#musicbrainz-devel

i think with me and warp in similar time zones, i could help mentor

I have some experience with haskell already, which would make warp seem the real choice

ianmcorvidae's task is a most 50% working on non-get-more-money-in the door stuff

eep!

:)

if we want to do ingestion in haskell, that may be a good project for me

ianmcorvidae: agreed.

given that it's related to getting more money in *and* something we're talking about doing in haskell :)

I think we should use haskell for the ingestion stuff

yes, please. support 1000%

OT, vaguely related: ruaok: rollup replication stuff is on codereview if you feel like looking at that :)

ianmcorvidae: I have no chance. I'd like to, but my eyes are bleeding.

email comes in faster than I can read it, much less respond.

no worries, just figured I'd inform you

thanks. :)

so, so we have two paths

1. Delay NES while warp works out a prototype in haskell.

2. use haskell for ingestion stuff and proceed with dal on python.

does that sum up our feelings atm?

those seem to be the 2 suggested paths, yep

ok

well, haskell's getting used for ingestion either way, but yes :)

I think from a business and a finance perspective #1 is not good at all.

considering we're going to be stuck with whatever language we choose for DAL/NES for quite some time. I think we need to have properly evaluated all options. so I say #1.

1 month isn't that much in the grand scheme of things

:)

you're not in charge of keeping everyone paid. a month is a major pain.

and I know exactly the costs of that.

honestly, I think #1 as well; NES doesn't need to block working on ingestion stuff, which is one of our major finance routes

ocharles: if we choose #1, will you promise not to hassle me on where NES is?

you fuck all this.

you know.

I know exactly what is going to happen here.

ruaok: i'm only hassling to make sure these conversations happen and everyone is clear where we are on the map. i'll certainly know where we are on the map, so i'm fine.

we spend a month on the prototype and then everyone decides to use haskell.

seems tall order to get warp developing a prototype in a month if he has no knowledge of haskell to start with

so lets just skip the fucking prototype and get on with it.

but then I object to ingestr also being in haskell.

two projects is too much. too many eggs in one basket.

DAL/NES goes forward as haskell.

ingestr goes forward in python so I can help.

then ingestion in python? this may be my project, which means if not haskell python or perl is best

ok

ianmcorvidae: that last part was unclear to me.

ingestr in python, DAL/NES in haskell is good for me

are you saying perl or python?

ok.

ruaok: he is saying yes, perl or python.

ruaok: I'm saying that if it's not haskell, it should be either perl or python, I'm fine with either; since you said python python is fine :)

but we are NOT using haskell for anything else until we review.

i'd still like to do the prototype of Data::Label. if it's good, it's not entirely throw away - we'll have a framework for serving haskell functions over http

ok, final summary: DAL in haskell. ingestr in python. all new projects are subject to review.

and obviously labels are not going away either

ruaok: I'm ok with that outcome.

ocharles: this DAL in haskell better fly faster than an armada of pigs.

we waste a bit of work by writing a NGS view of labels, but nes still needs more planning before that truly starts

ruaok: african or european?

pocket.

smaller footprint, ergo faster in air.

I think its a complex set of bad choices.

but, motion carried.

what is our next step?

are you really happy with carrying motion on those grounds?

I'm not writing that much code.

you guys are.

I can force my will and make you guys unhappy.

one is much more likely to get things moving faster.

ijabz: of being unhappy? yeah.

but this isn't about me.

I just hope the DAL in haskell is going to walk on water, chew gum at the same time and be hot has shit to look at.

ocharles: what is our next step on DAL/NES?

is wasn't about me either, I have the interests of musicbrainz at heart

in the grand scheme of moving parts fitting together, i'm not sure

to recap...

ijabz: agreed. we are in the same camp. lets go have a beer in london. :)

we have a schema and a clear model of the various bits of data, and also a transformation from where we are now, to the nes schema

:) sounds good

the steps though... that might be getting it in read only form?

ocharles: on a more meta perspective.

for moving directly forward with nes, i probably need to give it a more formal look

who is going to drive the next steps towards the DAL. you or I?

i'm happy to provide suggestions and have them approved or rejected

i suppose ultimately there's still (far too much) a lot in my head

we dont really have a requirements doc for NES, do we?

we have the requirement slides :)

we have a requirements doc for the DAL that needs to now have questions answered and re-worked.

and the requirements-y blog post

I think you should take the first stab at editing the DAL doc.

get all the various bits of info we already have and see if we can make a cohesive doc.

then next week I can go in and make sure we're on the same page.

have the requirements or anything changed much?

then when we're in SLO, we can spend the entire time designings things and coming up with a clear roadmap, requirements and a spec for DAL.

perhaps some designs stuff as well.

ocharles: not really. there are just a lot of open things we could not answer based on the questions we've just now answered

and some lines of code!

so, we need to go back to that and edit the doc with the knowledge we have now.

warp: we may just want to spend our time talking.

we have tons to do and talk about.

should i re answer the questions or something?

ruaok: yes, I know.

ocharles: if there are open questions that can now be answered, answer them.

collect open questions and we will tackle those early next week.

lets see if we can have the dal requirements done early next week.

is that an acceptable goal?

i'll give it a shot

ok.

how do we communicate our decisions today?

can someone summarize this conversation and post it to mb-devel?

sure.

great, thanks warp.

nothing from me

then we're done.

thanks for the battle, gentlemen.

ruaok: it's really late though, so I'll post that tomorrow (european) morning.

warp: ok

thanks ruaok

ruaok the fact you've managed to throw out Perl despite some opposition is at least something to be positive about

ijabz: yes.

if that was the compromise that was needed to put a boot up perl's ass, so be it

maybe one day I'll be writing in it for mb :3

I just hope people aren't like me :P

CatCat: no offence but do you know much about that sort of stuff ?

hawke: now with cloud haskell, you can be excused :)

ijabz: I know nothing. but I have a mind for to leanr code. it is what I will probably end up doing. why do yo usay "no offence"? that automatically assumes that offence could be made?

honeslty I'm to tired to read between the lines right now

natta reoafk

natta ;)

hi pbryan!

ruaok!

you coming to BCN?

No Barcelona for me... :-(

fuss. :(

Sorry.

Whaa?

the CAA was officially launched yesterday

Ah, sweet, it's official!

Is http://archive.org/details/coverartarchive down?

archive.org seems to be down

they seem to be having some trouble, yes :(

Doh.