Recently started Fallout 3 on steam deck, am a few hours in, pretty good game! Love the freedom, the exploration, but it can occasionally be a challenging game with the fighting and various encounters. Story is good so far.

My number one gripe with organic maps is how fragile the search is. If you don’t write it exactly right, you get no or irrelevant results. Also, it seems to have no clue of what is popular and what people expect when they search for something. I’m not talking about personalized results but for example the following: searching for “Eiffel”, leads me to minor roads, restaurants and all kinds of results unrelated to the Eiffel tower. This is what is troubling me the most.

Exactly, this is about compression. Just imagine a full HD image, 1920x1080, with 8 bits of colors for each of the 3 RGB channels. That would lead to 1920x1080x8x3 = 49 766 400 bits, or roughly 50Mb (or roughly 6MB). This is uncompressed. Now imagine a video, at 24 frames per second (typical for movies), that’s almost 1200 Mb/second. For a 1h30 movie, that would be an immense amount of storage, just compute it :)

To solve this, movies are compressed (encoded). There are two types, lossless (where the information is exact and no quality loss is resulted) and lossy (where quality is degraded). It is common to use lossy compression because it is what leads to the most storage savings. For a given compression algorithms, the less bandwidth you allow the algorithm, the more it has to sacrifice video quality to meet your requirements. And this is what bitrate is referring to.

Of note: different compression algorithms are more or less effective at storing data within the same file size. AV1 for instance, will allow for significantly higher video quality than h264, at the same file size (or bitrate).

To be fair, resolution is not enough to measure quality. The bitrate plays a huge role. You can have a high resolution video looking worse than a lower resolution one if the lower one has a higher bitrate. In general, many videos online claim to be 1080p but still look like garbage because of the low bitrate (e.g. like on YouTube or so). If you go for a high bitrate video, you should be able to tell pretty easily, the hair, the fabric, the skin details, the grass, everything can be noticeably sharper and crisper.

Edit: so yeah, I agree with you, because often they are both of low bitrate…

I think that’s what we see with apple silicon, right?

ChatGPT is already free, even GPT-4o since recently.

Granted, you do need an account which requires a phone number, but there are no financial costs.

I don’t know if the content is of particular interest to everyone, but I did want to follow the topic as it was a source of disagreement between FUTO and Rossman and I wanted to see whether they had come to terms.

Do you have a copy of the now deleted/removed video, or are you referring to a videoviseo which is now publicly available? If it is now publicly available, is it at the same link as my edited original post or something else? Thanks!

I have this one sitting on my digital shelf, should probably play it!

Perhaps it would be worth checking with the !homelab@lemmy.ml community and see if they have any advice? They are more likely to have dealt with such situations, by nature.

The Witcher 3. It’s not far from being 10, but got a very nice graphics update for free and has 2 DLC. The game and the DLC and the free graphics update and a very recent mod kit, all for around 10-15 USD right now on GOG . it’s a steal! I highly recommend it. It became my favorite game of all time, very fast. And it will offer around 100h, and it will also offer replayability. What is there not to like?

Edit link

An equalizer does not have to sum up to any specific number. Each frequency range is basically being amplified or attenuated individually. You are boosting or reducing specific frequency ranges. If you reduce them all equally, then the end result is that your song is lower volume. If you boost them all, your song is louder.

Of note: boosting songs may cause occasional crackling sounds. If this is the case it is because the boosting is clipping the top end of the amplitude of your signal at various frequencies. So boost moderately. You are better off reducing some frequencies and leaving the rest normal and increase the volume of the source to compensate whenever possible.

Why? Do Russians not deserve a good browser?

As Mozilla, you probably don’t have much choice. What’s the alternative? Get yourself banned? How is that better? At least this way, locals can use Firefox despite not having access to those extensions. And people find clever ways to get those extensions anyway, so it’s the lesser evil I guess?

Yes, actually it can be quite straightforward. What you are probably best off doing is requesting a google takeout and upload that takeout to Ente directly. I have not followed this process myself so I can’t say much about it but it is described here and is probably the easiest way to migrate:

https://ente.io/faq/migration/from-google-photos/

I’m a customer and have moved over multiple family members, everyone seems happy. Their face recognition and smart search are still WIP, but they are impressively present, despite being all E2EE, by leveraging local processing. They are making very good progress.

I’m sorry, I don’t have any specific suggestions for you, but I am wondering: is there no open source app you yourself wish existed because you would need it?

Working on an open source app because some else (and not you) needs it, is not a good way of staying engaged and caring about the solution. Being the user and target of a project yourself is usually a much netter way of caring and proposing something tailored to at least one individual, maybe more.

Of course, if you are looking for a programming exercise, go for it, but then you don’t need ideas, you can reimplement something which already exists, perhaps which you like, but in your own way. But if you want to have an impact in the open source, it starts by needing something which you don’t really find anywhere and taking matter in your own hands to fix it :) this is not meant to disincentivize you, quite the opposite! I hope you stay attentive to your digital ecosystem to see which holes can be plugged :)

I maintain a private list of ideas I just think of as I go about my day, of things I would like to write/create for myself and while I won’t be going through with all of them, I hope to be able to pick up one or several of them whenever I have time. I can through some ideas here, not as a hint that you should do it (I’ll probably do them myself regardless), but just to inspire you, maybe:

• I am subscribed to a teachable program which has no app and the program is just static information. I want to pull it all and represent it to me offline, not requiring internet to manage my progress. It is also intended to help me archive what I paid for and not depend on the goodwill of teachables to allow me to continue access the resource.
• an RSS feed manager which uses embeddings to automatically organise the content by topic rather than by source.
• an anki plugin to highlight content in the browser based on words from anki that I have and have not learned, to improve my language learning and reading ability.

I have a few more, but this should give you some hints, I hope! Good luck!

To be honest, I’m not sure. Signal does not have an official f-droid build. They distribute an APK on their website for degoogled devices which allows for websockets to be used as the google notifications systems is not used on degoogled devices, but I’m unsure what it means for the rest of the dependency. Some dependencies can be shipped with the apk so do not require google services to be installed, but an expert should probably shime in to shed some light in this situation specifically.

The SMS based pone number verification is one. Another one is google for reCaptcha used to create a signal account. There is also some google dependency for the face blurring. There may be more.

You’re putting a very large amount of trust on something which may simply require the flip of a switch to add the specified information to be sent back to Google along with all the regular heavy telemetry already feeding back…

How much responsibility would a service like Signal have, if they were to inadvertently host a private group for pirated content? I believe signal groups can have up to 1000 members, and these members can be pretty anonymous given the need to only share an ephemeral username which can not be linked to a phone number or any other identity? Can they claim plausible deniability and not do anything?