diff --git a/doc/forum/git-annex_parallel_invocation_and_the_git-annex_branch/comment_6_7d6d0da2243967b9660d8222c546112d._comment b/doc/forum/git-annex_parallel_invocation_and_the_git-annex_branch/comment_6_7d6d0da2243967b9660d8222c546112d._comment
new file mode 100644
index 0000000000..7558377b49
--- /dev/null
+++ b/doc/forum/git-annex_parallel_invocation_and_the_git-annex_branch/comment_6_7d6d0da2243967b9660d8222c546112d._comment
@@ -0,0 +1,18 @@
+[[!comment format=mdwn
+ username="Ilya_Shlyakhter"
+ avatar="http://cdn.libravatar.org/avatar/1647044369aa7747829c38b9dcc84df0"
+ subject="comment 6"
+ date="2019-05-06T22:38:43Z"
+ content="""
+During parallel running of git-annex commands, I also get errors like
+
+git-annex: .git/annex/othertmp/inge59014-3: getFileStatus: does not exist (No such file or directory)
+failed
+
+git-annex: .git/annex/othertmp/ingest-assemble_den59014-8: removeLink: does not exist (No such file or directory)
+failed
+
+Is this also now fixed by [[todo/only_allow_one_git_queue_to_be_flushed_at_a_time]], or is it a separate issue?
+
+Also, is it correct that linked worktrees must be on the same filesystem as the main worktree, when using git-annex, because the linked worktrees must be on the same filesystem as .git/annex/othertmp ?
+"""]]
diff --git a/doc/todo/key_checksum_from_chunk_checksums/comment_2_aaff896ff99366d3b96c523284c1248e._comment b/doc/todo/key_checksum_from_chunk_checksums/comment_2_aaff896ff99366d3b96c523284c1248e._comment
new file mode 100644
index 0000000000..9a02bff2de
--- /dev/null
+++ b/doc/todo/key_checksum_from_chunk_checksums/comment_2_aaff896ff99366d3b96c523284c1248e._comment
@@ -0,0 +1,8 @@
+[[!comment format=mdwn
+ username="Ilya_Shlyakhter"
+ avatar="http://cdn.libravatar.org/avatar/1647044369aa7747829c38b9dcc84df0"
+ subject="comment 2"
+ date="2019-05-07T00:59:31Z"
+ content="""
+Thanks for the BLAKE2SP224 pointer; it does solve (1).  I'm still looking for the best way to solve (2): registering large remote files with git-annex, without downloading them.  `addurl --fast` does that, but creates a non-checksum key.  If I can get an MD5 without downloading, I can use `setpresentkey`.  But often I only have the MD5 for the fixed-size chunks of the file, not for the whole.  Adding a backend variant computable from MD5s of the chunks would solve the problem.  Maybe, there are other solutions?
+"""]]