summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorIan Jackson <ijackson@chiark.greenend.org.uk>2018-01-31 18:26:01 +0000
committerIan Jackson <ijackson@chiark.greenend.org.uk>2018-06-16 12:25:49 +0100
commit880af0a90f1b02db4607e94d38555f2e86af47a3 (patch)
treefe4f4b6bde56de430d4c136589e629ac3aa2e1ae
parentcdb3884ed9c0a404c956fd0bc3df420ee545debe (diff)
git-debrebase: docs updates
Signed-off-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
-rw-r--r--NOTES.git-debrebase31
-rwxr-xr-xgit-debrebase124
2 files changed, 76 insertions, 79 deletions
diff --git a/NOTES.git-debrebase b/NOTES.git-debrebase
index aa725b1..c268d75 100644
--- a/NOTES.git-debrebase
+++ b/NOTES.git-debrebase
@@ -1,3 +1,34 @@
+
+#
+# git-ffqrebase start [BASE]
+# # records previous HEAD so it can be overwritten
+# # records base for future git-ffqrebase
+# git-ffqrebase set-base BASE
+# git-ffqrebase <git-rebase options>
+# git-ffqrebase finish
+# git-ffqrebase status [BRANCH]
+#
+# refs/ffqrebase-prev/BRANCH BRANCH may be refs/...; if not it means
+# refs/ffqrebase-base/BRANCH refs/heads/BRANCH
+# zero, one, or both of these may exist
+#
+# git-debrebase without start, if already started, is willing
+# to strip pseudomerges provided that they overwrite exactly
+# the previous HEAD
+# xxxx is this right ? what matters is have we pushed
+# I think in fact the right answer is:
+# git-debrebase always strips out pseudomerges from its branch
+# a pseudomerge is put in at the time we want to push
+# at that time, we make a pseudomerge of the remote tracking
+# branch (if raw git) or the dgit view (if dgit)
+# for raw git git-ffqrebase, do want preciseley to record
+# value of remote tracking branch or our branch, on start, so we
+# overwrite only things we intend to
+# the previous pseudomerge check for tags and remote branches ?
+
+
+=========
+
workflow
git-debrebase blah [implies start] strips pseudomerge(s)
diff --git a/git-debrebase b/git-debrebase
index 0e2e0b7..85cf331 100755
--- a/git-debrebase
+++ b/git-debrebase
@@ -3,7 +3,7 @@
# Script helping make fast-forwarding histories while still rebasing
# upstream deltas when working on Debian packaging
#
-# Copyright (C)2017 Ian Jackson
+# Copyright (C)2017,2018 Ian Jackson
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
@@ -18,94 +18,60 @@
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
-# git-debrebase new-upstreams-v0 \
-# NEW-VERSION ORIG-COMMITISH
-# [EXTRA-ORIG-NAME EXTRA-ORIG-COMMITISH ...]
-# [<git-rebase options>]
# usages:
-# git-debrebase status
-# git-debrebase start # like ffqrebase start + debrebase launder
-# git-debrebase new-upstream [stuff] # see below
-# git-debrebase <git-rebase options> # does debrebase start if necessary
#
-# git-debrebase analyse
-# git-debrebase launder # prints breakwater tip
-# git-debrebase create-new-upstream-breakwater [-f] <upstreaminfo>...
+# git-debrebase [<options>] new-upstream-v0 \
+# <new-version> <orig-commitish>
+# [<extra-orig-name> <extra-orig-commitish> ...]
+# [<git-rebase options>...]
#
-# <upstreaminfo> is
-# [,][<subdir>:][+]<commitid>[,...]
+# git-debrebase [<options> --] [<git-rebase options...>]
+# git-debrebase [<options>] analyse
+# git-debrebase [<options>] launder # prints breakwater tip etc.
+# git-debrebase [<options>] downstream-rebase-launder-v0 # experimental
+
+# problems / outstanding questions:
+#
+# * dgit push with a `3.0 (quilt)' package means doing quilt
+# fixup. Usually this involves recommitting the whole patch
+# series, one at a time, with dpkg-source --commit. This is
+# terribly terribly slow. (Maybe this should be fixed in dgit.)
