LINUX Unplugged - Episode 69: Perfect Linux Laptop | LUP 69
Episode Date: December 3, 2014The founder of Purism Librem 15, a laptop that promises to respect your freedom and be the perfect Linux machine joins us to discuss the hardware, software & goals of the project & how he hopes to enc...ourage manufacturers to free the entire stack. But are the goals of this project too ambitions? We’ll ask!Plus CoreOS announces Rocket, a new Docker competitor that we’re very excited about & more!
Transcript
Discussion (0)
Hey, Eric, was it you that submitted the imp to the subreddit?
That was me, yes.
Did you see this?
We have another crowdsourced Linux computer.
I'm a sucker for these.
I love, like, you get the promo video.
It's like you get the will it make its funding.
Like, it's not gambling, but there's an element of it.
It's a lot of fun, and it's interesting to see all these independent hardware guys, too.
They've been doing a lot of social outreach, shall we say.
If you follow the Twitter feed, just go and have a look at the tweets and mentions of their marketing person.
It's pinging basically everyone in the free software community, asking them, hey, what do you think of our thing?
Can you reshare it, please?
So I said to them, well, it's not open source.
It says open source, but it isn't because there's binary blobs in there.
And they were like, yeah, yeah, yeah, that's binary, but everything else is open source. And I said, yeah, but at the top it says fully open source it says open source but it isn't because there's binary blobs in there and they were like yeah yeah that's binary but everything else is open source and i said yeah but at the
top it says fully open source and it isn't so let's so we should talk about that because uh
if everything works out the uh the uh libram 15 uh creator the uh laptop that uh i crowdfunded
a couple of weeks ago in the last news segment is going to join us today and i i think this is a
question we need to ask how do you promote yourself as like do no evil, all open, but yet still
manage to ship something that is competitive, right? Which is my nice way of saying, you
know, an actual GPU that has a binary blob driver and closed source drivers. So I think
that's a really good question. I want to play this video. Wimpy, I'm going to toss the link
to that guide in the show notes for people that are listening
after the fact. But let's play this a little bit
and then we'll
talk about it because I love the fact that all these come with
videos. That's like the best part for me.
Oh, we got some music! Yeah!
In the first decade, technology has
virtually exploded.
Surprisingly, only one thing has stayed
the same. The desktop.
Imp is here to change that imp resolves your entire
home computing needs in one ready to use unit that fits in the palm of your hand just plug it in and
see what it can do for you imp is a full-blown desktop computer that suits all of your everyday
home use needs whether it's browsing the web checking your email being social needs. Whether it's browsing the web, checking your email, being social,
or even getting some... Man, if it had this
holographic interface, that would be a slam dunk.
Oh, no kidding!
So it's looking...
I'm trying to figure out what the desktop environment is. Do you guys
know? It's a
custom desktop environment that they have built
on top of Unity.
Oh, no. It's Cinnamon, I thought.
It's Cinnamon on top of Ubuntu. So no. It's Cinnamon, I thought. It's Cinnamon on top of Ubuntu.
Yeah.
So it's a Cinnamon theme, perhaps?
Because I don't recognize the look of it.
I've seen or heard about this somewhere.
So it talks to my head where it will...
I think it's interesting
they're talking about media streaming capabilities, too.
It can even back up all your phone's media.
That's the only cord I've seen
connected to this thing so far.
So it streams to your
television and it backs up
your phones and it uses
the Cinnamon desktop.
And it's called Imp.
So, oh, this is
we got more video
but what do you guys think?
Was it implying that there was some sort of NFC when you set your phone
next to it, it's going to start syncing?
No, in the video they showed him hooking up a USB cable.
I got you.
Yeah.
So it's raised $33,000 USD.
It's trying to get to $100,000.
33% funded, 35 hours left.
Only 35 hours at 33%.
That's stretch goal.
Yeesh, yeah, that's tight.
I guess.
See, what you have to be buying here is their custom software solution that sits on top of this,
because otherwise you'd probably just buy a NUC, although the prices are cheaper than a NUC.
Yeah, but it's way lower spec than a NUC.
It's an O-droid.
Yeah, exactly.
And that's one thing Q5Sys was pointing out.
O-droids are not legal in the U.S. because the FCC will not recognize them.
What?
What?
Yeah.
No, I think there's devices here that have Odroid processors.
Isn't there?
Well, it took Q5Sys forever to get his Odroid machines into the U.S.
There's like a licensing issue?
It'd be like, oh.
Because they are FCC certified.
I've never had a problem.
I've ordered seven or eight Odroid products,
and I've never had a problem with any of them.
So, yeah.
I wonder if it depends on the model.
So this is using the Odroid U3.
Oh, they're not going to hit their target at all, are they?
No, I don't think so.
Probably not.
But it's still interesting.
Is this the wrong approach?
Though, I mean, maybe the better.
I mean, is it possible to do?
Is the crowdfunding required because you have to reach scale?
Because if I could just buy one of these, I mean, maybe you could just sell them one off.
I don't know.
I guess that never really works, I suppose.
Well, the problem is that crowdfunding is a way to go from zero to hero in a very short period of time.
crowdfunding is a way to go from zero to hero in a very short period of time.
What they don't want to do is they don't want to put in the work that every other company that exists right now has gone through,
which is sell a small number of one item,
then sell a slightly larger number of your next revision of that item,
then the next, and then the next, and go through iterations.
They want to go from nothing to big bang and sell all of that one device.
However, here's the devil's advocate to that.
So in their specs, right, they're going up against the Chromebox.
Now, the Chromebox is a pretty hard competitor
because the hardware is literally being subsidized by Google and Intel, right,
on the x86 versions of Chromeboxes.
