It's scary, but even if you release something as open source, for free and free for people to modify as they please, you're still in a weird relationship with the people that use your software. You haven't exactly promised to do anything for them, and the code as-is is in itself a great gift... but on the other hand, there's an implicit acknowledgement that "yes, you can build on this".
(Both the open letter and the response from Tom Dale seem to acknowledge that there's two sides to this issue, which is refreshing. Usually you get these really boorish "dude, it's open source, just fix it yourself will ya" responses.)
It makes me wary of pushing any of my projects. I still open source pretty much everything I do, but usually with a big fat "for your edification, don't expect any support or ongoing development" disclaimer.
Maybe the open source community needs a sort of standard way of declaring/disclaiming how production-ready a project is, and how much effort we intend to put into supporting it.
People should assume that if there is no affirmative declaration of "This is production ready" then the code is not production ready.
If there is no affirmative declaration of "I will be actively supporting this beyond my own personal needs" then assume that the code will not be supported beyond the personal needs of the project owner.
Or people can just ask, instead complaining (however well-intended) that their assumptions have turned out false.
This is like people who build a business on some aspect of Twitter or Facebook then complain that Twitter or Facebook is shutting them out. If you are using someone else's code or API make sure you get some positive, unambiguous statement from the owner about what to expect.
Things can still go south, but then you at least have a legitimate gripe.
In the gist, even though the Ember Data README states that it's not production ready, the author complains that it 'should be' or something. We have that standard, it's called '1.0'.
Man this rings so true with me. Emberjs has so much promise but it is just so frustrating to work with. Every time I turn my head something fundamental seems to have changed (the router and now dependency injection). It does always seem like a good move but because the docs don't keep up it means you're suddenly googling solutions that don't work any more. Maybe a kickstarter would be a good way to get things stabilised? If yehuda etc. don't have the time maybe someone would like to step up and with some kickstarter money spend a few months getting up to speed with the project and getting it stable, docs done etc.
It's easy to use snark to dismiss someone's problem.
Why show the domain at all if you can hover over the links? For a minimalistic interface like HN you'd think there would be a reason for it. And if people are identifying content based on that URL, it's probably best to make it descriptive when subdomains have very different content.
Plus, many touchscreen devices don't have an easy way to see where a link is pointing to, even on devices with 'hover' functionality.
Problems that are clearly formulated are easier to solve than ones that are opaque.
At the very worst, it lets the devs know where the users want new features. At the very best, it helps crystallize aspects of the problem that make a solution easier to find.
User feedback is always a gift. Even if they are yelling at you, they're giving you valuable information.
I suppose there are some developers who are not advanced enough to contribute to the project, but smart enough to use a framework? I don't know: Tom Dale did reply to the gist, and mentions there are non-trivial CS problems being solved.
Ember is actively trying to compete with angular (backed by google) and others - and I think the people rooting for ember are just people feeling let down after deciding to move forward on it in production.
edit: Tom's actual quote: "trying to solve are novel computer science problems"
Contributions are not only in form of code right. The anonymous person can help in documentation, interact with the core devs on problems through back and forth communications etc.
To the ember.js team: don't worry, people like this are poison to open source. They demand so much, have such high expectations, and are not willing to contribute anything. They do not realize people have real lives outside of open source, and sometimes they need a break.
Pull requests can take more time to process than feature requests, because 1) The code is usually inconsistent with the rest of the project, 2) there are often merge conflicts that the submitter has to resolve 3) people rarely write tests or documentation about the feature.
You make good points, but each of these is solvable:
1) Require the code to be consistent. Tell the submitter how to do this.
2) Merges have to be straight forward, or you don't accept the submission. I think Linus Torvalds has the same strategy for the kernel - you figure out the merge, I'll accept it.
3) That depends a lot on how you set up your tests and documentation. If you make it simple to write both, it's simple to require them. If they're not simple, you should make them.
