solnic has quit [Remote host closed the connection]
solnic has joined #rom-rb
solnic_ has joined #rom-rb
solnic- has joined #rom-rb
solnic_ has quit [Read error: Connection reset by peer]
solnic has quit [Read error: Connection reset by peer]
solnic- has quit [Read error: Connection reset by peer]
solnic has joined #rom-rb
snusnu has joined #rom-rb
snusnu has quit [Quit: Leaving.]
snusnu has joined #rom-rb
solnic_ has joined #rom-rb
solnic has quit [Read error: Connection reset by peer]
solnic_ has quit [Quit: Leaving...]
postmodern has quit [Quit: Leaving]
cored has joined #rom-rb
<cored>
hello all
solnic has joined #rom-rb
<solnic>
kapowaz: hey
<kapowaz>
solnic: hey.
<solnic>
kapowaz: any luck installing ruby? ;)
<kapowaz>
bleh
<solnic>
it doesn't *require* ruby 2.0
<kapowaz>
I actually forget where I got to it
<kapowaz>
yeah I imagine I could downgrade the gemfile?
<solnic>
so whatever ruby you have will work
<kapowaz>
it looks like my ruby setup on this machine is kind of screwed now
<solnic>
why?
<solnic>
need help
<solnic>
?
<kapowaz>
if I type ruby -v I get:
<kapowaz>
rbenv: version `2.0.0' is not installed
<kapowaz>
twice
<kapowaz>
:|
<solnic>
did you try removing rbenv?
<kapowaz>
no… but I'm a little concerned about the implications of doing that
<kapowaz>
how do you use ruby-build to install a version?
<kapowaz>
I could have sworn it was ruby-build install <foo>
<kapowaz>
where <foo> is a version
<solnic>
no idea, I use ruby-install + chruby
<kapowaz>
ah, ruby-build != ruby-install
<solnic>
I bet you could switch to it in no time
<kapowaz>
probably. I'll upgrade homebrew and see about installing that
<cored>
solnic: saw your comment
<cored>
solnic: from the rom-vagrant PR
<solnic>
cored: yeah I just cherry-picked a couple of commits
<solnic>
I'll push it in a second
<cored>
solnic: do what you need, I'm planning to keep evolving it but I need to know some more stuffs about the entire team process of developing
<cored>
in particular the .dotfiles thingy
<solnic>
cored: it's my own thing :)
<cored>
oh
<cored>
he
<solnic>
I needed it, so I added it
<solnic>
:)
<solnic>
I work inside the vm, some people don't
<cored>
I thought everyone in the team use the same dotfiles and I was like this is awkward ;-)
<solnic>
lolwat :)
<solnic>
well, we all use vim
<solnic>
but our configs are not the same
<cored>
solnic: I really don't do that, still can't see the benefits of working entirely inside the vm
<cored>
solnic: nice :-)
<kapowaz>
I think I see the problem, although god knows what is causingit
<kapowaz>
rbenv has a command to list all known definitions for versions of ruby, which it ostensibly gets from ruby-build
<kapowaz>
but they don't have the same list available
<kapowaz>
so ruby-build has 2.0.0-p247 but rbenv only has 2.0.0-p0
<kapowaz>
:|
<solnic>
kapowaz: switch to ruby-install / chruby :)
<solnic>
it's dead simple
<kapowaz>
yeah I think I'm not going to do that at 4pm on a Friday on my work machine ;)
<solnic>
WHY?! ;)
<solnic>
it's the perfect moment
<solnic>
maybe wait 45 minutes
<kapowaz>
I am supposed to be looking up things to do this weekend with the missus
<kapowaz>
so far failing
<kapowaz>
also failed to achieve ANYTHING on my personal project today
<kapowaz>
which you might expect, at work and all
<kapowaz>
but today is day 12 of Nothing To Do
<kapowaz>
:|
<solnic>
gotcha
<kapowaz>
I think I have a solution though
<kapowaz>
you can install with ruby-build, it's just a little more complicated
<kapowaz>
think it's working
<kapowaz>
I'll have to actually, y'know, use ROM at some point.
<kapowaz>
I'm expecting there will be quite a bit of re-learning to switch over from DM
<solnic>
kapowaz: yes, quite a bit...
<kapowaz>
just when I was starting to feel like I understood DM!
