Page MenuHomePhabricator

marmoute (Pierre-Yves David)
User

Projects

User does not belong to any projects.

User Details

User Since
Jun 29 2017, 2:56 PM (162 w, 21 h)

Recent Activity

Today

marmoute added inline comments to D8633: share: introduce config option to store requires in .hg/store.
Fri, Aug 7, 4:40 AM
marmoute added a comment to D8633: share: introduce config option to store requires in .hg/store.

I left a handful of comments.
Changing repo opening semantics and requirements handling is a bit scary.
I'm generally in favor of separating working directory requirements from non-working directory requirements so I think this patch is a step in the right direction. There are a few alternative implementations worth considering. (Although I'm unsure if they are superior.)

  1. Write the new requirements file in the .hg directory as .hg/store-requires (or similar). Conceptually, we're talking about working directory and backend requirements. The "store" is just today's layout for storing backend data [which can be shared between working directories].
  2. Write a new requirements file for working directory specific requirements.
Fri, Aug 7, 4:24 AM

Yesterday

marmoute added a comment to D6776: bookmarks: validate changes on push (issue6193) (BC).

As @valentin.gatienbaron pointed out, the now avoid adding more semantic to bare --force with an associated --force-bookmark especially because we want to be able to select the bookmarks that get force pushed here.

Thu, Aug 6, 1:06 PM
marmoute added a comment to D8859: keepalive: Do not append _rbuf if _raw_readinto exists (issue6356).

Looks legit. Thanks.

Thu, Aug 6, 7:25 AM

Fri, Jul 31

marmoute abandoned D8789: infinitepush: remove unused import of tempfile in store.py.
Fri, Jul 31, 1:51 PM
marmoute added a reviewer for D8789: infinitepush: remove unused import of tempfile in store.py: pulkit.

Changeset fc54f52779dd is doing the same thing.

Fri, Jul 31, 1:51 PM
marmoute added inline comments to D8653: copies: handle more cases where a file got replaced by a copy.
Fri, Jul 31, 12:58 PM
D8624: wip: this does not help but seems less wrong now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:04 PM
D8572: merge: remove special case for in-memory merge now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:04 PM
D8504: diff: add experimental support for "merge diffs" now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:03 PM
D8515: filemerge: add a hacktastic version of internal:merge3 for merge diffs now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:03 PM
D8570: rebase: use context to load mergestate instead of loading it directly now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:03 PM
D8569: merge: get mergestate from context instead of directly now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:03 PM
D8568: overlayworkingctx: implement mergestate() using in-memory mergestate now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:03 PM
D8567: mergestate: implement trivial in-memory mergestate now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:02 PM
D8564: fakemergerecord: get mergestate via context now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:02 PM
D8563: localrepo: get mergestate via context now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:02 PM
D8626: memctx: add mergestate method now requires changes to proceed.

According to IRC discussion, this series is not ready for review, so moving it out of yadda.

Fri, Jul 31, 12:02 PM
marmoute committed rHG0041a42c6f28: commitctx: gather more code dealing with copy-in-extra.
commitctx: gather more code dealing with copy-in-extra
Fri, Jul 31, 9:14 AM
marmoute committed rHGb3040b6739ce: commitctx: extract copy information encoding into extra into commit.py.
commitctx: extract copy information encoding into extra into commit.py
Fri, Jul 31, 9:14 AM
marmoute committed rHGefe8a67793b6: commitctx: rename files to touched in a couple of place.
commitctx: rename files to touched in a couple of place
Fri, Jul 31, 9:14 AM
marmoute committed rHG4cde23ba076e: commitctx: create the new extra dict on its own line.
commitctx: create the new extra dict on its own line
Fri, Jul 31, 9:14 AM
marmoute committed rHGe15416c95b25: commitctx: explicitly pass `manifest` to _commit_manifest.
commitctx: explicitly pass `manifest` to _commit_manifest
Fri, Jul 31, 9:14 AM

Wed, Jul 29

marmoute accepted D8841: cleanup: fix bad formatting of state.py from D8811.
Wed, Jul 29, 12:17 PM

