cdist, branch kill-docker Tool used to build DCP-o-matic packages https://git.carlh.net/cgit/cdist/atom?h=kill-docker 2022-10-04T06:53:45Z Kill docker image on SIGINT. 2022-10-04T06:53:45Z cah cah@ableton.com 2022-10-04T06:53:45Z urn:sha1:c9a6e16bfffad15db84b3a85e18fef45cace677d Another attempt to fix unnecessary building while packaging. 2022-02-19T21:34:45Z Carl Hetherington cth@carlh.net 2022-02-19T21:34:45Z urn:sha1:0b1c33732bcc66d00d4486c2ccf7f8111a076ae2 Revert "Don't build a tree that is being packaged; the package() method must do it (if required)." 2022-02-19T20:06:10Z Carl Hetherington cth@carlh.net 2022-02-19T20:06:10Z urn:sha1:0472bc3bab6557ff356b2cb313402b12b86c96e6 This reverts commit 1b9fcb5ab32d501a24f603d8a2e4fe8aff282e95. On reflection this approach just gets too complicated considering the override of package() for macOS. Don't build a tree that is being packaged; the package() method must do it (if required). 2022-02-19T16:32:01Z Carl Hetherington cth@carlh.net 2022-02-19T16:31:54Z urn:sha1:1b9fcb5ab32d501a24f603d8a2e4fe8aff282e95 This is to make Linux builds more efficient; previously cdist would build() everything (including the whole of DoM) and then most linux package() methods would end up building it again (making the build in build() redundant). Now the cscript in DoM can call build() if necessary. I'm a little bit worried that the environment is no longer preserved during the final build(), but I'm unsure if this is necessary anyway. Remove duplicate call to cleanup(). 2022-02-18T13:43:10Z cah cah@ableton.com 2022-02-18T13:43:10Z urn:sha1:a862b364888ff7b0cd30f6267309bab6ac56cdce Little cleanup. 2022-02-17T23:02:04Z cah cah@ableton.com 2022-02-17T23:02:04Z urn:sha1:b32b040544df8a84b89eb351764a456ed4884c15 Cleanup: rename git_commit -> commit. 2022-02-17T23:00:53Z cah cah@ableton.com 2022-02-17T23:00:53Z urn:sha1:d3d8e6763614a0f18b62eb0209acd99dc95ae343 Cleanup: rename specifier -> commit_ish. 2022-02-17T22:58:57Z cah cah@ableton.com 2022-02-17T22:58:57Z urn:sha1:1a13f315211066abe2f1bd0c5529154301bd9f61 Out-of-date comment. 2022-02-17T22:55:26Z cah cah@ableton.com 2022-02-16T13:01:23Z urn:sha1:2712596be3a8dc25636e202c9d9b3e9e8ba8283d Try to use finally better to terminate subprocesses when we die. 2022-02-14T12:38:43Z Carl Hetherington cth@carlh.net 2022-02-14T12:38:43Z urn:sha1:114c97a47cbb84d655259b757d5c015e2906f849