So if you're an independent manufacturer
and you're going up against these industry titans
that can subsidize the hardware cost, thereby lowering the cost of the chromebook or the chromebox
then maybe your only way to be competitive is to have some scale so that way you can buy in
at lower prices and charge lower rates right and the only way to get scale is crowdfunding
at when you're not their only way of having a usb they've got other usbs that the chromebox
doesn't have like all this home syncing stuff that that i haven't seen on any computer out of the box so that's a usp microsoft office so you're saying that that that
service value add would justify a higher price i think not justify a higher price but justify
existing in a normal retail channel and not having to go the yeah uh crowdfunding yeah
built-in apps and services, they've got the
Microsoft Office suite listed
and OneDrive, so I assume
they're linking to
Office 365. Yeah, I think so.
This particular
doesn't, this is
on Indiegogo and it doesn't matter if they
reach it or not, they're keeping them from hunting
regardless. That's right.
They get a larger cut taken from Indiegogo though.
Right, right, right.
No, it's a fixed funding.
It's flexible funding that you get it all, isn't it?
No, Indiegogo is flexible, isn't it?
No, it's a fixed funding campaign.
Yeah, it is fixed funding.
If you hover it, it says that if you hover the question mark next to it,
it says they're still going to keep it.
I don't know how it's fixed.
This campaign will receive all funds raised even if it does not reach its goal.
So they're already going to make $33,000.
Maybe they use that as investment seeding to go out and make a product that does go to retail.
I don't know.
Welcome to Linux Unplugged, your weekly Linux talk show that has a totally legitimate reason for wearing pajamas all day.
My name is Chris.
And my name is Matt.
Hey there, Matt. You wouldn't believe the day I've had today.
Oh, no.
So, I mean, we've got a great show. I mean, not to overshadow episode 69. It's going to be great.
Todd from Purism is joining us to talk about the ultimate free Linux laptop.
And then there is a major Docker competitor that has been announced by the CoreOS project,
and I'm super excited about it.
It's really geeky, but I'm really excited.
We're going to talk about that, plus we've got feedback.
But, Matt, Matt, Matt, this is the thing, Matt.
All right.
I woke up this morning to the sound of somebody pouring water,
a glass of water, maybe like a jar of water on the floor
upstairs.
Oh, no.
And here's how it goes.
My wife reaches over.
She's like, hey, wake up.
Hey, wake up.
I'm like, what?
Do you hear that?
And my wife's first conclusion is somebody has broken into the house, and the whole purpose
of breaking into our house is to pour a glass of water on the floor.
Right?
And I immediately discredit that theory.
I'm like, no, that's not what's going on.
So I go out there.
My three-year-old daughter didn't quite make it to the bathroom.
Oh, no.
Yeah, so that was my morning,
is I had to clean up a little bit of a mess right there.
You're good.
Well, actually, you know what I did.
I'm going to be honest.
I just put a towel on it.
But, you know, somebody had to clean up that mess.
I don't know who.
And then I leave, Matt.
I drive to the studio.
And, you know, it's Arctic, 21 degrees Fahrenheit And then I leave, Matt. I drive to the studio.
It's Arctic, 21 degrees Fahrenheit out here.
Oh, yeah.
And I get out of the truck.
And the JB1 Studios is a fairly public street.
There's not much privacy to be had there, right?
I step out of the truck, and my favorite pair of jeans, Matt,
not like a pair of jeans I like to wear, but a pair of jeans that I've worn for years.
Love it.
Take care of them.
I barely wash them.
You know, all of the things you do with your favorite pair of jeans.
I step out of the truck, and they're old, so it snags the step on my truck.
Oh, no.
Rips the jeans all the way up to my butt cheek.
All the way up to my butt cheek, Matt.
It rips them all the way up.
Oh, perfect.
And I'm standing out there with half of my ass hanging out in the road in 21 degree weather. And I'm
thinking to myself, I haven't even made it into the studio yet. What could possibly go
wrong? However-
And this is where you cue Swedish volleyball team driving by in a bus. As they go by.
I'll have to, you know, I kind of have a story along those lines I'll have to share on another
show one day. I'll have to have a few drinks before I share that story because it's scarring.
But so I realized, though, this was an opportunity to make some lemonade.
And, you know, I'm always looking for life's lemonade opportunities.
Oh, yeah.
Because I had, as every good podcaster does, let's be frank, I had a backup pair of pants in the studio, a pajama pants.
I love that. frank i had a backup pair of pants in the studio a pajama you know i had pajama pants in the studio
so i've legitimately had a had a solid excuse to wear pajama pants all day long today and at first
i was kind of upset about it because my favorite pair of jeans then i realized pajama pants no
guilt that is almost worth it plus i think the wife might be able to fix the pants so all in all
that's cool yeah it's not so bad well, we do have a lot to get to.
I probably shouldn't mess around too much because tons of stuff to cover.
So I'll bring in our mumble room, and they can help us comment away on things as they come up.
But time-appropriate greetings, mumble room.
Great to have you here.
Hello.
Happy Tuesday.
Hello.
Is that a drafty?
It was a little drafty.
Yeah, you're right about that.
And welcome back, Colonel Linux on the Cheapy. Yeah, you're right about that.
And welcome back, Kernel Linux on the cheap rig.
Good to have you here, sir.
How's the cheap rig sounding today?
I think it sounds pretty good still.
Wow.
I think it sounds better than I do. Wow, that sounds great.
All right.
I need that microphone.
It's official.
We're doing a segment on Kernel Linux's cheap podcasting Linux rig setup.
We've got to do it before the end of the year, Kernel Linux.
We've got to figure it out.
People can make it their end of year purchase.
All right, so we've got to get into some follow-up
per tradition on the Linux Unplugged show.
We like to keep the continuity,
so those of you who listen to every episode,
you get rewarded by some continuity for every show.
And those of you who don't get a chance to listen,
you kind of hear some fragments of stuff we've covered
so you get an idea if maybe it's worth jumping back into the back catalog.
So Chris, not me, writes in with our first email.
