.de> Gerrit-Reviewer: Harald Welte <laforge at gnumonks.org> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: Neels Hofmeyr <nhofmeyr at sysmocom.de> Gerrit-HasComments: No