| gregcoit | hi. how do I undo a "bzr rm"? no other commands have been run since then (like "bzr ci"). | 00:02 |
|---|---|---|
| maxb | bzr revert pathname should do it | 00:06 |
| gregcoit | maxb: ty - that worked! | 00:06 |
| lifeless | mgz: deryck | 00:33 |
| poolie | good morning | 00:50 |
| poolie | hi lifeless | 00:50 |
| lifeless | hi poo | 00:50 |
| lifeless | hi poolie | 00:51 |
| poolie | snort | 00:52 |
| mgz | you really want those last three letters | 00:53 |
| mgz | lifeless: should I bug him on irc or subscribe him to the bug or what? | 00:54 |
| lifeless | mgz: overloaded word that | 00:55 |
| lifeless | I'd discuss with him | 00:55 |
| lifeless | and see if he wants a unique bug or what have you | 00:55 |
| mgz | thanks. | 00:55 |
| === jam changed the topic of #bzr to: Bazaar version control | try https://answers.launchpad.net/bzr for more help | http://irclogs.ubuntu.com/ | Patch pilot: poolie | bzr 2.1.1 is out | bzr 2.1.2 is having binaries built for it | Launchpad down/read-only from 23:00 - 00:30 UTC for a code update | ||
| spiv | Good morning. | 01:04 |
| lifeless | hola | 01:06 |
| === Chex changed the topic of #bzr to: Bazaar version control | try https://answers.launchpad.net/bzr for more help | http://irclogs.ubuntu.com/ | Patch pilot: poolie | bzr 2.1.1 is out | bzr 2.1.2 is having binaries built for it | ||
| poolie | hi there spiv | 02:25 |
| thumper | in the bzrlib source | 02:37 |
| thumper | is there a method to do interdiffs? | 02:37 |
| poolie | thumper: i don't think so- | 02:38 |
| lifeless | mmm | 02:38 |
| lifeless | depends on what you mean | 02:38 |
| thumper | I'd really like to not to have to shell out to interdiff binary | 02:38 |
| lifeless | ok, that, no - we haven't done that yet. | 02:38 |
| thumper | I have an old review diff, and a new review diff generated from newer commits | 02:39 |
| thumper | I want to see the difference between them | 02:39 |
| thumper | (and email it out) | 02:39 |
| lifeless | right, lp mp incremental changes ? | 02:39 |
| thumper | yep | 02:39 |
| lifeless | if I can suggest a different way | 02:39 |
| lifeless | you have the old tip O and the new tip N | 02:40 |
| lifeless | do a preview-merge with O and get a tree - not a diff | 02:40 |
| lifeless | do a preview-merge with N and get the tree. | 02:40 |
| lifeless | then do diff between the two preview trees | 02:40 |
| lifeless | I'd expect that to be more useful, and its not going to be an interdiff, so its easier to read. | 02:41 |
| thumper | are preview trees up to that yet | 02:42 |
| thumper | ? | 02:42 |
| lifeless | If they aren't it will be straight forward to fix I suspect | 02:42 |
| lifeless | if they aren't I'd really like to know anyhow :) | 02:42 |
| poolie | lifeless: re ease of reading, the output of interdiff is not a diff of diffs | 02:54 |
| poolie | iirc | 02:54 |
| poolie | it's just a diff | 02:55 |
| lifeless | ah yes | 02:55 |
| lifeless | I'm thinking debdiff | 02:55 |
| lifeless | which is interdiff | 02:55 |
| lifeless | of patches | 02:55 |
| lifeless | and known to blow up spectacularly with some packages | 02:56 |
| poolie | jam (if any), lifeless, we need a way to get debug info for out of memory errors | 02:56 |
| lifeless | poolie: nevertheless the failure modes in two previews seem better to me, IMBW. | 02:57 |
| lifeless | poolie: yes, agreed. | 02:57 |
| lifeless | hmm, -Derror should give us something | 02:58 |
| poolie | can anyone confidently answer about bzr-svn in https://answers.edge.launchpad.net/bzr/+question/116868 | 03:00 |
| poolie | there is a bug for this | 03:00 |
| poolie | may need jam jam | 03:00 |
| lifeless | jam jam ? | 03:01 |
| jam | lifeless, poolie: you rang? | 04:08 |
| lifeless | I didn't; poolie may have | 04:08 |
| jam | poolie: what sort of mem info do you want? | 04:08 |
| lifeless | jam: there is an OOM error occuring on a 30MB file according to some guy in some bug report. | 04:09 |
