Post

🧵 The Attie backlash feels familiar. Not the "AI is bad part" but the "Company absorbs community feature" part. I spent 10+ years in the Magento ecosystem and watched eBay/Adobe do this multiple times. Were people upset? Yeah. Did we survive? Also yeah. Grandpa tells a story 👇

Adobe absorbed Page Builder (community had BlueFoot CMS), multi-warehouse inventory (Amasty, Magestore had extensions), B2B features (Cart2Quote, Aheadworks), search (competing with Klevu, Algolia). Platform eats ecosystem feature on repeat.

Almost none of those vendors died. Some went deeper: Amasty built "Advanced MSI" on top of the basic native version. Others diversified: Mageplaza expanded to Shopify. Yet others pivoted: Magestore became a Point-of-Sale company. Some got acquired into larger agencies.

What we saw: core integrations are usually more basic/generic. The platform team has to support every edge case for the broadest audience. Community products can specialize, go deeper for users who actually want the feature. The platform version validates the category. It grows the total market.

That doesn't mean nobody got hurt. Cart2Quote had their core product replicated nearly 1:1 by Magento's Negotiable Quotes. Some vendors had to make painful pivots. Being sherlocked forces you to differentiate or die. That was real and it was uncomfortable.

And nothing the company says can easy that pain

And yes, the comms around Attie were bad. Celebrating community projects on stage and then demoing competitor with no heads-up is clumsy at best. The chilling effect on investor confidence is a legitimate concern. Both deserve better from Bluesky.

The interesting part is what the Magento community did when the company sherlocked their projects, and hopefully where we can also learn.

Adobe was too slow to fix the frontend. Hyvä built one and became the de facto standard: most new Magento projects now choose Hyvä over Adobe's own frontend. The community wanted meetups and conferences Adobe wasn't organizing. So the community did and build whole programs around them.

We wanted independent stewardship of the open-source project. So we founded the Magento Association. (Disclaimer: I'm a founding board member of this Magento association)

The original Magento co-founders thought B2B was being handled poorly, so they left and built OroCommerce from scratch as a dedicated B2B platform.

The community didn't wait for Adobe to do the right thing. It built what was needed, often building things Adobe couldn't or wouldn't build themselves

Now of course ATproto ≠ Magento. And I think this is greatly in ATproto's favor. In Magento, Adobe owned the core. They controlled the APIs, the roadmap, what was even possible. Everyone built at their mercy.

We don't have that issue with ATproto (or at least much less) As @danabra.mov said: Attie "could have been made by anyone, today, it doesn't rely on any special access." The ATProto community has more power than Magento developers ever had, not less. And it applies to everyone in the ecosystem

The framing seems that the community needs Bluesky to NOT build things in order to succeed. That's not decentralization but dependency with extra steps. (I think someone else said that first but don't recall who, sorry for the missing attribution).

On ATProto, you should EXPECT that anyone (including Bluesky) will build on "your" lexicons and data. It's a deviation of what we That's the whole point. That's what we signed up for. If your moat is "nobody else builds what I built," you're in the wrong ecosystem.

That means we need new moats. Deeper specialization. Better UX. Domain expertise. Community trust. Optional premium features. Enterprise integrations. The old thinking of "my proprietary access to data protects me" is exactly the Big Tech model we're here to leave behind.

Yes, this is hard. Yes, it's uncomfortable. But isn't that discomfort exactly why we're here? Because we don't like what Big Tech does with their closed platforms and misaligned incentives?

The Magento community's best moments weren't when Adobe played nice. They were when the community stopped waiting and built what it needed at breakneck speeds (well, for a time without AI coding assistants that is). ATProto gives us a better foundation for that than any other system. Let's use it.