He says, hey, Chris, Matt, and the Mumba Room.
I just finished listening to Linux Action Show and Unplugged episodes around Linux in
the Penn Manor schools.
And I got to say, wow, I am in IT for a district school system in WV.
And that's, I think, if I'm right, that's
the Volkswagen bug, and I have been trying to introduce Linux for 15 years.
In my district, it's been very hard.
We have 22 sites, 11,000 students, 4,000 staff, and only seven in the IT staff.
I was envious of Penn Manor with their nine.
Anyways, last year we introduced some Chromebooks, and I thought maybe Linux was finally going to be a possibility
until this school year the State Department made a deal with Microsoft.
Every employee and student in our state gets unlimited storage on OneDrive.
Yes, unlimited.
It started at one terabyte, but it's now actually become unlimited for them.
Every employee and student also gets five installs of Office 2013.
Oh, there's a winner.
That they can download and manage from Office 365.
And our state has
unlimited licenses on
Windows 7 and 8 installs.
The price of Office and Windows
is now gone. We only pay for
server licenses and CALS,
client access licenses. So, Microsoft
is fighting back. I know this cannot be
free, and the state has
to be paying something to Microsoft,
but districts never see that money, and it doesn't even cut into the district budget.
This, I think, cut into our plans for more Chromebooks this year.
But I have now seen more Microsoft services come through the shop.
Just thought you might want to know that Microsoft is fighting back and doesn't want to die.
Boy.
It sounds like they're not just fighting back, but it sounds
like they're going long.
I think they realize that
there's no sense in trying to drill down on the school
districts. Go long, get people brainwashed
into using the product, and you'll get them
later. They're going to buy the services.
They are trying to go hardware. Sounds like a little bit like
maybe get the state to bear the cost.
You know, I got to wonder though, is this a sign that Windows 10 is going to be free?
Oh, it will be.
I would be – it would be stupid at this point if they don't.
It's the only – it's because it works.
It's hard to argue with free, even if it's crap.
Yeah, and you can see here they're responding to Chromebooks and devices like it by selectively making Windows free as they can.
You know, not making a big deal about it, not making it public.
But when it comes down to it, they're making it free.
And you've got to think they're doing that specifically because they know that the cost of the operating system is like, look, Apple gives it away now, right?
Even though it's a commercial OS, they're giving it away.
Linux is free.
Chrome OS is quote-unquote free.
Android is free.
So what's Microsoft going to do?
They're going to have to do this.
That's right.
And it's – honestly, I'm kind of surprised it took them this long.
Well, believe it or not, there have been miscellaneous leaks from multiple executives at Microsoft that have said that Windows 10 will be free.
No.
Yeah.
I'm not surprised. executives at Microsoft that have said that Windows 10 will be free. No. Yeah. Yeah.
I'm not surprised.
I've read a lot of articles that basically
there was somebody,
like one of the executives in Indonesia
did say it would be free.
Free as in cost, we should point out.
Yes, free as in cost.
Yeah.
But surely that will be some special version
of Windows that's free.
Oh, no, they're talking free upgrades
from 8 and 7.
I think I'm thinking
that cash cow is Enterprise, isn't it?
With all of the extra Enterprise servers.
And surely they can continue to charge for that.
Otherwise they've got no business.
Yeah, it's going to be probably not as limited
as a starter edition, but
there must be some restrictions in order to get
Enterprise to pay.
I don't know.
Were you going to say something rotten?
I was going to say that windows 10 home sounds like it'd be free,
but they're like the,
like the pro version would be the expense.
I could totally see that.
And windows and probably for most consumers,
it'd be fine.
I don't like it.
They rip us off and then they make it free.
They rip off the ideas of compass and,
and,
and desktop management and,
and all this crap. And then they also rip off the whole
pricing structure. Damn it.
I just don't like it at all. Oh, but are
their windows wobbly?
Probably.
You could probably edit the registry to make it happen, man.
There you go. Oh, man. I remember registry
hacks. Yeah, right. Exactly.
Good old registry hacks. Start menu, say
stop. Things like that. Hey, why don't. Good old registry hacks. Start menu, say stop, things like that.
Hey, why don't we do one more bit of follow-up?
So a big thank you to Gabriel3 in the subreddit,
who's been a very active poster in there recently
with some really great content,
especially for those of you trying out OpenSUSE.
So Gabriel mentions that we covered the SyncThing GTK front end
on the Linux Action Show this Sunday,
and since then there's actually been some updates.
And he points us to the web update article.
Web updates have really been doing a great job
of following the SyncThing project.
And they do a great job of breaking some of this out.
For those of you who maybe aren't familiar with this,
SyncThing, and you might have heard of Pulse,
they're now two different sync solutions
that are similar to Dropbox or BitTorrent Sync
or a Sparkle Share.
It's totally open. It has a really interesting trust model, which works really well if you're
just syncing in between some of your machines. And now there's a GTK front end. It's under heavy
development, and in a couple of weeks there's been several updates. It now has an auto updater
for the binary of the daemon, which is really cool. It's not enabled by default on Linux,
but WebUpdate will tell you how to turn it on.
I'll link to that in the show notes. It now has a first-time
run wizard to help get you all set up and what
directories you want to point at. Speed throttling
options are now exposed through the
notification icon menu.
And I thought this was kind of neat.
It's doing, and I talked about this on Linux
Action Show, but it's actually using the
file system IO notify to
know when files have been changed
and then making sure they get
synced up, which is just nice and slick and integrated.
So the reason
why I'm talking more about sync thing these days is
I think people are getting more and more upset with
BitTorrent sync, and I just want to make sure people
are aware of a free alternative in
every sense of that term, free, that's
out there. And now it's nice to see it getting a GTK
front end, and WebUpdate is linking you right here to the PPA if you're on an Ubuntu install.
And it's also available in the AUR.
Looking good.