| poolie | jam, when bzr crashes with MemoryError i'd like us to get some useful debugging info back from the user | 04:09 |
| lifeless | be nice to be able to see whats really oging on. | 04:09 |
| poolie | there's a bug asking for this | 04:09 |
| jam | poolie: I remember chatting with you about this a bit. One of the problems is that stuff like 'str' objects aren't in the gc list | 04:12 |
| jam | strs and ints/floats/etc being the most obvious not present | 04:13 |
| jam | and now StaticTuple :) | 04:13 |
| poolie | so i guess the question is: | 04:15 |
| poolie | if there is any data we could easily report we should do it | 04:16 |
| poolie | even if it's not 100% complete or if it's useful in only a subset of cases | 04:16 |
| jam | poolie: well, you can certainlly call trace.debug_memory() | 04:16 |
| jam | which would at least tell you vmpeak and vmsize, etc | 04:17 |
| === poolie changed the topic of #bzr to: Bazaar version control | try https://answers.launchpad.net/bzr for more help | http://irclogs.ubuntu.com/ | Patch pilot: poolie | bzr 2.1.2 is out | ||
| poolie | mm, we should put that into the apport file ,if we don't already | 04:17 |
| poolie | i don't know if that will tell us _why_ we ran out though | 04:18 |
| poolie | perhaps the traceback may give some clue | 04:18 |
| poolie | can we take the question that you'd ask the user experiencing an oom problem | 04:18 |
| poolie | and automatically answer it when they first hit the bug? | 04:18 |
| jam | poolie: I'd have to think about what the first questions would be. Things I would usually ask are stuff like "what action were you taking" which should already be in the log file | 04:58 |
| jam | what the size of files are you working with | 04:58 |
| jam | I think that would be a bit hard to automatically answer | 04:58 |
| jam | knowing the sizes of objects in the frames on the stack could be useful | 05:00 |
| poolie | jam i guess we could walk up the stack and print either the actual contents of the locals, or some data about them | 05:12 |
| poolie | such as their estimated size | 05:12 |
| jam | poolie: I think contents is risky, but size could be reasonable | 05:12 |
| poolie | some risk of leaking private data there | 05:12 |
| jam | well, I'm thinking more about 100MB files | 05:13 |
| poolie | that too | 05:14 |
| poolie | there is a cgitb module that does some of this | 05:14 |
| poolie | including trimming large values | 05:15 |
| poolie | so there's a function (in meliae?) that estimates the size of an object? | 05:15 |
| jam | poolie: in python 2.6 you can use sys.getsizeof() | 05:30 |
| jam | meliae also exposes something like this | 05:30 |
| poolie | ok | 05:30 |
| poolie | so this won't tell us if we're using lots of memory in a global not referenced by the stack | 05:31 |
| poolie | and it would probably be too noisy to print all the globals | 05:31 |
| poolie | but that may be an unusual condition | 05:31 |
| poolie | i guess we could print globals using a lot of memory | 05:31 |
| jam | we could walk gc.get_referents() at a cost of 1 pointer per object | 05:34 |
| poolie | cost in terms of needing to allocate more memory to do that? | 05:37 |
| jam | yeah | 05:53 |
| jam | for the list that holds it | 05:53 |
| jam | it could be a problem if we are already oom | 05:54 |
| lifeless | jam: for when you awake. | 06:21 |
| lifeless | https://bugs.edge.launchpad.net/bzr-builddeb/+bug/552950 | 06:22 |
| ubot5 | Launchpad bug 552950 in bzr-builddeb "Additional changelog entries are modified when merging another branch (affected: 1, heat: 10)" [Medium,Triaged] | 06:22 |
| lifeless | please comment | 06:22 |
| yao | if I commit 10 times, from rev 1 to rev 10, in my local branch, and didn't push them to remote branch yet, | 06:57 |
| yao | I want to uncommit commit #5, what should I do? | 06:57 |
| yao | 'bzr uncommit' doesn't help, | 06:57 |
| poolie | yao: it depends on exactly what you want | 07:05 |
| poolie | do you want to pretend #5 never happened, or do you want to just reverse it's effect? | 07:05 |
