I’d really like to do this, but it’s not just AutoForm. It’s the combination of that with Collection2 and SimpleSchema2 that bring in the value.
To me SimpleSchema is a real gem.
Where do I start when trying to figure out client side schemas, client side validation, default values, and error messages?
@yuukan, do you have a simple app that use your package =) i’m too is looking for a simple and lightweight autoform
Kind of an old thread, but I just wanted to say that we should resist the impulse to just jump ship to another package as soon as the one we use gets too big or too unwieldy.
Autoform can seem complex and hard to understand, but it’s still an amazing package that gets a ton of things right. I know @aldeed is planning on simplifying some things, making it more modular, and improving the docs.
So I think it’s worth trying to help him out and contribute to the effort, instead of looking elsewhere for a different solution that will probably end up having the exact same issues a couple months down the road.
Thanks @sacha. And thanks to all for bearing with me on issue backlog. Will be getting caught up soon.
Hey aldeed. Thumps up. I like your work. At least I use SimpleSchema and I love it. But I also love simplicity and flexibility, so when I had to choose a form package for my app, I went for templates:forms. And the beauty is that templates:forms uses SimpleSchema for validation. And this is great for the community. We have options, and we love the options. If you haven’t tried out templates:forms, just give it a try. You will love it.
+1 for templates:forms, although the current doc is not so good and the new doc, despite @jamz’s promises, is still not there 
I’ve made a lot of progress towards removing autoform, simples schema, and collection2 (and heck even Iron Router subscriptions) from my project (with all due respect to aldeed).
I removed all the getFieldValues and use ReactiveVars/ReactiveDict instead. Also, I used this vid and was able to get reactive validation working just the other day. It’s a long road, but every chance I get, I remove more dependencies.
I’ve actually learn a lot from the process of removal.
Thanks to aldeed for helping me get started with forms in Meteor.
This.
I’m looking at meteortemplates:forms and typical of many projects, the activity follows a similar inverse parabolic pattern (1/x^2). Last activity 25 days ago on devel, last comment a month ago.
It’s disheartening to hear the more advanced meteor devs drop a open source project to ‘roll their own’ rather than contributing to it. Maybe someday they’ll start a new and awesome forms project which everyone buys into, ask for features, gets large, flexible and complex and then gets dropped because there are too many bugs and no one wants to contribute.
There is no response to issues on the autoform repo from aldeed anymore. Currently almost 300 open issues. Aldeed has state he thinks autoforms is ‘complete’. PRs seem to be ignored, 6 open at this time. It doesn’t seem like he’s open to loosening control to others (no one else is abled to accept PRs AFAIK). No viable forks of autoform that I can see at this time.
There are bugs I’ve posted and supplied a repo example of almost a year ago that still are solved.
For example,
Radio controls. A very common form control. Yet validation doesn’t work on them when the form is in ‘update’ mode (only in ‘insert’) mode. This has been an issue for over a year. I recently created a new ticket (even though there’s an old one too), just to see if aldeed would respond: https://github.com/aldeed/meteor-autoform/issues/1270 and here’s the old one from almost one year ago: https://github.com/aldeed/meteor-simple-schema/issues/182
This is why it’s really hard NOT to search for alternatives. But there seem to be none that come close at the moment.
For me, I’ve tried to just do forms without any aldeed tools. But it’s really hard, mainly because of reactive validation and the schema.
If there was an easier way to do reactive validation on any number of controls, I think I’d make jump to doing ‘forms’ completely by hand.
Aldeed is a great guy that has help many, I just hope that he will truly ‘release’ this product to the community to keep it going in his absence.
Right, this sounds like a good compromise to me also. I might try this someday. But my main problem is the validation in simple schema custom function not working on select-radio controls in update mode.
Have you tried https://github.com/usefulio/forms ? I saw a devshop presentation on it, looked pretty good
AutoForm “state” is a recurrent pattern or anti-pattern. We have the best form package that any platform can dream but we are bound to his bad part too. Is like Pink Floyd, the best to hear but always going a hear the same songs.
I think Eric, @aldeed, have the same problems than us. Wainting for Blaze future or not future, a lack of resource to mantain and support packages.
Maybe we have to wait a little for MDG official Blaze complains and then decide to start from scratch with other form package.
If you considered React then there is a nice package: formsy-react. I have had good experience with it so far. It’s very flexible, does validation and allows you to speed things up significantly
In my early AutoForm I fill some issues too and got a little attention. One days late not found any package form that can replace this. So I decide to fix AutoForm by my self and contrib to this project, my bussiness depend on this package I have to learn how it works. I fill some PR´s, Eric take her time to review the code and merge it. Hi is careful with any line of code.
I you see some of the old not merge PR, you gone see that don´t fix the issue at all or have other kind of problems.
AutoForm is a big and complex package and have many Meteor and JS incomers, like me :-). You cannot simple do ‘meteor add aldeed:autoform’ like many others package. This must be study at code level, cloning it on your local project and put some console.log here and there and see what happend.
I invite to every one reading this that clone AutoForm and contrib with anything, even a README typo fix. This package deserves it.
By the way there have been two serious updates to templates:forms within the past few days and almost all of the open feature requests have been addressed. It is still being maintained, I just took a break for a few months. I merged in any PRs during that time though. Everyone is welcome to contribute.
@jamz what’s your plan for future? Now when Blaze is practically deprecated we well need to move into React or React based ‘Blaze2’. Any ideas toward this direction.
The same question is for @aldeed.
In January we’re planning to start new product which will be form heavy. We’re thinking to do it with React and SimpleForm or Astronomy, depending on form library we will choose for it. Any suggestion what to choose? Or we will have to start our own form package?
In terms of templates:forms I did small package which is adding SemanticUI for it. Planned to release when add more widgets (which I don’t need for my project), but feel that now all templates:forms and my packages is already deprecated. Does anybody needs SemanticUI for templates:forms in current situation with Blaze?
The future of templates:forms involves decoupling it from any specific framework, UI component library, or schema library, and providing some officially supported integrations (React being one of them). Such a version has been in private development for a while now, and I will be releasing it under a different name early next year. The API will feel very similar to what exists now.
I intend to build a community where people can share form components, and also to offer some official component sets that span the supported integrations. These ideas were initially discussed almost a year ago in GitHub issues that are still open.
In summary, big things should be happening with this library in the near future. Anybody who wants to get involved or be notified of progress can reach out to me directly.
@aldeed Any word on the
And thanks to all for bearing with me on issue backlog. Will be getting caught up soon.
Not much has happened (or catching up) since 6 months. Its clear autoform is a side project which doesn’t necessarily pay the bills, but at least leaving it hanging as it is now doesn’t benefit anyone on the other hand.
Pull requests aren’t accepted, neither issues resolved. Well…