I don't know if I'm going to switch yet, but it's looking good.
I don't know why I'm so gun-shy about it.
I'm becoming an old man.
Young Chris would just jump whole hog. And, you know, part of it is, I don't know if you guys saw this, but SpiderOak had a one terabyte deal for Cyber Monday,
and I grabbed that because it was a great one terabyte storage deal for SpiderOak.
Some kind of, I don't know.
The problem I have with SpiderOak is the client is not my favorite.
Yes, true.
It's called SyncThing GTK, and you just need Python and GTK on your system,
and you can start playing with it.
If you haven't switched anything yet, check out SyncThing.
I think it's the one to get behind for a little while,
at least for most of us Linux users.
Unless you need to distribute publicly, then you might want to look at BitTorrent Sync.
Hey, I'll tell you about something else you might want to look at for a little bit.
That's our first sponsor this week, and that's Linux Academy.
tell you about something else you might want to look at for a little bit. That's our first sponsor this week, and that's Linux Academy. It's Linux courses created by people that truly understand
you and Linux, because they're Linux users themselves. And I think we all know the difference
that can make. I'm going to recommend you do something. Just take a minute and go over to
Linux Academy right now. In fact, if you go to Linux Academy, check out their tour. They have a
tour of all of their great features.
But go to linuxacademy.com slash unplugged.
Start there.
linuxacademy.com slash unplugged.
It's going to get you a 33% discount.
Take a look at some of their training courses.
They've got everything.
And I've been hearing from some of you who have been signing up and trying out different things, stuff I never even think to mention.
They have, I think, a good one. And if I ever get around to actually hosting all of my config files on my own GitLab server, this course that they have on Git basics and all of the
introduction to Git, installing and configuring Git, working with GitHub or GitLab, working with
Atlassian or Bitbucket, just so I have, I like this just so I have an idea of what the alternatives
would be
like. That is super valuable for me. And it's so nice to know that Linux Academy has just got this
content available. It's part of my subscription and I get access to it. So when I get around to
finally getting like my home director or whatever great sync solution I get, or maybe it's our
publishing system for Jupyter Broadcasting that I stick up on a GitLab server, I'm going to want
to make sure I take this course. And it's like that for everything.
Linux Academy has seven plus Linux distributions you get to pick from.
And they automatically adjust the courseware.
So that way it reflects the distribution you've chosen.
They have on-demand training.
You go in there and say, I need a learning plan for this amount of availability.
And it will custom build that training for you.
It'll give you the timeline, the estimates.
If you're at work and you know down the road,
boy, 2015 is finally the year we're going to get off our butt
and get Puppet implemented.
We're going to finally do this right.
Well, why not use this time now to go over to linuxacademy.com
slash unplugged and take their Puppet training?
It's a great opportunity to sort of get in there,
do their scenario-based training,
use their automatically generated VMs that go along with the courseware.
You get public access to those.
It's so neat.
You get to keep track of your progress as you go along.
They'll give you a good visualization and display,
so you get a really good sense of where you're at.
You can get exactly how much time it's going to take.
If you need to expand out to AWS for some scaling
or something you might want to load on there,
you can go take all of their courseware available on AWS.
They have a bunch of courseware on OpenStack.
If you find yourself in that gray area of DevOps
that I like to tease people about,
but it's a real thing, you know?
It's something you kind of...
Sometimes you go into a job that's DevOps
and sometimes you realize your job is DevOps.
It can go both ways.
And Linux Academy has you covered.
It's so great.
They also have team accounts for groups.
So if you're part of a team or a small business or even a large business
and you want to all work together, they've got that.
They have a community support infrastructure to give you that motivation you need.
And they're always adding new labs, two to three a week at this rate.
It's really awesome.
And I think the scenario-based training is not only unique to Linux Academy
because you'll work with all of these technologies.
Let's take anything, for example.
Maybe it's an implementation of Nginx. Maybe it's an AWS implementation. But the great thing is with
these scenario-based trainings, you'll actually use the technology you're going to be supporting,
the technology you're going to be responsible for, the technology you're going to be on the
line for. Well, with Linux Academy, you'll actually do it. You'll implement it. That's
the kind of confidence that you can't
just go out and get. You have to use somebody who, you have to go with a training platform.
You have to work with a group that really knows this stuff, that knows the scenario you should
be building out, and that it actually is one that you would face in the real world. That difference,
that's Linux Academy. And you can take advantage of it by going to linuxacademy.com
slash unplugged. It supports this show, and it also helps you get ahead.
It's a great way to bump up that resume, fill some time, or even just challenge yourself.
linuxacademy.com slash unplugged.
They've got courses on pretty much everything.
If you're not really sure where to start, why not go in there and scratch an inch?
Maybe this is the time to finally make that backup solution for your photos. Holidays are coming up. You're
going to be snapping a lot of pictures. Wouldn't it be great going into that to know that you
could create your own backup solution for that? They've got coursewares on that. It's
really so much opportunity, and they're always adding new stuff. Linuxacademy.com slash unplugged,
and a big thanks to Linux Academy
for sponsoring Linux Unplugged programs.
You guys rock stars over there.
I can't believe how hard they work.
They make me look like a slacker.
I don't even like it.
Okay, before we bring Todd on from Purism,
I just want to get to a couple of more emails
just so we can get through these
because I feel bad when we don't get to all the emails.
And this next one is a perspective that it's hard for us to sometimes take on the show.
And yes, I know we talk about SystemD a lot,
but I think we should make a little more space up for it
considering that on Sunday's Linux Action Show,
it's going to be a lot of positive stuff because we interviewed Lennart Pottering
about, among many, many things, SystemD.
And we're going to hear a lot of his side of the story.
So Mr. P writes in from the Slackware side of the camp.
He says, hey, Chris and the rest of the crew,
thanks for all the hard work on the shows.