| yao | poolie: former, I think, | 07:06 |
| * yao is wondering whether it is possible to pretend #5 never happened | 07:06 | |
| aidalgol | Google is turning up nothing on this: is it safe to edit .bzrignore? | 07:45 |
| spiv | Yep. | 07:49 |
| aidalgol | OK, good. | 07:52 |
| aidalgol | Why might this line (in a .bzrignore) have *two* stars? | 07:53 |
| aidalgol | lisp/**/loaddefs.el | 07:53 |
| spiv | aidalgol: see 'bzr help ignore', it explains the patterns | 08:00 |
| spiv | (Also, 'bzr help patterns') | 08:00 |
| spiv | To answer your specific query: "/**/ - Matches 0 or more directories in a path" | 08:01 |
| poolie | hi there spiv, vila | 08:59 |
| vila | hey poolie | 08:59 |
| poolie | sorry yao i missed you there | 09:00 |
| poolie | are you still here? | 09:00 |
| poolie | have a look at the bzr- | 09:00 |
| poolie | have a look at the bzr-rewrite plugin | 09:00 |
| yao | poolie: I am here... | 09:00 |
| poolie | do you want to keep the later commits as distinct commits? | 09:01 |
| poolie | or are you happy to fold them into just one? | 09:01 |
| yao | poolie: I want to keep later commits as distinct ones... | 09:03 |
| * yao is looking at https://launchpad.net/bzr-rewrite | 09:03 | |
| poolie | try bzr help rebase | 09:05 |
| poolie | we should document that specific case | 09:22 |
| yoroy | How do I find where bzr is intalled on OS X? Trying to get the Eclipse plugin working. | 10:36 |
| C-Keen | yoroy: which bzr will tell you on a Terminal | 10:45 |
| yoroy | C-Keen: forgot to mention I don't want to use a Terminal | 10:46 |
| yoroy | :) | 10:46 |
| yoroy | but I could handle it if I knew the command | 10:46 |
| C-Keen | I don't understand | 10:46 |
| GaryvdM | mgz: Ping | 10:47 |
| yoroy | C-Keen: the eclips plugin asks me for the path to the bzr executable. I don't know how to find that path | 10:47 |
| C-Keen | yoroy: well type 'which bzr' on a Terminal and you know | 10:47 |
| yoroy | C-Keen: thank you | 10:48 |
| mgz | pong garyvdm, but I'm leaving in five minutes, hopefully be around a bit later though | 10:52 |
| GaryvdM | Hi mgz | 10:52 |
| GaryvdM | The .pyo in the installer were fine | 10:52 |
| GaryvdM | for plugins | 10:53 |
| GaryvdM | but the .pyo that were created for plugins that I had in BZR_PLUGIN_PATH got doc strings removed | 10:53 |
| mgz | did you see the log? | 10:54 |
| GaryvdM | So if we could get the files in library.zip to be -OO | 10:54 |
| mgz | you've not got the final part of my don't-recompile-plugin-pyo-files fix | 10:54 |
| GaryvdM | but leave the exe as -O | 10:54 |
| GaryvdM | Oh - were is that? | 10:55 |
| mgz | you need the recent change to bzr setup.py and/or the third rev in the bzr-windows-installer merge proposal | 10:55 |
| mgz | <GaryvdM> but leave the exe as -O <- if one of us can work out how to do this, it would also be good. | 10:55 |
| GaryvdM | Ok - I find that an try a build... | 10:55 |
| mgz | I saw the bug with recompilation and stopped investigating there, but getting the exe as -O would be good. | 10:56 |
| GaryvdM | mgz: maybe build the library.zip ourselfs | 10:56 |
| GaryvdM | Ok cool. | 10:57 |
| mgz | there's probably a neater way, just didn't get as far as looking for it. | 10:57 |
| GaryvdM | Thanks | 10:57 |
| mgz | okay, leaving now. | 10:58 |
| bobslee | hi, does bzr explorer install without a glitch on mac osx? | 14:02 |
| GaryvdM | bobslee: The installer works well | 14:12 |
| GaryvdM | bobslee: There are some polish issues on mac though - like the app name in the menu shows as "python". | 14:15 |
| GaryvdM | bobslee: I'm also not sure if it installs a app icon, so either run "bzr explorer" from a terminal, or create a shortcut icon. | 14:17 |
| bobslee | GaryvdM: ok thanx for info. But it requires a bunch of dependencies Qt, QBzr.. takes a while to install? | 14:24 |
| GaryvdM | bobslee: Maybe you are reading a old info. The latest installers have every thing you need. | 14:26 |
| GaryvdM | bobslee: http://wiki.bazaar.canonical.com/MacOSXBundle/SnowLeopard | 14:27 |
