gotwebd: typo; fix closing of dd tag

tog: make patience the default diff algorithm As discussed on IRC. It makes nicer diffs and the performance hit should only be noticeable with the odd large diff. It might also be a motivator in trying to speed up the patience implementation. ok stsp@

replace "(cd path && git cmd)" with "git -C path cmd" This matches the existing use of "got -r path cmd" and "git_commit path args".

honor fetch_all_branches configuration again Fix a regression in db82695e31945b3ce001a5fb69674c3638622368 where the fetch_all_branches configuration setting from got.conf was dropped and always defaulted to "no". Add corresponding regression test. ok stsp

put actions in test_fetch_all into more logical order ok stsp

vertical align briefs age and author

portable: add --enable-cvg flag Just while cvg is being developed, mask the building of this via a flag to ./configure -- this is not intended to be used by any packagers, but rather, provide a means for developers for testing. This flag will go away once cvg is considered releasable.

portable: don't yet build cvg cvg is still under development, and not ready to be installed yet.

in diff error message, say what was being diffed When trying to run "got ci" adding a large file, this changes the following error message: got: diff_atomize_file: Cannot allocate memory into: got: /dev/null vs /path/to/the/really/big/file: diff_atomize_file: Cannot allocate memory ok stsp@

handle NULL argument to got_diffreg_result_free ok stsp@

portable: set next version

portable: release 0.93

bump version number

CHANGES for 0.93

fix expected output; with previous change fetch -l no longer lists HEAD twice

make gotsh handle poll timeouts when checking for gotd closing the pack pipe This will hopefully fix issues seen on servers where gotsh exits early for no good reason.

actually detect modified files when checking out with -E When checking out over a pre-existing directory tree with -E, got would store the checkout timestamp in the fileindex and so not detect local edits until those files are touch(1)ed again. Instead, detect that a file already exists during checkout and don't update its {c,m}time, leaving them to zero (the epoch), so a subsequent `got status' will look at its contents. While here, change the status char for such file from ? to E, which is what the manpage actually documents. Issue reported by naddy@ ok stsp@

make 'got send' detect connections that are unexpectedly closed by the server ok op@

sync files from diff.git b5a9c15f4d68c06ec3bf839529b3ed2def0a6af6 Fixes 'No newline at end of file' message appearing in unexpected places. Problem noticed by op@

fix test_send_config test directory name

gotwebd: move the buffering from the fastcgi layer to the template Reduces the indirection in fcgi.c, starts to make the struct template opaque, simplifies the template usage. All with a net negative :-) reads fine to stsp@ (thanks!)

gotwebd: don't loose track of the file in commit listing When browsing the history of a file, the more button should generate a link containing the file= parameter, otherwise we switch back to the 'global' history. Spotted by stsp@

TODO: gotwebd should reply with 404 on some kind of errors

regress/template: use template_free() rather than free(3)

template: join subsequent tp_puts() calls This greatly reduces the number of tp_puts() calls: gotwebd' pages.tmpl goes from 444 to 190. Otherwise, this doesn't produce other observable changes.