<namelessjon>
mbj: Looks good to me. Though, you refer to a LICENSE file, but don't provide one
<mbj>
namelessjon: thx fixed!
<dkubb>
mbj: hehe, that's not a bad problem to have
<cored>
dkubb: oh you are around
zaidan has quit [Ping timeout: 276 seconds]
<dkubb>
cored: yeah I haven't left yet
<dkubb>
cored: how is the refactoring going? I checked it yesterday and it looked really good. I suspect those specs for Attribute itself will need a stand-in example Attribute with .type ?
knowtheo1y has joined #rom-rb
knowtheory has quit [Ping timeout: 252 seconds]
djsell has joined #rom-rb
dkubb has quit [Quit: Leaving...]
djsell has quit [Quit: Leaving.]
postmodern has joined #rom-rb
<mbj>
postmodern: hola
<mbj>
postmodern: Time to checkout new mutant?
zaidan has joined #rom-rb
<postmodern>
mbj, not yet
<postmodern>
mbj, let me install it
brandon7 has quit [Quit: Leaving.]
brandon7 has joined #rom-rb
knowtheory has joined #rom-rb
knowtheo1y has quit [Ping timeout: 246 seconds]
brandon71 has joined #rom-rb
brandon7 has quit [Ping timeout: 252 seconds]
<cored>
shits, dkubb went away
brandon71 has quit [Quit: Leaving.]
djsell has joined #rom-rb
<cored>
I thought that if I use Virtus
<cored>
and add something like attribute :name, String
<cored>
and after instantiate it that object and add something like o.name = 1
<cored>
that will throw an exception
kleech has joined #rom-rb
kleech has quit [Remote host closed the connection]
travis-ci has joined #rom-rb
<travis-ci>
[travis-ci] mbj/mutant#414 (master - 3666139 : Markus Schirp): The build was broken.
<zaidan>
mbj: you mean merging config.yml + generated content + Gemfile with rom-rb/rom-rb.org
<mbj>
zaidan: Jo, rom-rb/rom-rb.org is build with github-pages (a service by github)
<mbj>
It supports static content only, but we dont have a really need for dynamic content.
<zaidan>
mbj: should we rename rom-status, giving it a more general name?
<mbj>
zaidan: Ideally we remove rom-status and merge it with rom-rb.rg
<zaidan>
mbj: yeah i am familiar with gh-pages
<mbj>
zaidan: I think they support the generation of content "once". So we still could use that stuff behind config.yml
<mbj>
But I'm not sure, never used gh-pages for myself.
<zaidan>
mbj: you mean removing the "old" rom-status or the new one? we do not need to remove the new one, because it became an indipendent tool.
<mbj>
zaidan: Sure we can still create a new software that ships rom-status features.
<mbj>
But for rom-rb I think a static status side is enough.
<mbj>
Or we add some features, such as twitter tracking, open issue count etc, something we need serverside logic fore.
<mbj>
But all that stuff I said is also possible via javascript.
<xybre>
What prompted the name change?
<mbj>
xybre: The need for a fresh namespace.
<mbj>
xybre: We had DataMapper::Repository etc in dm-mapper at this time.
<mbj>
Nobody could have run dm-1 and dm-2 together.
<mbj>
Or we'd have to workaround with stupid names.
<xybre>
That seems a simple enough fix. lib/datamapper.rb and lib/datamapper2.rb, where 1 sets the namespace to "Datamapper" and 2 to "Datamapper2" or even make it configurable.
<mbj>
xybre: tbh, no.
<mbj>
I'd dislike to have a single char difference between that two namespaces
<mbj>
And to make it configurable etc.
<mbj>
Also the rom project is not related to the datamapper-1 sources in any way.
<mbj>
It is a compleate rewrite.
<xybre>
I think a fresh start could be good actually.
<mbj>
It is ;)
<xybre>
Datamapper was super lightweight, and ROM seems even more flexible.
<xybre>
I discovered a horrible ActiveRecord "feature" today too. It does coercions ("type casting") eagerly when you set the value. And I don't see any way to turn it off.
<mbj>
xybre: ActiveRecord (pattern) is by design a single reponsibility violation.
<xybre>
So for a boolean column, as soon as you set the value on the Ruby object it will immediately try to force it into the expected type. As a result the string "true" is converted to a true object, but "True" is converted to false.
<mbj>
The implementation ActiveRecord does far to much at one spot.
<mbj>
Okay, we could argue the pattern can be used in a way we dont have an SRP-violation.
<xybre>
So I can't even do validations against the actual originally set value!
<mbj>
But the implementation stuffs far to much into your models.
<zaidan>
mbj: ok great, will think about a good name
<xybre>
ActiveRecord::Base's namespace is *full* of stuff.
<xybre>
Dumping it with `ls` in Pry reveals just an obscene number of methods and objects.
<mbj>
zaidan: jo
<mbj>
zaidan: Also talk to snusnu, he wrote such a tool some years ago
<mbj>
zaidan: We could reuse his name, and probably win him as a contrib
<mbj>
xybre: ack
<zaidan>
mbj: ok good idea
<xybre>
394 methods, not included whatever it monkeypatched onto Object.
<xybre>
+ 70 constants.
travis-ci has joined #rom-rb
<travis-ci>
[travis-ci] mbj/concord#23 (master - 974bbc9 : Markus Schirp): The build was broken.