MediatR is going commercial

[jbogard]

Hi folks,

I'm letting everyone know here that I've decided to take AutoMapper and MediatR commercial. It was a long decision in the making, but one I feel is necessary to both ensure the future sustainability and growth of these projects and especially the communities. You can read the full announcement on my blog:

AutoMapper and MediatR going commercial

I don't have any more details to share, as I just don't know, but I will be as transparent as possible as I figure things out. Thanks everyone to your support over the years!

[JGMelon22]

As long as you do not take the same approach as the FluentAssertion drama, I'm happy, mate. Wish you all the best, Mr Bogard 🖖

[sudheshg]

I wish you all the best @jbogard .

[GabrielHSFerreira]

I believe the MassTransit model, paid support, is a good choice whereas Fluent Assertions strategy can kill the projects. Anyway, good luck Jimmy.

Disclaimer: MassTransit is going commercial too :s

[guibranco]

Totally agree

[byjokese]

Thanks for all those years of amazing work. I hope we end up with a great mix that allows us to support your work while still keeping it a viable solution for the community.

I fully understand the challenge and difficulty of making such a decision with a project of this scale.

Will keep eyes on the final decision and form of the new era.

Have you thought about make an Open Collective project to be able to receive donations, or a similar approach?

[mansellan]

Sorry to see this increasingly happen in the .net OSS world, but I understand the challenge. You've handled this far better than the FluentAssertions debacle, and I wish you all the best.

[altbrot]

Again another library I have to turn my back on. Seems like we are ending up implementing everything ourselves. MediatR was one of my favorite libraries. This is sad.

[nbarnwell]

This is fine. OSS is free as in "free to choose to use something else", not free as in "won't cost me anything". Maybe people don't want to pay with money, but many don't even pay with contributions of their own.

All these people who have been relying on the kindness of strangers for so long, who are now upset that the free ride is over, should feel ashamed. This isn't price gouging like when certain non-SQL database vendors double their license fees and you are forced to pay up or make a costly move to an alternative. The free version people are already relying on today will continue to exist, with the same feature-set. Jimmy's even admitted the support (not that we're entitled to any, really) hasn't been amazing for a while, so it's not like anyone was used to 5-star enterprise-level SLAs.

So there are several options:

1. Maintain a fork. You have been contributing back to the project all along, haven't you? Haven't you?
2. Build your own and maintain that. Maybe open-source it? Maybe have a commercial version one day - who could blame you?
3. Try and find a still-free replacement. But you've been here before, I hope you learned your lesson.
4. Realise the game is up and start paying for the thing you're resting key parts of your business on.

Now, if the price is exorbitant, then option 4 is off the table for smaller shops - that would definitely be a shame and I hope the commercial ownership doesn't result in that. But people deserve to be paid for the time they give up, and surely we were all waiting for the day when the free lunches came to an end. In the meantime, it was worth every penny.

[jcfnomada]

The issue here isn’t the fork; you can bet a strong one will show up. The real problem is that now devs need to be extra careful when starting new projects, because when they search for Mediatr, the commercial version will show up without any clear warning. These kinds of moves are misleading, and GitHub/NuGet should have stricter policies to prevent this.

[jbogard]

NuGet already has policies for this in place - you can force license acceptance. I already see this with MS packages and commercial ones. NServiceBus for example pops up a dialog to download a developer license key, adds errors to the logs etc. No matter what route I go, I would want to do similar - don't affect any runtime behavior but let people know somehow that you need a license.

[nbarnwell]

Besides, checking whether you were about to take a dependency on a commercial version of something was always something you had to do, for every new dependency, so again there's no change here. Being on nuget.org doesn't mean it's free - that's only a convenient delivery mechanism.

[jcfnomada]

Besides, checking whether you were about to take a dependency on a commercial version of something was always something you had to do, for every new dependency, so again there's no change here. Being on nuget.org doesn't mean it's free - that's only a convenient delivery mechanism.

Thank you for that clarification. It's now everything clear.
What kind or argument is that? Did you even read my post? Don't worry because I see @jbogard understood, and it doesn't appear to have your mentality.

[nbarnwell]

You said "devs need to be extra careful" but I'm suggesting they always had to be careful.

