lothiraldan (Boris Feld)
User

Projects

User does not belong to any projects.

User Details

User Since
Jul 11 2017, 12:05 PM (67 w, 9 h)

Recent Activity

Today

lothiraldan closed D5185: storage: update sqlitestore to use the new `deltamode` parameter.
Tue, Oct 23, 3:13 PM
lothiraldan committed rHGa0e7fa019290: storage: update sqlitestore to use the new `deltamode` parameter.
storage: update sqlitestore to use the new `deltamode` parameter
Tue, Oct 23, 3:13 PM
lothiraldan created D5185: storage: update sqlitestore to use the new `deltamode` parameter.
Tue, Oct 23, 1:34 PM
lothiraldan committed rHG49c7b701fdc2: phase: add an archived phase.
phase: add an archived phase
Tue, Oct 23, 1:00 PM
lothiraldan committed rHG808b762679cd: changegroup: add a option to create bundle with full snapshot only.
changegroup: add a option to create bundle with full snapshot only
Tue, Oct 23, 11:01 AM
lothiraldan committed rHG968dd7e02ac5: changegroup: allow to force delta to be against p1.
changegroup: allow to force delta to be against p1
Tue, Oct 23, 11:01 AM
lothiraldan committed rHG6a917075535a: storage: also use `deltamode argument` for ifiledata.
storage: also use `deltamode argument` for ifiledata
Tue, Oct 23, 11:01 AM
lothiraldan committed rHG59a870a4ad6e: changegroup: refactor emitrevision to use a `deltamode` argument.
changegroup: refactor emitrevision to use a `deltamode` argument
Tue, Oct 23, 11:01 AM

Yesterday

lothiraldan closed D4820: logtoprocess: sends the canonical command name to the subprocess.
Mon, Oct 22, 11:57 AM
lothiraldan committed rHG106adc261492: logtoprocess: sends the canonical command name to the subprocess.
logtoprocess: sends the canonical command name to the subprocess
Mon, Oct 22, 11:57 AM
lothiraldan updated the diff for D4820: logtoprocess: sends the canonical command name to the subprocess.
Mon, Oct 22, 11:54 AM
lothiraldan closed D5180: logtoprocess: fix message formatting.
Mon, Oct 22, 11:49 AM
lothiraldan committed rHG6bd477ee7294: logtoprocess: fix message formatting.
logtoprocess: fix message formatting
Mon, Oct 22, 11:49 AM
lothiraldan committed rHGdb0dba2d157d: profiling: move default mode to "real" time.
profiling: move default mode to "real" time
Mon, Oct 22, 11:49 AM
lothiraldan added a comment to D4820: logtoprocess: sends the canonical command name to the subprocess.
In D4820#77397, @yuja wrote:

+ # extract the canonical command from the passed opts.
+ # we need to pop it up from the opts else the msg will try
+ # being formatted using the passed opts.

Oops, that's probably a bug. IIUC, **opts is just ancillary data, and
the msg[0] should always be parameterized by msg[1:]. That's how blackbox
works.

Can you send a separate patch to fix the issue?

Mon, Oct 22, 9:57 AM
lothiraldan added a dependent revision for D5180: logtoprocess: fix message formatting: D4820: logtoprocess: sends the canonical command name to the subprocess.
Mon, Oct 22, 9:56 AM
lothiraldan added a dependency for D4820: logtoprocess: sends the canonical command name to the subprocess: D5180: logtoprocess: fix message formatting.
Mon, Oct 22, 9:56 AM
lothiraldan updated the diff for D4820: logtoprocess: sends the canonical command name to the subprocess.
Mon, Oct 22, 9:55 AM
lothiraldan created D5180: logtoprocess: fix message formatting.
Mon, Oct 22, 9:55 AM
lothiraldan abandoned D5179: command: reintroduce support for commands starting with ^ (issue6005).
In D5179#77396, @yuja wrote:
Do we have performance concerns around the deprecation, how often is called `loadcmdtable`? My gut feeling is telling me it shouldn't matter but I would prefer to be sure.

I don't think my change would cause measurable slowdown. It's called once
per extension.

