stpedgwdgfhgdd

You can use git worktree to work in parallel in multiple terminal tabs. It does give a higher cognitive load.

Test driven development, there is nothing more to say for the coming year

Anthropic acknowledged there were bugs that are now resolved, see their status page for latest info:

https://status.anthropic.com/

I did not notice a degradation in quality last weeks. Not saying it is perfect, but the quality is similar (using Sonet) for the last month.

Using only 2 MCP servers and not extending claude.md.

I hear that multiple MCP tools eat up your context. See last video Indy Dev Dan.

Maybe one day this becomes now relevant, but today? Isn’t it better to stick to CLI tools? Lesser chance of supply chain attack if you stick to the vendor’s cli.

(I use 2 MCP servers in daily life, hesitant to add more)

This RCA is too vague: ‘a bug’

I want to know how i could have been impacted.

Yes! This is a big problem!

However it seems, the change is reverted. Just today /model gives 3 options instead of 4:

│ 1. Default (recommended) Use the default model (currently Sonnet 4) · $3/$15 per Mtok │ │ 2. Opus Opus 4.1 for complex tasks · $15/$75 per Mtok │ │ 3. Opus Plan Mode Use Opus 4.1 in plan mode, Sonnet 4 otherwise

XP

Last week I asked Claude to improve a piece of code that downloads all AWS RDS certificates to just the ones needed for that AWS region. It figured out several ways to determine the correct region, made a nice tradeoff and suggested the most reliable way. It rewrote the logic to download the right set, did some research to figure out the right endpoint in between. It only made one mistake, it fallback mechanism was picking EU, which was not correct. Maybe 1 hour of work. On my own it would have taken me close to a working day to figure it all out.

People are already doing waterfall with their quest for the perfect prompt that will produce a complete, fully functional implementation.

(These people dont realise that there a lot of tradeoffs to be made that pop up during implementation)

Try Crush

When you start to refactor 100k+ lines code base it adds up even more… Some Docker builds, run the linters and the tests, you quickly exceed $40

“TDD failed because it assumed you could design a perfect systems before implementation”

TDD states the opposite.

TDD is very hard to do right and takes a lot of discipline. If i hadn’t worked for a company that did 100% XP, i would not have either believed it could be so effective. (Best Enterprise software i’ve ever seen and written)

In a way, it is funny. You can practise XP with your AI as pair.

Mmmm… read the book on Extreme Programming.

The idea is that you let the code drive the system and do not optimize prematurely. Sometimes developers design parts that are not needed, and often not in the first phase.

It is a way to evolve a system unbiased. Of course there is a trade-off. If the refactoring/change is very expensive, like a database schema change in production, it is good to spent a lot of upfront design. Takes experience to see where you can relax and where you need to be careful.

An AI tool like CC requires lots of experience to be used effectively. With today’s technology it still & also requires a lot of old fashioned coding experience. In the hands of a power user it can give a big productivity boost.

That said, I wonder how many dormant bugs get introduced at this moment by the less talented. Just a matter of time…

I did not understand ‘hg’ and asked Claude:

The “hg” in this passage refers to Mercurial, a distributed version control system. `hg` is the command-line tool for Mercurial (just like `git` is the command for Git).

The Go project itself was historically stored in Mercurial repositories before migrating to Git. Some Go modules may still be hosted in Mercurial repos.

There is a big difference between Claude Code today and 6 months ago. Perhaps the LLMs plateau, but the tooling not.

And relying on (another) 3rd party provider that indirectly has access to your code….

I do not how it is implemented, but if I can press ‘continue’ from my phone, someone else could enter other commands… Like export database…

There is an interview with the architect and he explained that Go language maps quite nicely onto the Typescript language (e.g. garbage collector). And of course a lot faster due to concurrency and native binary. 6 times increase iirc.

For enterprise software development CC is definitely there. 100k Go code paas platform, micro services architecture, mono repo is manageable.

The prompt needs to be good, but in plan mode it will iteratively figure it out.

You need to have automated tests. For enterprise software development that actually goes without saying.

The deeper problem are the custom commands, hooks and subagents. The time has come that you need to make a strategic choice. Once you have heavily invested into CC, it is not easy to turn to an alternative.

Side remark: CC is very expensive when using API billing (compared to e.g. GPT-5). Once a company adopts CC and all developers start to adapt to it at full scale, the bill will go out of the roof.

I give the MR id to CC and let it review. I have glab cli installed so it knows how to pull and even add a comment. Unfortunately not at all specific line number afaict. I also have Atlassian MCP, so CC can also add a comment in the Jira work item (fka issue).

Based on my own experience and reading a ton of HN posts, i would summarize it as:

- vibe coding is fun, but not production-ready software engineering

- LLMs like CC today moderately boost your performance. A lot of attention is still needed.

- some TDD style is needed for the AI tool to converge

- based on the growth of the last few months, it is quite likely that these tools will increase IC productivity substantially

- fully autonomous agentic coding will take more time as the error rate needs to decline significantly

Aside that space invaders from scratch is not representative for real engineering, it will be interesting to see what the business model for Anthropic will be if I can run a solid code generation model on my local machine (no usage tier per hour or week), let’s say, one year from now. At $200 per month for 2 years I can buy a decent Mx with 64GB (or perhaps even 128GB taking residual value into account)

My suggestion is to try CC, use a language like Go, and read their blogs how they use it internally. They are transparent what works and what does not work.

Remarkable that an official White House plan uses the term “America” instead of United States of America.

It probably depeds on the kind of software you create, but in our case:

When i add non-trivial code I want others of my team to review it first. When it is on trunk, it gets less easier to refactor as others already might use it. In this sense using a MR is really good way to get some quality checks before it goes to main and needs to be supported (in most cases) indefinitely. Next, we have system tests that i really want to avoid breaking on main and getting EVERYONE stuck.

The problem arises if I keep that branch for more than a week. Merge conflicts get harder, the financial investment is not leveraged, etc etc.

The problem is with long lived branches. This is something you should avoid, you can take it to the extreme, no branches. But extreme is almost never good.

PS CC is really good at rebasing and resolving merge conflicts.

This is similar to what people said about chess bots. Look at GothamChess and how amazed he is when stockfish plays another bot or human.

As a CC user I can say there there are at least two reasons why CC dominates.

One, it is very good except for its prise. Because of the cost, I tried Codex, but after a few simple questions I gave up. Nowhere close to CC when it comes to tool usage. Gemini cli asked for a Cloud project id, no idea what is meant with this. Cursor is not an option because I think long term they can’t compete with the integration that Anthropic can offer.

Second, CC offers capabilities you will learn as power user, in that sense it is like a game; addictive.

That is also what Anthropic recommends. In edge cases use Opus.

Opus is also way more expensive. (Don’t forget to switch back to Sonnet in all terminals)