My best guess is that Ember is marketed as "omg look awesome!" and Angular is marketed as "hey, this is useful", and this causes the user community compositions to differ just enough for this to be the case (though there are still crazed fanboys on both sides, hence the Ember<->Angular comment thread wars).
This is not even wrong...it's just an absurd statement. Using angular makes it no easier or more difficult for Google "to collect data and increase advertising revenue" compared to any other JS framework.
I am surprised by these threads too, and in general the ember vs angular turf wars. Maybe it's just that I don't work with this sort of thing often, but to me they look strikingly similar.
I'm trying to come up with similar software schisms: vi vs emacs? GNOME vs KDE? Linux vs BSD? C vs C++? C++ vs ObjC? All of these differences sound more substantive to me.
Well, Angular and Ember actually do have substantial differences - they have surprisingly different architectures and general philosophies driving them.
Unfortunately, that seems to lead to a lot of acrimony between them. I wonder if a big part of it is that each side is convinced there won't be space for both in the future, and that a clear winner has to be declared - which I don't think is true at all. This town is certainly big enough for the both of them :)
As for drama within the Ember community itself, that's a little more complex. I think a big part of it is that Ember seems to have a lower barrier to entry than Angular at first glance, but certainly has its own catches and hang-ups, and running into those when you thought you had found a holy grail can be extremely frustrating.
He's probably upset at investing in a pre-1.0 software (apparently he spent "weeks" making pagination work) and then finding out -- gasp! -- it isn't completely stable yet.
It's a big issue that any successful open source project faces. It's natural for the team to want to promote all the great things the product will do. But the core team doesn't scale as fast as contributors, let alone the userbase, and there is great risk of over-promising.
I think the approach of something like jQuery works best: do one thing well. Add polish, not features. Let other people start their own projects that extend yours.
A way to scale back expectations even more is to provide sample code, not a drop-in library. If you want to use it, copy and modify it so you know how it works and can fix the bugs yourself. This ensures that people don't think of themselves as just users.
These are great ideas for things to improve in Ember, but I don't understand how someone could start an Ember project without understanding 1) It's very new, and therefore very raw. And 2) You're gonna have to dig into the Ember source if you want to do anything beyond what you can see in the tutorials.
If you want a framework that can be used for a large, complex project without you having to do infrastructure work yourself, then you have to use a framework where someone has already done that, and there's an established organization that's in the "fixing little bugs caused by slight variations in developer use cases" stage. End of story.
http://emberjs.com/ doesn't say anything about being that unstable on the homepage and the list of users has enough well-known companies that you would reasonably conclude it's ready for production.
Various posts have made the front page here and I don't recall many (any?) caveats saying it's only for thrill-seekers, either.
Not to sidetrack the conversation, but why is Backbone "a joke"? It seems a bit different in approach from Ember, but I've used it for a pretty big application and it worked great.
IMHO Backbone is in a weird spot between plain old jQuery apps with no structure, and frameworks like Ember/Cappuccino/Angular/Knockoutthat have "bindings" constructs that can drastically reduce glue code.
When you choose a framework, surely you have a responsibility to thoroughly review it to ensure it meets your needs? Your needs might include professional support. If none was advertised and you still chose it, why bemoan the fact it's missing?
This, back at 0.8.x, I choose Ember (over an array of frameworks), 4 weeks wasted fighting the framework. Switched to angular in under a day. Using ember is and continues to be filled with surprises. I don't like surprises in a framework.
Tom Dale said it himself, ember doesn't have enough manpower. I find it odd, that Dale & Katz, are unable to raise funds to increase the level of manpower on it. A simple kickstarter may help.
(Both the open letter and the response from Tom Dale seem to acknowledge that there's two sides to this issue, which is refreshing. Usually you get these really boorish "dude, it's open source, just fix it yourself will ya" responses.)
It makes me wary of pushing any of my projects. I still open source pretty much everything I do, but usually with a big fat "for your edification, don't expect any support or ongoing development" disclaimer.
(Also see fat's "What Is Open Source & Why Do I Feel So Guilty? at http://www.youtube.com/watch?v=UIDb6VBO9os)