| GaryvdM | bobslee: What version of Mac OS do you have | 14:27 |
| bobslee | 10.5 (leopard) | 14:28 |
| GaryvdM | bobslee: The 10.6 installer has everything, but for the 10.5 installer, you have to install Qt | 14:28 |
| bobslee | GaryvdM: allrighty, thanx I'll give it a try | 14:29 |
| GaryvdM | bobslee: http://wiki.bazaar.canonical.com/MacOSXBundle | 14:29 |
| bobslee | GaryvdM: do you whether bzr supports reverse merging (like i.e. in svn: merge -r1000:993) ? | 14:31 |
| GaryvdM | bobslee: yes | 14:31 |
| bobslee | GaryvdM: I prefer to install bzr (explorer) via macports. Is that a good choice? - because I want to prevent dependency/upgrade nightmares.. | 14:34 |
| GaryvdM | bobslee: my experience no, but I don't use a mac much, so I would say investigate yourself. | 14:35 |
| GaryvdM | bobslee: check if they have the latest versions... | 14:35 |
| bobslee | GaryvdM: thanx again | 14:56 |
| detritux | hi all. I've been trying to push my local modification to my launchpad project for a while but never succeeded: Transport operation not possible: readonly transport | 15:43 |
| detritux | I added my ssh keys to launchpad and logged into launchpad from my bazaar explorer as described in one of the tuto | 15:43 |
| detritux | any ideas? | 15:43 |
| GaryvdM | detritux: Have you done 'bzr launchpad-login'? | 15:51 |
| detritux | yes I did that in the bazaar explorer | 15:51 |
| detritux | if I run it without any arguments it prints my login so I assume it worked | 15:51 |
| GaryvdM | detritux: What is the url that you are pusing to. | 15:52 |
| detritux | bzr+ssh://bazaar.launchpad.net/~ugocupcic/sr-ros-interface/trunk/ | 15:53 |
| GaryvdM | detritux: That's odd. | 15:53 |
| detritux | :S yeah, I only found tutorials / forums and it seemed quite straightforward each time | 15:54 |
| GaryvdM | detritux: could you please pastbin the relevant entry from ~/.bzr.log | 15:54 |
| detritux | 0.173 bazaar version: 2.1.1 | 15:57 |
| detritux | 0.173 bzr arguments: [u'commit', u'-m', u'added dependency to robot_state_publisher for easier build.', u'shadow_robot/sr_hand/manifest.xml'] | 15:57 |
| detritux | 0.175 encoding stdout as osutils.get_user_encoding() 'UTF-8' | 15:57 |
| detritux | 0.210 ssh implementation is OpenSSH | 15:57 |
| detritux | 8.869 opening working tree '/home/genugo/Projects/st-ros-interface' | 15:57 |
| detritux | 8.934 Traceback (most recent call last): | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/commands.py", line 853, in exception_to_return_code | 15:57 |
| detritux | return the_callable(*args, **kwargs) | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/commands.py", line 1055, in run_bzr | 15:57 |
| detritux | ret = run(*run_argv) | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/commands.py", line 661, in run_argv_aliases | 15:57 |
| detritux | return self.run_direct(**all_cmd_args) | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/commands.py", line 665, in run_direct | 15:57 |
| detritux | return self._operation.run_simple(*args, **kwargs) | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/cleanup.py", line 122, in run_simple | 15:57 |
| detritux | self.cleanups, self.func, *args, **kwargs) | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/cleanup.py", line 156, in _do_with_cleanups | 15:57 |
| detritux | result = func(*args, **kwargs) | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/plugins/qbzr/lib/commands.py", line 788, in run | 15:57 |
| detritux | return run_subprocess_command(cmd, bencoded) | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/plugins/qbzr/lib/subprocess.py", line 789, in run_subprocess_command | 15:57 |
| detritux | return commands.run_bzr(argv) | 15:57 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/commands.py", line 1055, in run_bzr | 15:58 |
| detritux | ret = run(*run_argv) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/commands.py", line 661, in run_argv_aliases | 15:58 |