Mon, Oct 22, 9:39 AM
lothiraldan added a comment to D5179: command: reintroduce support for commands starting with ^ (issue6005).
In D5179#77391, @yuja wrote:
Mon, Oct 22, 8:56 AM
lothiraldan added a comment to D4820: logtoprocess: sends the canonical command name to the subprocess.
In D4820#77055, @yuja wrote:
  • a/hgext/logtoprocess.py +++ b/hgext/logtoprocess.py @@ -40,6 +40,9 @@ import sys

    from mercurial import ( + dispatch, + encoding, + extensions,

Unused.

@@ -110,6 +113,12 @@

script = self.config('logtoprocess', event)
if script:
    if msg:

+
+ # extract the canonical command from the passed opts
+ canonical_command = opts.pop("canonical_command", None)
+ if canonical_command is None:
+ canonical_command = ""
+

    1. try to format the log message given the remaining
    2. arguments try: @@ -134,7 +143,8 @@ for key, value in opts.iteritems()) env = dict(itertools.chain(procutil.shellenviron().items(), msgpairs, optpairs),
  • EVENT=event, HGPID=str(os.getpid())) + EVENT=event, HGPID=str(os.getpid()), + LTP_COMMAND=canonical_command)

As I said before, LTP_* doesn't sound great. Can't we simply pass in
canonical_command as OPT_CANONICAL_COMMAND? If we really want to name
it other than OPT_*, please document it.

Mon, Oct 22, 8:53 AM
lothiraldan updated the diff for D4820: logtoprocess: sends the canonical command name to the subprocess.
Mon, Oct 22, 8:51 AM
lothiraldan created D5179: command: reintroduce support for commands starting with ^ (issue6005).
Mon, Oct 22, 6:24 AM

Fri, Oct 19

lothiraldan added inline comments to D5147: config: make 'experimental' the section instead of 'experimental.server'.
Fri, Oct 19, 3:22 AM

Wed, Oct 17

lothiraldan accepted D5145: tests: fix "running x tests using y ... " output in a few more places.

LGTM

Wed, Oct 17, 4:38 PM

Tue, Oct 16

lothiraldan updated the diff for D4820: logtoprocess: sends the canonical command name to the subprocess.
Tue, Oct 16, 11:10 AM
lothiraldan committed rHG4aa04d009167: revset: drop special case of 'revset(...)' function in analyze.
revset: drop special case of 'revset(...)' function in analyze
Tue, Oct 16, 6:45 AM
lothiraldan committed rHGd894d2372ffe: revset: document the `revset(...)` syntax.
revset: document the `revset(...)` syntax
Tue, Oct 16, 6:44 AM

Thu, Oct 11

lothiraldan committed rHG21261fc0fb2b: perf: extract result formatting in its own function.
perf: extract result formatting in its own function
Thu, Oct 11, 4:23 AM
lothiraldan committed rHGd5d28d360a19: pref: support negative indexing in perfrevlogrevisions.
pref: support negative indexing in perfrevlogrevisions
Thu, Oct 11, 4:23 AM
lothiraldan committed rHGacf560bc9b56: perf: extract the timing of a section in a context manager.
perf: extract the timing of a section in a context manager
Thu, Oct 11, 4:23 AM
lothiraldan committed rHG25ce80ce84c5: perf: accept formatter option for perfmanifest.
perf: accept formatter option for perfmanifest
Thu, Oct 11, 4:23 AM
lothiraldan committed rHG46f9b1d2daf0: formatter: more details on assertion failure.
formatter: more details on assertion failure
Thu, Oct 11, 4:23 AM
lothiraldan committed rHG8643219146e1: perf: fix -T json.
perf: fix -T json
Thu, Oct 11, 4:23 AM

Wed, Oct 10

lothiraldan added a comment to D2679: [PoC] obsolete: config option to enable local only obsolescence mode.

