software engineering

Frontend Code Review Policy. Frontend Code Review Policy

Speed of iteration over perfection - Frontend development is inherently iterative. It is often better to ship something and then iterate on it than trying to get it perfect from the get go.

Not all changes are equal - CSS changes shouldn't require review. With the advent of LLM's, a lot of NextJS code feels solved as well.

Review Requirements by Change Type

CSS/HTML-only changes: No review required

• Pure styling changes (margins, colors, fonts, layouts)

• HTML structure changes without logic

• Exception: Changes to key components (e.g. StoryCard, IndexContent)

• Developer is responsible for visual QA before merging

Very Small Component logic changes: No review required

• Changes to the frontend tests

• Piping some variable around in order to display it (e.g. adding a date)

• Adding/modifying some simple logic

• One-line bug fixes

• Developer is responsible for visual QA and LLM-core review before merging

Component logic changes: Relaxed review requirements (see below)

• Data transformation and display logic

• Any React or NextJS component change that follows standard React/NextJS best practices

  • Adding hooks

  • State management changes

  • Event handlers and user interactions (useEffect)

  • Props and type definitions

  • etc.

Critical paths: Regular review requirements

• Authentication/authorization flows

• Payment or checkout flows (in the future)

• Data mutations that affect other users

• Any code in common or backend (i.e. the change is not purely a frontend change)

Relaxed Review Requirements

1. Before requesting a review, the submitter is expected to have ask an LLM to review the code and find any code smells, bugs, things that might bite us later on, gotchas, or design problems, and address them. This process should be repeated until the LLM is satisfied.

2. Review should be up for a day to give others a chance to respond

3. Code can be pushed if either of these two conditions is met:

   1. Approval from another UpTrust dev

   2. 24 hours + passing the LLM review

Thoughts on AI coding an effifciency?  would love yalls thoughts on this. Here’s a provocative start: 

This whole thing is bullshit. So if you're a developer feeling pressured to adopt these tools — by your manager, your peers, or the general industry hysteria — trust your gut. If these tools feel clunky, if they're slowing you down, if you're confused how other people can be so productive, you're not broken. The data backs up what you're experiencing. You're not falling behind by sticking with what you know works. If you’re feeling brave, show your manager these charts and ask them what they think about it.

From a Mike Judge substack article

that’s one take.

There are a few other takes that I’m aware of, but I’d really love to hear the developers first

Thoughts on AI coding an effifciency?  would love yalls thoughts on this. Here’s a provocative start: 

This whole thing is bullshit. So if you're a developer feeling pressured to adopt these tools — by your manager, your peers, or the general industry hysteria — trust your gut. If these tools feel clunky, if they're slowing you down, if you're confused how other people can be so productive, you're not broken. The data backs up what you're experiencing. You're not falling behind by sticking with what you know works. If you’re feeling brave, show your manager these charts and ask them what they think about it.

From a Mike Judge substack article

that’s one take.

There are a few other takes that I’m aware of, but I’d really love to hear the developers first

Thoughts on AI coding an effifciency?  would love yalls thoughts on this. Here’s a provocative start: 

This whole thing is bullshit. So if you're a developer feeling pressured to adopt these tools — by your manager, your peers, or the general industry hysteria — trust your gut. If these tools feel clunky, if they're slowing you down, if you're confused how other people can be so productive, you're not broken. The data backs up what you're experiencing. You're not falling behind by sticking with what you know works. If you’re feeling brave, show your manager these charts and ask them what they think about it.

From a Mike Judge substack article

that’s one take.

There are a few other takes that I’m aware of, but I’d really love to hear the developers first

"You can't not-have-resentment," they tell me.
"You must have conflict. You must process resentment, or else you're bypassing."

Right on the first count, wrong on the rest.

You can't not-have-resentment. But only because not-having things isn't a thing.
In order to not-have resentment I have my hands full HAVING something else.

If you are tired of managing and processing resentment and would like to simply not-have it, pick something else. Pick something specific, some other relational stance or tone, other than a resentful one.