Sun, Jul 26

marmoute committed rHGbd7515273fd6: commitctx: gather more preparation code within the lock context.
commitctx: gather more preparation code within the lock context
Sun, Jul 26, 12:17 AM
marmoute committed rHG13814622b3b1: commitctx: extract all the file preparation logic in a new function.
commitctx: extract all the file preparation logic in a new function
Sun, Jul 26, 12:17 AM
marmoute committed rHG39f4ec4039a8: commitctx: move a special case about files earlier.
commitctx: move a special case about files earlier
Sun, Jul 26, 12:17 AM
marmoute committed rHG0c468fef09b3: commitctx: extract all the manual logic to process the files.
commitctx: extract all the manual logic to process the files
Sun, Jul 26, 12:17 AM
marmoute committed rHGb65b4b09859c: commitctx: treat `filesadded` more like `filesremoved`.
commitctx: treat `filesadded` more like `filesremoved`
Sun, Jul 26, 12:17 AM
marmoute committed rHG595307e14140: commitctx: move `writechangesetcopy` business at the end a code section.
commitctx: move `writechangesetcopy` business at the end a code section
Sun, Jul 26, 12:17 AM
marmoute committed rHG4eb6466e6889: commitctx: no longer use the `writecopiesto` variable in the function.
commitctx: no longer use the `writecopiesto` variable in the function
Sun, Jul 26, 12:17 AM
marmoute committed rHGada51c1b6916: commitctx: move copy meta config reading in a dedicated function.
commitctx: move copy meta config reading in a dedicated function
Sun, Jul 26, 12:17 AM
marmoute committed rHGf0d4d1343cb4: commitctx: extract the function that commit a new manifest.
commitctx: extract the function that commit a new manifest
Sun, Jul 26, 12:17 AM
marmoute committed rHG5d0998ccedbb: commitctx: stop using weakref proxy for transaction.
commitctx: stop using weakref proxy for transaction
Sun, Jul 26, 12:17 AM
marmoute committed rHGd056a131c93f: commitctx: document the None return for "touched" value.
commitctx: document the None return for "touched" value
Sun, Jul 26, 12:17 AM

Sat, Jul 25

marmoute committed rHG559ebfb5a58e: push: test for checks preventing publishing obsolete changeset.
push: test for checks preventing publishing obsolete changeset
Sat, Jul 25, 6:12 AM
marmoute committed rHG5ff77aaba1e4: push: test for checks preventing publishing obsolete changeset.
push: test for checks preventing publishing obsolete changeset
Sat, Jul 25, 6:12 AM
marmoute committed rHGfc751824ae97: push: another test for checks preventing pushing orphaness to a server.
push: another test for checks preventing pushing orphaness to a server
Sat, Jul 25, 6:12 AM
marmoute committed rHG187cc478e18f: push: test the checks preventing pushing orphaness to a server.
push: test the checks preventing pushing orphaness to a server
Sat, Jul 25, 6:12 AM
marmoute added a reverting change for rHGc26335fa4225: exchange: check actually missing revs for obsolete / unstable revs (issue6372): rHG6063c1857d0a: exchange: backout changeset c26335fa4225.
Sat, Jul 25, 6:12 AM
marmoute added a reverted change for rHG6063c1857d0a: exchange: backout changeset c26335fa4225: rHGc26335fa4225: exchange: check actually missing revs for obsolete / unstable revs (issue6372).
exchange: backout changeset c26335fa4225
Sat, Jul 25, 6:12 AM

Thu, Jul 23

marmoute committed rHGbfc6e75c0114: infinitepush: remove unused import to tempfile.
infinitepush: remove unused import to tempfile
Thu, Jul 23, 11:50 PM

Wed, Jul 22

marmoute closed D8710: commitctx: extract _filecommit too.
Wed, Jul 22, 8:18 AM
marmoute closed D8709: commitctx: extract the function in a dedicated module.
Wed, Jul 22, 8:17 AM
marmoute committed rHGce9ee81df9ff: commitctx: extract _filecommit too.
commitctx: extract _filecommit too
Wed, Jul 22, 8:17 AM
marmoute closed D8700: commitctx: document a fast path in _filecommit.
Wed, Jul 22, 8:17 AM
marmoute committed rHGae5c1a3bc339: commitctx: extract the function in a dedicated module.
commitctx: extract the function in a dedicated module
Wed, Jul 22, 8:17 AM
marmoute committed rHG31393ec06cef: commitctx: document a fast path in _filecommit.
commitctx: document a fast path in _filecommit
Wed, Jul 22, 8:16 AM