Here's a case I just ran into where I would have liked to de-obsolete a commit:

  1. Someone queues a patch and pushes to central repo. Repo looks like this:
Wed, Oct 10, 4:55 PM

Fri, Oct 5

lothiraldan committed rHGcf01616f8d96: copies: add time information to the debug information.
copies: add time information to the debug information
Fri, Oct 5, 8:16 AM
lothiraldan committed rHG25b2868206e2: copies: add a devel debug mode to trace what copy tracing does.
copies: add a devel debug mode to trace what copy tracing does
Fri, Oct 5, 8:16 AM
lothiraldan committed rHG50700a025953: context: reverse conditional branch order in introrev.
context: reverse conditional branch order in introrev
Fri, Oct 5, 8:16 AM
lothiraldan committed rHG6ed53b194124: context: drop a redundant fast path in introrev.
context: drop a redundant fast path in introrev
Fri, Oct 5, 8:16 AM
lothiraldan committed rHGccf4d808ec4c: context: fast path linkrev adjustement in trivial case.
context: fast path linkrev adjustement in trivial case
Fri, Oct 5, 8:16 AM

Thu, Oct 4

lothiraldan committed rHG91f7402537a0: copies: add time information to the debug information.
copies: add time information to the debug information
Thu, Oct 4, 6:55 PM
lothiraldan committed rHG50a2f16394d4: copies: add a devel debug mode to trace what copy tracing does.
copies: add a devel debug mode to trace what copy tracing does
Thu, Oct 4, 6:54 PM
lothiraldan committed rHGee7c1352fbd6: context: reverse conditional branch order in introrev.
context: reverse conditional branch order in introrev
Thu, Oct 4, 1:30 PM
lothiraldan committed rHG865e4e2784e5: context: drop a redundant fast path in introrev.
context: drop a redundant fast path in introrev
Thu, Oct 4, 1:30 PM
lothiraldan committed rHG3858f622d1c5: context: fast path linkrev adjustement in trivial case.
context: fast path linkrev adjustement in trivial case
Thu, Oct 4, 1:30 PM
lothiraldan committed rHG208303a8172c: obsolete: explicitly track folds inside the markers.
obsolete: explicitly track folds inside the markers
Thu, Oct 4, 11:41 AM
lothiraldan committed rHG9f439959fca2: obsolete: explicitly track folds inside the markers.
obsolete: explicitly track folds inside the markers
Thu, Oct 4, 9:09 AM
lothiraldan committed rHGca9d0c93acea: cleanupnodes: update comment to drop mention of filtering.
cleanupnodes: update comment to drop mention of filtering
Thu, Oct 4, 9:09 AM
lothiraldan added a comment to D4820: logtoprocess: sends the canonical command name to the subprocess.
In D4820#73256, @yuja wrote:

+def extsetup():
+ # this is to get the canonical name of the command: "commit", not "ci"
+ def wrapdispatch(orig, *args, kwargs):
+ encoding.environ.pop("LTP_COMMAND", None)
+ return orig(*args,
kwargs)
+
+ def wrapruncommand(orig, *args, kwargs):
+ encoding.environ["LTP_COMMAND"] = args[2]
+ return orig(*args,
kwargs)
+
+ extensions.wrapfunction(dispatch,'dispatch',wrapdispatch)
+ extensions.wrapfunction(dispatch,'runcommand',wrapruncommand)

It's scary to update the environment variables globally and pass
LTP_COMMAND in to any child processes.

Can you add a proper way to teach the command name to the logtoprocess
extension, by ui.log() for example?

Also, the word LTP isn't used anywhere. It'll need a better name, and
should be documented.