| detritux | return self.run_direct(**all_cmd_args) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/commands.py", line 665, in run_direct | 15:58 |
| detritux | return self._operation.run_simple(*args, **kwargs) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/cleanup.py", line 122, in run_simple | 15:58 |
| detritux | self.cleanups, self.func, *args, **kwargs) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/cleanup.py", line 156, in _do_with_cleanups | 15:58 |
| detritux | result = func(*args, **kwargs) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/builtins.py", line 3138, in run | 15:58 |
| detritux | exclude=safe_relpath_files(tree, exclude)) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/decorators.py", line 192, in write_locked | 15:58 |
| detritux | self.lock_write() | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/workingtree_4.py", line 618, in lock_write | 15:58 |
| detritux | self.branch.lock_write() | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/remote.py", line 2377, in lock_write | 15:58 |
| detritux | remote_tokens = self._remote_lock_write(token) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/remote.py", line 2367, in _remote_lock_write | 15:58 |
| detritux | repo_token or '', **err_context) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/remote.py", line 52, in _call | 15:58 |
| detritux | return self._client.call(method, *args) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/smart/client.py", line 132, in call | 15:58 |
| detritux | result, protocol = self.call_expecting_body(method, *args) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/smart/client.py", line 145, in call_expecting_body | 15:58 |
| GaryvdM | detritux: Next time, please use something like http://pastebin.org/ when you paste :-) | 15:58 |
| detritux | method, args, expect_response_body=True) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/smart/client.py", line 81, in _call_and_read_response | 15:58 |
| detritux | expect_body=expect_response_body), | 15:58 |
| mwhudson | detritux: paaaaaaaaaaaste bin | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/smart/message.py", line 306, in read_response_tuple | 15:58 |
| detritux | _translate_error(self.args) | 15:58 |
| detritux | File "/usr/lib/python2.6/dist-packages/bzrlib/smart/message.py", line 355, in _translate_error | 15:59 |
| detritux | raise errors.LockFailed(*error_args[:2]) | 15:59 |
| detritux | LockFailed: Cannot lock LockDir(lp-70455952:///~ugocupcic/sr-ros-interface/trunk/.bzr/branchlock): Transport operation not possible: readonly transport | 15:59 |
| detritux | 8.934 Transferred: 1KiB (0.1K/s r:1K w:1K) | 15:59 |
| detritux | 8.934 return code 3 | 15:59 |
| detritux | 411.969 Traceback (most recent call last): | 15:59 |
| detritux | File "_dirstate_helpers_pyx.pyx", line 1852, in bzrlib._dirstate_helpers_pyx._loop_one_block | 15:59 |
| detritux | StopIteration | 15:59 |
| iocor | oh my lord | 15:59 |
| detritux | 417.318 return code 1 | 15:59 |
| detritux | a bit long, sorry :S | 15:59 |
| detritux | sorry guys, I'll use pastebin next time | 15:59 |
| mgz | is your launchpad username *not* ugocupcic? | 16:00 |
| mgz | because you can't push to other people's branches. | 16:00 |
| GaryvdM | mgz: /whois detritux = "Ugo Cupcic" | 16:01 |
| GaryvdM | Seems like that is his user :-) | 16:01 |
| detritux | yeah | 16:02 |
| mgz | okay, that's me out. anyone else? | 16:02 |
| GaryvdM | detritux: I just want to make sure about something. Please will you pastebin 'bzr info'. | 16:06 |
| mgz | on his local branch rather than the lp one? | 16:06 |
| GaryvdM | Yes | 16:06 |
| GaryvdM | Ah - maybe bug 412657 | 16:07 |
| ubot5 | Error: Could not parse data returned by Launchpad: The read operation timed out (https://launchpad.net/bugs/412657) | 16:07 |
| detritux | http://pastebin.org/385322 | 16:07 |
| GaryvdM | https://bugs.edge.launchpad.net/bzr/+bug/412657 | 16:07 |
| ubot5 | Launchpad bug 412657 in Bazaar "update fails when trying to lock master branch (in a readonly checkout) (dup-of: 412223)" [Undecided,New] | 16:07 |