<Gibheer>
kapowaz: when you understand something, move on ;)
<Gibheer>
or else it gets boring
<kapowaz>
that or it starts to get useful
<kapowaz>
;)
<Gibheer>
useful in IT? that would be so wrong :D
<kapowaz>
I have a suspicion that some people use a profession of building architecture-type tools as a career in the same way others spend a lifetime in academia…
<kapowaz>
present company excepted, naturally! ;)
<Gibheer>
hehe
kpwz has quit [Ping timeout: 240 seconds]
mbj has joined #rom-rb
<mbj>
hola
<solnic>
hola mbj
dkubb|away is now known as dkubb
<solnic>
dkubb: morning
<dkubb>
solnic: good morning
<solnic>
dkubb: looks like you're close to finishing relation::variable :)
<dkubb>
yeah, I think so
<dkubb>
it was easier than I thought
<dkubb>
I need to do some tweaking on the optimizer side
<dkubb>
or at least testing
<mbj>
dkubb: nice, I need to catch up with that rvariable hting!
kpwz has joined #rom-rb
<dkubb>
heh, I saw tenderlove talk last night at vanrb
<solnic>
dkubb: about?
<dkubb>
he showed a way to change a class of an existing object. it was nasty, and by no means should actually be used in production code, but it was still neat
<solnic>
lolwat
<dkubb>
hehe
<dkubb>
there was some interesting AR optimizations he was working on
<mbj>
lulz, that sounds the the most stupid mutation of a datastructure ever.
<dkubb>
hehe
<dkubb>
it was more or less to see if he could
<solnic>
dkubb: integrating memory adapter with relation::variable shouldn't be a big task right?
<mbj>
Okay, we could use this for mutant. It does dump thinkgs.
<dkubb>
solnic: no, I think I'll have that working by next week sometime.. in fact I think I may be able to slim down the memory adapter a ton
<dkubb>
solnic: I don't even know if it needs an #insert, #update or #delete method.. or even a gateway. it mostly just needs the #[] method on the adapter
<snusnu>
dkubb: good morning, and thx for pointing me to that anemic domain models link!
<snusnu>
dkubb: might well be that i change my mind, and add behavior to (some) of the "dtos" .. initially i was thinking to have the rich domain model only use those as data capsules under the hood, but we'll see how it goes
<dkubb>
snusnu: no problem. I knew that you were thinking about moving it to use ROM eventually, but I figured I'd mention it just in case
<snusnu>
also, hey solnic :)
<snusnu>
it will be so interesting to *finally* be able to develop an app using a real datamapper … i guess we're basically all trying to implement fowler's ideas (especially, poro domain models) for the first time
<snusnu>
while i always liked the theory, i have to admit that i never actually implemented any larger (web based) project using those ideas .. well, simply, because there wasn't feasible tool support in ruby
<mbj>
snusnu: heya
<solnic>
hey snusnu
<dkubb>
yeah, there'll be a period where we're experimenting with it to find the right balance
<mbj>
snusnu: I read this some time before and I have to agree.
<solnic>
yeah anemic model is a funny anti-pattern, happened to me a couple of times
<snusnu>
solnic: oh really? with web based rails projects?
<solnic>
snusnu: yes
<dkubb>
I've done it too
<snusnu>
solnic: you must've left the "golden path" then …. :p
<dkubb>
AR kind of encourages this
<mbj>
I have objects like this in most of my stuff. But:
therabidbanana has joined #rom-rb
<solnic>
data objects w/o logic can be handy
<mbj>
They are leaves and inputs and outputs of wrapping domain logic classes.
<solnic>
esp for presentation logic
<snusnu>
mbj: right … that goes in line with what i have and am thinking
<mbj>
Yeah, most of the basic nouns in the domain can easily be "behaviorless immutable domain objects".
<dkubb>
I'm more likely to use proxy objects for presenters
<mbj>
I more and more call them "domain dtos".
<solnic>
yeah that's what I meant, you build a data object and wrap it with a proxy used for presentation stuff
<cored>
dkubb: hello
<cored>
mbj: hi
<solnic>
so let's say you fetch a user data object and build UserJsonPresenter or something wrapping that original user data object
<cored>
dkubb: how can I know which is the priority of the list on the axiom-types refactoring?
