Why wasn't it always that way

From latest Hyperdev newsletter:

Version 1.0.60 changes this completely. Anthropic’s documentation now confirms that subagents operate in independent context windows with tailored system prompts and tool access, preventing context pollution while keeping each agent focused on its domain. A security review agent analyzing bash commands gets maybe 500 tokens of relevant context, not my entire conversation. A documentation agent updating API specs gets function signatures and changes, not debugging session transcripts.

Why wasn’t it working that way from the start? I keep reading articles about different products “now doing this amazing technique” (same applies to Cursor, Windsurf, etc) and I’m astonished, thinking… but Beyond Better has been working that way for months, and from the initial design.

So I keep returning the biggest problem with Beyond Better; how to solve the viral attention. BB is clearly an advanced product, but me & BB lack the ability to convey that to a wide audience. (Unfortunately the article about BB I was expecting got bounced in favour of an article about Apple - hopefully next week…)

Sorry Charlie, that sucks. I feel your pain and it is one a know all too well. It makes me think of is the passionate burger maker being put out of business by McDonalds. Unfortunately the best product doesn’t win the day, Microsoft, McDonalds, Coke, etc, etc.

I am not even sure what does win the day; a product that is good enough + boat loads of money? is that it?

Maybe the play isn’t fighting for viral attention directly, but positioning BB for when the market realises these architectural decisions actually matter. It’s often best to be seen as second or third player in a market, even if you have from a technological point of you been first all along.

Since BB isn’t being positioned for the coding market, but instead for “project” market, so there is still chance to be first. There is plenty of “browser automation” but I feel those are short-term plays. BB is taking a more direct approach with custom datasource integrations, effectively bypassing the pointy-clicky needed for browser automation.

BB works directly with the user’s data wherever it resides. Currently that is read-only, but having read-write is “very close” now. The foundries are doing similar efforts, but locked into a single provider ecosystem. MCP is doing similar, but in a very clumsy manner.