| ubot5 | Launchpad bug 412223 in Bazaar "bzr up locks master branch (affected: 2, heat: 4)" [High,Confirmed] | 16:07 |
| mgz | that bug doesn't seem to have anything for cause or workaround... | 16:08 |
| detritux | hm unbind / rebind I can try that | 16:09 |
| mgz | oh, wait, bind/unbind/redo | 16:09 |
| detritux | how do you unbind ? | 16:09 |
| mgz | bzr unbind && bzr bind | 16:09 |
| detritux | :D | 16:10 |
| mgz | if that fixes it for you I'd wham an "affects me to" on that bug. | 16:11 |
| GaryvdM | mgz: Can we chat re installer -OO? | 16:11 |
| GaryvdM | time ok? | 16:12 |
| mgz | yup, I'm back and lunched now | 16:12 |
| GaryvdM | Cool - so I was building with all your patches | 16:12 |
| mgz | ^or rather, affectsmetoo on the bug that one is duped against I guess | 16:13 |
| * GaryvdM get branch urls to prove | 16:13 | |
| detritux | yeah I'll do that if it fixes it | 16:13 |
| mgz | and... stuff still breaks? | 16:13 |
| GaryvdM | Yes. | 16:13 |
| mgz | meh, my timestamp checking lines are scrolled out of buffer | 16:14 |
| mgz | I think I pasted them in here though... | 16:14 |
| GaryvdM | mgz: The issues was that a plugin that was allready installed on disk from source in a manually set BZR_PLUGIN_PATH got compiled with -OO when bzr.exe ran. | 16:15 |
| mgz | basically, need see if the mtime of the py file which complains is the same as the stamp in the packaged pyo file | 16:15 |
| mgz | oh, that's possibly expected | 16:16 |
| mgz | if it was never actually installed with optimize | 16:16 |
| GaryvdM | mgz: when I did "set BZR_PLUGIN_PATH=" then it worked... | 16:16 |
| mgz | one of the things I wanted to find out with that patch is if it's common to use from-source plugins with bundled installers | 16:16 |
| mgz | which strikes me as dodgy compared to installing into the right plugin dir, but if non-devs actually do it... | 16:17 |
| GaryvdM | mgz: So the only way that I see -OO viable is if we can get it so that library.zip is compiled as -OO but the .exe is not . | 16:18 |
| mgz | if people actually do that, I agree (for the moment, a future way is just to mark up plugin cmd objects as well) | 16:19 |
| GaryvdM | mgz: Your patches to prevent pyo recompile are still valuable though... | 16:19 |
| mgz | I'll poke around in py2exe | 16:19 |
| GaryvdM | mgz: How much difference does -OO make compared to -O? | 16:20 |
| GaryvdM | mgz: I think it would be good to try measure. | 16:20 |
| mgz | not much, a couple of megs to library.zip and a fractional cold startup improvement | 16:20 |
| mgz | Alexander did a couple of tests at UDS for me when we were merging it | 16:21 |
| mgz | https://code.launchpad.net/~gz/bzr/support_OO_flag_installer/+merge/24484 | 16:22 |
| GaryvdM | mgz: the branches that I built with: lp:~garyvdm/bzr-windows-installers/maybe and lp:~garyvdm/bzr/2.2b3+packing | 16:23 |
| GaryvdM | The latter now has the org -OO patch reverted. | 16:23 |
| mgz | dod you know if there'll be a 2.2b4 from the current trunk? | 16:23 |
| GaryvdM | Bla - lp is sloooow at the moment. | 16:25 |
| GaryvdM | mgz: according to https://edge.launchpad.net/bzr/2.2 - yes | 16:25 |
| mgz | is that lp's fault or your undersea cable's? | 16:26 |
| GaryvdM | Oh yeah - I forgot about that. | 16:26 |
| mgz | okay, reading py2exe source, doesn't seem to be a configurable way of doing it, but could easily chop it in | 16:34 |
| GaryvdM | mgz: 7.7% faster cold start - Ok - that makes it worth it ... | 16:39 |
| GaryvdM | mgz: We could also maybe look at which plugins we can do a -OO on, and which not... | 16:40 |
| mgz | well, it's not a big job to just fix them all | 16:40 |
| mgz | but messing with py2exe looks amusing | 16:41 |
| GaryvdM | mgz: Would it work if we ran byte_compile(optimize=2) for everything going it to library.zip, and then did py2exe with optimize=1 . | 16:46 |
| GaryvdM | mgz: or does it compile from scratch? | 16:47 |
| mgz | nope. | 16:47 |
