Marvin

DAalseth said:

iqatedo said:

The_New_tonton said:

I was going to move when he reinstated Trump or if they ignored moderation requests against racists but I quit Twitter when that a hole fired everyone. Now things got way way worse with “Chinese and Russian bots are totally welcome!”

I’m on Tribel now.

Have you looked at Mastodon?

I keep looking at Mastodon but I haven’t found a server. The site says 4.6k servers are out there, but the ~hundred or so they list on their site haven’t interested me. I’m trying to figure out how to find the rest. 

The biggest server is the most like Twitter:

https://mastodon.social/explore

It pulls in comments from other servers so would be the best place to start.

https://mastodon.social/@[email protected]

Every other social media network outside of the top few is going to look like MySpace compared to Facebook though. Just now they have a fraction of the users, 1-2 million at most vs over 300 million on Twitter and it will probably stay that way for a long time. The fact people struggle to figure out how the alternatives work is one reason why. For mass adoption, they have to lower the barrier to entry. Another reason is high profile users/celebrities who want the most followers as they are promoting products and services. Most of them will be staying on Twitter.

Alternatives are also harder to scale because they don't have the investment and hardware to back it up. Mastodon servers were crashing under the load, even at under 1/100th the size of Twitter:

https://www.wired.com/story/twitter-users-mastodon-meltdown/

It seems like Twitter subscriptions to verify users isn't where the high demand is. Usually businesses monetize what's in demand. The demand on Twitter is for people to read high value comments and conversation topics. High value commenters should be verified free and the demand would be for people to pay to view all of their tweets and be able to message them. They could charge tokens/seeds per DM and share revenue with the high profile users and people can buy tokens via IAPs. Their payments/subs can also act as a verification mechanism because bots are much less likely to pay anything.

High value posters can get a gold verify tick (a-list celebrities, big companies, heavily followed/liked commenters etc), maybe silver for politicians/b-list celebs, normal subscribers can get a blue verify tick but won't be mistaken for a high profile user. Non-subscribers would still be able to view content but with more ads, maybe time-delayed tweets.

Twitter is one of those businesses where making money conflicts with their purpose. The aim of the business is to give everyone a part of the conversation but to operate at scale, they need a lot of revenue and they can only make money in ways that make the experience of taking part in the conversation worse.

Twitter revenue was $5b in 2021, $4.5b from ads, net loss of $0.2b. To make a reasonable income like $2b net profit, they'd need around 20m subscribers out of over 300m users. I think that would be easier to do monetizing the demand to read high value comments than the demand to write under a verified profile but it seems attainable one way or another.

AppleZulu said:

Marvin said:

AppleZulu said:

Marvin said:

9secondkox2 said:

But but didn’t hair force one give a number of RECENT interviews stating apple has no intentions to merge Mac OS and iPad?… LOL 

Federighi manages both macOS and iOS development and won't merge the systems. Both systems already run on M1, macOS can run iOS apps and iOS supports external displays. All that's needed is boot support for macOS on iPad hardware.

It doesn't make sense to run macOS on all the iPads because there's not enough storage and RAM on all of them but the iPad Pros go up to 16GB/2TB.

The Mac install and boot can be controlled by iPad OS, which can be running at the same time and require that an external display be plugged in to show macOS. iPad OS will continue to be displayed on the touch screen and macOS will display on the external display. macOS will be able to open supported iPad apps and files/email/calendar/music can be shared between both systems.

This makes a higher priced iPad Pro (1TB 12" is $1799) way more appealing for some people as they don't have to get both Mac and iPad. They can sketch in Procreate with the Pencil and import it into Photoshop on the external display on the same device.

While the iPad is connected to the display, the mouse would seamlessly switch between both systems and allow drag and drop. There's no adding touch to the Mac (which they can't due to the UI) and iPad OS stays the same but the user gets the best of both. It allows software development because macOS will allow compilation in the background. On the go, someone can code in an editor on iPad and compilation can run on the Mac system without breaking the iPadOS sandbox and the Mac can install it on iPad for testing if it's an iPad app.

It would also be entirely optional. People who are happy to use iPads with just iPad OS can do that. People who need the Mac side get this extra capability with minimal effort. There would be a macOS for iPad app that installs the OS and allows booting it in the background. The app will say to connect a monitor and it can switch between using as an extended display for iPad OS and showing macOS.

It would be the lightest Mac at half the weight of the Air. This is zero convergence between iOS and macOS, it's just allowing extra optional capability to the iPad hardware.

This just makes my head hurt. That’s not how any of this works. Here we have a bloatware, dual simultaneous(!) OS iPad that requires an external screen to use the second OS. It will not “be the lightest Mac” because you have to lug around a separate display. It will be less convenient than any MacBook because the Mac display is a separate device. Simultaneously running two operating systems, it will be slower than either individual device, when it’s not crashing from simultaneous but incongruous demands on the processor, memory, etc.