+#
+# * dgit push usually needs to (re)make a pseudomerge. The "first"
+# git-debrebase stripped out the previous pseudomerge and could
+# have remembeed the HEAD. But it's not quite clear what history
+# ought to be preserved and what should be discarded. For now
+# the user will have to tell dgit --overwrite.
+#
+# To fix this, do we need a new push hook for dgit ?
+#
+# * Workflow is currently clumsy. Lots of spurious runes to type.
+# There's not even a guide.
+#
+# * There are no tests.
#
-# if initial comma is supplied, entries are not positional. Unspecified
-# <subdir> means root (and there may be only one).
-# xxx want auto branch names
-# xxx too complicated
-# how about for now
-# [+]<commit> [<subdir/> [+]<commit>...]
-# ? plus options
-# --new-upstream-different-subtrees
+# * new-upstream-v0 has a terrible UI. You end up with giant
+# runic command lines.
#
-# automatic case
-# git-debrebase new-upstream
-# - previous breakwater merge must be gdr-generated
-# - orig set is the same as before
-# - implicitly uses upstream branches according to orig set
-# - not all upstream branches need be updated
-# - insists on fast-forward of each branch, unless
-# --force (or --force=<subdir>[/])
-# branch set adjustments
-# git-debrebase new-upstream --add <subdir>/
-# git-debrebase new-upstream --rm <subdir>/
-# git-debrebase new-upstream / [<subdir>/ ...]
-# - orig set is adjusted
-# - otherwise like auto (--add is not checked for ffness, obv)
-# - multiple --add and --rm may be specified
-# - --add makes new upstream the last contributor
-# explicit
-# git-debrebase / [<rootcommitid>] [<subdir>/ [<commitid>] ...]
-# - orig set is precisely as specified now
-# - previous breakwater merge is irrelevant
-# - no fast forward checks
-# for now only explicit with commitids
-
-# implicitly uses `upstream'
-# # (or multiple other branches)
-# git-debrebase new-upstream \
-# [<subdir>/]=<commitid>
-
-# UPSTREAM[,[[SUBDIR:]SUBUPSTREAM]
-# default for SUBDIR: is from previous upstream merge[xxx terminology]
-#
+# One consequence of the lack of richness it can need --force in
+# fairly sensible situations and there is no way to tell it what
+# you are really trying to do, other than just --force. There
+# should be an interface with some default branch names.
#
-#xxx
-# when starting must record original start (for ff)
-# and new rebase basis
+# * There should be a standard convention for the version number,
+# and unfinalised or not changelog, after new-upstream.
#
-# git-ffqrebase start [BASE]
-# # records previous HEAD so it can be overwritten
-# # records base for future git-ffqrebase
-# git-ffqrebase set-base BASE
-# git-ffqrebase <git-rebase options>
-# git-ffqrebase finish
-# git-ffqrebase status [BRANCH]
+# * Handing of multi-orig dgit new-upstream .dsc imports is known to
+# be broken. They may be not recognised, improperly converted, or
+# their conversion may be unrecognised.
#
-# refs/ffqrebase-prev/BRANCH BRANCH may be refs/...; if not it means
-# refs/ffqrebase-base/BRANCH refs/heads/BRANCH
-# zero, one, or both of these may exist
+# * Docs need writing and updating. Even README.git-debrebase
+# describes a design but may not reflect the implementation.
#
-# git-debrebase without start, if already started, is willing
-# to strip pseudomerges provided that they overwrite exactly
-# the previous HEAD
-# xxxx is this right ? what matters is have we pushed
-# I think in fact the right answer is:
-# git-debrebase always strips out pseudomerges from its branch
-# a pseudomerge is put in at the time we want to push
-# at that time, we make a pseudomerge of the remote tracking
-# branch (if raw git) or the dgit view (if dgit)
-# for raw git git-ffqrebase, do want preciseley to record
-# value of remote tracking branch or our branch, on start, so we
-# overwrite only things we intend to
-# the previous pseudomerge check for tags and remote branches ?
+# * We need to develop a plausible model that works for derivatives,
+# who probably want to maintain their stack on top of Debian's.
+# downstream-rebase-launder-v0 may be a starting point?
use strict;