I'm a Slackware user since 1995, both private and in my daily work.
He says I'm a programmer for Embedded Linux
and I do some system administration.
I've listened about all of the talk regarding SystemD,
and I'd like you to address part of this discussion that's missing.
We don't use SystemD in Slackware.
The big problem for us in Slackware is that many packages nowadays
are requiring SystemD to be able to build or to run.
The tone in the Slackware community is respectful towards SystemD,
but it's hard for us to see the impact on all
third-party applications.
And I think by that he means it's hard for us to see this when
we don't use SystemD, but yet these requirements
are coming up. When I listen to you on air,
it seems like you're surrounded by only pro-SystemD
people. Regards,
Pierre. Or Per, I'm not sure.
So,
I wanted to open the floor to anybody
in the Mumba room that has some concerns about SystemD.
Maybe we could just have an open discussion about them.
I don't know if we have anybody.
I think Fred was our biggest vocal anti-SystemD user last week.
Does anybody in the Mumba Room have concerns about SystemD?
Do they want a voice?
It's an open floor.
Not really.
I think this might be an opportunity for us to try and get one of the
V1 guys. Yeah.
He, Mr. P, recommended
somebody from Slackware that we could bring on, and I'm
going to forward that contact input to the production
staff. So, here's what I'm
trying to get to. Here's why I asked you guys.
Because we have a good-sized room here, and
you know, I suspect
that the reason why there's not a lot of anti-SystemD commentary on this show is I think it's reflective of the overall opinion.
I think a lot of us maybe have concerns, but we also see that it's sort of a competitive necessity for Linux.
Wimpy, you say you might have some concerns.
What would those be?
Yeah, I mean, in general, I'm very supportive of SystemD, and I think it's a good thing.
But my one concern about SystemD is scope creep.
It doesn't seem to have a well-defined roadmap of where it's heading,
and new facilities and features seem to be blossoming into the SystemD area
and isn't clear where the edges of that project are.
Right.
Well, I think, you know, not to speak for Lenart,
but I think he would define it as anything that kind of sits
between user land and kernel could be fair game.
And it's quite a big stretch of land.
Yes, it is, isn't it?
Yeah, it is. It is a very big stretch of land. Yes, it is, isn't it? Yeah, it is.
It is a very big stretch of land.
And I'm trying to think of how other operating systems do it.
And when I think of the commercial OSes,
when I think of Windows,
Windows has what comes to my mind,
and maybe this isn't a fair parallel,
but Windows has SVC host, right?
SVS host?
Am I getting that right?
Which is this huge parent process that controls ungodly amounts of things.
And it is totally opaque to me.
I have no idea what it does.
Now, here's the thing about...
Now, here's my response to what you're saying, Wimpy.
Is the way SystemD is going about and replacing these individual bits of functionality,
like take NTP,
for example, right? It's doing it with individual programs. It's not like it's all, there's not a
systemd binary, right, that has all this code in it. It is individual programs that are part of
the systemd project, but they are broken out in small bits of functionality, which I think, A, makes it easier to kind of – it gives you insight because you can see the tool.
B, it makes it easier to replace just that component if you replace it with something else, right, because it's just this NTP piece I have to replace or just this DHCP client piece I have to replace. It's not the whole systemd binary. So I agree
that that is a problem, but I think the balance that systemd strikes by replacing it with individual
small tools that just happen to all communicate really well to each other is a pretty reasonable
approach to that. And I think too, like you have to look at some of the things that have been replaced
and be honest and say, you know, some of these things were kind of neglected, right?
And not in all cases, but in some cases, they were just kind of a hot mess that were neglected.
Yeah, I appreciate how system D is put together and that everything isn't lumped in together
in one binary and that it is a set of tools.
And I don't, thinking about it, I don't think there's anything that's been added to system D
that I haven't found useful or can see a rationale for.
My concern is more looking into the future.
Where will it continue to grow and what will the scope of the project become?
Because that isn't clear.
Nobody knows.
Right. project become because that isn't that isn't clear nobody knows right dare devlin do you believe that
systemd taking over these more and more functionalities actually will make for a more
nimble linux system yeah i i think it's uh the closeness is actually something that i feel that
we're missing other departments as well actually what do you mean? Things that actually go and talk to each other
but are being developed in this
widespread
multiple repos, pretty much.
It's all over.
And I think the closeness that the SystemD project
made for the multiple
programs is essential
to actually keep the projects in check and actually
in pair with each other although i do kind of also agree with what peacemaker is saying in that uh
it doesn't leave a lot of rooms for distros like slackware or gen 2 or any of the others uh dev 1
that uh don't want anything to do with systemd Like there's not a lot of air left in the room.
So that is a difficult thing because I think most of us would agree monoculture is bad,
even if the monoculture is made up of a lot of little bits.
So one of the things that I – then we'll move on because we've touched on enough.
But one of the things that I think we forget
is that the future is changeable. And if we get to a point in the future where there is a big
downside to going with system D, the open source community will write itself. It will write that
ship and it will make a course correction. And the future is not set in stone. So if a lot of
things today are written for system D, that does not mean that in five years everything is going to depend on system D.
It might mean that we go through a period of time where a lot of things – but you know what?
We went through a period of time where a lot of things were written for ALSA, and now a lot of things are written for Pulse, right?
It just changes.
Things happen.
And technology is fundamentally changeable. And especially open source software. I totally legit get the concern if this was Windows or Mac OS X and it was happening behind closed doors and we as a community could do nothing about it.
nothing about it. But anyone can write code. Anyone can write a patch. And it just takes one person to start a change. We are not under the same constraints we would be in a totally
proprietary commercial system. But yet I feel like our reaction here is like the future is set in
stone. If we make this decision, everything is ruined. We can never write the ship. But that's
totally opposite of how open source works. So for me, it's like I totally respect like, ah, that does suck that the Slackware project is getting the air sucked out of the room.
