This is the second perfectly on-time major release in a row. Way to go team!
Django 1.9 Released
71–80 of 91 posts
Re: Django 1.9 Released
#72This is still better when you are able to experiment versions without quirks and workarounds. Hail virtualenv.
also, hail Docker.
Re: Django 1.9 Released
#73Not looking to start a fanboy-war here; but i would like to hear some opinions. If you were to start a new small to medium sized web app today, would choose Ruby on Rails or Django?
If you like Ruby, Rails. If you like Python, Django. Both are full featured with active communities. The 10,000ft view of each is very similar.
Right - if you're not familiar with either language, then have a look. For me, at least, I could never really wrap my head around some Rubyisms. Python on the other hand, just "made sense" to the way my brain works. I can parse Ruby mentally, it just takes more effort.
If you're familiar with both, and have no clear preference, the question becomes a lot less clear-cut.
Re: Django 1.9 Released
#74Not looking to start a fanboy-war here; but i would like to hear some opinions. If you were to start a new small to medium sized web app today, would choose Ruby on Rails or Django?
I'm shocked, frankly, at how lousy both are out-of-the-box. I just don't get why a simple CRUD app (which many/most apps are) has to be so difficult. I had trouble figuring out how to do email login on Django and just gave up. At first, Rails seems like it's going to be a lot better. I don't care for generating files. Don't like TDD. And found the migrations very confusing (if I make a typo in a field name, I delete…
Sorry about your troubles.
However, you can use Django AllAuth (https://github.com/pennersr/django-allauth) to achieve an email login system by tweaking its settings.
Specifically:
ACCOUNT_AUTHENTICATION_METHOD = "email"
ACCOUNT_USERNAME_REQUIRED = False (if you do not want to use the username field)
More here: http://django-allauth.readthedocs.org/en/latest/configuratio...
Cheers!
Re: Django 1.9 Released
#75Not looking to start a fanboy-war here; but i would like to hear some opinions. If you were to start a new small to medium sized web app today, would choose Ruby on Rails or Django?
I'm shocked, frankly, at how lousy both are out-of-the-box. I just don't get why a simple CRUD app (which many/most apps are) has to be so difficult. I had trouble figuring out how to do email login on Django and just gave up. At first, Rails seems like it's going to be a lot better. I don't care for generating files. Don't like TDD. And found the migrations very confusing (if I make a typo in a field name, I delete…
I think this is more a flaw of the authoritarian nature of larger web frameworks. They can't cater to every use case and have to be opinionated in certain areas.
In Django's case I would say their Authentication model was an opinionated choice that has since become a major staple of Django. Many frameworks will not include default ORM-models and will instead defer to the user to implement their own authentication scheme.
Personally I disagree with this choice, I think database schemas have to be flexible and an in-framework authentication setup will only encourage users to lean more on the framework even when it is working against them. I've burned plenty of hours in Django trying to work / extend the authentication model. While I have no doubt much of this was due to my inexperience with Django, there is a certain responsibility with having a framework used by many beginners who will struggle and perhaps pick up bad habits (like expecting your web framework to automagically solve everything) if you do not design it for them.
Overall I think Django is awesome. But it's not for every use-case, or every developer. About a year ago I moved over to NodeJS professionaly, some day I will perhaps go back but have yet to have any significant reason to do so.
Re: Django 1.9 Released
#76Is the Django team planning anything big for a 2.0 release?
The integration of Andrew Godwin's django channels [1] would be a big thing :)
Re: Django 1.9 Released
#77Earlier quoted context omitted.
Good call, this is very important! It's baffling that such a huge project, used by thousands of companies including at least one billion dollar unicorn, have such a hard time finding money to fund only one single developer. As a solo owner and happy Django user, I've donated what I could and encouraged others to do so but would love to find a solution to entice larger businesses to donate regularly.
The Django Software Foundation has just hired a part time person to organise fundraising [0]. There is also discussion happening around the community about trying to find a sustainable way of funding open source projects and Django in particular. Part of the problem is that it's hard to get companies to "donate" money. If they can purchase something like "support" or something else tangible, or "sponsorship" out of t…
Maybe I do not know some important details, but why DSF (Django Software Foundation), does not do Django consulting in the first place. They very highly likely can do the best Django consulting in the world.
Secondly, maybe DSF could bake some great products for/from Django? The first ideas I have is Django specific, and user friendly:
- PaaS;
- analytics (as in 'new relic');
- crash and performance logging (as in 'sentry' and/or 'opbeat')
- CI;
DSF can just simply make business, by making great infrastructure stack around Django.
Re: Django 1.9 Released
#78Before anyone says something about Django not being hip enough for realtime stuff, just know that you can layer Pushpin ( http://pushpin.org ) in front of it. This may sound like a bandage but it's actually good design, regardless of programming language. Disclaimer: author.
I didn't hear about Pushpin before, looks like a solid product, will definitely look into it. I'm not sure if that's a valid question, but how does the Pushpin compare to something like SwampDragon ( http://swampdragon.net/ ), besides not being tied to Django?
I'm doing my own thing called channelstream:
https://github.com/AppEnlight/channelstream
It's gevent based and supports websockets and long-polling, by design it integrates with any kind of application for real-time broadcasts, be it django, php, cron scripts etc.
Re: Django 1.9 Released
#79This is still better when you are able to experiment versions without quirks and workarounds. Hail virtualenv.
Django is really easy to deploy manually too.
Download the tarball, untar it, make a symlink from inside your project dir to the "django" subdir. If you want to try another version, you just change the symlink to point to the other version. If you are curious which version you're running, you can just do a "ls -l" in the project dir. That's it.
Re: Django 1.9 Released
#80Is the Django team planning anything big for a 2.0 release?
Expanding a bit: the plan is that 2.0 just switches us to a new versioning system. The tl;dr is: Versions start going 2.0, 2.1, 2.2, then 3.0, 3.1, 3.2, then 4.0, 4.1, etc. Every X.2 is a long-term support release. The deprecation cycle gets tweaked so that if you're on, say, 2.2 LTS and your codebase runs with no deprecation warnings, then you'll be able to just run as-is on 3.2 LTS. You can then clean up any deprec…
And as a special case of that: according to the current roadmap, 1.11 will be the last version to support Python 2.7, 2.0 won't.