There's been a lot of talk here about React and its impact on Meteor/Blaze--what about RethinkDB and GraphQL?? In other words, following React farther down what is clearly their intended stack?

What about making a strong (public) long bet on the future of web development through these adjacent tools and MDG rallying behind it? The result would be strengthening challenging undertaking, specifically in the coordination of the teams behind the other such projects, i.e. getting as much help as we can from them to make Meteor something they want to integrate with (and early on). Why not break the mold, take a stand, make a hard break from their past strategy, and embrace would-be competing products: Webpack? React? GraphQL? RethinkDB? Their might be a stack already evolving here that can integrate quite nicely with Meteor if we get appropriate buy-in from these teams.

To make clear what I'm saying, here's an example: the RethinkDB team right now is pondering deep integration of GraphQL (vs a higher level less-cool abstraction). The deep integration would come with so SO many benefits for developers using both tools. The Meteor community could help get them to make this difficult decision if we publicly make the decision to support it once built. Is that outlandish?

It seems right now the race to the Ultimate platform for javascript development is on the horizon and there are some clear winners--one MDG clearly and publicly has just realized: React. That's what most the hubub in this conversation is all about. Well, React is on a path to be tightly coupled to GraphQL. RethinkDB is thinking about getting a piece of the action--though I'm worried they fear making that decision, and ultimately don't make it.

We are all in search of the Ultimate Stack--that's why we fell in love with Meteor. And it still is the only full stack solution for us javascript developers. So from where I'm looking at, React, RethinkDB, and GraphQL (and of course React Native, and likely Webpack) could be a major boon to Meteor as a whole if tightly integrated. For some of these (most noteably React Native) this is a no-brainer. It's just gonna have to be done, and will be done.

For the ones I'm brining up (RethinkDB and GraphQL), the decision is less obvious. It's too early. That could be a blessing instead of a course. You have to think these companies/organizations/teams are too going through such decision processes. Meteor clearly has been struggling with developer resources, and they will continue to do so as long as they try to be everything to everyone. I'd much prefer some strong decisions on developing a more focused stack, even if it means pissing people off in the meantime. For example, kinda letting Blaze go (keep it as it is, and continue with bug fixes, but don't keep upgrading it; promote it as a way for new developers to get started), and focus all their energy on the React world. They clearly have already started on this path--I mean they now have React integration tutorials on their site for goodness sake; that must say something. So why not double down on that decision and follow React farther down the stack. In my gut I feel that is the right decision; and keep in mind I'm someone who still uses Blaze and am just as unhappy about this React situation as many of you.

We need to use “The React Situation” to our advantage wholeheartedly, rather than fight it--that's my main message. Something very smart and with more foresight than normal can be done here.

The RethinkDB stuff I'm mentioning might be pie in the sky, but that's only one part. With some foresight, some boldness, Apple-esque decision making (i.e. making the hard decisions to focus just on a few things) and coordination with other said teams, MDG could let everyone know what their plan is ahead of time and prepare users for these features/integrations, but more importantly inspire developers from other teams/organizations to make sure it happens efficiently. I believe Meteor is the best platform out for full stack javascript development, and the funny thing is it's likely to stay that way given how time-consuming we've seen it is to build such a thing. It's only really competitor is what Facebook has brewing. Not to phrase it as a false decision, but basically we have the choice to embrace what Facebook is doing with the “React Stack” or be killed by it. That stack is already going downstream and will continue to do so.

Meteor could fill in a lot of missing parts (particularly server side, DDP, etc) it will take Facebook years to replicate (during which Meteor would otherwise endure a slow and painful death).

Meteor should become an “official friend” of the “React Stack.” Meteor's always been about various tight couplings to be a complete stack tool. This wouldn't be that much of a departure from their current development process, besides the fact that they would now be married to core components run by outside teams/developers for once. I'd prefer to see Meteor double down on this tightly coupled way of doing things rather than continue to fragment out into this sort of power grid it's becoming that will support various flavors of SQL, various client side frameworks like React and Angular, different packaging systems like Webpack, etc.

It's been absolutely amazing the # of integrations with Meteor has come about. It's been absolutely amazing that such a full stack platform could support such a thing. That level of flexibility perhaps could only ever be possible thanks to the browser, javascript and Node. But anything that gets to big can also become a problem. That level of flexibility absolutely kills MDG's focus, the meteor community's focus, and definitely that of us developers/users.

The great thing about Meteor starting out was that there was very few decisions to make, besides the business logic of your actual app. Now we have so much power, but a disproportionate # of more decisions to make. If I was in the minority, it would just be one developer begging for his needs to be focused on. But I think I fit the profile of 80% of Meteor developers. For example, 80% of us probably could go without SQL. 80% of us would likely be very satisfied with super pro React and React Native integration (once we got over the pains involved with letting go of Blaze). So why not take it a step further and get married to RethinkDB and GraphQL. It's likely only a matter of time before Facebook decides to make their own server side aspect and packaging system. You think they are just going to stop with GraphQL? No, they won't. But I also doubt the differentiation they come up with in those departments will be that worth wild. I.e. they will likely be something comparable to Meteor. We could get to the point where we've been looking for a lot sooner. The corner stones--for the “Connected Client”--are going to be RethinkDB and React Native. We should follow their lead and get behind them.