Where's the vision gone? Where ist the soul gone?

Planet Earth
22

Actually I wanted to write on this topic but claiming completely the opposite… :wink: Namely why I do love Meteor and work with it over some years now. Maybe it seem the community is not that vibrant as it was before simply because many early bugs have been fixed and now we have a rather stable framework? For example I see no reason to ask any questions here because most of them have been already answered and what is left is just to sit down and do the work? Huh? :slight_smile:

And btw, Blaze rocks and until Vue is mature and integrated enough, there is no need to switch to React or any other gimmick… Peace. :v:

10 Likes
23

I think like that:

I wish, some of the origin magic beyond business and money comes back and they surprise us again with some really new and innovative meteor features that blow our mind, like it did in the beginning. That’s my wish for meteor. Let’s say for Christmas.

That does not mean, that they don’t do a good or great job. But that does mean, that i think, there could be more. Not more work.
I mean again more vision. Because in my eyes, which off course is only my subjective way to see it, they sacrificed a good part of it and also only in my eyes they sell themselfes somehow under their true worth.

They still have shown how big they can think. this guys from MDG are really great in my eyes in what they are capable of.

And thats how i now let it go.

2 Likes
24

If you don’t like the changes of MDG, nothing prevents you to stay stuck in the past by not upgrading your meteor version.
As a hobbyist, you are building toy apps that doesn’t have the same needs as people that do that for a living, and need more features.

The bigger picture here is that the learning curve of the early versions is so low that it enabled non engineers like you to build stuff as hobby. But enginners started to use it for the speed of development(indeed, it is ultra fast to prototype) and we need more from the framework, as we are building companies around it.

So as a Hobbyist, just keep using the early versions of meteor.
As a CEO, you should be happy that your company is able to build products based on open source(hear free) tech. If you are not knowledgeable enough to understand the tech your team is using, your role is to find à CTO who will.
If your dev team cant adapt to the evolution of the tech they are using, fire them and build a new team of qualified engineers that understand that the web is constently evolving and they need to learn and adapt accordingly.

Last but it least, if you are disappointed with the framework and think you can do better, nothing prevent you from building your own framework with your own time and money.

Cheers,

A Belgian.

1 Like
25

I agree with the original poster and can see what he is trying to say.

The problem is that Meteor’s philosophy, which is to take away the complications in building modern web apps and provide an integrated solution for everything, is great but not what hardcore developers seem to want. 2015 (and maybe 2016 too) was the year of ‘JS fatigue’ on which there are many articles and posts. Things move fast.

To attract and retain developer mindshare, Meteor had to change - e.g. add npm support, which killed Atmosphere, which was a great pro of Meteor since it had packages designed to work together. Soon we’ll have npm installable meteor and lose the pros of the meteor build chain. Same thing happened with supporting other view layers.

From day one there were people who demanded that Meteor be a very small lib instead of a framework and that it should work with other databases, other libs etc. Which just cannot be done losing the core advantages of a full framework. The same thing is happening to Angular, Ember, Auerelia etc. You buy in to the framework or build your own.

Right now is state of flux, lots of moving parts. Apollo is a huge effort. Earlier this year lot of effort went into Galaxy, and the free meteor hosting ended because it cost them too much.

But the core differentiator in Meteor has always been bringing together complex tech in an easy to use package. MDG has enough money I think, and smart people. I hope that with 1.5/1.6, we can get back a new Meteor based on GraphQL with the same features like realtime data, and tools to manage it.

I also think Blaze gets a lot of flack but is actually great and nothing is forcing people to use React etc. You can ignore ES6, ignore React, and switch to those things slowly if you want to, they are better but not forced upon you like most other stacks.

Setting up a new web app with React is a nightmare compared to Meteor. So even after all that has changed, Meteor still remains simple but lacks some of the hot buzzwords (like HMR). People in the JS community love to jump ship to the latest/greatest with no regard to compatibility, just look at what the Angular and React Router teams did. This is why its good when companies like Ember/MDG actually care about supporting older versions and not breaking apps.

Summary - nothing really is broken, don’t use new features till you want to. I hope Apollo actually makes Meteor better and I also expect other nice things.

4 Likes
26

I agree with Lukas. Whenever people get defensive in a forum like this, it is not a good sign. User feedback is exactly what the word says: user feedback. Negative feedback is often more valuable than praise, especially if taken seriously and without pushing back on the person asking the questions.

It is my sense that MDG would be well advised to clarify to the community their definition of a target user. Like with every product, you cannot do everything for everyone. So what problem are you trying to solve, assuming this is not just a technical exercise.

And what’s up with the increasing number of deprecated, abandoned, or under-maintained pieces? I mean perception is reality, and especially if you think this perception is factually incorrect, then an effort should be made to correct the perception.

