nijave

I think it also depends on how well defined requirements are. I can have Claude nearly one shot a Terraform provider but it's just taking 2 API contracts and combining them together (the provider contract and the API you're automating)

On the other hand, some things are incredibly product intensive and poorly defined and it seems LLMs will latch on to an arbitrary direction with much lower success rate.

It is also worse in "brown field" where there's lots of tribal knowledge and technical debt. Two systems may both exist that look very similar but oftentimes there's the one on life support that almost no one uses or cares about and the LLM doesn't know it should largely ignore that until it's thoroughly documented (which can be a significant undertaking)

Idk imo code is and has always been a liability. Like the toilet paper bear says "less is more"

Eh, I think they're still terrible at architecture. They'll find bugs but they'll try to make them go away instead of building a sustainable system.

They also have a propensity to try to rewrite gigantic amounts of code in one go making changes riskier and harder to scrutinize... without using more LLMs.

Maybe they forgot to run this one by the lawyers before posting. I'm surprised they'd publicly admit to that

It does have to do with GLM-5.3 but you're correct it doesn't have anything to do with the model.

It effects the _endpoint serving GLM-5.3_

I'm not sure how else you'd identify the GLM-5.3 serving endpoint without naming GLM-5.3.

That said, having the word "model" in the title is misleading imo

Changing "model" to "endpoint" might help since it's the endpoint, not the model

Also sandboxing and strong permission boundaries ie when things fail, contain the damage

I'm not sure that's true. It seems very popular in OSS in general but not much enterprise/commercial uptake.

I guess node has taken some market share it seems like

Yeah, I thought it was dying a bit but was looking for webmail software (Roundcube, etc) a couple days ago and all of it is PHP.

Haven't tried it yet but there's OpenMuse for the Athena S Muse headband.

I think there's an open source Whoop firmware now for that.

They are nice in cold weather to "hold you over" on short drives before the air heating warms the car up.

My Civic remote start with the key fob will automatically turn them on when it's cold.

I've had (what I think is) the Renpho Elis 1 for years and have been happy with it. Was like $20-25 from Amazon I think in 2018 and still works fine

Found out on 1.x it goes crazy with file watches. If you launch in a top level directory with lots of sub directories, iirc it can create millions of watches X2

Not sure if it's fixed in 2.x. There's some way to get it to not do that with ignore files or rules but not sure exactly how--I had Claude fix it...

Don't worry, it's janky on my Pixel 11 as well

ACP?

I generally trust benchmarks but my personal experience in the last 6-9 months is they have little relation to real LLM performance, unfortunately.

I've had better luck just running a few of my own simple tests.

Cloudflare Access does that. I suspect other SASE/zero trust solutions probably also do (iirc it does everything you've named)

You used to be able to also do a .conf or .js file and iirc the only thing different between platforms was the path.

I think it's policies.json and mozilla.cfg

Really Lowes and Home Depot have been doing better imo for anything they have. Their website actually has filters and decent search and most stuff is free next day delivery. On top of that, the Chinese knockoffs are fairly minimal or at least more established knock-off brands.

On top of that, that's without a $150/yr subscription.

Costco and Chewy are also pretty good.

I thought you could push a non user writable config that could control pretty much everything since... Forever?

Along with that, push extensions to add DLP bits, etc

Firefox works with uBlock--maybe that alone is enough incentive

Why stop at DB? May as well just write a purpose built OS from scratch.

Good storage won't save you if your schema and queries are bad enough, either.

Mostly agree although you should probably spend a bit more time testing on a more realistic setup.

At least run some plans on the real database with real sized data and real statistics.

Similar to how mocks/stubs are great until they aren't.

That won't magic away statistics problems, but it's still good practice to acknowledge the lower app environments tend to not reflect performance of the production one.

There's a little bit of latitude to push back in code review. "Show me x,y,z you've collected before and after your change"

Really you're just asking for manual or automated performance tests (and you might be requesting different criteria than some already provided)

I recently did a lower tech "write an async divide and conquer S3 lister in Python" and the results were fairly lackluster.

Opus 5.5 and GLM 5.3 Flash came out with pretty reasonable code that was fairly fast. However, all the models left aioboto defaults which have a 10 connection pool size so they ended up all artificially capped even though they added semaphores with higher limits.

My attempt I spent a couple hours AI assisted on ended up being faster, using less CPU time, and having roughly half the code.

Sure but Luna didn't correctly do the orbit paths. On Firefox Mobile (at least), the Luna version they're all sharing the same path.

Probably would work although if you release/distribute/allow others to use it, you're treading into copyright infringement. Afaik personal use if you own a properly licensed copy is usually okay.

Generally it's fair use if it's a legal copy and you have a legitimate reason such as backup/recovery or working around a bug and the models don't care.

If you stumble into vulnerability/cryptography territory sometimes they'll whine.

The ToS seem quite broad and arguably allow people to clone and modify the source (obligatory not a lawyer)

>The Photopea editor can be used by anyone for any purpose, for free. If Photopea allows you to do something, without asking you to pay, you can do it, and you don't have to pay.

Doesn't "edit the source code to remove ads" fall under "any purpose"?