I think they just decided to focus on what they are in-control of / exceptionally good at.
Providing and maintaining a cloud DB provider / infrastructure isn’t what “Meteor” does, plus there are plenty of alternatives out there that are more established, proven and feature rich… Compose.io & mlab.com
That may well be that case, but he mentioned running a local server, which also wouldn’t have oplog tailing support, as you need a replica set (I believe?). So if it’s just a “simple” collection, then the polling client should be fine.
Also, the reason why I am ready to pay for a push button deployment service like Galaxy is because I am not an IT guy. For instance I don’t even know what is oplog tailing. And I don’t want to know, which is why I pay for Galaxy.
What’s the point in providing me only half of the service ?
The relevant bit is only one environment variable:
Everything else in that document you would need to to regardless, you need to create the DB, setup DB users to provide access to your DB, create the Galaxy application, deploy it, etc. That’s as simple as it gets really.
While I appreciate that you may not be an IT guy, surely the person writing the Meteor application is?
Also, it’s not half of a service, it’s 100% of the service, they provide meteor application hosting, and not all Meteor apps use MongoDB, some use RethinkDB, some use Apollo now and interface with pretty much any data service. As such, if MDG did provide a MongoDB service, it would just add extra load on the servers for a service that may not even be needed.
I don’t remember having to do any DB configuration with the old meteor deployment (free) service.
What makes you believe I am not the person writing the Meteor application ?
You mean there is a service regression (with respect to the old free version) because some guys have been replacing Mongo with other databases before MDG ends working on it (which is in part what Apollo was meant for)
People wanted support for other databases long before Apollo came around. Also, Apollo may not be finished, but it’s certainly more than production-ready now.
You said you weren’t an IT guy, if you’re a programmer, I would say that you are.
That’s true, the original pre-galaxy hosting didn’t need that, but it also didn’t allow you to configure pretty much anything at all, doesn’t support all the new features and also shut-down your application automatically after a period of inactivity.
If you want something more like the old-style Meteor free-tier, maybe look at https://modulus.io/ though it’ll take more configuration.
I tried deploying on Modulus a couple years ago : demeteorizer was a mess and somehow bugged (actually more like always behind the Meteor releases and not always thoroughly tested).
Scalingo is a complete Platform as a Service, which can host your Meteor app and MongoDB database, and has now a long track record of successful Meteor projects.
I ditto the sentiments in this thread. Galaxy’s value proposition is simplicity and they’re dropping the ball on this by not having even a simple MongoDB option.
There’s a steep price to pay for data corruption. Some one steals your phone, you personally charge him with corrupting data and get return on investment like that, rather than the legal term for “damage or inconvenience to lifestyle”.
We deployed a new meteor demo app to Scalingo a couple of days ago. Can’t speak for the long term of course, but deployment and provisioning the database couldn’t have been simpler and so far we’re very happy with the service. Price is right too since we won’t be making any money from this app and it is otherwise dead cost to us.