<mbj>
cored: hola
<dkubb>
cored: I would generally just work top-down
<dkubb>
cored: although I think most of those are roughly equal in weight and independent. it'd be different if task #2 was dependent on #1 being done, etc
<dkubb>
cored: If you feel the order is dependent, then feel free to reorder it from top to bottom
<cored>
dkubb: got it
<cored>
dkubb: I think I will do the first one in a couple of inutes
mbj has quit [Ping timeout: 264 seconds]
mbj has joined #rom-rb
<cored>
dkubb:
<cored>
Float::MAX from axiom-types return Infinity
<dkubb>
cored: I had to create two objects to represent positive and negative infinity, because the built-in stuff wasn't comparing properly to BigDecimal
<dkubb>
cored: why even have DEFAULT_SIZE at all? afaik, there's no reason for it to exist in the end
<dkubb>
cored: maybe I screwed up in the dependencies, but the goal is to have any need for it to be gone
<cored>
got it
<dkubb>
cored: note where this constant is used is going to be changed so it only uses the option if the key exists, so it doesn't reference DEFAULT_SIZE anymore (I think I left a note with some example code to explain this in the PR)
<cored>
oki
<cored>
dkubb: in that case
<cored>
removing the default_size and use it from axiom-types doesn't imply that we should remove the unit test inside axiom?
<cored>
I'm guessing that the default_size stuff is been tested inside axiom-types
<dkubb>
solnic: nice work on the devtools binary. I'll start using that today
<dkubb>
solnic: I plan to beef that up to copy over other support files like .rspec, CONTRIBUTION.md, Guardfile and others
<dkubb>
solnic: thanks for bringing that logic together. I can probably do this fairly easily now
solnic has quit [Quit: Leaving...]
snusnu has quit [Quit: Leaving.]
zekefast has quit [Quit: Leaving.]
jessekempf has quit [Quit: Leaving.]
snusnu has joined #rom-rb
kpwz has joined #rom-rb
<therabidbanana>
Have any of you guys played with CQL (cassandra query language?)
<Gibheer>
therabidbanana: I did a bit
<Gibheer>
what is your question?
<therabidbanana>
It's SQL like, but has some weird restrictions. I don't really have a question, was just wondering if those restrictions would make it hard to build support for CQL into rom.
<therabidbanana>
Gibheer: did you have a gem you were using?
<therabidbanana>
I'm writing raw CQL queries for this project I'm working on and it just feels dirty and wrong. :)
<Gibheer>
I think mbj did a spike for cql some time ago
<Gibheer>
I did use a gem, but it was a pretty heavy work in progress between cql 2 and 3
<Gibheer>
so some stuff was not working at all :/
<Gibheer>
therabidbanana: I was using cassandra-cql
<dkubb>
therabidbanana: rom shouldn't know too much about cassandra. it'd be axiom that would know about it..
<dkubb>
what we'll probably do is something similar to what we're doing with sql support. we'll write a cql gem to generate CQL from an ast, and then have an axiom-cql-generator map an axiom relation to the CQL ast
<dkubb>
there are some limitations in CQL though
<dkubb>
it doesn't support OR conditions for example
<dkubb>
however, we can simulate that by transforming it into an equivalent query
<dkubb>
or rather, sets of queries which will need to be combined in memory
<dkubb>
any query with an OR in the WHERE clause can be rewritten as a union of two queries with a WHERE clause of the lhs and rhs of the OR
<dkubb>
if that makes sense
<dkubb>
axiom allows us to combine stuff in-memory
solnic has joined #rom-rb
<solnic>
dkubb: thanks, I just put it together quickly for a good start, there's a ton we could do to improve our dev workflow
<solnic>
I want to automate as much as possible :)
<dkubb>
me too
<dkubb>
therabidbanana: from everything I've seen working with CQL and cassandra, there's no reason we can't support it in rom. it'll take time and effort for someone interested in doing it, but it shoudl be just as powerful as the RDBMS adapters
<solnic>
oh cassandra! that would be cool
<solnic>
I'm hoping to see all major nosql dbs supported in ROM at the same time when we add RDBMS support
<dkubb>
solnic: cassandra is pretty awesome actually. I was using it on my last contract and it was neat. I think it has the ability to stand-in for databases in many situations
<therabidbanana>
Yeah, everything I've seen so far is pretty cool.
<solnic>
I've never used it
<dkubb>
the only downside is there's no constraint system, which I guess is a big one, but it's ability to be distributed is really neat
<solnic>
I'm esp looking forward to mongodb, riak and cassandra support
<therabidbanana>
And the distributed nature introduces some interesting limitations for people used to standard RDBMS queries.
