Attention is currently required from: laforge, neels, pespin.
4 comments:
Patchset:
Just an idea: […]
I suggested this improvement to the customer (see note-16 in SYS#6913), but they refused. I would still like to implement this, but I ran out of the estimated time, so I'll leave this as a potential improvement for a separate patch.
If the SMSC has a response message to the readyForSM saying "ack, i had nothing to submit", then I would again like to remove this timeout, as soon as osmo-msc supports those messages.
This sounds logical and would be nice, but there exists no such thing in MAP protocol and thus should not be in GSUP, unless we really want to deviate from the former. See 3GPP TS 29.002, Table 12.4/1.
File src/libmsc/msc_a.c:
Patch Set #4, Line 164: static void msc_a_put_use_lu(struct msc_a *msc_a)
(i'd prefer the name msc_a_put_lu_deferred() […]
Done
File src/libmsc/ran_infra.c:
Patch Set #4, Line 47: { .T = -5, .default_val = 0, .unit = OSMO_TDEF_MS, .desc = "Additional delay for LU connections" }, \
(two separate things in one thread making it a bit hard to respond)
doc string change: ack, updated in the new patchset.
you need to add a new timer entry to the list in the wiki: [...]
Done.
I would encourage to avoid re-using the same timer numbers.
By that scheme, currently the next available X timer is X 36 or -36.
I remember this was discussed somewhere, but I am not sure if we ever reached an agreement on the numbering scope of those Osmocom specific X-timers. According to that table, timer numbers like X1 and X2 are re-used in several projects, but AFAIU this is because they were added before you suggested to make X-timer numbers globally unique? Also, there is nothing on that wiki page saying "please avoid re-using the same timer numbers"...
I don't have a strong opinion here, so I'll just follow your suggestion.
To view, visit change 36760. To unsubscribe, or for help writing mail filters, visit settings.