=== Logan_ is now known as Guest76402 | ||
AfC | I just did my first Pull Request on GitHub. | 08:23 |
---|---|---|
AfC | I feel dirty. | 08:23 |
fullermd | Did you wear gloves? | 08:53 |
gmarkall | sometimes when i try to rebase, bzr identifies almost an entire file being in conflict when only a small portion has changed - does that sound like I'm doing something wrong? | 08:56 |
AfC | In bzr land, using rebase is doing it wrong, but anyway. | 09:16 |
gmarkall | i've just noted that if i try a merge instead, it seems to pick out the lines that have actually changed | 09:16 |
gmarkall | AfC: can I quote you on that? | 09:17 |
gmarkall | (i'm not a fan of rebasing with bzr at all! :-) ) | 09:17 |
=== yofel_ is now known as yofel | ||
=== mmrazik is now known as mmrazik|lunch | ||
=== mmrazik|lunch is now known as mmrazik | ||
AfC | gmarkall: sure | 11:22 |
AfC | gmarkall: Bazaar takes a much stronger view of merging, and doesn't encourage people to throw history away. Some Git communities think that merging (by anyone other than the maintainer) is bad, so rebase. | 11:23 |
AfC | gmarkall: my impression is that GitHub has effectively made it such that people don't have to think about this anymore, but we'll see | 11:23 |
LeoNerd | I hardly ever rebase in bzr | 11:25 |
LeoNerd | The only time I do it is when I accidentally diverged history on a branch, say, between my laptop and my server by not being up-to-date first before committing locally on the laptop | 11:26 |
mgz | I cherrypick specific changes from time to time, which is pretty much the same thing | 11:26 |
mgz | but easier to do safely as you have two branches at all points | 11:27 |
gmarkall | AfC: thanks. when you say "people don't have to think about this anymore", are you referring to the fact that you can just click a button to merge a pull request as long as its rebased on the branch being merged into? | 11:28 |
gmarkall | or do you mean that people are rebasing without thinking? | 11:28 |
AfC | gmarkall: no, that it's provinding better UI around commit selection and selectively viewing history | 11:28 |
AfC | gmarkall: meaning you don't have to rebase | 11:29 |
gmarkall | ah, right | 11:29 |
gmarkall | so anyway, if a project is using bzr and wants clean history (hence the rebasing), it's acceptable to say "you're not using a good strategy, could you please start using merging instead of rebasing?" | 11:35 |
gmarkall | since i'm relatively inexperienced with bzr, i'd be nervous about telling people how to use it without being sure of what i'm saying | 11:35 |
LarstiQ_ | gmarkall: hmmm, telling others how to work is perhaps not the way to go | 11:49 |
fullermd | Not? Well, what's the fun in THAT? | 11:49 |
* LarstiQ_ stresses the perhaps | 11:49 | |
gmarkall | it does seem rather brazen | 11:50 |
* LarstiQ_ isn't sure what to suggest | 11:51 | |
LarstiQ_ | gmarkall: how attached are they to this workflow? | 11:52 |
LarstiQ_ | gmarkall: rebasing can be useful, but there are reasons not to do it all the time. Perhaps dig up a post by Linus where he complains about this. | 11:53 |
LarstiQ_ | gmarkall: I guess I don't understand this desire for a "clean history" over real history | 11:53 |
* LarstiQ_ back to math | 11:54 | |
fullermd | Hey, you eliminate terms in math. That's just like making a clean history ;p | 11:55 |
gokr | I know its a bit silly to enter and ask this question but I do have quite a bit of experience of SCMs but can't really find a proper "modern" comparison of bzr and hg. If anyone can share their view on why bzr is preferrable I am interested. | 12:08 |
gokr | I have used git, SVN, CVS, darcs and even hg a bit a few years back. Only dabbled with bzr but now I am contemplating giving it a more serious try. | 12:09 |
=== mmrazik is now known as mmrazik|otp | ||
=== dpb_ is now known as Guest46508 | ||
=== lifeless_ is now known as lifeless | ||
=== BasicOSX is now known as 17WAARE39 | ||
=== 17WAARE39 is now known as BasicOSX | ||
=== spiv_ is now known as spiv | ||
wilx | Hi. | 23:05 |
wilx | What does 'bzr rmbranch' do exactly? | 23:05 |
wilx | What if I just do 'rm -rf foo' on the branch checkout and directory inside my shared tree? | 23:06 |
wilx | What is the difference? | 23:06 |
jelmer | wilx: rmbranch removes the branch, not the tree or the control directory | 23:08 |
wilx | Hmm, like in the shared repository's metadata? | 23:09 |
jelmer | wilx: in older style bzr branches it basically removes .bzr/branch | 23:10 |
wilx | Or to put it another way around: Can I safely do 'rm -rf foo work-foo' to remove the temporary foo branch from my shared repo? | 23:10 |
wilx | Ah. | 23:10 |
jelmer | wilx: yes | 23:10 |
wilx | Cool, thank you. | 23:10 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!