From e9b0cf08eb6e3aa0e80600c0e8de079bface27ae Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Wed, 6 Oct 2021 17:08:57 -0400 Subject: [PATCH] branch --- ...ent_11_fe04d3da8859101ba1649fdd9d5ee39e._comment | 13 +------------ 1 file changed, 1 insertion(+), 12 deletions(-) diff --git a/doc/bugs/borg_special_remote_memory_usage_high_for_large_borg_repo/comment_11_fe04d3da8859101ba1649fdd9d5ee39e._comment b/doc/bugs/borg_special_remote_memory_usage_high_for_large_borg_repo/comment_11_fe04d3da8859101ba1649fdd9d5ee39e._comment index b234f769cb..ed5c7e2b12 100644 --- a/doc/bugs/borg_special_remote_memory_usage_high_for_large_borg_repo/comment_11_fe04d3da8859101ba1649fdd9d5ee39e._comment +++ b/doc/bugs/borg_special_remote_memory_usage_high_for_large_borg_repo/comment_11_fe04d3da8859101ba1649fdd9d5ee39e._comment @@ -24,16 +24,5 @@ committed. For borg, each archive would be a subtree; 500k filenames will fit in memory or at least fit better than `365*500k`. -The interface I'm thinking about is something like this: - - data ChunkedImportableContents info - = ImportableContentsChunk - { importableContentsRoot :: ImportLocation - , importableContentsSubTree :: [(ImportLocation, info)] - -- ^ locations are relative to importableContentsRoot - , importableContentsContinuation :: Annex (ChunkedImportableContents info) - } - | ImportableContentsComplete (ImportableContents info) - -This is a promising idea! +This is a promising idea! Started working on it in a `borgchunks` branch. """]]