Closed JFoxUK closed 1 year ago
Instead of a time limit, I'd suggest keeping the resource limit, ie, "dyno hours". Many of us (including me) used to have some sort of personal project that we just wanted to host somewhere and not really care about good uptime, so even if we get a little fewer dyno hours, I guess it'd still be good.
Without the ability for a free trial or similar, talented individuals will simply learn alternative technologies that do offer this
Or just leave the features of the old plan free but at an affordable price, because I'm not ready to put 7$/month in a dyno
Decision has obviously already been made. So I'm basically crossing out a brand off my list. Tainted names won't make it far in the post- digitalization world. Ain't it embarrassing to claim something in the spirit of we can no longer ensure prevention against malicious players?! Totally disappointed.
Also some of us feel offended of accusations of "abuse of the Heroku free product plans". I believe we were all developers, experimenters, innovators.
@ASayre @capeterson how often will these issues/contributions get reviewed and will we hear Heroku's feedback on this one directly here?
Another idea would be to offer a dyno priced similarly to the hobby tier ($7/month for full uptime) but with the auto-off functionality of the current free tier. So if it is only being used a few hours a month for a proof-of-concept/class project/simple staging, the charge to users would be very low but Heroku could still cover its own expenses and prevent abuse of the free tier (maybe even make it a slightly higher rate/hour vs the hobby tier to account for the on-demand nature of the tier).
This is really bad timing for me. I just finished a Discord bot and need to host it. My default choice has always been Heroku, but now it's difficult to find a hosting provider who is as flexible as Heroku in terms of what you can host. The free tier needs to return.
@wsoeltz this can be done with autoidle addon. The problem is the extra $9/month for the db.
For right this moment, #55 is where we're focusing effort to keep a lower cost option for this kind of scenario, but I have to agree with the potential for time-limited "scratch"/"trial" environments to be extremely useful here. It's admittedly vague for right now as we work with legal, compliance, and other folks to figure out what options are actually commercially viable, but it's a major focus to lock in decisions there ASAP.
Some kind of dyno-hours pack is something I'm cautiously optimistic we can bring based on the current state of conversations, along with some way to keep auto-idle of dynos out of the box (as @wsoeltz suggests).
@JFoxUK - I'll try and keep this issue updated, but #55 is going to likely be more authoritative as we lock in decisions. I'm going to keep them distinct since this goes above and beyond what's in scope for that more immediate move though, and I really want to keep longer term feedback going even once we have some more immediate answers there.
We have now announced our new low cost price plans on the Heroku Blog https://blog.heroku.com/new-low-cost-plans. Meanwhile I also agree we need a means to try and explore/play with Heroku - related to this have - https://github.com/heroku/roadmap/issues/56 - we should consider merging these ideas.
We hear you on where we can make the Heroku platform more accessible for folks who want to learn, experiment and prototype with our services.
Since the original posting, Heroku has introduced #55 to provide Eco dynos, #126 for OSS projects, and #45 for Students.
There's still plenty of work for us to do, of course. #18 for Non-Profits, #58 for Educators, and #56 for broad experimentation are still very much on our radar and top of mind for us.
In order to help keep us organized, this specific #51 issue will be closed, and the feedback merged into the aforementioned. We're going to be using the aforementioned issues for tracking and calls for feedback and discussion. And if there are other areas of focus that we are may have missed, we want to hear that feedback too.
Feedback is a gift - and rest assured that we do hear you and continue to make strides to make the platform better each day!
Required Terms
What service(s) is this request for?
All developers who previously used this platform to learn and then master this platform.
Tell us about what you're trying to solve. What challenges are you facing?
The removal of the free hobbyist tier is a disaster. Heroku was used by many to learn how to develop, practice news skills and build proof of concepts before embarking on the real Heroku journey - building paid for enterprise apps.
Much like Salesforce Dev orgs - my proposal is that you bring a fashion of the hobbyist tier, in the style of a trail org - whereby it is only valid for 'x' days before expiring. This means that developers can build their app, practice, test, learn but it has an expiry and can not be used endlessly for malicious purposes. The timer 'kicks in' once the app is switched on and is live, and has a short life of only a couple of days. This could either be a timer which decreases only while the app is active meaning you can turn it on, use it and then turn it off, conserving a 'X' hour time limit for the app OR a simple 'Once activated, the time starts indefinitely until the clock experies and the app is rendered permanently deactivated'.
This could also be enough for those who used Heroku free tier for staging and UAT environments as those use cases are often short lived too.
Here, with this approach, you still allow people to learn, thest the platform, build out their proof of concepts and test their code, without allowing the app to live forever and consume an infinite amount of resource over a long period of time.
Added to this, for security, in case of malicious use in the short life of which the trail environment is live, verification of ownership, using an ID&V provider such as Onfido or Jumio to track who is using the trail environment or even a verified card holder (much like AWS do for their free tier) - though reassuring the user that they will not be charged as Heroku can enforce a maximum usage during the trial period to ensure no extraordinary resources are used.
Those who read this proposal, please react with a 👍 to support it.