All of the supposed benefits of this arrangement are achieved way more effectively by using an iPad and a MacBook Pro. You can already transfer documents from one to the other seamlessly, and each has the full attention of its processor and memory to carry out tasks. 

Also, Apple operating systems are not “free.” They are included in the price of Apple hardware. Apple cannot simply give away MacOS as a second operating system for free as an iPad app.  

I left the MS Windows world behind long ago, in no small part because I really hated the ubiquity of the “workaround” as a means to make your computer do things. 

Apple’s ‘It just works’ ethos is the antithesis of that. It’s reliable and does the things it’s supposed to do, and you’re not meant to ‘get under the hood’ to force it to do other things. The hardware and software are designed together to do the things that it does well, and not to chase every possible permutation of bells and whistles in an effort to be all things to all people. It’s how Apple avoids bloatware, hardware and software conflicts and system freezes and crashes. 

It’s mind-boggling that supposed Apple enthusiasts are oblivious to that ethos or don’t understand at all how it’s achieved. So they come up with “ideas” to make Apple become complicated, buggy, and with this gem of an idea, primed for an awkward, clunky workaround on the most fundamental level. 

You wouldn't have to lug a display around - one at home, one at work - and there would be nothing stopping them displaying the Mac view on the iPad without touch support but it's more intuitive for it to be on an external non-touch display and leave the iPad fully controllable.

The commands to the processor and storage don't have to be entirely separate, both systems share the same base architecture.

It would be slower than separate hardware but not unusably slow. Both iPad Pro and Macbooks are way ahead of basic needs in terms of efficiency and people buy both with 8GB so running two systems with 16GB RAM is fine. iPad OS runs fine with under 2GB RAM.

The proper secondary display support that was added to iPad OS has improved things a lot so continued improvement there might be enough but it will always be lacking full-featured apps and software devs are one of Apple's biggest customers:



Apple's marketing pushes iPads for students and they are pushing software development as being one of the most important skills young people can learn but they don't allow code compilation on iPads, no C/C++/NodeJS/Python. A dual system setup is an instant fix for this.

Maybe it doesn't need to be the whole OS running. Mac apps can run in full-screen mode. There may be a way to have a Mac app run in a pane similar to an iPad app. It would just need to be clear that the view wasn't touch enabled. This would be more like running Mac apps on the iPad than running macOS in a similar way to running iOS apps on macOS. The user wouldn't have to think about the OS part and would just open a Mac app like Final Cut and iPad OS takes care of the background system requirements.

You’re describing a combo device that would sort of deliver both operating systems, but in a compromised package that is markedly inferior to the things it’s supposed to replace, supposedly to offer consumers a cheaper option than using separate iPads and MacBooks. 

When has Apple ever introduced a new or expanded (but compromised) device in order to pursue the low-end of the market? Sure, they’ll offer a version of a prior year’s watch or phone at a lower cost, but that’s not what this is. 

You’re suggesting that Apple cannibalize their MacBook line to offer a slower, inferior MacOS experience as a free add-on to a MacBook Pro. There is nothing in their history to indicate they would ever do something like that. 

They allow iOS apps to run on Macs:

https://developer.apple.com/macos/iphone-and-ipad-apps/

This is just the reverse.

It wouldn't cause much cannibalization. People haven't stopped buying iPads now that they can run iOS apps on Mac.

Apple also had Bootcamp before, this would be much less involved than Bootcamp.

netrox said:

Marvin said:

Apple's marketing pushes iPads for students and they are pushing software development as being one of the most important skills young people can learn but they don't allow code compilation on iPads, no C/C++/NodeJS/Python. A dual system setup is an instant fix for this.

Maybe it doesn't need to be the whole OS running. Mac apps can run in full-screen mode. There may be a way to have a Mac app run in a pane similar to an iPad app. It would just need to be clear that the view wasn't touch enabled. This would be more like running Mac apps on the iPad than running macOS in a similar way to running iOS apps on macOS. The user wouldn't have to think about the OS part and would just open a Mac app like Final Cut and iPad OS takes care of the background system requirements.

So, maybe if they could modify iPadOS to allow programmers to run those popular programming languages in a small generic VM (like Docker), that would be a huge step. Especially with web development.  

