Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How many GNU/Linux users are needed to change a light bulb?

gnu.org

Post reply on HN
Print view View original

171–180 of 184 posts

Re: How many GNU/Linux users are needed to change a light bulb?

#171

Post by salawat »

whalesalad wrote: post #103

One to tell you to use arch. One to tell you arch is not necessary and Ubuntu works great. One to tell you that why not just use Debian, because it's the original pure form of Ubuntu? Then a FreeBSD hacker will come in and tell us about the manual pages. Soemone from a crypt in a dark forest of Germany will come out and ask us our thoughts on Open SUSE. The macOS user in the back meanwhile has already changed the bul…

The Slacker will look on forlornly from amongst his binders full of printed out Slackbuilds, and a lonely dot matrix printer.

Re: How many GNU/Linux users are needed to change a light bulb?

#173

Post by rolandog »

renewiltord wrote: post #157
the8472 wrote: post #144

Earlier quoted context omitted.

Single-user deployments are fairly common in the gnu lightblub crowd.

Very insecure to run all your light bulbs in the same user context. I use SELightBulb to isolate light bulb actions. If you add PAM and pam-fingerprint you can require a fingerprint to alter lightbulb state for better security. Advise two-factor unless you really don't care for electricity at all.

Instructions unclear; burnt fingerprint on incandescent lightbulb.

Re: How many GNU/Linux users are needed to change a light bulb?

#175

Post by yetihehe »

rusk wrote: post #134
yetihehe wrote: post #63

Earlier quoted context omitted.

After debugging configuration error for last 7 days, where even logs didn't help because manufacturer decided to use reference clock from different pin which was not documented officially to be able to do this - moder IoT sucks, because chip manufacturers release documentation as confidential and then firmware is implemented by hardware guys. As a programmer who can design some simple digital circuits, I could very m…

What I’m finding is the components themselves are so cheap ... if you’ve a little knowledge you can rig up your own IOT setup that you actually have control over for pennies. Raspberry Pi is a great place to start.

You are VERY underestimating knowledge required to do this. Most people don't even know what is an ip address (and google wants to even remove url bar). How will they setup a server to control their devices?

Re: How many GNU/Linux users are needed to change a light bulb?

#176

Post by yetihehe »

keeganpoppen wrote: post #170
segfaultbuserr wrote: post #96

Earlier quoted context omitted.

One quote from the list of BOFH excuses (of the early 2000s, https://gist.github.com/chuckwagoncomputing/ff980f81482b5894... ) was: > Firmware update in the coffee machine It isn't funny any more... > https://support.smarter.am/hc/en-us/articles/115005135369-My... > In case your Smarter Coffee machine has become stuck while updating, please follow the steps below to manually perform the firmware update... - Turn prod…

alright, that's it-- can we just burn it all down to the ground and start over?

Regarding that burning - my home wood pellet furnace needed software updates after installation to start fully working. Good luck that I had mobile internet available.

Re: How many GNU/Linux users are needed to change a light bulb?

#177

Post by Jaruzel »

saalweachter wrote: post #132
Jaruzel wrote: post #120

Earlier quoted context omitted.

Light-as-a-service could work... where you outsource all your light switches to an AWS lambda function, and turn them on/off via a subscription based "AI Powered Easy To Use App!" AI powered of course in that it knows when you've entered and exited a room, unlike dumb humans who have problems with this... or something.

This is actually sounding like a perfect startup proposal. The startup puts a full suite of smart bulbs/switches in your house for free, and then charges you a monthly subscription fee (flat fee + a per-candela-second rate). Sure, there's a large up-front cost for the startup followed by a long amortization period before you start actually making money, but that's what VC is for.

What's the URL for YC funding again ? XD

Re: How many GNU/Linux users are needed to change a light bulb?

#178

Post by segfaultbuserr »

yetihehe wrote: post #176
keeganpoppen wrote: post #170

Earlier quoted context omitted.

alright, that's it-- can we just burn it all down to the ground and start over?

Regarding that burning - my home wood pellet furnace needed software updates after installation to start fully working. Good luck that I had mobile internet available.

> needed software updates after installation to start fully working

Not too bad, at least it's a one-time job, better than some IoT thermostats that crash randomly in winter midnights and freeze everyone to death (figuratively)...

> https://www.nytimes.com/2016/01/14/fashion/nest-thermostat-g...

> Nest Thermostat Glitch Leaves Users in the Cold

> The Nest Learning Thermostat is dead to me, literally. Last week, my once-beloved “smart” thermostat suffered from a mysterious software bug that drained its battery and sent our home into a chill in the middle of the night.

Re: How many GNU/Linux users are needed to change a light bulb?

#179

Post by segfaultbuserr »

aidenn0 wrote: post #127
FrankSansC wrote: post #10

> 1 to remind everyone that the right name is GNU/Lightbulb it still cracks me up everytime I read it :D

I always wanted to port a BSD userspace to Linux just so I could counter-correct someone who corrects me to GNU/Linux.

Alpine Linux already does it, musl + busybox. Not BSD though. Nowadays I use "GNU/Linux" vs "Linux" to distinguish between an ordinary Linux distro (which comes with gcc, glibcs, coreutils, and all the other tools you expect) vs a minimum or embedded distro or system (OpenWrt, Android, Alpine, etc).

Re: How many GNU/Linux users are needed to change a light bulb?

#180

Post by segfaultbuserr »

fierarul wrote: post #115
taffer wrote: post #59

Earlier quoted context omitted.

Light bulbs and cars have significant marginal costs that software does not have. Some types of software (operating systems, DB, browsers) also have a network or platform effect, so people actually have an incentive to get other people to use or even improve their software.

Even so, geek themselves are a pretty nice market. Would you not buy an open source printer? I know I hate HP after the last one died just after warranty ended and they have no spare parts whatsoever. Same with other things. I would buy those things. And if they really are good they can become a mass market product. Yet, we only get a phone or something and that's it.

My answer would be: The cost of distributing or running a software package is almost zero, same is not true for hardware, followed by an argument of how hardware manufacturing depends on the economics of scale, unlike software.

However, I soon realized your question is much deeper. It is not "Would you not buy an open source printer?", but "If you can buy an open source phone, would you not buy a printer?" - I think it's a good thought-provoking question. I'm not sure about the answer yet.

Post reply on HN