+++ /dev/null
-[[!comment format=mdwn
- username="jkniiv@b330fc3a602d36a37a67b2a2d99d4bed3bb653cb"
- nickname="jkniiv"
- avatar="http://cdn.libravatar.org/avatar/419f2eee8b0c37256488fabcc2737ff2"
- subject="`git annex sync --no-commit --content` takes double the time of `git annex get .`"
- date="2021-08-20T02:05:53Z"
- content="""
-Hi Joey! Could you think of a reason why a `git annex sync --no-commit --content` takes pretty
-much double the time to retrieve an annexed file than a `git annex get .` in the same directory /
-git remote with this one file added in origin? Naturally I drop the file inbetween measurements
-and my files are of multi-gigabyte size so the delay is really noticeable between two HDDs.
-It seems that the extra time (and disk activity) takes place after the retrieval has hit 100% so
-to me it feels like git-annex is doing an extra verification pass after copying the file. This is
-on Windows and git-annex is a fresh `8.20210804-g492036622` but I seem to have noticed it
-happening even before your latest development activities above (as in months ago). Should I
-file a bug report about this?
-
-"""]]
+++ /dev/null
-[[!comment format=mdwn
- username="jkniiv@b330fc3a602d36a37a67b2a2d99d4bed3bb653cb"
- nickname="jkniiv"
- avatar="http://cdn.libravatar.org/avatar/419f2eee8b0c37256488fabcc2737ff2"
- subject="it turns out I had to file this as a bug"
- date="2021-08-27T01:38:29Z"
- content="""
-@joey I don't know what you mean by scanning in this case but I now have proof that git-annex really is
-doing an extra read pass over the whole file after it's been transferred from one regular remote to the
-other in the case of `sync --content[-of _file_]`. I filed a bug with details:
-[`sync -C` takes longer to get file than `get`](https://git-annex.branchable.com/bugs/__96__sync_-C__96___takes_longer_to_get_file_than___96__get__96__/)
-(btw. I can't seem to be able to make this into a WikiLink as ikiwiki thinks of the underscores of the page name as denoting strong emphasis).
-
-"""]]