<dkubb>
mbj already did an arrangodb adapter
<dkubb>
maybe he can get those companies to fund adapters for their dbs too
<therabidbanana>
Like they won't let do range filters unless it's part of the primary key.
<solnic>
yes I thought about that the other day
<solnic>
dkubb: btw I'm thinking about running a huge fundraiser after we release 0.0.1
<solnic>
somewhere around september/october
<dkubb>
solnic: yeah, I would support that. I would want something to show before asking for anything
<solnic>
I think reaching out to companies behind those dbs would be a good idea
<solnic>
dkubb: I'm in touch with bountysource people, they offered me help in organizing it
<dkubb>
yeah, the arrangodb people also helped fund axiom-fuzzer too. it would be nice to see those utilities get extra attention and help
<dkubb>
they also tend to find bugs in the db too
<solnic>
dkubb: now *that* is an awesome argument
<dkubb>
all this stuff feeds back into improving mutant, axiom, rom, etc
<dkubb>
I have not failed to find a bug in a database when fuzzing axiom
<dkubb>
in fact, I'd bet some of the nosql dbs out there will have dozens of bugs we can find and report
<dkubb>
the rdbms' seem to be pretty tight for the most part.. only a few bugs for each of mysql, postgresql and sqlite
<dkubb>
sqlite is arguably the best tested open source db, and I even found bugs in that
<dkubb>
they fixed them in less than 24 hours which was amazing
<solnic>
dkubb: do you have cases like that written down somewhere?
<dkubb>
solnic: no, but I could easily bring them up from memory if I sat down and wrote them down
<solnic>
ok
<dkubb>
my name is on the sqlite mailing list for one of the bugs I reported, so I could just search for that
<dkubb>
the cool thing too is everytime we create an adapter we make sure it passes with the fuzzer, improving the fuzzer in the process, and then we can re-run it against any existing adapters to find more bugs in the db or adapter implementations
<dkubb>
it's like having a shared integration test between every db
<solnic>
sounds great
<dkubb>
I plan to also add an abstract adapter with some shared specs that exercise each adapter to make sure they handle all the queries
zekefast has joined #rom-rb
<solnic>
dkubb, snusnu: I'm gonna remove ice_nine dep on devtools...
<solnic>
dkubb: how do you find rubocop?
<dkubb>
solnic: cool
<dkubb>
solnic: I like it so far. *most* of the recommendations I agree with or have no super strong opinion on
<dkubb>
solnic: I have a few customizations in axiom's rubocop config where I disagree
<solnic>
dkubb: I want to try it in rom
<dkubb>
solnic: atm I'm just learning what I like and don't like. at some point I'll update devtools' config to include exceptions for stuff I'm pretty sure we all disagree on
<dkubb>
solnic: yeah, I'd recommend it. you don't have to fix everything, just fix the stuff you agree with and turn off everything else.. then if you ever do add something that doesn't meet your conventions it'll tell you
<solnic>
dkubb: hmm ok
<dkubb>
that's what I did in axiom. I have a few TODO notes for stuff to fix, like the maximum line length
<dkubb>
I basically set the threshold to the longest line, with plans to reduce it
<dkubb>
I would prefer per-file overriding of course
<solnic>
rite
<dkubb>
then I can have the nastier code not setting the bar for the normal code
<solnic>
dkubb: thanks for comments on devtools commits
<dkubb>
np
<dkubb>
I love Pathname so much, I like to remind people about it
<solnic>
I wanted to work on this feature partially because I needed to refresh my memory; it's been a while since I write this kind of stuff in ruby
<dkubb>
solnic: oh maybe because it was referring to the gateway branch?
<dkubb>
solnic: when I merge a feature branch into master I typically remove the branch
<solnic>
whoops
<solnic>
thanks yeah it does point to gateway
<dkubb>
jessekempf: cool. I'll ping mbj about it.. it looks fine to me. I wanted to merge it in 3 days ago ;) I'll ping mbj about it
<dkubb>
jessekempf: mbj may have suggestions to change the predicate to match his examples but otherwise I don't see why we can't merge this in asap
<jessekempf>
dkubb: kk. FWIW as a result of mbj's saying "no" I think this is a much better place to locate the concern of whether we should dump output on failure.
<solnic>
jesus christ yardstick in devtools project causes build failures now :P
<dkubb>
did YARD coverage drop?
