Subnostr

I am a strong believer that 99% of codebases have a 2 year life span. Whatever a dev worked on in 2022 sucks right now.

New languages, new stacks, new coding styles, new libraries make what was done last year faster, simpler and better.

Which means, you have one year from idea, to execution to collecting enough revenue from that codebase to pay for the time you spent on it.

This is why Product Velocity is the most important metric of any new endeavor.

Make it simple, make it valuable, ship it fast and enjoy the revenue while you can.

Reply to this note

Please Login to reply.

Discussion

sommerfeld 1y ago

Even the other 1% only survive past it because they go under regular major rewrites.

Diacone Frost 1y ago

or it's SAP 😂

Derek Ross 1y ago

Hmm. I agree. I don't know if two years is too short of a timeframe, but it definitely happens over time.

When will you be nuking Amethyst and rewriting it with new tools?

Vitor Pamplona 1y ago

Already thinking about it. 👀

choss 1y ago

Be that one percent then

Jeroen Ubbink 1y ago

Strangely enough most of the unix tools do not seem to suffer from this.

do one thing and do it well.

It does not yield nostr clients but then again, nostr clients hardly do one thing. It would not be super complex to pipe a few of these ancient tools and have a very tiny and simple nostr client. Most of these tools have had no major changes in the last few decennia. They do not rely on modern frameworks that are outdated in 2 years and replaced by something better that, again, does do many thing and therefore makes itself prone to becoming outdated very fast.

Vitor Pamplona 1y ago

Yeah, that's the 1%. It's the same for some of the crypto implementations. But apps in general will to be rewritten or they just stop working.