BitSound

joined 1 year ago
 

They're type of band that this could be a prank, but it seems legit. Going to be hard to replace him

[–] BitSound 2 points 1 day ago

Not sure how ollama integration works in general, but these are two good libraries for RAG:

https://github.com/facebookresearch/faiss

https://pypi.org/project/chromadb/

[–] BitSound 3 points 1 day ago (1 children)

I like the strategic aspect of knowing your chances with what spells are remaining, but I already have a hard time coming back to a run and forgetting details like that. Maybe if the book could show what spells are remaining.

[–] BitSound 3 points 2 days ago (1 children)

I've been corrupted by too many of these posted over in [email protected], because I was really expecting to see something like this:

[–] BitSound 2 points 2 days ago* (last edited 1 day ago) (1 children)

Hell yeah 🤘 It all kicks ass, but that's an impressive vocal range. Wonder if he's going to get any haters for the little bit of pig squeal at the end. Also not really the shirt I expected to see on the vocalist

[–] BitSound 1 points 2 days ago

That's a great line of thought. Take an algorithm of "simulate a human brain". Obviously that would break the paper's argument, so you'd have to find why it doesn't apply here to take the paper's claims at face value.

[–] BitSound 1 points 2 days ago

There's a number of major flaws with it:

  1. Assume the paper is completely true. It's just proved the algorithmic complexity of it, but so what? What if the general case is NP-hard, but not in the case that we care about? That's been true for other problems, why not this one?
  2. It proves something in a model. So what? Prove that the result applies to the real world
  3. Replace "human-like" with something trivial like "tree-like". The paper then proves that we'll never achieve tree-like intelligence?

IMO there's also flaws in the argument itself, but those are more relevant

[–] BitSound 2 points 3 days ago (1 children)

Not in general, sorry. Best bet is to make sure you're using the most recent kernel, which Ubuntu tends to lag on. You can also try checking out the arch wiki entry for it. It's a different distro, but the wiki is good and commonly has tips relevant for any distro.

[–] BitSound 1 points 4 days ago

What kernel are you running? From what I understand, that should be the major differentiator if you're not using S3.

[–] BitSound 1 points 4 days ago

Couldn't tell you unfortunately. It looks like AMD is also on board with deprecating S3 sleep, so I would guess that it's not significantly better. The kernel controls the newer standby modes, so it's really going to depend on how well it's supported there.

[–] BitSound 14 points 4 days ago (12 children)

Sleep kind of sucks on the original 11th gen hardware. They pushed out a bios update that broke S3 sleep, so now all you've got is the s2idle version, which the kernel is only OK at. Your laptop bag might heat up. S3 breaking isn't really their fault, Intel deprecated it. Still annoying though. I've heard the Chromebook version and other newer gens have better sleep support.

Other than that, it's great. NixOS runs just fine, even the fingerprint reader works, which has been rare for Linux

[–] BitSound 12 points 5 days ago

Meshuggah:

https://www.youtube.com/watch?v=m9LpMZuBEMk

Listened to them before I got into metal, came back to them later and now love them. That's from probably one of their more accessible records, they also have more experimental stuff like this:

https://www.youtube.com/watch?v=Dw3SdOFmubU

[–] BitSound 11 points 5 days ago (1 children)

Do you have any links to read up on him? I know this is a very contentious topic, but I haven't heard much about him and I'm curious. What would you hold as his worst actions?

 

Mindustry dev has had enough

 

I've encountered some conflicting usages of Tag:landuse=residential. Some areas are very specific, and broken down into individual blocks, while some areas cover multiple blocks. Here's an example of both styles adjacent to each other:

https://www.openstreetmap.org/way/653823458

https://www.openstreetmap.org/way/652122607

The wiki doesn't really say much on the topic. Does anyone have opinions/rules of thumb on how to tag them exactly? It seems like all adjacent areas not separated by major highways should be joined together?

I've encountered some residential areas that are broken down into mapping each block, and literally follow the curb, rounded corners and all. That seems too specific?

