diff options
Diffstat (limited to 'doc')
| -rw-r--r-- | doc/bugs/Bloom_filter_capacity_too_large_to_represent/comment_2_9b74457549e2739ae45dccd128de946f._comment | 12 |
1 files changed, 12 insertions, 0 deletions
diff --git a/doc/bugs/Bloom_filter_capacity_too_large_to_represent/comment_2_9b74457549e2739ae45dccd128de946f._comment b/doc/bugs/Bloom_filter_capacity_too_large_to_represent/comment_2_9b74457549e2739ae45dccd128de946f._comment new file mode 100644 index 000000000..803465e01 --- /dev/null +++ b/doc/bugs/Bloom_filter_capacity_too_large_to_represent/comment_2_9b74457549e2739ae45dccd128de946f._comment @@ -0,0 +1,12 @@ +[[!comment format=mdwn + username="http://joeyh.name/" + ip="209.250.56.132" + subject="comment 2" + date="2014-09-12T16:34:56Z" + content=""" +However, in Greg's case he had no such configuration. Instead, I think something is broken with the use of floating point or bit math that bloomfilter uses, on the NAS where he's using git-annex. + +I have made git-annex not crash when this happens, just show a warning and fall back to a reasonable default bloom filter size. If the problem is with the bit math, then the bloom filter may not work either, which would probably show up as false negatives, so `git annex unused` not finding things that are unused. + +I need to update the armel build with this so Greg can test it.. +"""]] |