@f0623ab9 @ba2e762b 
Looking forward to it. Sometimes the only way up is through. We can't fix what we don't understand. 

 @527c3522 those look great! What a great use for your printer 

 @516e4ada couldn't agree more. As with so many things, it will likely follow Garners hype cycle and eventually be used but in far more modest ways 

 @cf8c0364 @90431d67 @c4691757 

Thank you! As a UX designer who thinks about coding-syntax-as-design, this feels much cleaner as the core use case, basic nesting, now can look much cleaner. The & character is there for more advanced cases. 

 @90431d67 @cf8c0364 @c4691757 Similar question. I had been reading that the & operator was there to that you could declare things like:

&:hover
&.foo

I'm *not* saying it should, just saying it *did* and now that seems to be taken away. Or am I misunderstanding this change? 

 @0ffd42b0 But that's kind of my entire point, if something as basic as simple nesting requires "reading the spec" to get it right, something is wrong. 

CSS is overly agressive with brevity hacks which causes readability problems and bugs. It's optimizing for writers not readers.

Too many times I've stared at CSS thinking I was looking at a REGEX-style expression, it was far too baroque.

Easy to fix in this case, just use the & all the time. Will likely prevent a few bugs in the process 

 @0ffd42b0 
Thanks! But what about
a {
 b {}
}

That's the same as 
a b {} right?

That seems counter to your example with c & a as the c is outside the scope but in my example b is inside the scope. Both don't use the &.

I'm saying only that not using the & is indeed shorter but leads to inconsistency, which makes the coffee harder to read, especially for novices 

 @e9777705 Yeah, you're not the only one that said this. My point was a bit bigger more like:

How will companies change when we don't need them for wifi as much?

(clearly there will always be dead spots, but for example, it may get to the point that I get a sim card in my laptop and just not care about wifi 95% of the time. 

 @4af4148c 19000.... hours....? That's 2 YEARS of playtime 

 @d73ddf98 This is great, thanks for sharing. 

 @4af4148c The BDO sign up web page fires off every 'micro transaction scare warning' for me. As a general rule I never play micro transaction games. Am I over reacting in this case? 

 @b4acb571 the general work is now in the public domain. There is nothing stopping any other os from implementing something similar. The actual code for my particular prototype is not at all ready for public consumption. My hope is that this encourages new work to happen 

 @367fbbb9 Well said 

 @90431d67 I must thank you yet again. But alas, CSS proves yet again that it eschews clear syntax for compactness.

The UX designer in me would say that div.foo is clear and so is div > .foo 

However allowing a space is needlessly compact, creates ambiguity, and hurts readability. (he says VERY naively) 

 @90431d67 So if you don't mind, is it easy to explain the difference between "div .foo" and "div.foo"? 

 @90431d67 Thank you, that's very helpful. But this makes me realize my issue isn't with & but the fact that div .foo {...} is different from div.foo {...} 

That's even MORE obscure! 

BTW, I want to stress that I'm NOT a professional CSS author. Just someone that is trying to not drown when reading it. 

 @716d58e8 I have been well, thank you. Always up for a chat but no rush of course. Whenever your free. Hope you're well 

 @716d58e8 Oh, and yeah, you CANT migrate content which drives me up the wall. This is especially frustrating as 4.2 just turned on 'allow public search to find my stuff' option so I can actually FIND my older posts now! 

 @716d58e8 The latest release (4.2) has added a few things that help, e.g. you can export lists and such so porting is a bit more bullet proof. 

(I'm not saying it's easy! Just a bit better) 

 @7fe7dd72 I've just confirmed it's working! I searched "from:me <some word>" and it found a post from me 4 months ago with that word in it. So happy it's for all my posts! 

 @7fe7dd72 I just did the same thing. Not quite clear how to test it though. Do you know? 

 @20a8cbf3 The majority of the code is in my prototype app written in a language called Processing. There were a few experiments moving it into Android but I don't believe it made it to AOSP 

 @f8a6718e 
This is what my eyes did when I first saw it. This is so staggeringly bad...

https://social-coop-media.ams3.cdn.digitaloceanspaces.com/media_attachments/files/111/133/145/799/903/295/original/22e13fab59117b47.png 

 @892ebc65 So excited to get this! (and so happy for you of course) 

 @b8c911e8 Something a lot more direct. e.g. the number isn't critical, the project isn't either. A good title sets a hook: What is this?

Here are two to just starters:
1. Open source Survivor Lessons 
2. Hard won strategies from open source 

 @1f8d831d It's a big world out there, with a wide variety of people out there (much more than 'clickers' and 'keyboarders')
My point was to rethink the foundations of text editing, get the core targeting/selection goals down so EVERYONE can benefit. If you want to have something keyboard, focused, that's great! It can be layered on top. I was just starting with the basics and showing one path where users could become more proficient. 

 @12f6653a @7906d04d I would think there are two types of fine tuning: Domain tuning (giving it the right general skills/knowledge for my device) and Personal tuning  (giving it the right skills/knowledge for *me). Obviously interrelated.

My point being a phone can do so many things a chat interface can't so tuning it to dial the phone, open apps, and generally do 'phone things' seems a domain goal.