Then again, it doesn't mean there will always be no air in the room.
There could be air that comes back in the future.
It's not – it's just we should move forward.
I'll just leave it at that.
And the interview with Lenar goes into some detail about this.
I think it's a good preset up for that.
Lenart goes into some detail about this.
I think it's a good pre-setup for that.
So if this topic interests you at all or any of the community trolling topics or future of software distribution on Linux and how that might be accomplished, all of those were covered in the interview with Lenart.
And that will be in Sunday's Linux Action Show so far.
That's when we have it scheduled to air.
So hopefully we can just sort of air that. And I totally am still willing to take input and and talk about the topic but also i have to respect if the audience is done with done with it i don't want
to dwell on it but i find it to be a fascinating change and uh there was an article that i talked
about linux action show about how net bsd community reacted to the same kind of thing 14 years ago
and it does really feel a lot like history kind of repeating itself just that you know a lot of
us don't remember there was a huge debate over X, right?
There's a lot of things that have happened in the community
that have kind of gone by the wayside now.
There was a huge debate over, not as big
as SystemD, but over AGLX
or XGL
for composited desktops.
And we don't even use either of those anymore. And there was a
big debate about it. So
the future is not fixed. That would be
my ending piece for our system
discussion. All right. So we're going to talk to Todd here in just a second. Why don't I do
one more sponsor break, and then we'll bring Todd in. And it's kind of a special sponsor break. So
everybody knows Ting. We love Ting. Go to linux.ting.com. You're going to get a $25 discount
off your first device. Ting is mobile that makes sense. And Matt and I have been using Ting for, geez, man, almost two years now.
Wow, already. Good grief.
It is nuts.
And I got some stats right here from Ting about the Jupyter Broadcasting audience
that I thought I'd share as we approach our second year.
So the average bill for Linux Action Show audience members and Linux Unplugged
and any JB audience member that has one device, the average monthly bill, $26.
Whoa.
Now, let that sit for a second.
This is a full-featured smartphone.
You only pay for what you use.
There's no contract, no early termination fee.
$26 is the average Jupyter Broadcasting audience member's monthly bill. Now, what are you paying today for your cell phone bill? Is it more than $26 is the average Jupiter Broadcasting audience member's monthly bill.
Now, what are you paying today for your cell phone bill?
Is it more than $26?
Do you have a contract on top of that?
Go to linux.ting.com.
But right now is truly the best time to ever switch to Ting if you've been thinking about it.
So you've probably heard me mention that Ting has an early termination relief program.
And they're going to give you 25% of the ETF, which is about $75.
So if you figure the average bill is $26 and you're going to get a $75 credit,
that's almost three months of free Ting service.
That's a really good deal.
Yeah, I would take that deal.
But Ting is upping the game for a little while until January 5th.
They're going to take that ETF
relief program and they're going to make it 50%. They're going to give you $150 in maximum credit
when you switch to Ting if you have a contract right now. So noodle this over. You go to
linux.ting.com. You're going to get a $25 service credit if you bring a device, or you get $25 off
your first device if you don't have one. You bring a device, you're going to get $25 of credit.
If you have a contract to terminate, you're going to get another $150 in credit.
That means you get a max of $175 in Ting credits just by going to linux.ting.com
and by taking advantage of the early termination relief program.
$175 in service credits for no contract only paying for what you use
just your minutes, your messages and your megabytes
plus the flat $6 for the phone
no hold customer service
their incredible dashboard
full access to your phone
great community if you want to run one of the
alternative operating systems like Ubuntu Touch
Sailfish OS or Firefox OS
they have an active, enthusiastic community
around that
and you can turn it on right now you don't have to call anybody if you don't want touch, Sailfish OS or Firefox OS. They have an active, enthusiastic community around that.
And you can turn it on right now. You don't have to call anybody if you don't want.
You can use their web interface. You can get up to $175
in credit. Now, if you go to
linux.ting.com and you don't have a contract, you're still
going to get our $25 service credit, which will still
get you probably more than your first month.
This is such a great time.
If I were you, and by the way, if you
want to read more about it for yourself,
you can go to their blog.
They've got all the details up on their blog like they always do.
Ting is super great about that.
It's really easy to track what this company is up to because they're really transparent.
They put it all up there, and you can read it through it.
It's such a great way to get your mind around what this company is about.
But check this out.
I think one of the best smartphones ever made, the HTC One, the M7,
with those awesome front-facing speakers, which are amazing for podcast listening, you can own it outright.
$340 from Ting right now.
No contract.
$340.
And you only pay for what you use.
With this phone, so I don't have, like, Bluetooth in my truck.
I don't have, like, an aux-in jack in my truck.
But I don't listen like Bluetooth in my truck. I don't have like an aux-in jack in my truck. But I don't listen to the radio.
I only listen to podcasts when I drive.
And I use this phone.
I did.
With these front-facing speakers.
I didn't need a soundbar.
Now that I have my Nexus 5, I'm back on a Bluetooth soundbar.
This is the phone.
Like this is the one that got away.
I gave it to Rikai.
I gave it to Rikai.
Rikai's got my HTC One.
And I covet it.
And you can get it now for $340. Go to linux.ting.com. I gave it to, ReekEye's got my HTC One, and I covet it.
And you can get it now for $340.
Go to linux.ting.com.
They'll give you a $25 discount on that device.
If you have a contract you need to terminate, they're going to pay more than they've ever paid until January.
This is only good until January 5th, so why not do it now while you get it in this year, right?
$75 used to be the old deal.
Now, now they're doing $150.
If you have a contract, you got to get it.
And then you're contract free.
You can turn on hotspot tethering when you need it.
It's so slick.
Linux.ting.com.
Seriously, do Chris's all.
Go get yourself, treat yourself.
It's the holiday season.
The HTC One is the Cadillac.
I mean, the M8s, they have the M8 too.
