On 1/14/26 12:22, Mark Millard wrote: > ampere4 building 143arm64-quarterly: > Host OSVERSION: 1600007 > Jail OSVERSION: 1403000 > > > electron37-37.10.3 32:47:14 > qt6-webengine-6.10.1_1 21:44:03 > > See: > <https://pkg-status.freebsd.org/ampere4/build.html?mastername=143arm64-quarterly&build=6fef647fdec5> > > > ampere4 building 150arm64-quarterly: > Host OSVERSION: 1600007 > Jail OSVERSION: 1500068 > > > electron37-37.10.3 34:38:40 > qt6-webengine-6.10.1_1 30:27:11 > > See: > https://pkg-status.freebsd.org/ampere4/build.html?mastername=150arm64-quarterly&build=6fef647fdec5
(I should have put that in <>'s.) > > > One or more things about 1403000 vs. 1500068 makes a big difference in > qt6-webengine-6.10.1_1 times (in comparison/contrast to > electron37-37.10.3 build times). > > Something about 15.0-RELEASE greatly adds time to the > qt6-webengine-6.10.1_1 build. Change of quote order to avoid confusions, moving the below from the bottom of the original message: > > > Note: All that is for builds of the same 6fef647fdec5 ports tree materials. > > Such is not true for the below comparison/contrast. > > It looks likely that the grand total times for the port-package builds > for 14.3 vs. 15.0 will also show a large difference once the > 143arm64-quarterly completes. Actually, since the context for the 2 above is an incremental build overall for each, such an comparison/contrast would not be so good. The prior (nearly?) from-scratch builds available for getting times: 143arm64-quarterly built 35544 failed 118 : 60:06:37 Host OSVERSION: 1600007 Jail OSVERSION: 1403000 <https://pkg-status.freebsd.org/builds/default:quarterly:143arm64:fce61163b43e:ampere4> 150arm64-quarterly built 35501 failed 154 : 78:25:36 Host OSVERSION: 1600000 Jail OSVERSION: 1500068 <https://pkg-status.freebsd.org/builds/default:quarterly:150arm64:0d8fad0cea6d:ampere4> So: 15.0-RELEASE had fewer built, more failed, and the longer time by a notable factor compared to 14.3-RELEASE. Both had chromium build/timeout its build after about 48 hrs. -- === Mark Millard marklmi at yahoo.com