 

I'm looking at Tag:crossing=marked, and it's a little vague. It says:

Set a node on the highway where the transition is and add highway=crossing + crossing=marked.

If the crossing is also mapped as a way, tag it as highway=footway footway=crossing crossing=marked or highway=cycleway cycleway=crossing crossing=marked as appropriate.

Doesn't that violate the principle of One feature, one OSM element? For example, here's a crossing from where overpass-turbo defaults to showing:

https://www.openstreetmap.org/node/7780814396

https://www.openstreetmap.org/way/833493479

You've got a way with these tags:

crossing=marked
crossing:markings=yes
footway=crossing
highway=footway
surface=asphalt

And the intersection node with the street it's crossing has these tags:

crossing=marked
crossing:markings=yes
highway=crossing
tactile_paving=no

Shouldn't that be one or the other? It makes sense to me to represent the crossing as a way with all the tags, and leave the intersection untagged. I noticed though that StreetComplete doesn't really like that, and will give you quests to add tags to the intersection node even if the way is properly tagged.

 

Original comment:

I don’t know much about voting systems, but I know someone who does. Unfortunately he’s currently banned. Maybe we can wait until his 3-month ban expires and ask him for advice?

Previous discussion

 

I've got a patio for a restaurant tagged as leisure=outdoor_seating. That page says you can add operator=* as a string, but I'm wondering if I can add a Relation between the patio and the restaurant. This is really for semantic reasons, because if the restaurant changes its name or gets a new owner, it would be nice if the patio didn't then have out-of-date information.

I don't see a Relation type that's relevant. I don't want to just start doing my own thing, so does anyone know of a way to use a Relation here, and if not, is that something that can be proposed?

Thanks for all of the responses on my other questions, btw. This community has been very helpful.

 

I'm taking a look at traffic circles like this:

https://www.openstreetmap.org/edit#map=19/33.790043/-118.142392

The main traffic circle has been split up into 8 different segments, so that individual segments can have Relations added to them, such as the "Long Beach Transit 174" bus route. I'm new to mapping, so I don't really know what to expect, but it seems odd to split it up like that. It ends up adding noise to StreetComplete, in that I can't just say "yep, this traffic circle is asphalt", I have to go to a bunch of tiny segments and mark each one of them as asphalt.

I've also seen this for items generated from Lyft data, where a single road gets split into tiny segments so that one part can be marked as "no u-turn" or "no left turn". StreetComplete wants me to mark each tiny segment individually.

 

I'm looking to tag a simple 4 way stop with typical US red/yellow/green traffic signals. I was wondering what the difference between signal and traffic_lights is in iD, and the wiki page just says this about traffic_lights:

A typical traffic signal. This value was the second most common value as of 2021-09-15 despite being undocumented until that point.

Looking at the talk page there, it links to this post, where an iD dev seems rather annoyed at the wiki:

I took a look at https://wiki.openstreetmap.org/wiki/Key:traffic_signals and now I'm furious.

Forget it.

There is no way I'm going to support traffic_signals=yes for pedestrian signals, after the wiki folks aren't even ok with iD using traffic_signals=signal for a normal traffic signal - a tagging that was accepted just not very widespread before iD started doing it.

The OSM Wiki needs to end. Seriously. It's ruining this project.

I'm using iD, so should I just leave it as the default signals and leave the fighting up to the devs? As an aside, does anyone know why there seems to be so much animosity there? Kind of surprising TBH

 

I've encountered a bus stop that still exists, but has a sign from the city saying that no busses stop there. There's the disused tag on the wiki which seems relevant, but I'm not sure how to tag it exactly. There's lots of tags like ref, route_ref, operator:wikidata and so on. Should all of those tags get prefixed with disused:?

 

I'm trying to correct local buildings on OSM. I've noticed that some of the buildings were traced before according to one set of satellite images, but are off according to others. One of the options for a background while editing that I've got is called orthoimagery. Can I assume that that is the best set of satellite images for tracing buildings from?

view more: next ›