I think it's more, probably.
Who needs it?
This is such an amazing phone. The HTC One M7, $340.
Linux.ting.com
and a huge thank you to Ting
for sponsoring Linux Unplugged. You guys are
crushing it for the holiday season. Just a
great opportunity, too. If you wanted to give
somebody a phone and then not have to stick them to
a contract, Ting's a great way to go.
You avoid that nastiness.
Well, Matt, let's
welcome Todd on the show.
Todd is from Purism, and they are currently crowdfunding the Librem 15,
a free Libre software laptop that respects your essential freedoms,
and I've crowdfunded it.
It's currently 29 days left on the crowdfunding effort.
They've raised $41,000 with a goal of $250,000. And Todd, welcome to Linux
Unplugged. Why don't you start by telling us why are you doing a crowdfunding campaign for a Linux
laptop? What's the goal here? Yeah, well, thanks for having me on. And so what we're doing is we
are clearly crowdfunding to reach a minimum order quantity to manufacture a motherboard that has the best free software
support on any laptop.
So this quantity of $250,000 is essentially to meet the obligations of a custom motherboard?
That's correct.
Yeah.
And actually, the way that it typically works when you manufacture a motherboard is you're
talking about minimum order quantities of 5,000.
What we've done is worked with the manufacturer of the motherboard to prepay the NRE fees so that we can get a small enough order quantity in,
so that way we can do a smaller run of units.
So that's why we can actually do it for just a quarter of a million dollars as opposed to what's typically a $5 million minimum order quantity.
So, Todd, you pitched the Librem 15 as the first high-end laptop that ships without any
mystery software, and that obviously resonated pretty well with me.
I like the look of it.
It's a metal housing.
Is it aluminum or is it plastic?
What is it?
It's aluminum housing.
Well, it looks really sharp and it's got a decent range of specs.
And it ships currently with Triscoll is the distribution of choice, right?
So if it does meet its goal, okay.
So the first thing that jumped out at me, and we can get into this maybe,
because I think this is probably the biggest questions on people's minds is uh it it sort of uh struck me a little bit as google's don't be evil
uh um slogan where as soon as you say it right then everybody's looking for every little evil
thing you do so in this one you know you sort of pitched it as the totally free libre laptop
but there's an asterisk in this sense that it does ship with a BIOS that does require a binary blob.
Can you tell us a little bit about that?
A little bit about that?
Yeah, correct.
So it's something that we try to be very clear about, actually.
So we put it right on the front page and also the certification and pretty much everything we put out there.
Is that we have everything from the software on, meaning bootloader, the kernel, the entire operating system, as well as all software,
as of course, by running Triscoll, all software we even install is free. And we also promote free
software by doing so. The area that we want to free, but is not currently free, is the BIOS.
There's actually a binary, the Intel FSP, which also includes the Intel memory engine ME, that is a binary provided from Intel.
Actually, it's technically provided through a subcontractor of Intel.
And what we want to do is have that freed.
That is our goal.
And then we have a, and that actually, once we reach that goal, we would actually be able to get free software foundation respects your freedom certification.
But we want to go beyond that and actually certify and free firmware that actually goes
into microcontrollers at the hardware level.
So really what it comes down to for us is we are as free as humanly possible right now,
but we want to push upstream into the manufacturing,
which no one has done, to be able to free the microcontrollers as well as that BIOS.
So that seems like a pretty ambitious goal. How realistic is a goal of that level of ambition
for a company that's not the size of an Apple or a Dell? I mean, to be honest,
for a company that's not the size of like an Apple or a Dell.
I mean, to be honest, that seems almost like a fight that would be hard for an HP or Dell to win.
So how does Purism plan to win that battle,
to convince them to open these and free these?
Yeah, so that's a great question.
So initially, it's really about business negotiations.
So initially, those large companies, like let's say Apple and Dell,
Apple and Dell have no incentive nor no belief system
that will ever care to actually free anything upstream.
So when we sit down at the table and negotiate with purchasing a hard drive
or purchasing memory, then we can, the number one thing we're discussing is,
are you a company who's going to back providing us with your source under NDA or where we can scrub it and release it or providing that source under an acceptable license?
And then we will purchase from you or in some cases actually work with them to provide the documentation to the Linux kernel developers to free the hardware.
So it's really mostly about that our number one topic when we're negotiating is freeing the source code.
So is it your observation now perhaps that perhaps other OEMs are not really concerned about this matter
and not really pressuring the manufacturers to do this? There's, well, Google clearly has, and they've actually made
some decent progress on freeing the BIOS, working with Corboot and a number of engineers who work
for Corboot. So that's, so Google is an exception to the rule. Clearly, they have other concerns for privacy and freedom. But as far
as pushing upstream, Google has actually made attempts to do so. And in some cases, has done
quite well. The main thing is that the push upstream has really not gone all the way to the
manufacturer who actually manufactures
the actual microcontroller.
It'll be to whomever's reselling it, and then it'll kind of fall on deaf ears and not actually
make it upstream enough.
I see.
That makes sense.
So the next thing that I think would probably jump out at the sharp-eyed viewer is it chips
with an NVIDIA card.
Now, we are aware of the Nivoo driver, but tell me a little bit about the choice,
because if this is a custom motherboard,
that must have been an explicit choice on your part
to include not just an Intel graphics,
but also a separate NVIDIA chip.
So why include that when the Intel's there?
Yeah, it's funny.
We actually had an ongoing debate about this,
and we could save about $75
if we just throw out the NVIDIA chip.
So it was a conscious decision to include it, which actually the debate was if we're going to
include hardware that can run with a free kernel level driver. But I also want to make it clear
that if you even use a free kernel level driver, you're still loading a binary from NVIDIA into the BIOS to actually be honor our purest beliefs that you can run completely
free software and just discard the NVIDIA. But those people who really want to include the
NVIDIA have that option. So it was really more from a customization standpoint.