2 Likes
27

Why are you talking down to Lukas? What makes you believe it is a hobby and what is so bad with hobbies? This is just not useful.

1 Like
28

There was a lot of language from MDG early on about democratizing web development, i.e. making it easier for beginners. From what I’ve seen this idealistic streak still exists, but not all good ideas make money, and for various reasons that have been discussed at length (evolving JS ecosystem and business requirements), MDG chose to prioritize professional developers over beginners (see Geoff Schmidt talk about this choice in May 2016 https://youtu.be/SFOnKtaeps0?t=11m4s).

Perhaps at this point in the JS ecosystem, professional developers are more interested in Apollo than Meteor.

I’d say that just means there’s an opportunity right now for someone with vision to fill that gap — the road between beginner and professional, from startup to scale — with a great developer experience, the kind that we once got from Meteor. Maybe MDG will come back and fill the gap at some point, maybe not.

1 Like
29

One problem is that the top end - “professional” and “scale” is very diverse, and getting more complex every year. So if most beginners need essentially the same thing - a way to put HTML on the page - advanced users often have very specific and different requirements, which is why a lot of the more “professional” web dev tools focus on flexibility and configuration over convenience. Meteor is simple because it makes a lot of decisions for you, but if those decisions don’t work down the road then it’s easy to regret the very thing that made it easy to get started in the first place.

4 Likes
30

If you see MDG as a whole including Apollo I would say the vision is not gone, I would say they are highly successful with Apollo already and I’m guessing they’ll make decent revenue with Optics. In terms of Meteor, I am highly concerned, on the brink of deciding whether I should abandon Meteor.

Especially now that I’m looking into GraphQL/Apollo and doing stuff with the node.js tutorials I’m really asking myself if I really need Meteor. Yep, I get Meteor gives you more out of the box but node.js really doesn’t seem super rocket science (I know, the devil is in the details). What I’m mostly concerned is not the lack of commits, one small decision I disagree with. What scares me is the lack of transparency. Your roadmap is ancient at this point and it’s absolutely unforseeable what will happen to Meteor in 2017 aside from the fact that we’ll have an Apollo release in the near term future (It doesn’t help that most of your learning material doesn’t focus on Meteor :confused: ).

I think what you’ve done with Apollo is phenomenal, I’ve just started diving into it in the past few days and I love it. I would definitely pay for Optics in the future! However, with Meteor I’m uncertain, everything’s just very uncertain. Meteor itself has lack of transparency, how to host Meteor aside from Galaxy always seems a “figure it out yourself” kind of thing, Galaxy’s docs are pretty bad and there’s no information what’ll be upcoming or Galaxy as well. If MDG wants to drive adoption of Meteor & especially Galaxy, they really need to step up their game. I’ll evaluate in a few weeks if Galaxy + Optics is a good combo though.

@sashko I really hope you’re working on a high level roadmap and it’s announced soon. Nobody likes to make such a big investment in a framework without knowing what the future will bring. Thanks for your ongoing great work :slight_smile:

1 Like
31

Do you have an example of a project which has done a good job of presenting a long-term roadmap, and then followed through with it? Would be interesting to take a look and maybe there’s a lot we can learn there.

1 Like
32

You should be able to give a 6 months roadmap at least. Yea, I get it, priorities change and you need to be agile but let’s look at your current roadmap.

MongoDB updates.
Done

Support for Node 4 and beyond
Done

Full transition to npm
Nobody knows what’s going on, except that it’s delayed

Testing updates
No idea what’ll happen here

Project Governance/Community Contribution
This one’s good

Data (SQL, REST, Performance)
In the works but when will it finally come out?

Out of the 4 points on your roadmap that are not yet done or I’m not sure they are done, I only know of one that’ll happen anytime soon and that one’s Apollo. And even there I have no idea when it’ll happen. But that’s not the main issue though. My thought process is this: “Ok, npm is delayed for unforseeable future, Apollo is probably coming soon but still not there yet and I have no idea what else will come. Will I ever get my code splitting? What are best practices to connect my databases to Apollo? What do I need to do now and how will my decisions in Meteor today affect my workflows in 3-6 months from now?”

Your roadmap should be able to answer these questions. If some of your priorities changed like move to npm, that is fine, just reflect it on your roadmap. Nobody’s asking for a 1 year roadmap or even longer, but if you can’t make a rough outline what’ll happen in the next 6 months I’d be worried. Not me, you as an organization should be. I doubt you don’t know what’ll happen 80% accurately in 6 months from now (I’ve messaged you privately before to ask about one of my concerns and you were able to answer honestly). I’m sure you can reflect some of your internal decisions a bit better externally :slight_smile: .