Resentment is a tone that radiates from the entire body. It is the shadow of Having: Burden.

Resentment is my way of looking upon all I have for the way it causes me problems and gives me more work to do, for the way it is a pain in my ass. I get to see my loved ones only in the context of me, I see how they behave around someone who's constantly irritated with them, critical of them, and overwhelmed by their existence. In my experience of resentment, those I love the most are the most irritating people to be around, I've signed up to have them around a lot, and I'm overwhelmed and overworked. Nobody's happy, and in lieu of that, I cling to how right I am with every muscle in my body clenched.

Ow. No thanks (she says, like someone who gave it up freely rather than having it wrenched brutally from her desperate grasp).

Devotion is what I Have instead, it's my way of looking upon all I have for the way it is ONLY HERE NOW FOR A LIMITED TIME, blessing me in particular. I get to see that there is only this now moment in which I might appreciate and enjoy my loved one, and I would not squander that for an opportunity to express irritation toward them; I'm too busy appreciating what is to even judge that something is going "wrong." I Have love now, I don't leave love to try to get love, I don't stay out of love believing there is some journey back to love. I Have devotion always, I have intimacy always, because I am in relationship to the now moment.

I don't process resentment OR bypass resentment. It doesn't come up because I Have something else.

Some integral philosophy/analysis behind TreeView. Forums like Reddit and Facebook are essentially "Orange" interaction paradigms—there's a static "original post" and a fixed hierarchy. Some benefits of this are that you get some sense of coherence to a conversation, and if you're ten layers deep in replies, you can quickly navigate your way back to the original post to try and reclaim some context. Some drawbacks are that really cool content deep in threads hardly sees the light of day. Also, it's really confusing to navigate, and especially confusing as the tail winds up. (Facebook tries to avoid the contextlessness and lostness of orange by imposing a limit of three deep, but this curtails conversational flow which often goes in strange directions to find the good stuff).

Twitter and all of its clones are essentially "Green" interaction paradigms—you get a personally tailored feed of content that comes from anywhere in a given lineage/conversation. The first post you see could be from twelve layers deep in a convo, the next a reply to a post, and the next an original post. Some benefits of this is that deep content gets surfaced—a reply to a reply to a reply to a reply might be the most interesting bit of the conversation, and on twitter it gets to your feed immediately. Some drawbacks are that it often lacks context, and sometimes is purposely removed from context to be more inflammatory. I believe this contributes to the sound-bite post-truth world we're in today, where context, nuance, history don't matter, only quick dopamine hits.

We believe TreeView is the start of a "Teal" interaction paradigm which allows deep content to be surfaced while retaining the context of the conversation it is birthed from. We believe this will integrate the benefits of both hierarchical forum interactions and pluralistic twitter feeds. I say "start" because right now it still leans Orange; we haven't yet built out the bits where whenever you click onto a reply, it becomes the top-level thingy in the TreeView; nor the feature where you can link "original posts" to each other in playlists, or to multiple originals, and see that there's not really such a thing as an "OP" because all of us are proverbially standing on the shoulders of giants. So where we're going is you'll be able to get a relative hierarchy (a holarchy) of what conversational lineage content is nested inside of. We believe this is models more of "in person" conversation actually works than any of the current paradigms. And even just this start has helped me understand longer threads!

What are your sci-fi TV show recommendations? Some i loved that jump to mind:

• The Expanse
• Most of the Marvel stuff like Loki, What If?
• Rick and Morty
• Legends of Tomorrow Etc

Only looking at roadblocks is self-defeating. I feel like friends including some here have been gesturing at this to me for years but I got it on a deeper level today in therapy.

I was talking about how unhappy I am about my work lately.

Therapist suggested what would be my ideal work like instead? And so I envisioned it but then immediately started thinking about the roadblocks to getting there.

But only focusing on the roadblocks is focusing only on the negative and where we put our attention only grows. Rather, I could also include more of what’s drawing me and how great it’d be. Ideally some combo of both.

A distinction between want (etymology linked to lacking) vs desire (await what the stars will bring)