Created attachment 86370 [details] [review]
automake: use serial-tests

Created attachment 86371 [details] [review]
test-log-manager: make sure to pass a (gint64) as timestamp

Created attachment 86372 [details] [review]
sqlite: deal with g_date_time_new_from_unix_utc() failing

Created attachment 86373 [details] [review]
tpl_log_manager_search_async: fix introspection

Comment on attachment 86370 [details] [review]
automake: use serial-tests

Review of attachment 86370 [details] [review]:
-----------------------------------------------------------------

::: configure.ac
@@ -87,4 @@
>  AC_CONFIG_SRCDIR([configure.ac])
>  AC_CONFIG_AUX_DIR(.)
>  
> -AM_INIT_AUTOMAKE(1.9 dist-bzip2 no-define no-dist-gzip tar-ustar -Wno-portability)

++

Comment on attachment 86371 [details] [review]
test-log-manager: make sure to pass a (gint64) as timestamp

Review of attachment 86371 [details] [review]:
-----------------------------------------------------------------

++

Comment on attachment 86372 [details] [review]
sqlite: deal with g_date_time_new_from_unix_utc() failing

Review of attachment 86372 [details] [review]:
-----------------------------------------------------------------

::: telepathy-logger/log-store-sqlite.c
@@ +329,4 @@
>    gchar *date;
>  
>    ts = g_date_time_new_from_unix_utc (tpl_event_get_timestamp (event));
> +  g_return_val_if_fail (ts != NULL, NULL);

Would it be better to return "9999-12-31" or something?

I'd be OK with declaring the answer to be "no" though.

Comment on attachment 86373 [details] [review]
tpl_log_manager_search_async: fix introspection

Review of attachment 86373 [details] [review]:
-----------------------------------------------------------------

++

(In reply to comment #7)
> Comment on attachment 86372 [details] [review] [review]
> sqlite: deal with g_date_time_new_from_unix_utc() failing
> 
> Review of attachment 86372 [details] [review] [review]:
> -----------------------------------------------------------------
> 
> ::: telepathy-logger/log-store-sqlite.c
> @@ +329,4 @@
> >    gchar *date;
> >  
> >    ts = g_date_time_new_from_unix_utc (tpl_event_get_timestamp (event));
> > +  g_return_val_if_fail (ts != NULL, NULL);
> 
> Would it be better to return "9999-12-31" or something?
> 
> I'd be OK with declaring the answer to be "no" though.

I think that's fine. I spot this only because of the other bug I just fixed.