Hello, Luis Felipe <luis.felipe...@protonmail.com> writes:
> ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ > On Friday, October 16, 2020 10:36 AM, zimoun <zimon.touto...@gmail.com> wrote: > >> Hi, >> >> On Fri, 16 Oct 2020 at 12:28, Ludovic Courtès l...@gnu.org wrote: >> >> > > https://video.hardlimit.com/videos/watch/c0dfb36a-a84b-4363-8b1b-17aeadd4aaaf >> > >> > Thanks! I think 7 minutes is too long; IMO we should aim for ~1mn–1.5mn >> > at most as is currently the case. Also full screen, large enough fonts, >> > no “guix pull” warnings, ‘--max-jobs=1’ on the daemon side to reduce >> > verbosity. Slick and to-the-point! :-) >> >> The video is really long mainly because each command takes literally >> ages. Well the "XDG mine" step to be precise; maybe Joshua you have >> something misconfigured. > If the slowness at the end can be avoided by configuring something, > I'd like to know, because that slowness is always the case in my > experience. For example, installing the program Joshua mentioned > earlier, wf-recorder, which is quite small, took ~7 minutes in my > computer (Intel® Core™ i3-8100 CPU @ 3.60GHz × 4, 4 GiB RAM, 1 TB > HDD). I also experience poor performance on profile generation when using Guix on an older machine equipped with spinning disks. This kind of operation appears to be IO hungry, and performs much faster on solid state drives (SSDs). I've opened an issue for the poor performance on spinning drives here: https://issues.guix.gnu.org/44053. Thanks, Maxim
