Now for the columns! What mcu are you using?
August 4, 2025 at 7:38 PM
Now for the columns! What mcu are you using?
"...learning to operate a textile loom in 1820"
❤️
❤️
August 2, 2025 at 1:37 PM
"...learning to operate a textile loom in 1820"
❤️
❤️
Oof. Extra nightmarish if you're putting together a monorepo.
January 24, 2025 at 6:10 AM
Oof. Extra nightmarish if you're putting together a monorepo.
I think the trick is to get the TS language server as well as a bundler and Node happy with the setup.
I like some kind of combination of TS project references and conditional package exports.
colinhacks.com/essays/live-...
moonrepo.dev/docs/guides/...
I like some kind of combination of TS project references and conditional package exports.
colinhacks.com/essays/live-...
moonrepo.dev/docs/guides/...
January 23, 2025 at 6:15 AM
I think the trick is to get the TS language server as well as a bundler and Node happy with the setup.
I like some kind of combination of TS project references and conditional package exports.
colinhacks.com/essays/live-...
moonrepo.dev/docs/guides/...
I like some kind of combination of TS project references and conditional package exports.
colinhacks.com/essays/live-...
moonrepo.dev/docs/guides/...
If you're working with a monorepo type thing then tsBuildInfo might help?
www.typescriptlang.org/tsconfig/#ts...
www.typescriptlang.org/tsconfig/#ts...
January 16, 2025 at 10:04 AM
If you're working with a monorepo type thing then tsBuildInfo might help?
www.typescriptlang.org/tsconfig/#ts...
www.typescriptlang.org/tsconfig/#ts...
January 16, 2025 at 9:47 AM
That tracks. Especially considering the impact LLM's are having on writing code.
January 13, 2025 at 7:09 AM
That tracks. Especially considering the impact LLM's are having on writing code.
Do you think it's because of the complexity of modern web-apps? Will the needle swing back as our abstractions mature?
January 13, 2025 at 6:56 AM
Do you think it's because of the complexity of modern web-apps? Will the needle swing back as our abstractions mature?
I think it's frowned upon because it strays too far from plain JS. The line being: If we were to replace all the type annotations with whitespace, will it still work?
January 13, 2025 at 6:50 AM
I think it's frowned upon because it strays too far from plain JS. The line being: If we were to replace all the type annotations with whitespace, will it still work?
Perhaps using a global error page would do the trick?
nextjs.org/docs/app/bui...
nextjs.org/docs/app/bui...
January 10, 2025 at 10:48 PM
Perhaps using a global error page would do the trick?
nextjs.org/docs/app/bui...
nextjs.org/docs/app/bui...
What? This isn't related to TypeScript! i've been duped!
January 10, 2025 at 8:22 PM
What? This isn't related to TypeScript! i've been duped!
Have you tried the Vitest VSCode extension? Dunno if I'm just lucky but debugging tests just worked out of the box!
January 9, 2025 at 8:11 PM
Have you tried the Vitest VSCode extension? Dunno if I'm just lucky but debugging tests just worked out of the box!
Losing it? Did they ever have it?
January 7, 2025 at 11:48 AM
Losing it? Did they ever have it?
That's interesting - did you reach for a custom adapter in better-auth?
January 3, 2025 at 7:19 AM
That's interesting - did you reach for a custom adapter in better-auth?
Quick question: what sets better-auth apart for you?
January 2, 2025 at 9:32 PM
Quick question: what sets better-auth apart for you?
I get the feeling the industry is still working on standardising what constitutes an Agent.
Anthropic seems to make a distinction between Agents and Workflows: www.anthropic.com/research/bui...
It kinda makes me think that agents are self-directed; they involve an LLM in choosing what to do next
Anthropic seems to make a distinction between Agents and Workflows: www.anthropic.com/research/bui...
It kinda makes me think that agents are self-directed; they involve an LLM in choosing what to do next
January 2, 2025 at 6:53 AM
I get the feeling the industry is still working on standardising what constitutes an Agent.
Anthropic seems to make a distinction between Agents and Workflows: www.anthropic.com/research/bui...
It kinda makes me think that agents are self-directed; they involve an LLM in choosing what to do next
Anthropic seems to make a distinction between Agents and Workflows: www.anthropic.com/research/bui...
It kinda makes me think that agents are self-directed; they involve an LLM in choosing what to do next
Oh that sounds interesting- I'll take a look
December 29, 2024 at 7:27 PM
Oh that sounds interesting- I'll take a look
Ooh. What's the plan to avoid the suckyness?
December 29, 2024 at 6:04 PM
Ooh. What's the plan to avoid the suckyness?
DHH (the Rails guy) was espousing the same idea a few years ago - I'd like to think it helped shift the Overton window in favour of SSR.
December 28, 2024 at 10:57 PM
DHH (the Rails guy) was espousing the same idea a few years ago - I'd like to think it helped shift the Overton window in favour of SSR.
Array.prototype.reduce gets quite close: The callback's 3rd arg is the index. Return the accumulator and the current value spreaded to flatten.
Eg.
[[1], [2,2], [3,3,3]].reduce((acc, curr, idx) => idx > 1 ? [...acc, ...curr] : acc, [])
I'd argue chaining filter and flatMap is usually clearer tho
Eg.
[[1], [2,2], [3,3,3]].reduce((acc, curr, idx) => idx > 1 ? [...acc, ...curr] : acc, [])
I'd argue chaining filter and flatMap is usually clearer tho
December 28, 2024 at 10:39 PM
Array.prototype.reduce gets quite close: The callback's 3rd arg is the index. Return the accumulator and the current value spreaded to flatten.
Eg.
[[1], [2,2], [3,3,3]].reduce((acc, curr, idx) => idx > 1 ? [...acc, ...curr] : acc, [])
I'd argue chaining filter and flatMap is usually clearer tho
Eg.
[[1], [2,2], [3,3,3]].reduce((acc, curr, idx) => idx > 1 ? [...acc, ...curr] : acc, [])
I'd argue chaining filter and flatMap is usually clearer tho
🦆 If it quacks like a duck...
December 28, 2024 at 8:24 AM
🦆 If it quacks like a duck...
About completeness: yeah, this only handles type-annotations. Enums and the like are behind another flag. Node is also waiting for the JS decorators proposal to land before adding support for them.
December 28, 2024 at 8:15 AM
About completeness: yeah, this only handles type-annotations. Enums and the like are behind another flag. Node is also waiting for the JS decorators proposal to land before adding support for them.
My take: prefer safeParse when using TS.
Form validation failing isn't exceptional - it's a thing that can (and probably should) be modelled using the type-system.
Caught errors are of type `any` - using that for flow control is a shlep.
Form validation failing isn't exceptional - it's a thing that can (and probably should) be modelled using the type-system.
Caught errors are of type `any` - using that for flow control is a shlep.
December 27, 2024 at 8:39 PM
My take: prefer safeParse when using TS.
Form validation failing isn't exceptional - it's a thing that can (and probably should) be modelled using the type-system.
Caught errors are of type `any` - using that for flow control is a shlep.
Form validation failing isn't exceptional - it's a thing that can (and probably should) be modelled using the type-system.
Caught errors are of type `any` - using that for flow control is a shlep.