#pyramid

hmm

time to figure out the best way to integrate raven/sentry with Pyramid

Ergo: almost have a cookiecutter ready with those options

currently i have some views with a json renderer, in which i return model objects that have a __json__ method.

what's the sanest way of being able to have different renderings of the model objects depending on the view (like, a summary rendering and a detail rendering)?

since __json__ takes the request object, i could set some property on request saying i want the fooModel to render in summary mode or in detail mode, i suppose

but it seems like there should be a better way

inklesspen: Not quite sure I understand the question but seems like one option would be to have a property on the object return another proxy object with the data you want, such as `return user.with_details()`. But I usually define a method such as to_dict/to_dicts that handles the serialization of models and takes an include param that determines what I want in the serialization. For example, Sale.to_dicts(sales, request, include=None) ... Sal

hello, anyone online?

nobody here but us zombie processes.

In defunct process land no one can hear you ctrl-G.

in pyramid or application development, when do we use sqlalchemy and when do we use ZODB?

whenever it's appropriate and we feel like it ;-)

afidegnum: that depends entirely on what you want to do.

afidegnum: typically you use sqlalchemy. You would seriously consider ZODB when you have a dataset that matches very strongly to objects - content management systems often meet this criteria, as they contain a diverse range of user-driven entities and templates that don't match well to a strong set of relations.

tree-ish stuff works well with zodb, table-ish stuff works well with sql

The "tyranny of choice"

let someone else choose it for you, and you can always blame them

ok

but what is the best to choose for large and complex database records with complex relations?

That question answers itself. Stick with the DB

i m told db can handle that,

zodb I mean

zodb doesn't deal with records

can you please explain more of zodb utility ?

If you already have data in a relational DB, just use SQLa and stick with it.

Please ask specific questions related to something you want to do.

I am building complex application which will interract a lot with the database as well such as saving and retrieving entires... the applicaiton will be social oriented,

that's why I want to find out about the appropriate database tool to use,

Then use a relational DB

afidegnum, check http://zodborg.readthedocs.org/en/latest/docume... for basic info on zodb. pay attention to chapter about search

+1 to relational db for your case

possibly with some caching in mongo for speed

hi everyone! have some question about pyramid. in setup.py we have requires - as I understood - it's require packages for my project doesn't matter production or development and in config file (development.ini) we have pyramid.includes - packages for environment only?

I've tried hynek's structlog, its quite nice, even got it integrated with app enlights log tag support nicely :-)

[This message was brought to you by Ziviec at PyCon PL]

ROTFL :P i can see that memory of cheap beer is strong in you my young padawan

still conflicted about Free vs. Almost Free

Ergo: well go on then, tell us what structlog is.

Something is really wrong with my Terminal. Echo keeps disabling.

Charlie_X: hyneks project - http://www.structlog.org/en/stable/

Charlie_X: https://youtu.be/LklGF1rcJII?t=21m44s ;)

(timestamped to the relevant part)

[redacted]

dowwie: whatever you find valuable for you i guess

obvious things are remote ips and users. but depends on your context. a scientist once wrote me how they use it to do some science shit. he just logged into a list.

bah! why does the US telecom industry suck donkey nards so hard?

my ISP's upstream provider (AT&T) has been jacking up its rates for shitty DSL service, and they finally had to pass it along to its customers

meanwhile the only viable option to DSL is Comcast, a soulless media giant

and they want $70 for 16/3Mbs plus $20 for a static IP

i could go residential instead of business service, but then i can't have a static IP

and they throttle

stevepie_: i think the only answer is they can do this and people pay :D

damn me for giving a shit about this stuff

almost makes me want to move into an urban environment

i'm in suburbia without the benefits of modern technology

effing copper

and AT&T no longer maintains its copper wires. ditto for verizon. let it rot on the poles!

and they don't provide a replacement other than DishTV

stevepie_: don't you just love "free markets"!

Charlie_X it's not even a "market"

I know

USD 70 for 3 MB upload is pretty good, to be honest but 16 MB downlink is poor.

Here in socialist Europe most countries have open access to the incumbent's network with a regulator setting wholesale prices.

Some exceptions where the incumbent is investing heavily such as FTTH.

My cable company put up charges by about 2% this year but that's the first time in four years. I could switch from 50MB/2MB to 100MB/4MB for a couple of Euros more a month but I just don't need it.

i'd welcome fiber here

This is residential but no one caps in Germany, competition is too fierce.