krzyk

> As long as the company states the time period

But they don't, they have the "30 days", but just after that they add "unless ....". So the time period is vague.

> how long they keep it

AFAIR it is not clear, because they write it is "30 days, but ...":

> After 30 days, the data is deleted automatically, except in the rare cases where it's part of a safety investigation or we're legally required to keep it.

So you have a vague clause saying "when" and vague clause saying for "how long". If it will fly I would be surprised.

> Note that Anthropic has committed not to train models on logged data, so I don’t understand some of the concerns here. What exactly is your threat model? That Anthropic would train models contrary to their terms of service? That you trust them enough not to log your data prior to this, but not enough to trust their stated limits on how logged data will be used now?

It is a different thing when they say they don't store your data.

And when they say they store your data for 30 days and review it for "issues", it makes your "spider sense" tingle. Who and how will review it, what are the "issues" they are looking for, etc. It is to vague and they can keep it this "dangerous" model for themselves.

If it was more painful for the suing party, it was a good thing.

Now all it needs is send some letter to github and they take down content, no questions asked, how is that better? It is for court to decide.

old Siri was always limited in that regard.

And Apple Intelligence supports just a fraction of languages: English, Danish, Dutch, French, German, Italian, Norwegian, Portuguese, Spanish, Swedish, Turkish, Vietnamese, Chinese (simplified), Chinese (traditional), Japanese, and Korean.

Danish, Dutch, Norwegian and Swedish have quite small populations compared Poland, while I don't see Polish there (37M). I also don't see Romanian, it is slightly bigger amount of people than Netherlands, and the rest from that list are ~< 10M.

Oh, well, at least I don't see Russian in that list.

With ChatGPT, or Claude.ai (or Deepseek, or local models) I can speak with languages that are outside of (traditionally) limited set of Apple. Because it all depends on what is on the web and web has magnitude larger set of languages compared to what Apple provides.

Unfortunately there are more and more users on Linux and as a result windowsization/macosization of Linux is in progress (systemd, wayland, some scary stuff Poettering is doing with boot, snap/flatpack).

Oh, I hoped it would be macos contained in those containers. Container for Linux are in millions, while I don't know if there are any with macos inside.

Has anyone seen a recent LLM release that supports just one language?

Not quite. It is up to Apple to design a system in which operators (even Apple) can't see your data. Apparently they designed it in a way that operator can see it (so it is cool if it is Apple, but not cool if it is someone else).

Core not, but here it is. Apple designed the system in a way that the operator can invade your privacy. So if only Apple is the operator it is "OK", but if they allow other operators it is not.

> As a small developer, the cost to support something like this would be so overwhelming I wouldn’t consider supporting the EU officially.

Would you consider supporting US laws?

Doesn't "DMCA", make, well, the DMCA notice a thing?

If you install LSP. AFAIR their first versions used some kind of treee/structure for easy search, but they found out that grep was better/similar but with less complications (they now ship some kind of grep).

But EU has a pretty uniform laws, so this comes to just one issue: time. Did Apple start working on this feature before EU implemented the law? This might be the case, but even if it was after, they could start working on implementing that sooner.

It would be good for US companies to know that EU laws are not "guidelines", just as US enforces their laws on companies from outside.

Yeah, Siri was such a poor solution compared to Google (and Google's is also poor in EU) that no one would make a campaign.

If Siri wants to be seen as anything it should first support every EU language and they can work from there.

Some features don't land in Europe because US companies can't handle the amount of languages. For them it is English and maybe Spanish or Chinese because they don't care how heybmake money.

Why does systems are not designed take into account that compliance work?

Oh nice, maybe they'll finally make Siri learn foreign languages?

> The main reason must companies don't publish Linux electron apps is fragmentation.

Unfortunately they do, instead of implementing a proper native application (don't hire web developers to do native apps).

Look at e.g. Cisco Webex, or Telegram. You can make a proper native application that just works and is not electron. A lot of people don't like electron apps, and it is more important in current economy (memory prices are too high to invest in more RAM).

Yeah, for people that work in orgs that are < 150 people. For others it pay by token (API pricing).

I'm curious how does it work, you handover the tools you wrote, .bashrc/.zshrc, etc?

When I'm hired in a company (not contract), they wipe the harddrive when I leave (well, I also do it before I hand it over sometimes). So they don't get the tools (I take them with myself, it would be a waste to loose them)

Results also depend on the prompt. You get different results if you ask to review the PR and focus on particular file than if you don't make it focus.

Or if you make it "be a security engineer" with particular focus points.

Or make it a grammar nazi, it will find way more typos than without such focus.

Of course all of those "focuses" needs to be in a separate context (agent/subagent) to make it work.

It can be used outside of local machine.

We built something similar, it looks for new PRs where the bot is added and does reviews. Makes the code more tuned toward similar rules. I can't assume that a developer run a code review tool himself (just as I don't assume he/she run a build - so we run builds also).

It is just another perspective for code review, besides human. Unfortunately it uses a lot of tokens, and considering that Anthropic, OpenAI and Github Copilot all moved to token based pricing, it is quite a money burner.

I think that in most cases you either agree on a PR comment or you don't. But it has to leave a mark in PR. This is how we do reviews, ignoring PR comment is one of the worst offenses one can make. I don't let it go.

I wouldn't call input a vast majority of token usage. In my experience output is 25%-33% of the cost.

Opus is $5 per mtok of input tokens, but $25 for output.

It depends what OS you need/want.

Some people don't want macos.

I can install Windows or Linux on Framework.

QQQ?

Well, they are where I am. But LLM reviews doesn't solve that. It just adds another perspective, which catches different issues.

Companies I remember: CD Project RED, but they are now switching their newest game to Unreal Engine.

id Software, the new Doom series uses highly performant engine (as if there was some legacy there for that).