Tue, Jul 21

marmoute updated the diff for D8710: commitctx: extract _filecommit too.
Tue, Jul 21, 6:10 PM
marmoute updated the diff for D8709: commitctx: extract the function in a dedicated module.
Tue, Jul 21, 6:10 PM
marmoute updated the diff for D8700: commitctx: document a fast path in _filecommit.
Tue, Jul 21, 6:10 PM
marmoute added inline comments to D8700: commitctx: document a fast path in _filecommit.
Tue, Jul 21, 1:00 PM
marmoute committed rHG55464c0b3a89: commitctx: move a tiny else clause above the very long one.
commitctx: move a tiny else clause above the very long one
Tue, Jul 21, 12:28 PM
marmoute committed rHGc068e355cd71: commitctx: consider removed as touched.
commitctx: consider removed as touched
Tue, Jul 21, 12:28 PM
marmoute committed rHG025ee25b5b23: commitctx: rename "changed" to touched.
commitctx: rename "changed" to touched
Tue, Jul 21, 12:28 PM
marmoute committed rHG6979a20ff004: commitctx: more filesremoved assignment closer to removed computation.
commitctx: more filesremoved assignment closer to removed computation
Tue, Jul 21, 12:28 PM
marmoute committed rHGe64c35dfc6cb: commitctx: compute files added from _filecommit returns.
commitctx: compute files added from _filecommit returns
Tue, Jul 21, 12:28 PM
marmoute committed rHG509f5b6c0b7e: commitctx: return "touched" status from _filecommit.
commitctx: return "touched" status from _filecommit
Tue, Jul 21, 12:28 PM
marmoute closed D8708: commitctx: move a tiny else clause above the very long one.
Tue, Jul 21, 12:18 PM
marmoute closed D8707: commitctx: consider removed as touched.
Tue, Jul 21, 12:18 PM
marmoute committed rHG1386a2292d93: commitctx: move a tiny else clause above the very long one.
commitctx: move a tiny else clause above the very long one
Tue, Jul 21, 12:17 PM
marmoute closed D8702: commitctx: return "touched" status from _filecommit.
Tue, Jul 21, 12:17 PM
marmoute closed D8706: commitctx: rename "changed" to touched.
Tue, Jul 21, 12:17 PM
marmoute committed rHGd167f9fbb61d: commitctx: consider removed as touched.
commitctx: consider removed as touched
Tue, Jul 21, 12:17 PM
marmoute committed rHG765c3b4e2176: commitctx: rename "changed" to touched.
commitctx: rename "changed" to touched
Tue, Jul 21, 12:17 PM
marmoute closed D8704: commitctx: more filesremoved assignment closer to removed computation.
Tue, Jul 21, 12:17 PM
marmoute closed D8703: commitctx: compute files added from _filecommit returns.
Tue, Jul 21, 12:17 PM
marmoute committed rHG25fc080eed1a: commitctx: more filesremoved assignment closer to removed computation.
commitctx: more filesremoved assignment closer to removed computation
Tue, Jul 21, 12:17 PM
marmoute committed rHG7fce7d1eeacc: commitctx: compute files added from _filecommit returns.
commitctx: compute files added from _filecommit returns
Tue, Jul 21, 12:16 PM
marmoute committed rHG4bc65fb8d6d1: commitctx: return "touched" status from _filecommit.
commitctx: return "touched" status from _filecommit
Tue, Jul 21, 12:16 PM
marmoute added inline comments to D8700: commitctx: document a fast path in _filecommit.
Tue, Jul 21, 11:36 AM
marmoute added inline comments to D8702: commitctx: return "touched" status from _filecommit.
Tue, Jul 21, 11:35 AM
marmoute accepted D8656: localrepo: load the share source .hg/hgrc also in share-safe mode (API).
Tue, Jul 21, 9:47 AM
D8764: localrepo: deduplicate code around reading requires file now requires changes to proceed.
Tue, Jul 21, 9:46 AM
marmoute added a comment to D8764: localrepo: deduplicate code around reading requires file.

That is a bad reason to use a close. Make the function private and move it at the module level.
(Mercurial used to be full of this strange closure and this was a pain)

Tue, Jul 21, 9:12 AM
marmoute added a comment to D8771: localrepo: only use 'bookmarksinstore' requirement if we have 'store'.

The tricky part is that all the requirements has default value. If requirement B depends on requirement A, and both are enabled by default, we cannot simply automatically enable A if B is true, because B is true by default. Warning would be good, but might end up a bit verbose for the same reason.
However if the user explicitly request A=no and B=yes, it seems bad to silently drop B since it was explicitly requested.
So, I don't see a good solution until we detect explicit activation vs default value (which is a scope bloat).

