[13:16:06] <vvv>	 Oh wow
[13:16:19] <vvv>	 The complete update of core and extensions took 15 minutes
[13:30:09] <Reedy>	 It's not quick, but I don't think it takes it that long for me
[14:22:07] <vvv>	 Why does profiling data takes forever to open?
[14:26:47] <Platonides>	 vvv, git update takes me about 11m
[16:35:00] <Amgine>	 Takes just under 20 for me, Reedy. I blame it on the netbook, though.
[16:35:22] <Reedy>	 real    10m3.694s
[16:35:24] <Reedy>	 On a VM
[16:36:23] <Amgine>	 <hands Reedy an Atom N570 Aspire one> Okay, now again.
[16:36:55] <Reedy>	 My eee 901 has 2 SSDs :p
[16:37:43] <Reedy>	 This vm host is low powered, Dual core Athlon II  Neo 36L (2 x 1.297GHz)
[16:38:26] <Amgine>	 <is timing it>
[16:38:57] <Amgine>	 Most of the time, realy, is in the network, but I'm on a wire.
[16:40:21] <vvv>	 Is there any hypothetical way to optimize that?
[16:40:32] <vvv>	 Probably by reusing SSH connection or something like that
[16:41:11] <Amgine>	 paralell.
[16:43:41] <Amgine>	 Reedy: warning: remote HEAD refers to nonexistent ref, unable to checkout. (in mediawiki)
[16:44:08] <vvv>	 That may be a good idea
[16:44:17] * YuviPanda  gives GNU parallel a hug

[16:44:38] * vvv  uses xargs for that

[16:44:50] <YuviPanda>	 vvv: xargs does parallel?
[16:44:57] <YuviPanda>	 without tricks?
[16:45:29] <vvv>	 Yes, if you use -n and -P
[16:45:38] <YuviPanda>	 hmm, this I did not know
[16:46:04] <vvv>	 -n 1 to run stuff in seperate processes, -P N to run N processes simultaneously
[16:49:52] <vvv>	 Oh wow
[16:50:03] <vvv>	 find -maxdepth 1 -type d | sed -e'/^.$/d' -e'/.git/d' | xargs -n 1 -P 16 ./gitupdate
[16:50:10] <vvv>	 That updates stuff in ~30 seconds
[17:34:25] <vvv>	 Okay, I wrote a script for that
[17:43:04] <Amgine>	 Oh, forgot to report to Reedy: Total time: 1567.6870670319 seconds
[17:43:04] <Amgine>	 real	26m8.154s
[18:09:31] <vvv>	 https://gerrit.wikimedia.org/r/#change,6072