However, tuning to my friends, word choice, preferences is very much a personal one. 

 @7906d04d @12f6653a thanks! The article discussed putting smaller versions of LLMs on a phone, that's a first start.

However getting one tuned, both to device tasks and to the user's life feels like the next big step. Having it running on a home server has even more potential 

 @18f2a82b @ca5977db Yeah, step 5 is by far the hardest 

 @18f2a82b @ca5977db I just want to back up Tyrone here. This is what I do, I picked a company that just turned it on for me (and setup HTTPS) I had to do my own domain name hacking though (configuring CNAME records) which isn't hard but it LOOKS hard)

I wouldn't recommend the company I chose but I assume there are many, some which I've been told even do the domain thing. Wordpress just auto updates to it's really easy once it's up. 

 @5d116069 This is fascinating, thank you for sharing. Even if it's 10 seconds today, that clearly will fall over time. 

If we wanted an Alexa-like product, I like the idea of a local base model to handle all speech-to-text requests. Out of the box, it would likely be quite bad but if we could train it with a long series of home focused queries (in the form of docs?), that would be interesting. I wonder how hard it is to make it (or likely improve it) to handle a new topic? 

 @63063633 @27e3ee9e 
This is very exciting. I've been looking to volunteer my time to an OSS project and something in this area sounds very interesting as technology like this always has UX issues. 

 @63063633 @27e3ee9e I'm familiar with TinyML, it can run on RPis but they are significantly smaller than the industrial ones running in the cloud.

I realize there may be no clear ways of measuring this yet but even using model size as a proxy, I've heard that ChatGPT 4 is in the terabytes.

Of course, it's not even clear we WANT something as comprehensive at that for many local tasks but I would expect even a reasonable language model is likely to be pretty large 

 @63063633 @9589bde7 @27e3ee9e 

This is amazing thank you! I expected OSS versions but not so quickly. Do we have the terminology to discuss the differences in these models? For example, I can imagine a small LLM running locally but something like Bard or ChatGPT 4 would likely be significantly larger and more CPU intensive. 

 @a6886a04 I didn't realize Grammarly had such sophisticated rephrasing tools. I'll have to give it another look 

 @a6886a04 That was exactly what I was looking for. I'd still not take the rewrite verbatim but it's very close. Thank you! 

 Here is an example, when I asked an LLM to 'clean up' my previous post, this is what I got back:

"I'm aware that LLMs can be powerful tools for editing and revising text. However, I'm also aware that they can sometimes be too eager to make changes. I'm looking for an LLM that can be used to make light rewrites without altering the overall meaning or tone of the text."

It's fine, but WAY too much 

 @64e56151 Because I'd like to go beyond either and do light rewrites as well. I just can't control it 

 As I've already got "just use a grammar checker" replies, I should add that I don't ONLY want to do grammar checking, I'd like to do light rewrites as well. The problem is that any LLM I've tried goes much too far. 

 @a60eb0ca The more personal data gets sucked into LLMs, the more the need for privacy (and likely local use) will grow. We're already seeing LLMs that will process my online documents and answer questions about my work. That level of personalization is only going to grow. But as that data is fairly slow changing  my 'mostly local' LLM could easily add that type of personal information to it's corpus yet still stay both local and entirely in my control. 

 @a0f4d5b3 As I said above, I agree. I'm just saying that many of my queries are totally trivial (e.g. setting a timer, simple math, basic knowledge trivia) and there is a huge UX benefit to these working entirely locally.

There is also a privacy issue but that goes even further. My point isn't the local is BETTER than networked, but that it's possible AT ALL. That strikes me as at the very least interesting (and likely disruptive) 

 @4af4148c It's exactly the same issue with global warming, we can just curl up and wait for death, or roll up our sleeves and actually make the future happen that we want.

I'm not saying this is a given or even easy. I'm just saying the first step is knowing WHAT I want and make sure we agree. There is strength in numbers. 

 @a0f4d5b3 Oh, Icompletely agree, things like weather, traffic, news, will ALWAYS need to have access to the the cloud for up to date information. But there is a very large subset of tasks that will easily be handled locally. And as people become more concerned about privacy, the ability to run the majority of my queries locally will be much more important, with specific 'real time queries' going out to cloud services.

BTW, I'm making a technical point, there is a social/economic dimension as well 

 @4af4148c Two points: first, a 'greedy tech company' could still try to disrupt one of the incumbents by trying to go mostly local. 
Second, what to say that in 10 years ONLY big tech companies are going to be running LLMs? I realize training an LLM is a HUGE effort, today. But they will likely get easier to create over time. 
I hope it's clear I'm not disagreeing with you, you're instinct is correct. I'm just saying we can WANT a different result and work hard to make it happen. 

 As a few replies have asked, I don't mean "never cloud" only that it's possible to run "mostly local" this seems transformative (and already in the direction that Siri and Google are headed)

Just a simple fall back, this would be very helpful. I'm just saying that as privacy concerns grown (and alternatives appear) the ability to be *mostly* local strikes me as a rather disruptive position.

Products like @cbff34d4 are well positioned to take advantage of this. 

 @4af4148c Great to hear! (and something that should be rewarded)
... as soon as I finish Baldur's Gate 3