@benstr This might warrant a separate topic, but I’m curious where you’re seeing evidence of this. I’m releasing a SaaS this summer, and was planning on a very limited free tier. But if that’s a failing model, I’d love to hear why.
I’m still interested in this idea of “spot containers” I mentioned in a post awhile back. It wasn’t what our production users asked for: they were very clear that they wanted access to the Galaxy Pro features, so we prioritized that work. But in the future, I think it’d be quite interesting to be able to offer “cost-optimized” containers in cases where it’s okay for them to recycle and move to other hosts more frequently.
It won’t happen overnight though. The engineering considerations in building such stuff are not trivial.
Freemium is good but too much can harm the company. With MDG offering free engineering for Meteor and Apollo is huge plus the burden of free hosting could be too much (which it sounds like it was).
Here are some articles that will help you with your own startup.
https://baremetrics.com/blog/freemium-saas-implode
http://blog.trak.io/freemium-vs-free-why-we-ditched-our-free-plan/
This is my understanding as well. Although it would be great to have MDG confirm this. If so, it would be worth retroactively updating the blog post and thread topic.
I don’t thing it’s a good solution. Even if you’ll get some free credits for Galaxy, where will you host database?
Memo: Regarding Meteor Free Hosting
To: Meteor Developers Everywhere
From: Fake Geoff Schmidt, CEO MDG
I screwed up and I want to apologize to all our passionate developers, partners and customers. We obviously did not fully think through the impact of our recent announcement on ending support for Meteor.com free hosting.
Therefore, we have modified our plans as follows:
-
We will keep free hosting on Meteor.com available “as is” through June 10, 2016. This will give everyone 90 days to transition their hosted apps and websites to either Galaxy or another hosting provider of your choice. See our guide on how to transition your application and MongoDB.
-
We will offer a limited, free hosting Galaxy solution for all our key developer partners and teaching partners. Details on this new plan will be communicated to all of you within two weeks. The goal of this plan is to offer all our package developers a place to freely host their example Meteor apps and package sites and to make it easy for people spreading the word on the benefits of developing modern apps with Meteor through teaching, meetups and free tutorials.
In the meantime, you can send us an email to partnerhosting@meteor.com with links to your Meteor.com sites that you think should be eligible for this new free Galaxy hosting.
- We will continue to investigate making a non-restricted, limited, free tier available on Galaxy. I can not promise a timeframe for this solution, but we will do our best to honor our original communications in this regard.
Please feel free to contact me directly with any questions or concerns at fakegeoffschmidt@meteor.com and I hope this gets us back on track with supporting our most important constituency, all of you as our passionate developers.
Humbly yours,
Fake Geoff Schmidt, CEO MDG
I was under the impression for some reason that Developer edition uses spot instances. Is that not the case?
I agree. Packages demos should have a dedicated free tier, only for them. Indeed they benefit all.
They got rid of concept of having limited functionality spot instances "… we’ve unified the flexibility of utility pricing with every Galaxy feature into a single product that scales with your app and business needs.
I just leave it here.
Quick and simple redirect guide. Found somewhere. For me it’s handy…
1). $ meteor create redirect-app
2). remove all files, leave just .meteor
folder
3). create server.js
file
4). paste the code:
REDIRECT_URL = 'http://i-want-my-app-here-now.com';
WebApp.connectHandlers
.use(function(req, res, next) {
var location = REDIRECT_URL;
res.writeHead(301, {'Location': location});
res.end();
});
5). $ meteor deploy my-current-app.meteor.com
I do welcome paid service because that means more ‘for-profit’ Meteor projects will be born.
However, as I am only checking out Meteor I am already stuck at MongoDB deployment stage.
- I have subscribed for Galaxy.
- I have subscribed for MLab.
Now MLab asks me: Provide URI for a cloud-accessible MongoDB
I really expected it to be that simple as:
meteor deploy myproject
(with config pointing to mLab)
or
meteor deploy myproject --dataprovider=mLab
As I suspect it is not a matter of weeks of months to make that easy it would be great to have Meteor To-Do Tutorial have clear Step 13 - Deployment. Meteor Guide talks about Deployment but it does not provide example of pushing database to cloud and connecting it with Galaxy either.
At the moment
https://galaxy.meteor.com/help/deploying-to-galaxy does not work for Windows and meteortips.com and other tutorials like http://meteortips.com/deployment-tutorial/galaxy/ are empty
Hopefully educational resources will catch up with the pace of change soon.
Update: I have taken Windows advice from Galaxy deployment DEPLOY_HOSTNAME is not recognized but my login/pass do not work. Maybe it is related to the fact that I am registered as part of organization?
If “it’s never go away” refers to deploy
and not the free tier, then this is about as poorly worded as the Second Amendment. But surely I’m not misreading the part in blue too?
At this point I think we can all agree that MDG pulled a 180 on this one. Is it a permanent 180, who knows. Meteor is really tight lipped about their plans, especially infrastructure plans. While I thoroughly support their decision because I think it is what is best for everyone, they really went about it wrong!