Very much agreed. Since we know that store is enabled by default and bookmarks in store is not enabled, we can be aware whether user is explicitly requesting something.

Tue, Jul 21, 6:10 AM
marmoute added a comment to D8771: localrepo: only use 'bookmarksinstore' requirement if we have 'store'.

The tricky part is that all the requirements has default value. If requirement B depends on requirement A, and both are enabled by default, we cannot simply automatically enable A if B is true, because B is true by default. Warning would be good, but might end up a bit verbose for the same reason.

Tue, Jul 21, 6:03 AM

Mon, Jul 20

D8742: merge: introduce 'commitinfo' in mergeresult now requires review to proceed.

More serious question: Maybe we should stop passing around tuple of dictionnary and return an object holding all that data with a better API. What do you think ?

Mon, Jul 20, 5:00 PM
D8764: localrepo: deduplicate code around reading requires file now requires changes to proceed.

I resigned (and commented) on the wrong revision.

Mon, Jul 20, 4:59 PM
D8764: localrepo: deduplicate code around reading requires file now requires review to proceed.

More serious question: Maybe we should stop passing around tuple of dictionnary and return an object holding all that data with a better API. What do you think ?

Mon, Jul 20, 3:44 PM
marmoute added a comment to D8742: merge: introduce 'commitinfo' in mergeresult.

I can confirm phab eat my comment.

Mon, Jul 20, 3:42 PM
D8742: merge: introduce 'commitinfo' in mergeresult now requires changes to proceed.

This is mostly a test of phab eating comment

Mon, Jul 20, 3:41 PM
marmoute added inline comments to D8764: localrepo: deduplicate code around reading requires file.
Mon, Jul 20, 3:29 PM
marmoute added a comment to D8766: manifest: kill one more instance of the old merge hash hack.

This seems good and if the test are happy, we should be too.

Mon, Jul 20, 1:29 PM
marmoute accepted D8673: config: add a .hg/nonsharedrc which won't be shared in share-safe mode.
Mon, Jul 20, 1:23 PM
marmoute added a comment to D8656: localrepo: load the share source .hg/hgrc also in share-safe mode (API).

I would be more confident if we get some explicit test about trust. However unshare is not working.

Mon, Jul 20, 1:18 PM
D8764: localrepo: deduplicate code around reading requires file now requires changes to proceed.
Mon, Jul 20, 1:11 PM
marmoute added inline comments to D8633: share: introduce config option to store requires in .hg/store.
Mon, Jul 20, 1:05 PM
marmoute updated the diff for D8710: commitctx: extract _filecommit too.
Mon, Jul 20, 12:52 PM
marmoute updated the diff for D8709: commitctx: extract the function in a dedicated module.
Mon, Jul 20, 12:52 PM
marmoute updated the diff for D8708: commitctx: move a tiny else clause above the very long one.
Mon, Jul 20, 12:51 PM
marmoute updated the diff for D8707: commitctx: consider removed as touched.
Mon, Jul 20, 12:51 PM
marmoute updated the diff for D8706: commitctx: rename "changed" to touched.
Mon, Jul 20, 12:51 PM
marmoute updated the diff for D8702: commitctx: return "touched" status from _filecommit.
Mon, Jul 20, 12:51 PM
marmoute updated the diff for D8704: commitctx: more filesremoved assignment closer to removed computation.
Mon, Jul 20, 12:51 PM
marmoute updated the diff for D8703: commitctx: compute files added from _filecommit returns.
Mon, Jul 20, 12:50 PM
marmoute updated the diff for D8700: commitctx: document a fast path in _filecommit.
Mon, Jul 20, 12:50 PM
marmoute added a comment to D8709: commitctx: extract the function in a dedicated module.

(reply written ages ago)

Mon, Jul 20, 12:26 PM

Thu, Jul 16

marmoute updated subscribers of D8666: locks: expect repo lock, not wlock, when writing to .hg/strip-backup/.

I am a bit late to the party, but does this mean that and old client and
a new client could do conflicting write to strip backup because they
would be respectively holding wlock and lock, thinking that make them safe?

Thu, Jul 16, 1:42 PM
marmoute added a comment to D8596: merge: mark copies in in-memory context when merging.

Can you mark them as "changes planned" in the mean time to clear up yadda?

There are no actual changes planned and I guess I figured that the base patch being marked as "changes planned" would be sufficient, so sure, I'll update the other ones to that status too.

Thu, Jul 16, 11:24 AM