diff options
5 files changed, 46 insertions, 0 deletions
diff --git a/doc/devblog/day_27__locking_fun/comment_2_e8b1dfe1b0641e031d05733448b7bc8b._comment b/doc/devblog/day_27__locking_fun/comment_2_e8b1dfe1b0641e031d05733448b7bc8b._comment new file mode 100644 index 000000000..c073f735a --- /dev/null +++ b/doc/devblog/day_27__locking_fun/comment_2_e8b1dfe1b0641e031d05733448b7bc8b._comment @@ -0,0 +1,8 @@ +[[!comment format=mdwn + username="http://joeyh.name/" + ip="4.152.108.243" + subject="comment 2" + date="2013-10-04T20:13:37Z" + content=""" +Sorry about that. It turns out my release script was broken. +"""]] diff --git a/doc/forum/bash_completion.mdwn b/doc/forum/bash_completion.mdwn new file mode 100644 index 000000000..649ad0457 --- /dev/null +++ b/doc/forum/bash_completion.mdwn @@ -0,0 +1 @@ +I wrote more than 2 years ago a bash completion file for git-annex (did I miss an official version?) I just updated it today to match current version of git-annex, it is far from perfect, still incomplete, and probably buggy, but works for me. I tried to attach the file in case it could be useful to someone, but it seems that I can't; is it ok to paste the code directly? diff --git a/doc/todo/wishlist:___96__git_annex_drop_--relaxed__96__/comment_2_353fbc2bcc40cb8c9af42907a34c6e5a._comment b/doc/todo/wishlist:___96__git_annex_drop_--relaxed__96__/comment_2_353fbc2bcc40cb8c9af42907a34c6e5a._comment new file mode 100644 index 000000000..40f299fa0 --- /dev/null +++ b/doc/todo/wishlist:___96__git_annex_drop_--relaxed__96__/comment_2_353fbc2bcc40cb8c9af42907a34c6e5a._comment @@ -0,0 +1,11 @@ +[[!comment format=mdwn + username="http://joeyh.name/" + ip="4.152.108.243" + subject="comment 2" + date="2013-10-04T20:17:07Z" + content=""" +I don't think that a third state would be necessary. Actually dropping the file when it happens would need to do the same numcopies verification that `git annex drop` does now. + +I agree it might be simpler to first improve the power of preferred content expressions. Unfortunately one thing that cannot be put in them is anything that probes the current state of the system. This is because repo A on machine X needs to be able to calculate the preferred content of repo B on machine Y. +But I could certainly add file size as a preferred content term, since that info is known throughout the network. +"""]] diff --git a/doc/todo/wishlist:_unify_directory_scheme_for_the_store/comment_1_44da58beaaab359ecaba7fb905ca4ae1._comment b/doc/todo/wishlist:_unify_directory_scheme_for_the_store/comment_1_44da58beaaab359ecaba7fb905ca4ae1._comment new file mode 100644 index 000000000..86c33d434 --- /dev/null +++ b/doc/todo/wishlist:_unify_directory_scheme_for_the_store/comment_1_44da58beaaab359ecaba7fb905ca4ae1._comment @@ -0,0 +1,10 @@ +[[!comment format=mdwn + username="http://joeyh.name/" + ip="4.152.108.243" + subject="comment 1" + date="2013-10-04T20:50:33Z" + content=""" +Using the mixed case hash directory names is not desirable because people want to use git-annex on a variety of filesystems and operating systems that treat them in a variety of broken ways. However, migrating to the all lower case hash directory names would require changing every git-annex symlink in every git repository, and I do not want to inflict that on my users. + +It seems to me that the best solution to your problem is to install git-annex on your server, which should not be very hard. +"""]] diff --git a/doc/todo/wishlist:_unify_directory_scheme_for_the_store/comment_2_bc698c501ecdb56df57171f4f3bb831a._comment b/doc/todo/wishlist:_unify_directory_scheme_for_the_store/comment_2_bc698c501ecdb56df57171f4f3bb831a._comment new file mode 100644 index 000000000..5c5269f8c --- /dev/null +++ b/doc/todo/wishlist:_unify_directory_scheme_for_the_store/comment_2_bc698c501ecdb56df57171f4f3bb831a._comment @@ -0,0 +1,16 @@ +[[!comment format=mdwn + username="https://www.google.com/accounts/o8/id?id=AItOawnpdM9F8VbtQ_H5PaPMpGSxPe_d5L1eJ6w" + nickname="Rafaël" + subject="comment 2" + date="2013-10-05T10:45:16Z" + content=""" +Thank you for prompt answer. I didn't know there were tarballs, +and indeed I managed to install them easily (although I had to +manually install glibc version 2.9, only 2.7 was installed). + +I found that bare git repos also use lower case hash directory +names... I still would be happy with an optional migration to all +lower case, with a new key-value backend, to avoid this little +complication which happens sometimes (say when converting a repo +from non-bare to bare). +"""]] |