Thu, Oct 4, 1:58 AM

Mon, Oct 1

lothiraldan committed rHGdfca83594145: logtoprocess: add a test to show pager and ltp bad interaction.
logtoprocess: add a test to show pager and ltp bad interaction
Mon, Oct 1, 12:29 PM
lothiraldan committed rHGc4a3d3c67c4f: logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager.
logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager
Mon, Oct 1, 12:29 PM
lothiraldan committed rHG61f39a892168: cleanupnodes: pass multiple predecessors to `createmarkers` directly.
cleanupnodes: pass multiple predecessors to `createmarkers` directly
Mon, Oct 1, 12:29 PM
lothiraldan committed rHG6335c0de80fa: obsolete: allow multiple predecessors in createmarkers.
obsolete: allow multiple predecessors in createmarkers
Mon, Oct 1, 12:29 PM
lothiraldan committed rHG99f0f052b6bf: obsolete: preindent code in createmarkers.
obsolete: preindent code in createmarkers
Mon, Oct 1, 12:29 PM
lothiraldan committed rHGa8318c9cb2ad: rebase: use tuple as `replacement` keys.
rebase: use tuple as `replacement` keys
Mon, Oct 1, 12:29 PM
lothiraldan committed rHGa8ccd9523d40: rebase: expand a long "one-liner".
rebase: expand a long "one-liner"
Mon, Oct 1, 12:29 PM
lothiraldan committed rHGbae6f1418a95: obsolete: explicitly pass relation items to effectflag computation.
obsolete: explicitly pass relation items to effectflag computation
Mon, Oct 1, 12:29 PM
lothiraldan committed rHG0428feb1f0d7: rebase: explicitly track collapses as fold.
rebase: explicitly track collapses as fold
Mon, Oct 1, 12:29 PM
lothiraldan committed rHGb153ca77a52b: histedit: don't cleanup nodes already disposed of.
histedit: don't cleanup nodes already disposed of
Mon, Oct 1, 12:28 PM
lothiraldan committed rHG7198cdbbbde1: rebase: don't try to prune obsolete changeset already in the destination.
rebase: don't try to prune obsolete changeset already in the destination
Mon, Oct 1, 12:28 PM
lothiraldan committed rHG1857f50a9643: cleanupnodes: drop special casing around prune markers (API).
cleanupnodes: drop special casing around prune markers (API)
Mon, Oct 1, 12:28 PM
lothiraldan committed rHGa89dd6d01df0: pullreport: rev duplicated and extinct into account.
pullreport: rev duplicated and extinct into account
Mon, Oct 1, 12:28 PM
lothiraldan committed rHGb5e12039e6e0: pullreport: skip or rework some early return.
pullreport: skip or rework some early return
Mon, Oct 1, 12:28 PM
lothiraldan committed rHGf9232b0310ef: pullreport: issue a message about "extinct" pulled changesets.
pullreport: issue a message about "extinct" pulled changesets
Mon, Oct 1, 12:28 PM
lothiraldan committed rHGa477679f6323: pullreport: skip filtered revs instead of obsolete ones.
pullreport: skip filtered revs instead of obsolete ones
Mon, Oct 1, 12:28 PM
lothiraldan committed rHG642cbc228bad: pullreport: add a test to show misreporting of visible changeset.
pullreport: add a test to show misreporting of visible changeset
Mon, Oct 1, 12:28 PM
lothiraldan committed rHG1c3f1491965f: scmutil: expand long "one-liner".
scmutil: expand long "one-liner"
Mon, Oct 1, 12:28 PM
lothiraldan committed rHGb99903534e06: scmutil: accept multiple predecessors in 'replacements' (API).
scmutil: accept multiple predecessors in 'replacements' (API)
Mon, Oct 1, 12:28 PM
lothiraldan committed rHGa3095bc47217: changelog: keep track of duplicated node in the transaction adding them.
changelog: keep track of duplicated node in the transaction adding them
Mon, Oct 1, 12:28 PM
lothiraldan committed rHGd9ba836fc234: shelve: find shelvedctx from bundle even if they are already in the repo.
shelve: find shelvedctx from bundle even if they are already in the repo
Mon, Oct 1, 12:28 PM
lothiraldan committed rHG42a6b228dd2e: shelve: return the shelved node as part of bundle application.
shelve: return the shelved node as part of bundle application
Mon, Oct 1, 12:28 PM
lothiraldan committed rHGdebc4ee597e7: revlog: add a callback "tracking" duplicate node addition.
revlog: add a callback "tracking" duplicate node addition
Mon, Oct 1, 12:28 PM
lothiraldan created D4820: logtoprocess: sends the canonical command name to the subprocess.
Mon, Oct 1, 12:13 PM
lothiraldan closed D4816: logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager.
Mon, Oct 1, 7:42 AM
lothiraldan committed rHG030f40677db6: logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager.
logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager
Mon, Oct 1, 7:42 AM
lothiraldan closed D4815: logtoprocess: add a test to show pager and ltp bad interaction.
Mon, Oct 1, 7:42 AM
lothiraldan committed rHG5f040397fcb6: logtoprocess: add a test to show pager and ltp bad interaction.
logtoprocess: add a test to show pager and ltp bad interaction
Mon, Oct 1, 7:42 AM
lothiraldan abandoned D1703: logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager.

