thejosh

except the fact that claude code in plan mode does destructive operations, requiring a hook to tell it not to do stupid things, so its not really a "read only" or "dont code yet" tool, imho.

Yes, the "okay if you want to disable AI, we're going to remove non-AI features" paradigm that Gmail is also doing.

I've been working on a demo database project, and have been really impressed by the UI. So glad they decided to put more effort into it, it has made building a "follow along" tutorial really nice.

also been a huge fan of zizmor (https://github.com/zizmorcore/zizmor) lately, basically: "am I going to footgun myself?"

it's been trash for years, their status page just doesn't reflect that.

so many ridiculous "how the fuck did this get through basic QA?" issues with Claude Code.

I can't believe how many critical bugs fall through.

My favourite one is the bug where Plan mode can execute destructive commands inadvertently.

Then all these get closed with `Closing for now — inactive for too long. Please open a new issue if this is still relevant.`. Awesome.

If I ever need anything more advanced than what Forgejo is offering me, I'd reach for Gitlab, for sure. :-)

For the forgejo side of things, this is for my own projects. Not hosting others (or for their code search, this was for mine).

For context, forgejo does have [code search](https://forgejo.org/docs/latest/user/search/code-search/), which is simple to enough.

I like forgejo/gitea - a single app.ini, postgres for the database. Done through a docker compose.

Last time I ran gitlab, it required oodles of memory - it's just me!

> You can just self-host GitLab CE and use self-hosted ElasticSearch integration to set up code search. I really don't understand the point of Forgejo, sounds like making your life harder

Sounds like more steps/maintenance/hassle and a more complicated than just using forgejo for my own personal projects, forgejo uses <120MB of memory for me, has runners, etc.

Will give it a go! Thanks for tooting! You've just made my afternoon!

I would love to use something that isn't GitHub for code hosting.

Codeberg sounds great, but from using it day-to-day, it's really slow sometimes, and frequently times out (from Perth, AU, if it matters). This is not just the SSH protocol for git pushes, but also for browsing.

(I do have a self-hosted forgejo (came from gitea), it's great, Postgres database backs up, files are easy to backup, it's also git so I can just mirror it elsewhere as well.). I love their self-hosted runners is super simple to host.

Bit random, but for code search, does anyone have any recommendations?

My current process is to just `git clone <REPO>` onto my server, then use ripgrep (Can't beat good ol' plain text search).

The result is that this is hundreds of GB -- I like to clone everything I use, dependencies etc, to then be able to look at source code. (I use quite a few languages/ecosystems, and sometimes reading the source code is the best way to figure something out).

I've tried Sourcebot (just for the code search, I don't care about "MCP"/"AI") etc, but was wondering if anyone has recommendations.

GitHubs code search is great, but it doesn't find everything, and it is also incredibly rate limited (annoying when trying to narrow something down, and you then get rate limited). I use this when trying to figure out random Apple issues - someone somewhere has figured it out usually.

php7 was such a great time period for PHP, honestly lots of great experimental projects around that time too (HHVM before that, etc).

nextest is one of those things I never have to think about, because it always "just works" and I have no dramas with it. <3

Bandwidth cost is also another major reason.

WOW.

Okay, this is really, really cool and is exactly my niche, as you mentioned it's kinda a combination of things like Stylus/uBlock Origin filters and custom filters/etc. This is really needed, as for example GitHub code preview is completely and utterly fucked, to put it lightly. Showing symbols, not being able to select code properly without weird visual glitches happening..... requires a bunch of scripts to fix. (https://github.com/orgs/community/discussions/54962).

What's your funding plan? You mentioned paid plan, but what's the actual benefit for users that they would pay for this? (I totally would, FWIW).

Do you foresee companies who need to build special widgets for whatever reason for random websites they use as kind of "Extension light" alternatives - your product reminds me of Retool (https://retool.com/), but for website tweaking.

Very cool product, love the ability to do "extra things" which will fix a whole bunch of websites I use everyday that I CBF'd either making an extension to fix or battling the uBlock/stylus filters.

Discoverability will also be needed, kinda like [karabiner elements complex modification rules](https://ke-complex-modifications.pqrs.org/)

edit: no firefox support, sadpants.

Yeah, performance depends.

I think a hybrid approach works best (store on Snowflake native and iceberg/tables where needed), and allows you the benefit of Snowflake without paying the cost for certain workloads (which really adds up).

We're going to see more of this (either open or closed source), since Snowflake has acquired Crunchydata, and the last major bastion is "traditional" database <> Snowflake.

After running Snowflake in production for 5+ years I would rather have my data on something like Parquet/Iceberg (which Snowflake fully supports...) than in the table format Snowflake has.

It's not that deep

Everytime you want to query your data, you need to pay the compute cost.

If instead you can write to something like Parquet/Iceberg, you're not paying for access your data.

Snowflake is great at aggregations and other stuff (seriously, huge fan of snowflakes SQL capabilities), but let's say you have a visualisation tool, you're paying for pulling data out .

Instead, writing data to something like S3, you instead can hookup your tools to this.

It's expensive to pull data out of Snowflake otherwise.

Sounds great until you're locked into Snowflake - so glad iceberg is becoming the standard, anything is great.

The trap you end up in is you have to pay snowflake to access your data, iceberg and other technology help with the walled garden.

Not just snowflake, any pay on use provider.

(Context - have spent 5+ years working with Snowflake, it's great, have built drivers for various languages, etc).

"May your SaaS never be purchased by Salesforce."

Autocomplete++ is the killer feature for me, especially for more tedious thingsike SQL column naming.

It's one of those things that if you need to ask why, you'll never understand :-)

I would argue webapps killed them.

Most applications now adays are just electron wrappers.

WebApps also allow a much more flexible update schedule, and other benefits.

Downside: apps now need gigabytes of memory for simple apps. Upside: Linux is now a lot more easier to use as a lot of things are web based. :-)

I'm not surprised. A few months ago I was reverse engineering their login flow (I was writing a driver for Elixir & Rust), and was getting MANY stack traces and weird bugs happening with their authentication flow, especially around OAuth.

Considering that when I asked about basic features, and bring told that the people interested/working on those left once Salesforce came in... Yeah.

I also love places that are named after someone and it's slightly amusing/horrifying.

Harold Holt was the Australian Prime Minister in the 1960's, and went for a swim and went missing (yes, there are many, MANY local conspiracies this :)). So what did we do? Named a swimming pool[0] after him (construction actually started before his death, I believe).

[0] https://en.wikipedia.org/wiki/Harold_Holt_Memorial_Swimming_...

to every random daemon that macos 100% needs to run RIGHT NOW.

I don't care about memory usage for editors as much I care about input latency and responsiveness.

Jetbrains (IDEA IntelliJ, Pycharm, etcetc) put a lot of effort into making their IDE low latency as it was getting to a point of being almost ridiculous. Their editor is built in Java, and they run on their own runtime as they have so many hacks and tweaks to make it work as a desktop app as well (font rendering, etc).

Pavel Fatin has a [great article](https://pavelfatin.com/typing-with-pleasure/) about typing latency and his work around implementing this in IntelliJ, well worth a read.

Also handy (regardless of framework, language, etc) is if you use feature flags, if you're using actual valid credentials against a service or a local testing service (eg S3 vs minio, etc).

Enabling this for non-technical users really helped also report any issues they would see, it was great for testing environments.

Plus you would probably end up needing all this information anyway if you do any error tracking, so it probably exists somewhere already :-).

Memory and CPU usage is still really high though.

Sane defaults are always good, (which firefox has, IMHO), but not allowing the user to tweak basic stuff is a pain.