| mgz | sec, I'll give you a diff to try | 16:47 |
| mgz | http://paste.ubuntu.com/460272/ < how about that for a horrible hack | 16:52 |
| GaryvdM | Not to bad. | 16:54 |
| detritux | GaryvdM: ah I found my real problem in fact. You can't commit on a branch that's being pulled from a svn. That's logical... thanks for your help | 16:54 |
| GaryvdM | mgz: Suggestion: In build_executable, remember what optimize was, and then restore it. | 16:56 |
| mgz | I started writing that, then I looked at the diff context and decided it was silly | 16:56 |
| mgz | really that's a sensible setting we want to keep, so would be best to get a proper version upstream | 16:57 |
| GaryvdM | mgz: +1 for upstream. | 16:57 |
| mgz | ...last release was 2008 | 16:58 |
| mgz | I've got theller to look at patches before though so can probably get it in svn at least | 17:00 |
| GaryvdM | Very useful: http://pastebin.org/385361 | 17:01 |
| mgz | yeah, methods are bound late, get a different object back each time | 17:03 |
| * GaryvdM goes to find food. | 17:03 | |
| lifeless | vila: hi | 19:15 |
| lifeless | vila: Is it true, a fix for the locking thing is in 2.1 branch? if so, we should get it into the LP tree | 19:16 |
| elmo | blah, what's the bzr push trick to push to where I just branched from? | 20:02 |
| elmo | I thought it was bzr push --parent, but it's not that | 20:02 |
| james_w | push :parent | 20:04 |
| elmo | james_w: thanks | 20:04 |
| lifeless | james_w: hi | 20:30 |
| james_w | hi lifeless | 20:30 |
| lifeless | I've pushed up a test | 20:31 |
| lifeless | I'm going to try and get some more love in place today and tomorrow | 20:31 |
| lifeless | I can has review? [or is it buried in my email overnight?) | 20:31 |
| james_w | lifeless: I'm in the middle of looking at it | 20:31 |
| james_w | it looks broadly fine | 20:31 |
| lifeless | \o/ | 20:32 |
| lifeless | thanks | 20:32 |
| lifeless | I realise the test style is a fresh one | 20:32 |
| lifeless | its not quite -right- but its broadly where I want things to be | 20:32 |
| lifeless | I think its fairly readable ? | 20:32 |
| james_w | I don't think you need to use dh-make, but I can understand that working out the needed steps might be a bit much | 20:32 |
| lifeless | james_w: if you want to give me a recipe to slot into that fixture, great. | 20:32 |
| james_w | yeah, it's fairly readable | 20:33 |
| lifeless | I was fighting a bit of an uphill battle at the time ;) | 20:33 |
| james_w | it's not what I would expect fixtures to be used for really | 20:33 |
| lifeless | thats interesting | 20:33 |
| james_w | well, at least say the FileMovedReplacedUpstream part | 20:33 |
| james_w | I would expect to have something like BranchBuilder, where you call methods, rather than compose fixtures | 20:34 |
| james_w | I'm happy to see how this develops though | 20:34 |
| lifeless | great | 20:37 |
| lifeless | I'll use these as a basis for the next patch too then | 20:37 |
| lifeless | its why I wanted to get this one seen to rapidly, to avoid a big pile of 'nooo' from you :) | 20:37 |
| james_w | heh | 20:43 |
| james_w | I just think that the method calls would make reading the intent of the test easier, but let's find out | 20:44 |
| lifeless | it would be nice | 20:44 |
| lifeless | (perhaps) | 20:44 |
| lifeless | to do @with_fixture(fixture_description_here) | 20:45 |
| lifeless | thats kindof the philosophic reason for making it all declarable rather than procedural | 20:45 |
| lifeless | one of the .. reasons | 20:45 |
| lifeless | another is to be able to switch out implementations very easily, a BranchBuilder is more monolithic, so you would switch out all at once rather than granularly. | 20:46 |
| === davidstrauss_ is now known as davidstrauss | ||
| davidstrauss | How can I make Loggerhead only bind to localhost? | 22:46 |
| davidstrauss | nvm | 22:47 |
| === oubiwann is now known as oubiwann-away | ||
| === oubiwann-away is now known as oubiwann | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!