For me the debate never reaches the end because different kinds of developers build fundamentally different kinds of products.
If you are building a website, a forum, or a generally document based application with little to no interactivity (beyond say, “play media”) then absolutely make a server rendered html page and sprinkle it with a bit of JavaScript for accordions.
If what you are building is a complex editor (image, text), is highly interactive (with maps, and charts and whatever) and users will generally spend a lot of time navigating between almost same pages. Basically when there would be no expectation that this should work with JavaScript disabled… then just build a purely client rendered application in the framework of your choice.
To me the dispute comes when one bleeds to another. I also think that mixed modes are abominations unless you truly have actual performance gains (maybe if you have 1B+ customers), which I’d argue is true for almost no one.
Given the displeasure a lot of developers have towards AI, I would not be surprised if such attacks became more common. We’ve seen artists poisoning their uploads to protect them (or rather, try and take revenge), I don’t doubt it might be the same for a non-negligible part of developers.
As a kid I had both lard and butter on bread. Bras with lard and onions is amazing. But also that’s roughly the only combination of that works. Butter is way more versatile.
I have not used Claude. But my experience with Gemini and aider is that multiple instances of agents will absolutely stomp over each other. Even in a single sessions overwriting my changes after telling the agent that I did modifications will often result in clobbering.
You should try Claude opus 4.5 then. I haven’t had that issue. The key is you need to have well defined specs and detailed instructions for each agent.
Op mentions in the follow up comments that he does a separate git checkout, one for each of the 5 Claude Code agents he runs. So each is independent and when PRs get submitted that's where the merging happens.
It worked for me for finishing my app (vps+shellfish+gemini-cli), I've done a lot of coding like this on the train and in between sets in the gym, picking up on the more complicated stuff when at home.
But also all of the changes I made from the phone were incremental.
In between sets!? I've found that if I do any activity in between sets (like watching Twitter) I'll just end up spending way too much and then make the exercise session super long. Also I can't focus and write a serious prompt or review serious results in just or 3 minutes. But maybe it works if the app is sonething you've recently worked on and you already have very clearly in your mind what you want, it just needs to be done.
For incremental changes 1-2 sentences are usually enough. Also, since the program itself is a workout app with live reload, I can actually fix bugs while I’m using it.
As for too long of a wait I agree, it makes the sessions longer. Ideal window is after a heavy superset where waiting for 3-5 minutes is not a waste.
(Note that I’m not doing this for my real job, just for my personal project)
^^ I probably rely on AI slop than most people on this thread. I've found with the gaps with waiting on Claude Code output match the frequency I'm already checking my phone out of addiction. By no means the healthiest way to spend my time, but if I wanted to spin up a simple website or build out the framework for a project doom coding works for me!
I don’t buy the “it changes” argument. Yes, stuff is added, but everything that worked at some point still works. Just because CSS grid was added doesn’t mean that using flexbox, or even properly using divs and box model is suddenly out of the window.
Depends ... are you using `display: table` or are you literally using HTML instead of CSS? If you are doing the former, you know CSS. If you're doing the latter, then you haven't demonstrated that you know CSS and we'll have to talk to figure out if you know it and are avoiding it or if you just don't know it.
Table based layouts were deprecated even 15 years ago. It goes against your point that things change frequently.
> do you consider yourself a person who 'knows CSS'?
Sure. They know an older version of CSS. They don't know the latest version. Shrug....
> You will get laughed out of a tech interview for it though.
I've never laughed at anyone in an interview. Can you share some more nuances here? What's the role and expectation? How good is their code? Did they just not learn some modern designs, but can learn it quickly? What have they been working on the past few years?
reply