2 Likes
33

Thanks for the detailed list - I wasn’t trying to call you out but honestly asking for an example of a good roadmap so that we can look at how other people with similar size projects do it.

34

Well, I honestly don’t have a good list. It’s always hard to show a sample of something that’s external I think.

Polymer is good. Why? This roadmap is very similar to yours, doesn’t change much BUT they add some updates to their roadmap, like “recently shipped” as part of a roadmap. https://github.com/Polymer/project/blob/master/Roadmap.md


This one is also nice but too fine-grained maybe and I think you tried that before and didn’t work for you.

35

As you Sashko are posting here, i decided to come back and post a thing, that for me really shows, what i miss.

But let me first give you my respect, for what you are doing from the technical standpoint of view. I think it’s extraordinary good.

But then there is another side with MDG and also i start to see it in Apollo.
As i heard the speak of Geoff Schmidt, where he explained about you asked, we delivered, i had a lot of things going on in me.

He said, Meteor would become more Mainstream and he said on what he will focus because of the feedbacks and so on.

But he was not speaking any more about His own Vision anymore.

But nobody can be mainstream focused and stay strong with a vision being different.
Innovation is to do things, before others know, that this visions will become mainstream.

You guys could have focused on the companies, that need development instead the development itself.

Instead of finding the meteor way of giving a tribute to the mainstream, it try 's now to implement the whole lot of things from there into the Meteor Ecosystem. But at the same time killing many of the really good things, that where way better in my eyes, than anybody else’s.

I think, it was to early, to let the idea down of being different.
I think it would have been better, to find ways to diversify the goodness of meteor with a more flexible backend. I think, you let your Baby(Meteor) down, before it was an adult and able to stand alone in this world.

And i think, thats why it won’t survive on the long term. It was your Baby. I think it could need a little bit more love. It’s not yet ready to be everybody’s baby.

I really can understand’ whats going on from a financial point of view. But not from the view of the initial idea and vision, which was, at least in my eyes, much bigger.

I apologize. Wether do i not respect the whole amount of effort and work, that you are putting into all this, nor do i think, i am right.
It’s really only the way i see it and nothing more.

I would love, to see meteor being a great way, to be a future way of development instead of being more mainstream.

That brings it to the point more or less.

2 Likes
36

This situation with Meteor/Blaze reminds me of how back in the day, Visual Basic 6.0 ruled the enterprise desktop development stack.
It was easy enough to get you started with developing Windows apps, yet powerful enough to work with all the MS back-end services.
Then, under pressure from the emerging Java standard, Microsoft decided to recreate it’s development stack with .NET, thus making VB.NET unrecognizable in comparison to VB.6. Most developers, including myself, bit the bullet completely and migrated to C#, leaving Visual Basic behind.
Had those been the times of Open Source Microsoft, maybe the community would steer things differently with regards to VB, who knows.

Additionally, I noticed that Microsoft’s recent move to .NET core (which I bet is Node.js inspired) is alienating a lot of ASP.NET devs who are fed up with Microsoft moving away from a technology just as it seems to have matured.

Personally, I have no trouble with Meteor+React, but that’s only because my day job required me to master React on a non-meteor project.

2 Likes
37

There was a saying about Microsoft vs Linux in the 1990s :

Microsoft makes the easy things easier and the hard things impossible. Linux makes the easy things hard and anything possible.

You could say that Meteor’s last 18 months has been a transition from a Microsoft world view to a Linux world view.

11 Likes
38

I don’t think the two are mutually exclusive. In an ideal API / framework the complexity of the task should be reflected in the complexity of its implementation.

Some “drift” towards higher complexity is to be expected with the introduction of new features. As long as the complexity is offset by the value the framework provides, it’s ok. Judging by the activity in the community, MDG isn’t doing so bad.

Disclosing a roadmap is IMO the ideal recipe for disappointment. If a roadmap is released, and your project really has a critical dependency on one of its items, be prepared to fund it.

I’m more interested in ideology, like “we are committed to building the easiest to use JS framework out there”. Living up to the ideology, plus some process to give the community some influence over general technical direction is about all you can expect for free.

39

How about debugging server code? I haven’t been able to use the node inspector for months.

1 Like
40

Yes, exactly. Standing still is not the option that so many of the whiners seem to be yearning for.

Red Queen Running

‘A slow sort of country!’ said the Queen. ‘Now, here, you see, it takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!’

Yes and no. A secondary benefit of a road map is clarification of where work focus will NOT be.

41

I love these discussions - I just wish they weren’t all started by threads with such ridiculous titles. Is there anything we can do about it?

4 Likes