Re: Some minor cleanups
Robert Krawitz <rlk-FrUbXkNCsVf2fBVCVOL8/[email protected]> Sat, 29 Feb 2020 22:20:41 -0500 (EST)
| Newsgroups | gmane.comp.kde.kimdaba |
|---|---|
| Message-ID | <20200301032041.BBB62424A53@localhost> |
On Tue, 25 Feb 2020 08:30:11 -0500 (EST), Robert Krawitz wrote: > On Tue, 25 Feb 2020 07:32:19 +0100, Tobias Leupold wrote: >>> I haven't changed maps to hashes at this time, but it's worth >>> considering. >> It definitely is, in each and every case the ordering doesn't matter (at least >> imho). >> >> Same for the Q_FOREACH removal I wanted to do recently. Tell me when your >> stuff is merged, so that I can redo it. > > I've merged everything but the change from maps to hashes, so you can > redo it. Maps vs. hashes is all localized in one place in > DB/ImageInfo.h, and shouldn't collide with what you're doing. The > parallel-md5 branch doesn't appear to contain any Q_FOREACHes. > > My repo's in flux right now since I started the merge last night, or I > would have done it myself. > >> Nice, the codebase gets cleaner and cleaner :-) > > Yes, this is a significant cleanup. OK, I've merged a slew of minor changes (including a fixlet to one of my minor perf improvements, related to thumbnail cell height), and the parallel-md5 branch. Last week I bought another 3.84TB SSD to replace the old 9.5mm 2TB HDD I had my non-7DmkII photos on (the shots from my 7DmkII were already on the 3.84TB SSD I used to develop the parallel-md5 stuff on). Almost surely coincidentally, within 6 hours of putting the new SSD into service, the old SSD died hard on me, basically just throwing errors on any command. So I wound up having to put the old HDD back into service and restore the 7DmkII images from backup, so the whole thing basically came out a wash. It was annoying because I shot an important basketball game yesterday evening and came home to find that the drive had failed on me. Anyhow, the new drive seems to be much less sensitive to the tuning parameters; I suppose its I/O latency is considerably smaller. The HDD setting is only modestly slower (450 MB/sec) than any of the SSD settings (about 500 MB/sec), vs. with the old drive, which was very sensitive to this. -- Robert Krawitz <rlk-FrUbXkNCsVf2fBVCVOL8/[email protected]> *** MIT Engineers A Proud Tradition http://mitathletics.com *** Member of the League for Programming Freedom -- http://ProgFree.org Project lead for Gutenprint -- http://gimp-print.sourceforge.net "Linux doesn't dictate how I work, I dictate how Linux works." --Eric Crampton