[ZeroAurora]

As long as the library is not closed source, I'm happy to accept that :) Paid support is probably a nice thing to have

[nbarnwell]

You could create your own fork right now, then you're safe. Future updates might not be open source but what's there now is still what it is.

[jcfnomada]

Why not keep the current NuGet package open source and release a new one with a different name, like Mediatr.Enterprise? That way, you're not misleading people into upgrading to a paid license by accident.

[jbogard]

I haven't decided anything yet, but will err on the side of "whatever's easiest for everybody involved." There's pros and cons to any approach.

Keeping the current package "open source" would simply mean I abandon it, archive the repo, and mark the packages as deprecated/no longer maintained.

[timctrahan]

Write, but you also don't screw users with a package upgrade suddenly going mit to a paid license?!?!?!?! Absolutely not! Thats some bs right here.

[Laoujin]

I've always disliked MediatR. Added complexity without benefit in my book. Like let's break 'go to implementation' for everyone; it hollows the actual Mediator pattern and guides 'architects' that don't know any better to use CQRS for their CRUD applications, confusing junior/medior developers like there is no tomorrow.

So I applaud this as I will see much less adoption in future projects where I'll have to maintain the resulting shit code.

Can't wait to see the reaction of colleagues that use it in their apps :D

Good luck to you though Mr Bogard, I'm sure you spent a lot of time on this, time to make some money. AutoMapper for example is just in another league compared to other OSS projects that commercialized (I'm looking at you FluentAsseritions)

[nbarnwell]

Tip: if you can't be nice, be quiet.

[GabrielHSFerreira]

You always had the option to not use it, bro. I don't see the reason to come here and try to thrash it.

[hanyfakhry]

Thanks for all those years of amazing work as @byjokese mentioned. its sad that maybe our projects will be required now to move to something else but thanks again for all the support @jbogard ..

[fluxys-willaerf]

Hi, just a kind request: it would be useful to give some indication of pricing or pricing model that you have in mind.
Reading your blog it sounds like it will be 'reasonable', but then again, what's reasonable, right.

Without this, for companies that are right now considering to start using MediatR, it feels a bit like signing a blank cheque 😅

[jbogard]

Ping me directly and I can share details. I'm launching the storefront site in the next couple weeks but happy to share what I have now.

[VZhuck]

Disclaimer: I express my opinion, which is legally;) protected by constitution and not saying what author should do or should not do

Shame, we are back in 2000s :) I thought that times when foundational libs are under monthly costs are gone. Unfortunately, many developers will think twice to contribute to next "free open source" project.

Considering big picture, I (assume not only me) can't recommend monthly payment for each library being used in solution for client, which potentially millions of users. I wish @jbogard chose different path, but we are where we are...

[timctrahan]

I just can't see how you can take the SAME project and go from one version being mit and the next in the SAME package being paid. This is a screw you, and you are right!?!? how many developers contributed to this project? feels very much like a money grab.

[GabrielHSFerreira]

MediatR never had many contributions besides Jimmy's. I know it is disappointing to lose such a great library as free, but it is the result of big companies not giving a penny to OSS developers.

[jcfnomada]

This is a significant issue affecting many projects simultaneously, and there must be an underlying reason for it. It could be easily resolved by frameworks that enable these libraries to succeed. Platforms like GitHub and NuGet could implement specific policies to prevent this, but I’m not sure that aligns with the interests of the companies behind them. It feels like we might be getting manipulated here. Regarding Mediatr, I’m unsure how many features you’re using, but I created my own version in a day that didn’t break my existing code, kept the namespaces, and is twice as fast as Mediatr because it avoids the overhead required for those extra features. I understand those enterprise-level features need support from a company offering them for free or, in this case, trying to monetize them. However, "he" shouldn’t be allowed to use GitHub and NuGet to distribute the package, but that’s just my opinion.

[jbogard]

My request to folks disappointed with OSS including commercial options (MediatR is still OSS btw) is to please PLEASE have your company sponsor the projects their systems depend on. We maintainers don't want personal donations. We all welcome company sponsorship. No one supports these projects for free, someone has to sponsor time. I had a sponsor for over 10 years, but then that was gone for over 5, this is why we're here.

OSS may be a hobby for some, for a period of time, but it was never mine.