but yeah if you notice the year, we were very early adopters. Had our first Meteor app in production on Meteor 0.6. There weren’t of a lot of hosting options at that time, so we had to make due. Later we migrated to Modulus, but that was too unreliable so ended up on the Galaxy “Early Access” Plan.
I understand Meteor has their reasons for shutting down the free deployment tier, but I think it’s very unfortunate as we’ve deployed to the free tier during Meteor demos, and it impresses people. It’s a good selling point. Too bad they can’t find some way to keep it around even if it’s in some more limited form.
Sad to see the free hosting go, but thanks for keeping it up for so long! I was wondering if there’ll be any facility for us to set up redirects on our old *.meteor.com URLs? One of my toy projects (spyfall.meteor.com) ended up with a largish following online and I’d love to be able to make sure people still end up at the right place
WOW! Disgusting news.
I could not expect MDG to burn bridges like that.
Come on, people, you could divide projects into Temp and Examples ones.
You could make a simple mechanisms to detect dead or outdated projects.
But you prefered to rather just wipe everything out.
Not just that, but also gave us only 2 weeks to deal with a broken deal.
Maybe you can reconsider and offer a compromise solution? As in - migrate the free service to Galaxy (as you promised you will), but without Mongo databases, which would be the user’s responsibility from now on. Then, for community project (like package demos, hackathons, etc.) we could use mLab or some other provider with free Sandbox Mongo DBs to host them, and everyone would benefit from this!
Also, I don’t think we need to migrate every app, because many of them were just experiments and don’t need to be migrated.
Here’s my idea how this could be sorted out to satisfy both the community and MDG:
create a free package for Galaxy
limit it to one active app at the same time by default
create a way to promote an account so it doesn’t have above limitation for package authors/Meteor evangelizers
I’d really like to reiterate that I think there is a space for the free tier. Demo apps etc. But this would only work of MDG made it impossible for people to abuse the service and run production apps.
Too many people are sad because their free stuff is getting taken away but I think the real issue has been what that free stuff has been capable of.
Reposition it:
give developers 1 free container on the current system for each package they have uploaded. Potentially limit it to packages with x downloads, stars etc. Limit the max the max they can have. Or have package developers apply.
non package developers can still use the current system but it gets torn down in 4hr, url can’t be reused for 24 hrs. Limit size of app and db.
This will stop abusers, from those attempting to run prod apps even to people trying to use it as a dev server (more than a 4hr uptime and people may use it all day during their work day.) but will also give packages a meteor supported demo space and give presenters, hack spaces, conferences the ability to show off how easy meteor is with meteor deploy.
On the other hand, I don’t see the finances and have no idea how many people are abusing the system with ‘production’ apps. Maybe what I just suggested would make no difference.
I’m that case there making the free tier paid for could work. I’d pay a couple bucks a month to be able to wind up a handful of servers on the current system. I just love the ability to show off Meteor and get people interested.
What do you mean “abusing” ?
Some people could have forgotten they had apps.
Some people are were not aware of apps they don’t currently need.
Some people did not bother to update outdated demos or projects.
But none of them “abused” the system. It was MDG’s promise to provide support for free-deployment environment, MDG did not request Meteor developers to somehow flag their projects or keep them relevant, did not restrict projects to “a kind-of-type” nor “per-user”.
If MDG first provided us a conversation “hey people, we can’t keep up like this, lets change something”, everyone would be alright, production apps would move to paid hosting, demos to separated one, but nope - “we’re done with free tier, get out”.
We all understand the reason. But, personally, I dislike the way they decided to finish it up.
How well you use the opportunity to turn free tiers into paid customers is all about how well you do your marketing homework and analyze the effectiveness of your campaigns to convert them. Obviously it also depends heavily on your product, if it sucks no one will pay for it, no matter how good your marketing conversion process is.
Just saying, for me it’s a missed opportunity but then again I said it before. Good coders aren’t good marketeers. @awatson1978 pointed this flaw in their communication skills in her first post, which was right on the money.
This is not a solution. Yes, you can spin a copy for most popular packages, but what about less popular ones / the ones, where the author didn’t see this post / new packages / useful demos of package combinations?
For me, this was the best part of meteor deploy.
Meteor will be soon just a set of npm packages with no benefits or differences from just another set of packages from npm, named whatever you like.
Galaxy is a small bonus and I think later it will be available for all js apps.
What if every package on Atmosphere would be entitled to a some Galaxy credit which would allow you to run one instance there. So effectivelly one instance per package.
Just two week notice… on official website was “meteor deploy (…) It’s never going away.” How about Galaxy, should we trust it’s enough to build business on top of it?
A lot of meteor.com websites are used to demo packages/tutorials/help with education. Shutting this down will hurt the community a lot. Not sure if it’s worth investing in meetups, conferences at the same time…
I saw a lot of AdSense ads about Meteor. Not sure if any developer pick framework based on ads, but likely it costs non-trivial amount of money (out of VC investment). Does it really have higher Return on Investment than *.meteor.com?
In general though a lot of work on Meteor goes to features that aims to pleased developers (Native NPM, New JavaScript), very few to make it possible to use it commercially. I believe the biggest missed opportunity is that there is not an easy way to start using Meteor in existing product/application. Angular/React/… can be added incrementally why Meteor is full-stack. Though it should be a pattern to add widgets (like chat) to existing applications.
Tl;dr: Build things that matter, don’t just live in startup bubble.
This is my definition of abusing based on how I view the free hosting’s use case. There weren’t any official limits or before but I think there is need for some.
How about if Meteor gives free hosting, at least to atmosphere packages, but adds a “Powered by Galaxy” logo across them all. Like Netlify does on it’s free tier?
Then every free meteor.com app is an advertisement for Galaxy!
Four days ago, @debergalis wrote about a “big improvement !” for the free tier that had reliability problems.
At the time, i didn’t guess that by “big improvement !” he was thinking shutting it down…
It’s very strange how intelligent people at MDG have been trying very hard for the last year to offend their historical users. It seems every new month brings a new problem.
I guess this is a managerial decision, with the CFO worried about not meeting profitability yet. thus pressuring to cut costs centers.
But 2 weekes of warning, just like with blaze, is a sign that the management doesn’t have a clear vision, and makes hasty decisions to respond to the investors concerns.
The thing is that you are hurting your credibility. AT LEAST give a 6 month warning. I would not be surprised if MDG will turn back on their decision.
On the other hand who think that free doesn’t convert I think it’s making a mistake. 1% is low? I’ve worked with companies that had a 0.05% conversion rate and they were hugely profitable. It’s your duty to raise that conversion rate, but consider that in the business it never gets higher than 5%, so you are already doing an excellent job.
Maybe you are saying that not enough people buy books, but that’s the market that is too shallow, not the freemium model that doesn’t work. Every book on Meteor has some free content available, I don’t think that all the publishers are wrong.
