Hacker Newsnew | past | comments | ask | show | jobs | submit | efortis's commentslogin

Is there a way to stream the ESI in a chunk? I'm asking for different use case.

Just so I understand what you are asking about. You have a ESI chunk which is a cache miss and the backend is fetching it, but you want the delivery of this specific chunked to have a streaming delivery?

Yes, but it can be regardless of being a cache miss.

If you are curious about the use case: https://github.com/ericfortis/aot-fetch-demo/blob/main/optio...


I think it could be streamed if compression is off.

> In contrast, bug reports to a typical end-user application go into a black hole from the moment they’re submitted.

True, but not always, I've heard PMs saying: "this was a user-submitted bug", meaning you better fix this ASAP.

But yes, for the most part they don't like it. I recall reporting a bug and then they obfuscated their SPA bundle.

Funny enough, today I spent like an hour thinking: should I report this, should I ever report bugs at all. It was a P0 but I didn't report it, and I noted it as my first time considering giving up on my career.

Last week I reported a list of medium-bugs and I ended with: "I'm not sure if you'll ever read this, so I'm not adding more details that could help you debug, but feel free to reach out…" They responded: "We will share this feedback with the appropriate team. No further action is required on your end. Your ticket 124443 is solved. Please rate us!"


> I've heard PMs saying: "this was a user-submitted bug", meaning you better fix this ASAP.

There are certainly developers with a practice of triaging and action user-submitted reports, but unfortunately it’s opaque to the one who reported it.


Two tips I learned from Sorin's yt channel:

- Bend the tips

- Don't buy an expensive iron, use a cheap one and assist it with a hot air gun

Sorin is incredibly skilled. Here he is fixing a ribbon cable:

https://youtu.be/P81IG0DZ1kg?t=494


> fetch, XMLHttpRequest : Nothing — these are the ones everybody checks.

That's one of the things I don't like about security scanners, they just check for fetch and friends, and without context, they downgrade your security score.

For example, I have a fully local client-server package that makes no external requests but it gets flagged with "Supply Chain Risk" because it uses `fetch`.


FWIW, YouTube, instead of SSR, simply includes the critical API data in the HTML document.

Here’s a PoC of that technique, see option 2.

https://github.com/ericfortis/aot-fetch-demo


Is recording for creating stubs possible with that local setup?


haven’t done it, but I’d try a rain sensor. and if the child is into music, a guitar tone knob.


An easy improvement could be offsetting the cuts a bit so the result doesn't feel choppy. See J Cut and L Cut.

A more complex improvement could be analyzing the movement and/or focal point so cuts flow better. For example, if in ClipN the focal point ends at 30%-top,30%-left try to get the starting focus point of ClipN+1 around that area too.


Currently, on ways to render code more legibly:

- Camelana. A proportional coding font with custom ligatures for `return` and `throw` keywords

- Tabular Eye. IDE plugin for tabularizing mono and non-monospace fonts without adding whitespace

https://github.com/ericfortis/camelana

https://github.com/ericfortis/tabular-eye


CSS Modules are a better solution. They compile classnames into unique identifiers, so there's no need for BEM or SMACSS conventions:

https://x.com/efortis/status/1888304658080256099


I disagree. I don't want to jump between two files constantly. With Tailwind I can stick to one file and never need to think about naming things.


Right so it is easier to write code quickly, but not to maintain a system. Tailwind is the right tool if you want to generate a prototype with LLM agents. It can be quickly deleted or can be used as visual prototype to create reusable components. And once the component is built for long term maintenance as priority, the collective knowledge has been built up so naming becomes simpler, important and elucidating.


The component itself is the semantics, adding naming over CSS classes adds no value whatsoever, only adds cognitive load and requirement of context switching to understand "what does this named class actually do?"


Css classes are not big issue with css nesting and if component boundaries are settled and the design is well defined before implementation. Tailwind is a shorthand notation which is harder to parse than a structured css file (at least I strongly believe this we can agree on) and thus has only editing speed gains which is largely not an issue if prior work is done on exploration. Another way to frame it: tailwind is a good prototype syntax but is not the best syntax to for production code (human parsability). I’d skip it altogether to simplify the flow.


this! a semantic class name that is scoped to only that component and only ever used in that component is pointless indirection only needed to be able the reference the actual css properties in that other file sitting next to your component by using some name.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: