will be in nyc for a week in early march with my fam (3 total including a baby). with airbnb no longer possible, anyone want to rent me their apartment?
Invoicing is a concrete example of how we design in prod.
We went from customer calls to prod in a few days. Once in prod, we ran two rounds of internal usage, feedback, and revision, in parallel with sharing with beta customers who had requested it. And the first customer-issued invoice has already gone out.
https://splits.org/blog/design-in-production/
imo bridging is a weird crypto-insider concept.
lots of work for us to do here, but one step towards making this obsolete is letting you pay someone in any token on any network from any token on any network.
we keep ~all our usdc earning interest on base, and pay out in different tokens on different networks.
this lays the groundwork for more flexible bill pay and invoicing flows.
how do we confidently get to a point where "the actual code written doesn't matter"?
dark software factory is inevitable, which creates a whole new set of challenges
What are the 10 year trends you're betting on, anon?
“If you base your strategy first and foremost on more transitory things — who your competitors are, what kind of technologies are available, and so on — those things are going to change so rapidly that you’re going to have to change your strategy very rapidly, too.”
https://splits.org/blog/10-year-narratives/
Sharing a few q's that have come in DMs:
What specific tools do we use for this?
Design is mostly in excalidraw these days - not using figma much at all. Haven't found a need for high fidelity mocks and excali is very fast so there's no need to maintain components outside of our FE. We rolled our own FE for the component library - not using storybook. And we use amplitude for feature flags.
How/where are we using AI?
Mostly to educate devs and ensure consistency. We have commands for ensuring component usage (props, language/tone, etc) is correct, and commands that get run whenever a component (in a specific directory) is updated to ensure docs are accurate. I think there's a lot more to explore here - eg better machine readable documentation on when to use which component/prop/view/etc.
Some thoughts on how we @splits think about design, and how speed leads to quality (instead of being at odds with).
For us it's a combination of:
- Performant, polished component library (code not Figma)
- Component discipline (must ship MVP to prod w/out modifying or adding components)
- Heavy use of feature flags (we are always our own first customer)
- Continuous feedback
- Staged rollouts
https://splits.org/blog/design-in-production/
Revisiting this convo with @rish in light of last week's news.
Some wonderful nuggets in here, gives you a sense for how Neynar team thinks.
- Combine intuition with data
- Just do stuff - FAFO is real
- Always be asking: are we focused on the right thing?
https://splits.org/blog/ship-first-commit-later/
Custom transactions are now live.
Call any contract function directly - paste in the address, choose a function to call, enter params.
No more connecting to block explorers.
