Hi all - I’ve got a production version 1.2.1 app using a form builder that is no longer supported, mesosphere for validation, collectionFS for uploads and iron:router.
I have started rolling autoform/simpleschema/collection2 into the app, but have noticed that these may also be “becoming old”.
What seems to be the best solutions for these lately?
Iron -> Flow?
CollectionFS(gridfs) -> Slingshot?
MyOldForms -> ??
If you’re going to stick with blaze then autoform/simple-schema/collection2, and FlowRouter are the way to go… FlowRouter may be in a bit of flux right now with arunoda announcing he would no longer be as active in the commnunity, but it’s still a great router and I think that some level of maintenance will still be afforded to the package. It may even be taken over by another member of the community. As far as autoform, simple-schema, and collection2, as the author of mesosphere, I personally recommend the switch. A small caveat though is that simple-schema has moved to npm as simpl-schema and collection2 package is deprecated in favor of collection2-core which uses simpl-schema.
As far as uploads go, I’ve switched to using cloudinary. lepozepo:cloudinary, and cosio55:autoform-cloudinary are a couple packages that might be of interest if you choose this route.
Basically Meteor is a mess, lots of libs are dead, and there is still a lot of churn going on. Seems like a nightmare for long term products built in Meteor.
I used to use Meteor a lot, but now, I’ve completely moved away and using other platforms ( which is actually working out a lot lot nicer ). But these legacy meteor apps I have are now a bit of a nightmare and still there doesn’t seem to be a good future proofing other than entirely embrace react … but if you do that, you don’t really need meteor. Also the more time I spend with Mongo, less I like it, so for the most I’ve abandoned Mongo as well.
The one thing Meteor does nice, that no one else seems to do nice is a baked in user / authentication system.
So my recommendation is to port off meteor if you can. Otherwise go react if you can, otherwise, delay the pain and muddle through with whatever is “current”
I really think that the state of meteor we have right now (including simpleschema & autoform) is fantastic and worth keeping alive.
So if you decide to go for blaze, be prepared to do some bugfixes on your own for existing packages, including creating pull-requests and forks, which in the end is what the open-source idea is all about. We are using free tools here, right?