<solnic>
yes :P
<dkubb>
jessekempf: yeah, I agree it's an improvement
<solnic>
why are we even using yardstick in devtools
<solnic>
these docs are just noise there
<dkubb>
probably because devtools is dogfooding it's own metrics
<solnic>
# return project root
<solnic>
def self.project_root
<solnic>
I fail hard at seeing how this is useful to anybody
<dkubb>
solnic: does ROM run with the new --warnings flag in the .rspec config file?
<solnic>
dkubb: nope
<solnic>
I'll add it
<dkubb>
I found some warnings in axiom-types a moment ago
<dkubb>
stuff where I didn't initialize an ivar before using it
<dkubb>
while I don't agree with *all* of ruby warnings, I do think it's important for libraries to run warnings free
<dkubb>
I dislike it when gems aren't good citizens in the ruby community and make it hard for peolpe to debug things with warnings turned on
<dkubb>
there was a good rant on this but I can't find it now
<solnic>
we're getting there
<dkubb>
yeah. I keep having to remind myself it's not where you are that matters, it's the direction you're moving in.. there will always be something that needs fixing, what's important is continuous improvement
<dkubb>
as long as we keep raising the bar this will be the case
<dkubb>
;)
<solnic>
dkubb: I can see you liked that raising the bar thing ;)
<dkubb>
hehe
<dkubb>
what's funny is I can recall myself using that term a bunch too
<solnic>
dkubb: meh, warnings in Concord /cc mbj
<mbj>
solnic: I'll look into it.
<mbj>
*both, mutant and concord
<mbj>
sorry I dont have that much time for opensource
<solnic>
nobody has, no worries :D
<mbj>
heh
<dkubb>
I do! only because I'm waiting to hear back from a few pending things
<dkubb>
but that could change tomorrow
<mbj>
dkubb: nice, btw merged that PR
<solnic>
dkubb: tomorrow is Saturday :)
<dkubb>
I wish I didn't though.. I'd rather be doing billable work right now, only because my wife will give me grief if I do oss all day
<dkubb>
solnic: heh
<solnic>
dkubb: I hear you haha
travis-ci has joined #rom-rb
<travis-ci>
[travis-ci] mbj/mutant#549 (master - 97f1231 : Markus Schirp): The build was broken.
<travis-ci>
[travis-ci] rom-rb/rom-mapper#40 (master - 0658075 : Piotr Solnica): The build has errored.
travis-ci has left #rom-rb [#rom-rb]
<dkubb>
weird that rbx would start failing all of a sudden
<solnic>
dkubb: is it? :P
<solnic>
gotta go, good night
<snusnu>
solnic: just read through the logs … i fixed concord to be warning free already, it needs a new release
<snusnu>
also, i noticed that mutant ERRNO:ESYSwhatever bug with substation on jruby too
<solnic>
blah
<solnic>
snusnu: thanks (re Concord)
<solnic>
snusnu: I'll be preparing code for the first release in the upcoming days and then it's only a matter of finishing memory stuff by dkubb and tweaking few things here and there on rom-side and...we're ready to push
<snusnu>
solnic: awesome!
<snusnu>
solnic: maybe i can spend a few hours help ironing out things too
<solnic>
cool
<solnic>
gotta run now
<solnic>
good night
<solnic>
(again)
solnic has quit [Quit: Leaving...]
travis-ci has joined #rom-rb
<travis-ci>
[travis-ci] rom-rb/rom-session#55 (relation-integration - cb36fb8 : Piotr Solnica): The build has errored.
<dkubb>
snusnu: how's the state of the rom code? warning safe, passing mutant, etc?
jessekempf has joined #rom-rb
<snusnu>
dkubb: i dunno exactly, i haven't been working on it since solnic nuked almost everything ;) i know that before that, it was passing devtools ci except mutant … i wouldn't be surprised if it's much closer to passing mutant now, as it's much less code now
<snusnu>
dkubb: it's probably not warning safe, but even if it isn't, i'd expect that making it so would take only a few minutes
<snusnu>
dkubb: it definitely will be, on first release
<dkubb>
yeah, warnings are pretty easy to remove
<dkubb>
I was mostly curious. I just wanted to make sure our projects are in sync, and if there's any stragglers I was going to help them
<snusnu>
dkubb: wow, i saw you already merged the relvar branch into axiom master! sweet!