I think we’ve seen in the last few months, that MDG has serious communication problems, an example we find in the Blaze/React thread where we’ve waited a long time for an official statement. I’ve just hoped that MDG learned from it and changed it’s communication level, but as we see, this didn’t happen (yet?).
I’m not using the free meteor service, but it is distressing to give all users only 2 weeks. Normally all people in management positions should know, that 2 weeks are ridiculous for a professional company which targets are professionals. What will happen if Galaxy isn’t profitable as expected? Will such users also get only 2 weeks until the service will be closed? If you say no, I think we can’t trust on that answer, because nobody knows, how
reliable MDGs statements are…
Completely understandable. When they suggested “it’s never going away” I’m pretty sure they didn’t anticipate such a huge drastic uptake in usage.
If they’re wasting a lot of money on it, for no reason other than showboating, it’s a good thing they removed it. Focus that money elsewhere and improve Meteor itself. We can pay for hosting ourselves.
I’ll clarify… “MDG doesn’t owe me anything” in that I haven’t paid them for a service or software at any stage. Therefore, I can’t expect too much in return.
I’m in the same position as you, I have taken more from the community than I have given. That’s partly a time constraint, partly because I found keeping pace and providing any real value challenging enough as to not bother.
I whole heartedly agree with your second point. That, is my fundamental upset.
I certainly don’t write that much here in the forums, but this time I feel I would like to highlight a few things about the latests MDG announments(including this one). All my opinions come from the perspective of a developer who understands what is to make a business sustainable and also from the perspective of someone who loves all the technology that MDG offers for free:
- Besides how good you are in bringing really cool tech to the table, I feel that your communication skills are somehow damaging the public opinion in some posts, and this has been spotted here in the forums multiple times:
- Ther’s a huge difference in terms of communication depending on the author of the post. Since you have many great and loved communicators at MDG, I would only allow them (one or two would be enough) to do all the important announcements, because they make sure to get to the point without being too pushy or cold, and knowing how to please most of the readers.
- It would always be good to have a small chat talk with a few community contribuitors before throwing this “bomb posts”. As someone already said, MDG is becoming popular for “what are they going to say in their next announcement?”(for good and bad). By talking to an external environment first, I think you would have to worry less about the final impact, and you won’t have to retreat from your original statements (like with Blaze or here giving hosting to well known package creators)
- Regarding the decission of shutting meteor.com apps down, I’m not currently affected, so it’s easier for me to understand and support your decission, while there are many things that in my opinion could have been done in a better way:
- Extend duration to at least summer.
- Give a free period of time(2-3 months) to every package uploader with a limit of 1 container and no database (can be created easily at mLab if needed) which is more or less like a special trial. Once the free trial ends, renew the subscription for another X months based on the package downloads/likes on atmosphere.
- Let X people from the community collaborate with MDG and make proposals for free “forever” test sites (those considered to help hugely in the adoption of the technology), rather than a list that might get spammed and filled with not always useful sites.
- Replace/Update the planned features Trello board. I think by now the Trello board is totally outdated, and I thought it was a terrific idea when it was used. Things like Uservoice or Trello can be really useful to match the needs of the industry/community/MDG etc. So if you are receiving input from many sides, I think all of them deserve to know what is being proposed or suggested at the overall level and your thoughts about it. (This is a simple as having X colums in trello one for community features, another from partners, another for MDG strategic features, … and then move the cards to a “Doing” column and sort them with priority tags). In this way, everyone could see what are you working on, where does the feature request come from and how likely is to be implemented. I don’t mean this board has to be the holy grail, but If people would have seen a card with “migration from free hosting to galaxy plan” in your column with a high priority, It would have helped to avoid such a dramatic reaction with this and many other matters.
P.s. Having said that, and after working with them for a a few weeks(checked many Meteor related things for improvement), I think NodeChef offers a cool hosting solution for those willing to pay less than the Galaxy entry point (and also eliminates the database creation hassle)
I only really used the free tier option for a couple of demo apps but it was definitely one of the things that attracted me to Meteor in the first place.
There’s already a lot of discussion about the short notice, the poor communication, all the great demo package apps that will be lost after this happens so I won’t go into this.
But I must say, I’m really surprised at this decision, simply from a business point of view more than anything else. I mean, obviously not being an employee, I have no idea how much the free hosting costs or what MDG’s future plans are so maybe if I had more knowledge it would make sense.
But as far as I know, MDG makes no money from Meteor. Their entire business model is in Galaxy which means that they’re positioning themselves to be the best company to host Meteor apps.
So with that in mind, why would they not have some sort of free tier, even if it’s a severely restricted one? It’s the perfect way to tie meteor to Galaxy and get every new developer using Galaxy from the get go. Now, every new developer is going to setup free hosting with something like Heroku while they learn and evaluate Meteor.
And then when they discover it’s awesome and start creating real Meteor apps that require professional hosting, well then they’ll probably just upgrade to a paid Heroku plan because it’s what they know and it already works for them.
Maybe the free hosting IS just prohibitively expensive, but I would’ve thought their entire current business model is dependent on it.
I wrote only one app on *.meteor.com - it was http://rtfd.meteor.com but it helped me many times to provide the best learning resources for Meteor framework to young Meteor adepts. It’s sad that it will go away!