<    March 2017    >
Su Mo Tu We Th Fr Sa  
          1  2  3  4  
 5  6  7  8  9 10 11  
12 13 14 15 16 17 18  
19 20 21 22 23 24 25  
26 27 28 29 30 31
04:02 glennpratt joined
05:28 plundberg joined
05:39 ta_ joined
05:51 filterfish joined
06:53 plundberg joined
07:03 filterfish joined
07:15 plundberg joined
07:45 ta_ joined
07:59 filterfish joined
11:28 GitHub7 joined
11:28 <GitHub7> [13sequel] 15iaddict opened pull request #1331: Add type conversion for MS SQL DATETIMEOFFSET (06master...06patch-1) 02https://git.io/vS3VJ
11:28 GitHub7 left
11:51 GitHub160 joined
11:51 <GitHub160> [13sequel] 15iaddict commented on issue #1331: This was not yet ready for prime time. Sorry! 02https://git.io/vS3rf
11:51 GitHub160 left
11:51 GitHub198 joined
11:51 <GitHub198> [13sequel] 15iaddict closed pull request #1331: Add type conversion for MS SQL DATETIMEOFFSET (06master...06patch-1) 02https://git.io/vS3VJ
11:51 GitHub198 left
12:45 GitHub184 joined
12:45 <GitHub184> [13sequel] 15iaddict opened issue #1332: Add type conversion for DATETIMEOFFSET 02https://git.io/vS3Xf
12:45 GitHub184 left
13:00 raz left
13:13 <Bish> each server got deprecated? i wonder why :o
13:21 <Caesium> https://github.com/jeremyevans/sequel/compare/06f91be62eef...89d73cd77be0#diff-369a4aafff78e230304b831eaaa8373dR242
13:22 <Caesium> server_block extension seems to replace it
13:24 <Bish> i see
13:25 <Bish> :O im having trueclass fields, and their accessor method returns nil
13:25 <Bish> why is that
13:27 <Bish> it returns nil, when it's false
13:27 <Bish> :o
13:28 <Bish> is that expected behaviour as usual?
13:37 <Bish> https://paste.pound-python.org/show/J0PFnUSW2BoKNf4H7Yap/
13:38 <Bish> => super || (@noncolumn_values[k] if !@values.has_key?(k) && @noncolumn_values)
13:38 <Bish> super is false.. so it tries to get the noncolumn value.. which is not there.. so nil
13:48 <Bish> that might be a bug, or isn't it?
13:51 <Rennex> yeah, seems like it
13:51 <Rennex> if @values has that key, it should return that value, now it doesn't do anything -> nil
13:51 <Bish> yay, i found a bug in sequel
13:52 <Bish> now i can die
13:52 <Rennex> and i ran into a https warning after googling what split_values is
13:52 <Bish> yes had those, too
13:53 <Rennex> jeremyevans: clearly this https issue isn't some random one-off isolated incident that you can ignore forever... :)
13:57 <Caesium> letsencrypt almost makes ssl a non-event these days
13:57 <Caesium> I throw it on everything now, its just too easy not to
13:58 <Caesium> even my munin has grown ssl recently ;)
13:58 <Rennex> Caesium: the trouble is that the domain points to github pages
13:58 <Caesium> yeah thats what I mean. get a letsencrypt one pointing at the right domain, solved :)
13:58 <Caesium> and free :)
13:59 <Rennex> as long as github doesn't support doing that, it's a no-go
13:59 <Bish> Rennex: what do i do with this bug, should i write a fix?
13:59 <Rennex> Bish: sounds like a good idea. :)
13:59 <Bish> it basicially looks like this
14:00 <Bish> res = super; return false if res == false; res||...
14:00 <Bish> is that tidy enough for sequel?
14:00 <Bish> or should i look at @values for that
14:00 <Rennex> Caesium: so there's 3 possibilities... host the content himself, change to a github domain, or set up cloudflare in front of github's hosting - which he didn't want to do because it's an extra dependency
14:01 <Rennex> Bish: i like that
14:01 <Bish> Rennex: the first one?
14:01 <Rennex> yeah, that code
14:02 <Caesium> Rennex: ah right, so github is hte issue
14:03 <Rennex> Caesium: yeah, or google for showing people https links even though they are served with a non-working certificate... but i wouldn't hold my breath for google to change that :)
14:04 <Rennex> but tbh i think it's a fair expectation that if domain:443 responds, clients could try to automatically switch to https. Doesn't "HTTPS everywhere" do something like that? unless it's really whitelist-based...
14:09 <Freaky> it uses a set of rules
14:11 <Bish> i am not using github very often.. how would that commit message look like
14:11 <Bish> >fixed a bug of Model#[] when using the split_values plugin
14:11 <Bish> should i describe the bug more?
