inglor

It's hard to call something virtually every other systems language does 'novel' given for example C++ has async/await though?

C# does this already https://learn.microsoft.com/en-us/dotnet/csharp/language-ref... - works pretty well. It's good that Java is finally getting more ergonomic these past few years.

AoC questions this year were _deliberately written_ to be confusing to LLMs, it's not failing because it's worse it's failing because the questions were written to make it hader for models :]

Edit: apparently not, the author is just really good at coming up with ai adverse puzzles. When testing ChatGPT did much better on last year’s puzzles.

Microsoft mostly gives different groups autonomy in what tools to use, where to host etc.

Most groups in Microsoft use the Microsoft stack as it's best supported and already compliant and works pretty well - so most groups use teams, office, Azure etc.

Groups are free to use other tech and may for example use zoom+slack (some groups do) or GSuite. This is also true for technology (e.g. a team can use yarn and flow instead of npm and TypeScript). It's just very rare they do.

Umm Microsoft employee here, we were never allowed to use ChatGPT on work related stuff (e.g. post code into it). Instead we have our own separate instance of ChatGPT we can use. Additionally CoPilot is fair game since we own that.

This isn't a new policy and has been the case for at least a year.

Good point, and conversely, web components can't since they must be registered in JavaScript before the side loads.

Hydration solutions exist but are usually worse and come with surprising tradeoffs and harm composabilitiy.

We actually ship a tap reporter in Node and it used to be the default until recently :)

We have at least ±5 more changes on the way to improve streams significantly (Node streams, then web streams), reduce the size of each stream and hot-path common cases.

There is already a PR on the way to fix the performance regression with what Jest s doing https://github.com/nodejs/node/pull/49950

Note we also ship a test runner ourselves pretty recently and it may suit your needs.

New contributors often have the opposite problem where after an initial contribution they're looking for areas to get involved with and time with a maintainer is valued. Usually the missing part is maintainers willing to do this.

(Unless you mean literally personal secretary and dealing with their personal life or setting meetings up and stuff, I assumed you meant tell you what bugs are worth looking at etc)

The linux kernel isn't a reward and while I empathize with the OP I would absolutely prefer a better solution to maintain than one from the community. So the maintainer was well within their rights.

That said, it's so easy to just give people who contributed to the fix attribution, in Node.js we make a point of trying to give credit in release notes and add `Co-Authored-By:` to PRs where we took _some_ work people did in PRs and adapted it.

When you maintain something for a while, credit often stops being important to you (you _are_ the maintainer of the area after all) so it's hard to remember that for new contributors it's often very important.

Totally a loss for OP and that part of linux that the maintainer wasn't more attentive to the fact sharing credit (especially when deserved like in this case) was important to OP.

They have an excellent LLM called Claude which is quite competitive with OpenAI offerings and has long context windows.

Other than OpenAI’s models they have been able to deliver the best results in projects I worked on and sometimes they outperformed OpenAI models too.

I’m happy to see this as this gives them funding to further train expensive models and compete - their product is really nice at the moment but could really benefit from bigger size and even bigger contexts.

If you could easily spin up hundreds of thousands of threads you wouldn't need any of those frameworks, callbacks or userland threads. You could use the OS level abstraction and not have your language/framework implement its own different abstraction.

ChatGPT but instead of learning from the internet and then producing text, learning from the internet + < your additional information > and then producing text.

It's an announcement about the availability of a feature to do that. The article doesn't mention the biggest issue with fine-tuned models though - cost.

My guess is that it makes it easier to work on large files which is common without downloading the whole file and it makes CoPilot related efforts much easier.

I work on Excel but not this feature and am happy to ask around if people are interested.

Hi I work on Excel, can you write a summary of why that's important to you and send it to bgruenbaum@microsoft.com ? I'll get it in front of a PM.

Alternatively PMs read all the feedback sent via the feedback UI on Excel web and desktop so you can use that instead if you prefer.

It's almost too good of a prediction since XAML is from 2008 and this article is from 2006 :)

(To be clear it means the article is likely _not_ from 2006 but from later like 2008-2010 since it mentions XAML)

You know, I suspect most people reading this weren't around for the "XML everything!" era and don't get the joke. They legitimately think it's a post about an experimental markup language.

We're very much stopping "thought pollution" all the time, we're not too great at it but the equivalent to someone forcing sound on you is someone forcing thoughts on you.

We have laws and automatic tools (again, not great ones) to find propaganda and lies and try to stop them. We absolutely should let people decide what information to be exposed to better.

How is enforcing the law to not make loud noises that make everyone else's life around you worse a bad thing?

What justifies someone suffering from PTSD, with a sleeping baby in their home or just enjoying life being put through noise pollution just because someone wants to make loud noises?

Amazing, it's so great to see this done and I wish sound pollution was taken nearly as seriously where I live.

Another cool thing is when I visited Japan and I learned a lot of the tools and drills used in renovations have more quiet/less noisy counterparts so when someone remodels a room in an apartment building or house next door it doesn't actually have to ruin your work-from-home day.

> people behind a firewall wouldn't be able to actually receive any audio/video streams from Jitsi.

That part isn't really correct? Most firewalls block incoming connections not outgoing ones so both parties agree on a port, hole punching happens and then there is a connection is formed.

It probably does STUN but not TURN and relies on the actual data going through P2P, that's what most services that are free do and for a lot of orgs/cases it works pretty well.

If no one answers I'll ask internally (I'm in Microsoft but just an engineer and not in teams), I can tell you that in our org we don't have access to customer data, actual customer data is encrypted anyway and every access to a production machine is recorded, audited and has to be justified legally (and is very rare to the point I haven't _actually_ seen it done except to literally learn how to go through the process on a machine with no dat).

I'm also curious what the ToS says about teams, if it allows it I'd be pretty disgusted and I was very surprised Zoom does this..

Intel has a lot of hidden unemployment, internal politics and people who honestly don’t really like working.

Intel hires about half of the graduates from my uni, everyone who likes coding fled to startups or super specific teams.

There are strong engineers in Intel for sure but it has a pretty awful culture.

Nvidia’a isn’t great but it’s heaps better in terms of actually taking results into account.

My personal experience is only with the local (Israel) branches of both.

Wokwi is free open source and neat with an MIT license https://github.com/wokwi/avr8js https://wokwi.com

Exactly, the boilerplate overhead isn't worth it when 95% of times you don't have anything meaningful to do with the exception but let it propagate.

I can explain how Microsoft makes money off its anti-virus if you'd like. It's only free for regular users but as soon as you need enterprise security measures (e.g. tracking exposure across your org, isolating parts of your network etc) it costs money.

So the built in anti-virus is in fact "free" because it makes selling the enterprise version easier.

I'm just a software engineer working for Microsoft and I'm on HN since I worked on 3 startups (one YC funded) and do a ton of open source (in my free time, Microsoft funds none of it).

I enjoy working for Microsoft (mostly) but I have _no idea_ how our sales looks like.

> You likely didn't have permissions to create branches

I did, it was just a very long and complicated process. You had to set up a lot of tooling for it and you were strongly discouraged from doing it so in the year of SourceDepot (on Office) I saw this option being used exactly once.