4 ms·

Shocker: A $75 multi-purpose device could not adequately replace his dedicated $1500 device. I don't know what this was intended to prove, exactly?

avip 6y ago
(̶4̶5̶$̶ ̶i̶n̶c̶l̶u̶d̶e̶ ̶s̶h̶i̶p̶p̶i̶n̶g̶)̶ Edit: wrong price quoted for the 2GB model.
- penagwin 6y ago
  That's the 2GB model
- hyperbovine 6y ago
  $75 for the 8gb model. Plus shipping
finnthehuman 6y ago
Yeah, I don't get it either. Has the Raspberry Pi marketing changed significantly that I'm unaware of? I own a few raspberry pis and never saw them as more as either a minimalist toy, cheap learning tool, or easy-entry DIY embedded platform.
- cannam 6y ago
  An 8G Pi-4 is a pretty powerful computer by Linux-running historical standards. This seems really about application compatibility, not power.
  - sukilot 6y ago
    Most people run modern software and data not historical software and data.
colesantiago 6y ago
I'm not sure why anyone would do this other than to waste time, to each to their own I guess...
- Avicebron 6y ago
  Sometimes tinkering with electronics for it's own sake is a form of reward. Building anything really is an experience in of itself
- AdmiralAsshat 6y ago
  From actually reading the content of the article, it seems like it would've been better titled, "I tried to use a Raspberry Pi 4 to do my dev work for a day." And that might have better captured the point, since I imagine a number of us have experimented with using a Pi for work, for personal projects, etc. in the past. But the current title to my mind is a tad clickbaity, for the implied comparison I noted already.
pizza234 6y ago
The point is different, on a different reading. > I guess one point I should make is that for almost every piece of software I wanted to use, I had to spend a lot of time just trying to find any that would work on Linux—then narrowing that to 'on Linux ARM64'. And then I had to usually spend a few minutes compiling it from source, placing my own shortcuts on the system (so I wouldn't have to open a Terminal every time I wanted to check Twitter), etc. > All of these issues (4K difficulties, having to compile apps, not finding apps) are exacerbated by the fact that the Pi runs on ARM, but it is still a problem in the wider Linux ecosystem. The point is that the above could apply to a $1500 ARM computer. However, I suspect that he's not very experienced with Linux, and ultimately makes a comparison that doesn't make much sense. > Anyways, on my Mac, I'm used to Shift + Option + dash for em, or Option + dash for en, but we're getting off the point. On GTK, one can use a compose sequence for that. There you go: https://askubuntu.com/a/31265 https://askubuntu.com/a/31265. > All of these issues (4K difficulties, having to compile apps, not finding apps) are exacerbated by the fact that the Pi runs on ARM, but it is still a problem in the wider Linux ecosystem. "Having to compile apps" and "not finding apps" are not problems in the wider Linux ecosystem. There's an enormous amount of stuff in the repositories of the main distros, and it's prepackaged. And there are external repos for what's not in there. The software one really has to compile manually is a very small minority (again: in the wider Linux ecosystem, that is, on x86). For the 4K, I can't speak, as I have no experience. Ultimately, using ARM for a desktop machine with specific software requirements doesn't make much sense; in particular, because his work(flow) revolves around closed-source software. However, that's what the explicit target of the article, so I can't really complain about it.
Scarbutt 6y ago
I don't know what this was intended to prove, exactly? Nothing, just another bad blog trying to get youtube viewers.
SquareWheel 6y ago
What is with Hacker News and these pessimistic hot takes? He tried it to see if it would work. That envisions the hacker spirit better than most ever accomplish.
Koshkin 6y ago
Looks like what they proved is that the difference is less than 20x.
- geerlingguy 6y ago
  The performance was the biggest surprise. Outside of 4K video, the Pi 4 was within at least javelin distance (not quite spitting) for a number of benchmarks I ran. Much better than the cost difference would indicate.

avip 6y ago

(̶4̶5̶$̶ ̶i̶n̶c̶l̶u̶d̶e̶ ̶s̶h̶i̶p̶p̶i̶n̶g̶)̶ Edit: wrong price quoted for the 2GB model.

penagwin 6y ago
That's the 2GB model
hyperbovine 6y ago
$75 for the 8gb model. Plus shipping

finnthehuman 6y ago

Yeah, I don't get it either. Has the Raspberry Pi marketing changed significantly that I'm unaware of? I own a few raspberry pis and never saw them as more as either a minimalist toy, cheap learning tool, or easy-entry DIY embedded platform.

cannam 6y ago
An 8G Pi-4 is a pretty powerful computer by Linux-running historical standards. This seems really about application compatibility, not power.
- sukilot 6y ago
  Most people run modern software and data not historical software and data.

colesantiago 6y ago

I'm not sure why anyone would do this other than to waste time, to each to their own I guess...

Avicebron 6y ago
Sometimes tinkering with electronics for it's own sake is a form of reward. Building anything really is an experience in of itself
AdmiralAsshat 6y ago
From actually reading the content of the article, it seems like it would've been better titled, "I tried to use a Raspberry Pi 4 to do my dev work for a day." And that might have better captured the point, since I imagine a number of us have experimented with using a Pi for work, for personal projects, etc. in the past. But the current title to my mind is a tad clickbaity, for the implied comparison I noted already.

pizza234 6y ago

The point is different, on a different reading. > I guess one point I should make is that for almost every piece of software I wanted to use, I had to spend a lot of time just trying to find any that would work on Linux—then narrowing that to 'on Linux ARM64'. And then I had to usually spend a few minutes compiling it from source, placing my own shortcuts on the system (so I wouldn't have to open a Terminal every time I wanted to check Twitter), etc. > All of these issues (4K difficulties, having to compile apps, not finding apps) are exacerbated by the fact that the Pi runs on ARM, but it is still a problem in the wider Linux ecosystem. The point is that the above could apply to a $1500 ARM computer. However, I suspect that he's not very experienced with Linux, and ultimately makes a comparison that doesn't make much sense. > Anyways, on my Mac, I'm used to Shift + Option + dash for em, or Option + dash for en, but we're getting off the point. On GTK, one can use a compose sequence for that. There you go: https://askubuntu.com/a/31265 https://askubuntu.com/a/31265. > All of these issues (4K difficulties, having to compile apps, not finding apps) are exacerbated by the fact that the Pi runs on ARM, but it is still a problem in the wider Linux ecosystem. "Having to compile apps" and "not finding apps" are not problems in the wider Linux ecosystem. There's an enormous amount of stuff in the repositories of the main distros, and it's prepackaged. And there are external repos for what's not in there. The software one really has to compile manually is a very small minority (again: in the wider Linux ecosystem, that is, on x86). For the 4K, I can't speak, as I have no experience. Ultimately, using ARM for a desktop machine with specific software requirements doesn't make much sense; in particular, because his work(flow) revolves around closed-source software. However, that's what the explicit target of the article, so I can't really complain about it.

Scarbutt 6y ago

I don't know what this was intended to prove, exactly? Nothing, just another bad blog trying to get youtube viewers.

SquareWheel 6y ago

What is with Hacker News and these pessimistic hot takes? He tried it to see if it would work. That envisions the hacker spirit better than most ever accomplish.

Koshkin 6y ago

Looks like what they proved is that the difference is less than 20x.

geerlingguy 6y ago
The performance was the biggest surprise. Outside of 4K video, the Pi 4 was within at least javelin distance (not quite spitting) for a number of benchmarks I ran. Much better than the cost difference would indicate.