14:13 banditron joined
14:26 GitHub31 joined
14:26 <GitHub31> [13sequel] 15bananarne opened pull request #1333: Fixed a bug of Model#[] when using the split_values plugin. (06master...06split_values_fix) 02https://git.io/vS3p6
14:26 GitHub31 left
14:26 <Bish> should've waited for #1337th pull request
14:26 <Caesium> I am totally gonna find a tiny doc bug to fix for that #1337 :)
14:29 glennpratt joined
14:30 GitHub112 joined
14:30 <GitHub112> [13sequel] 15jeremyevans closed issue #1332: Add type conversion for DATETIMEOFFSET 02https://git.io/vS3Xf
14:30 GitHub112 left
14:30 GitHub120 joined
14:30 <GitHub120> [13sequel] 15jeremyevans commented on issue #1332: There are many database specific types that the generic jdbc type handling doesn't convert to ruby types. It isn't a bug that the MSSQL-specific DATETIMEOFFSET is not converted to a ruby type. That being said, I'd certainly consider a patch that supported it, assuming it came with appropriate specs. If you would like it supported, please submit a working pull request with specs. Closing this no
14:30 GitHub120 left
14:34 GitHub197 joined
14:34 <GitHub197> [13sequel] 15jeremyevans commented on issue #1333: Thanks for the report. I agree that the current behavior has a bug in regards to handling false values. I'll merge this shortly and add a spec for it. 02https://git.io/vS3jU
14:34 GitHub197 left
14:44 GitHub19 joined
14:44 <GitHub19> [13sequel] 15jeremyevans closed pull request #1333: Fixed a bug of Model#[] when using the split_values plugin. (06master...06split_values_fix) 02https://git.io/vS3p6
14:44 GitHub19 left
14:44 GitHub27 joined
14:44 <GitHub27> [13sequel] 15jeremyevans pushed 2 new commits to 06master: 02https://git.io/vSse9
14:44 <GitHub27> 13sequel/06master 14fd020fa 15Arne: Fixed a bug of Model#[] when using the split_values plugin, which...
14:44 <GitHub27> 13sequel/06master 146226583 15Jeremy Evans: Add specs for correct false value handling in split_values plugin...
14:44 GitHub27 left
14:47 <jeremyevans> Rennex: It's a bug in GitHub (Custom Domains don't support HTTPS): https://github.com/isaacs/github/issues/156
14:48 <jeremyevans> Rennex: It's also a bug in Google for showing HTTPS pages with invalid certs when the HTTPS page has the same content
14:48 <Bish> how fast do commits end up in the gem?
14:49 <jeremyevans> Bish: Next release should be on April 1 (gem releases are around the first of the month, every month)
14:50 <Bish> great!
14:50 <Bish> i always had the problem, when patching a gem, how i could use my patches already
14:50 <Bish> do i specify my github repository as gem source? does that work?
14:51 <jeremyevans> Bish: I believe bundler supports git remotes, if you are using that
14:51 <Bish> well, if you don't you really soon will be forced to, isn't that true?
14:51 <Bish> i tried to live without.. always hurts
14:52 <jeremyevans> Bish: I think my only use for bundler is bundle install --local for preparing Gemfile.lock file for my heroku app
14:53 <jeremyevans> Bish: But I tend to have pretty good control of what dependencies are installed on my machines (many of which I wrote myself)
14:54 <Rennex> jeremyevans: i know. but only you have the power to fix their mess :)
14:55 <Rennex> (in the near term anyway)
15:00 <Bish> well, if i wrote dependencies, that would imply that i don't have control over them
15:46 AYGHOR joined
16:32 saltsa joined
17:17 plujon joined
17:42 glennpratt joined
18:26 filterfish joined
18:52 GitHub130 joined
18:52 <GitHub130> [13sequel] 15jeremyevans pushed 1 new commit to 06master: 02https://git.io/vSs9i
18:52 <GitHub130> 13sequel/06master 147379758 15Jeremy Evans: Correctly handle datasets with offsets but no limits used in compound datasets on MSSQL <2012...
18:52 GitHub130 left
19:25 mewp joined
19:27 <adam12> How does everyone handle duplicate migrations using the simple incrementing-number format? It's just me on this project and I still run into it. Could switch to timestamps but wonder if anybody has a trick before I do that.
19:28 <jeremyevans> adam12: you can always renumber the migrations. But I'm curious why you are hitting this. Different migrations in different code branches?
19:29 <adam12> Yah :| I feature branch quite a bit, especially on new projects
19:29 <adam12> I just renamed the one, and it works, but I wonder if that's prone to creating an inconsistent state?
19:30 <jeremyevans> adam12: I assume if you are feature branching, you are using separate test and development databases for each branch?
19:30 <adam12> I am not
19:30 <jeremyevans> adam12: or migrating down before switching branches and up after?
19:31 <adam12> Nope. :\
19:31 <jeremyevans> adam12: Well, I'm not sure what your workflow is, but you may want to switch to the timestamp migrations
19:32 <adam12> Yeah - I enjoy the simplicity of integer migrations, but maybe I'll switch.
20:02 Eiam joined
21:19 ta_ joined
21:26 filterfish_ joined
21:27 filterfish_ joined
21:28 d762b440__ joined
21:29 d762b440__ joined
21:30 d762b440__ joined
21:31 filterfish_ joined
21:32 filterfish_ joined
21:33 d762b440__ joined
21:34 d762b440__ joined
21:35 filterfish_ joined
21:36 filterfish_ joined
21:37 filterfish_ joined
21:37 d762b440__ joined
21:40 glennpratt joined