That's a possibility. Running a VM would get another step towards development support. If they allow the following to run, which somehow is allowed to run Python:

https://apps.apple.com/us/app/ish-shell/id1436902243
https://ish.app

they should be able to support their own VM for compilation and could allow apps like After Effects and Blender to work, as well as web development.

nubus said:

TheObannonFile said:

Perhaps Apple will start Mac Pro on 3nm, possibly branded as M3, or maybe just it’s own thing separate from the M series

Could Apple kill Mac Pro and let Mac be "limited" to Studio with Mx Ultra for now? Having an enclosure with space for PCI cards and storage might not make sense as the GPU is fused to the CPU. Add driver problems and there might not be any cards to use. USB 4v2 and "Next gen Thunderbolt" are not ready to ship but those could remove the need PCI-based cards. Doing a Mac Pro M2 with an enclosure and then kill it next year seems like a lot of work. Doing a Mac Studio with Ultra M3 and saying "this is the future"... could it be an option?

They could have done this but not now that they said they would make a Mac Pro. The Mac Studio is powerful enough to cover the same performance level that most Mac Pro buyers are buying. The highest end Intel Mac Pro is still another 2-3x faster.

tht said:

TheObannonFile said:

Perhaps Apple will start Mac Pro on 3nm, possibly branded as M3, or maybe just it’s own thing separate from the M series

They still have to answer how more GPU performance gets in the box. 172 cores isn’t enough. It should be however many that a 1400 W PSU can power. So about 600 GPU cores?

While 172 cores wouldn't quite reach high-end PC workstations, it would still make for a very fast Mac. This would be 50-60TFLOPs. A 4090 is around 80TFLOPs and a 4090ti around 95TFLOPs. Being within 50% of the highest-end competing GPUs is good enough IMO.

For marketing purposes, it would be nice to be able to claim the first 100TFLOPs GPU. An overclocked 4090 can reach this. This would need 320+ cores i.e quad Ultras instead of dual Ultras (around $17k+). I don't think the market is enough to justify it any more. It's worthwhile in some markets but it means designing a whole machine around this small use-case.

Dual Ultras can be done much more easily and 172 cores is likely based on M2 (N5P), 25% more cores than M1.

M1 (8-core) -> M2 (10-core)
M1 Pro (16-core) -> M2 Pro (20-core)
M1 Max (32-core) -> M2 Max (40-core)
M1 Ultra (64-core) -> M2 Ultra (80-core)
M2 Extreme (160-core)

M3 allows for 70% increase in density over this (272-core).

I think most Mac Pro buyers would be happy with 172-core GPUs for under $10k.

tht said:

blastdoor said:

If the Mac Studio cooling solution can handle the ‘extreme’ I hope they make it an option. Personally, I don’t need whatever extra expansion a Mac Pro might provide — I just want the CPU power. But I’m curious to see what else the Pro offers beyond the SOC. 

The biggest issue they have to work is the poor performance scaling with GPU cores, and putting in more CPUs and GPUs in the box than the native one. The Ultra really should have GB5 Metal scores around 120k to 140k. It scores 90k. That is very poor scaling.

There does seem to be something off in some scenarios. In the Blender Metal addition they have been making adjustments to make better use of the cache and getting a 15% gain. Hopefully they are finding where the software isn't scaling well and either fixing it in the drivers or different hardware design. M2 real-world performance jumped 40-100% vs M1 in some cases so it seems like they are fixing things beyond raw hardware.

It would be best having a user control for how much ANC to apply. Wireless ANC buds make really good ear plugs when you just want to block ambient noise like in the following video:



This could be controlled in the iPhone settings. If someone is drilling, banging, using a hedge trimmer, vacuuming and you don't want to be listening to music, strong ANC can block it. If the ANC isn't strong enough, it needs music playing on top to block out the noise.

Fidonet127 said:

As others have pointed out, this only took a year to port to Mac, vs a much longer time frame for other ports. This game uses metal 3, which isn’t out yet to the public. Ventura with Metal 3 should be out Monday and the game on Friday. There is no conspiracy here. 

Yeah, it uses the Metal upscaling feature for faster performance, the following suggests 2x performance and would be needed to be competitive with similar features like DLSS from Nvidia, FidelityFX from AMD and XeSS from Intel:

https://developer.apple.com/documentation/metalfx
https://www.nvidia.com/en-us/geforce/news/dlss3-ai-powered-neural-graphics-innovations/

According to Capcom, it allows 1080p on the Air and 4k on the Studio:

https://youtu.be/q5D55G7Ejs8?t=5081

The Nvidia 3050 can run this at 1080p/60 and is around the same performance as the M1 Air. The 3060 (M1 Max equivalent) can handle 4K.

There's also a final DLC coming on October 28th, I'd guess it will be available for Mac too:

https://www.ign.com/articles/resident-evil-villages-shadows-of-rose-dlc-will-conclude-the-winters-family-saga

With the Resident Evil engine ported and supporting Metal, they can port other games, including RE2/3 and the upcoming RE4 remake would be able to get a same day release.

https://en.wikipedia.org/wiki/RE_Engine