Replaced by https://phab.mercurial-scm.org/D4816

Mon, Oct 1, 6:22 AM
lothiraldan abandoned D1428: logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager.

Replaced by https://phab.mercurial-scm.org/D4816

Mon, Oct 1, 6:21 AM
lothiraldan abandoned D1427: logtoprocess: add a test to show pager and ltp bad interaction.

Replaced by https://phab.mercurial-scm.org/D4815

Mon, Oct 1, 6:21 AM
lothiraldan abandoned D1702: logtoprocess: add a test to show pager and ltp bad interaction.
Mon, Oct 1, 6:20 AM
lothiraldan added a comment to D1702: logtoprocess: add a test to show pager and ltp bad interaction.

I've finally reworked this patch and sends it as an independent series here: https://phab.mercurial-scm.org/D4815

Mon, Oct 1, 6:20 AM
lothiraldan created D4816: logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager.
Mon, Oct 1, 6:18 AM
lothiraldan added a dependent revision for D4815: logtoprocess: add a test to show pager and ltp bad interaction: D4816: logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager.
Mon, Oct 1, 6:18 AM
lothiraldan created D4815: logtoprocess: add a test to show pager and ltp bad interaction.
Mon, Oct 1, 6:18 AM

Sun, Sep 30

lothiraldan committed rHGf3b65098630f: cleanupnodes: pass multiple predecessors to `createmarkers` directly.
cleanupnodes: pass multiple predecessors to `createmarkers` directly
Sun, Sep 30, 9:43 AM
lothiraldan committed rHG7cc3ef24170e: obsolete: allow multiple predecessors in createmarkers.
obsolete: allow multiple predecessors in createmarkers
Sun, Sep 30, 9:43 AM
lothiraldan committed rHG0d7465635509: obsolete: preindent code in createmarkers.
obsolete: preindent code in createmarkers
Sun, Sep 30, 9:43 AM
lothiraldan committed rHGed7414e7eea0: rebase: explicitly track collapses as fold.
rebase: explicitly track collapses as fold
Sun, Sep 30, 9:43 AM
lothiraldan committed rHGe46ec770943f: rebase: use tuple as `replacement` keys.
rebase: use tuple as `replacement` keys
Sun, Sep 30, 9:43 AM
lothiraldan committed rHGee5c63f4ca48: obsolete: explicitly pass relation items to effectflag computation.
obsolete: explicitly pass relation items to effectflag computation
Sun, Sep 30, 9:43 AM
lothiraldan committed rHG795b7fe3c6b2: cleanupnodes: drop special casing around prune markers (API).
cleanupnodes: drop special casing around prune markers (API)
Sun, Sep 30, 9:43 AM
lothiraldan committed rHG412e14e321c7: rebase: expand a long "one-liner".
rebase: expand a long "one-liner"
Sun, Sep 30, 9:43 AM
lothiraldan committed rHGe66a4c206daa: rebase: don't try to prune obsolete changeset already in the destination.
rebase: don't try to prune obsolete changeset already in the destination
Sun, Sep 30, 9:43 AM
lothiraldan committed rHG46dfaac8a3ee: histedit: don't cleanup nodes already disposed of.
histedit: don't cleanup nodes already disposed of
Sun, Sep 30, 9:43 AM
lothiraldan committed rHGdfb34b2b0b2b: obsolete: allow multiple predecessors in createmarkers.
obsolete: allow multiple predecessors in createmarkers
Sun, Sep 30, 9:16 AM
lothiraldan committed rHGfe5a0ca516ea: cleanupnodes: pass multiple predecessors to `createmarkers` directly.
cleanupnodes: pass multiple predecessors to `createmarkers` directly
Sun, Sep 30, 9:16 AM
lothiraldan committed rHGdb971139d013: obsolete: preindent code in createmarkers.
obsolete: preindent code in createmarkers
Sun, Sep 30, 9:16 AM
lothiraldan committed rHG92c24473d317: obsolete: explicitly pass relation items to effectflag computation.
obsolete: explicitly pass relation items to effectflag computation
Sun, Sep 30, 9:16 AM