[15:06] <ovnicraft_> hello i want to debug bzr, so i get this error https://gist.github.com/ovnicraft/5717722, i read about it and it was caused by delete a file and i want to debug to identify what file_id
[15:06] <ovnicraft_> i activate BZR_PDB
[15:06] <ovnicraft_> i get postmortem but i don know what var use
[15:09] <ovnicraft_> i found this https://bugs.launchpad.net/bzr/+bug/1179527
[15:09] <ovnicraft_> there are many old bugs
[15:10] <ovnicraft_> and this http://permalink.gmane.org/gmane.comp.version-control.bazaar-ng.general/74886
[15:10] <ovnicraft_> looks like unversioned files cause the problem
[15:11] <ovnicraft_> i this post http://old.nabble.com/Error-during-merge%3A-No-final-name-for-trans_id-'new-70'-td34444552.html he use pdb to identify trans_id refer
[15:11] <ovnicraft_> but dont describe how to use pdb in bzr postmortem
[15:13] <jelmer> ovnicraft_: set BZR_PDB=1 in the environment
[15:13] <jelmer> and you'll be dropped into pdb when an error occurs
[15:13] <ovnicraft_> jelmer: yes i got pdb
[15:13] <ovnicraft_> jelmer: so i want to know how to identify trans_id
[15:13] <ovnicraft_> i dont know bzr code
[15:14] <ovnicraft_> jelmer: i am here https://gist.github.com/ovnicraft/5722285
[15:14] <jelmer> ah, that's a more complex question - I don't have time to help with that atm unfortunately
[15:15] <ovnicraft_> so first Q: i am in right channel ?
[15:15] <jelmer> yes
[15:18] <ovnicraft_> jelmer: there is any dirty solution ?
[15:20] <mgz> ovnicraft_: broadly speaking, you want to use 'u' and 'l' till you find something pathy
[15:21] <mgz> I'm not sure if this is the shortest path (ho ho) to finding out what the issue with your tree state though
[15:23] <mgz> running a diff between the branches prior to merge (if that's what's failing? you weren't explict) may be easier
[15:32] <ovnicraft_> mgz: i am running the diff now
[15:37] <barry> hey folks, perhaps mgz, jam, or abentley: have any of you seen crashes in bzr on saucy, e.g. doing `bzr info`?  i think the svn plugin may be broken as it's trying to import CommittedQueue from subvertpy.wc
[15:37] <jelmer> barry: yes, there is a bug
[15:37] <jelmer> barry: bzr-svn has been removed from saucy
[15:38] <jelmer> pad.lv/1187840
[15:38] <barry> jelmer: thanks, i shall purge it from my existence then :)
[15:39] <barry> jelmer: got it, thanks
[15:39] <mgz> hm, how hard is that sru with the dep to do I wonder...
[21:10] <thumper> morning
[21:16] <jelmer> hey thumper, how's life in juju land?
[21:17] <thumper> hi jelmer
[21:17] <thumper> interesting
[21:17] <jelmer> mgz: it should be a one line change
[21:17] <thumper> just like python, there is the "go way" of doing things
[21:17] <thumper> which often rubs with me ATM
[21:18] <jelmer> thumper: ah, you're working on Go juju ? is that version 2.0?
[21:19] <thumper> jelmer: well, 1.11.x right now, but yes, heading to 2.0
[21:19] <thumper> I wanted to get back into a dev role, and also learn go
[21:19] <thumper> I hadn't learnt a new language in a while
[21:21] <jelmer> thumper: what do you think of go, any improvement over C++?
[21:21] <thumper> jelmer: well, that depends...
[21:22] <thumper> C++ is horribly complicated, but powerful in the right hands
[21:22] <thumper> Go is much simpler, but is missing a lot of what I'm used to
[21:22] <thumper> like generics
[21:22] <thumper> I'd say 90% of C++ devs write bad code
[21:22] <thumper> just due to language complexity
[21:22] <thumper> go reduces that number somewhat
[21:23] <thumper> I've not done any real benchmark tests
[21:23] <thumper> but go does provide some nice library stuff that C++ is missing
[21:23] <thumper> around network and general process control
[21:23] <thumper> I guess it depends a lot on what you are writing :)
[21:26] <jelmer> yeah, the complexity in C++ is my main gripe with it
[21:27] <jelmer> I'm not sure what to think of go yet. The exception handling/return value stuff seems kind of arbitrary but some of the other features, like - as you mention - the libraries and goroutines are really nice
[21:32]  * thumper nods
[21:32] <thumper> I love the power of C++
[21:32] <thumper> but it is too easy for people to get things wrong
[21:32] <thumper> I have heard that D is better
[21:32] <thumper> but never used it
[21:43] <LeoNerd> goros are just CSP. "yaaay". That's been around since like, 1970
[21:43] <LeoNerd> mvreturn is cute, but eh.. Perl and most Lisps have those
[21:44] <LeoNerd> Python too come to think of it
[21:44] <LeoNerd> Sortof ish