But it was a debate that we went round and round. And one of the
issues actually comes from the Free Software Foundation, which is if you include hardware
that requires or only operates it at full capacity with a proprietary binary, then it's not
certifiable. So that was something we actually toyed around heavily with the certification for Free Software Foundation.
And the end result was sort of,
for somebody like me,
I think I might not have been inclined to fund it
if it didn't have the NVIDIA chip.
So for me, it was,
I might not use the proprietary driver,
but I'd like to know that in the future
if my workload needs it, I have it.
How will that work?
Say this meets its goal and we get the funding. How will that work
for me as an end user? How do I switch between Intel and NVIDIA?
Well, it's actually called a hybrid. So, actually
it's an NVIDIA 3D graphics controller. So, what happens
is when it boots and runs off of the Intel, but whenever it needs
to access 3D
acceleration, then it uses the NVIDIA graphics. So you can run full the entire system off the Intel
and then use NVIDIA when needed. You can't do it the other way around. You wouldn't be able to just
use, at least my understanding and our testing is that you would not be able to just use NVIDIA.
NVIDIA is added for the 3D control. And you can see from LSPCI output that you would not be able to just use NVIDIA. NVIDIA is added for the 3D control.
And you can see from LSPCI output that you can actually tell that the 3D controller is
on a separate PCI bus compared to the VGA.
Okay.
Okay.
Wow.
So it does the driver stack handles live switching.
So it sends what?
It sends OpenGL calls off to the NVIDIA chip and standard 2D calls off to the Intel chip?
That's my understanding, but that's actually outside of my range.
I have to bring in a developer to explain more about that.
Interesting, though.
So how are you feeling right now?
Looking at the funding, you guys have 29 days left.
It is the holiday season.
It's kind of probably a tough time to raise money.
How do you feel about right now, 29 days in, 16% funded?
Yeah, so clearly the trending for most crowdfunding is that you initially have a pretty good push,
and then it kind of plateaus for a while until the end because there's a lot of people who will sign up for the mailing list
or notify me when it's ending.
So we're getting an awful lot of interest, uh, specifically with, you know,
notify me as the, as the days start to, uh, dwindle. Um, but you're right in the fact that
launching during this time was actually a tough decision for us. We, um, we actually considered
running it through the middle of January, um, because, uh, getting after the holidays and
people clearly aren't buying for themselves. And this is going to ship in, you know, end of March, April timeframe anyway.
So we decided to go ahead with it now because just mostly trying to bring it to market.
The timing was right for our negotiations and trying to get some leverage with the manufacturers.
So we decided to take that leap.
But I feel really good with it right now to answer your question.
So I guess kind of what I'm grokking from what you're telling me is there's not really
a straightforward way to do a custom built for Linux laptop in a one, two, three, four level
quantity, maybe one sale, two sales, three sales every day. In order to have leverage with the manufacturer, you've got to be able to cut them a big fat check.
And so that kind of does that is that dictating the crowd supply funding approach instead of just taking an order here and there as people want them?
That's exactly right. So, I mean, clearly we can we can buy, you know, something off of the shelf and then install Linux on it and ship it.
And then that's really what I consider the old days of Linux.
I actually want to change that in the sense that I want to push the GNU Linux system upstream
where we're actually shipping it by default, manufacturing it specifically for that.
So to be able to do that, we can't just buy onesie twosie.
And if we did, we could just launch it on our site and sell them,
you know, one offs. The point here is to see if there's enough interest in producing
manufacturing and basically backing a company that wants to push the free software movement
up into the manufacturing chain, so that we can design and manufacture and ship products that's
specifically targeted towards this audience. Yeah, I mean, that seems like a total catch-22
because to make this experience better for future Linux laptops, you have to have that initial
leverage at the negotiation table to make it worth their time to go through the trouble of freeing
these binaries and this documentation and the source code up.
So it seems to be a unique situation.
So why CrowdSupply versus something like Kickstarter?
I'm just curious because I see a lot of hardware, open source hardware funded these days and
I'm seeing different sites use.
So what did CrowdSupply offer that maybe a Kickstarter didn't?
Well, there's a couple of things.
The first is actually I got to meet the founder of CrowdSupply and I was quite
impressed with their capabilities. They also did back Bunny's Novena project, which I think is a
great project, clearly from a maker standpoint, right, where he actually manufactured the
motherboard himself. That's fantastic. And so they had experience with
supporting the free and open source world. And they also actually have worked with the
Free Software Foundation through my introduction, actually, to make sure that people can buy without
JavaScript enabled. So they're going to become the first free software foundation endorsed
crowdfunding website.
So that's great. That's also great in my mind.
So those are a couple of decisions why we went with CrowdSupply over anything else.
So, Todd, give us a little bit of your background.
Have you been involved in free and open source software for a while?
Yeah, I've been ever since Debian 0.2 days, so about 1994, I've been a Debian developer.
I was actually a Debian developer before it took an arm and a leg to become a Debian developer.
And then since fell out of that and have been installing Debian forever, over 100,000 units installed. So when I was system architect at a couple different companies, I used fully automated installation to build, deploy Debian-based multimedia services or audio-based services.
So if you go into most retailers, you'll go into and you'll hear audio in the store.
hear audio over the in the store and that'll
run a system that's
60% of them are going to be a system that I
was a system architect on which are all
going to be a Debian based multimedia
system. Very cool
that's a pretty big runs Linux
and I suppose then have
you is this in some way
I'm imagining you're probably a bit like me where
you've finally been like none of these laptops
have satisfied me and I've I find myself to be pretty happy with a certain subset of laptops.
Mostly one of our sponsors, System76, makes some great laptops that run Linux really well.
But when I go out in the market and I look for laptops that are custom built for Linux,
there's not a lot out there.
There's plenty of room left in the market.
And something that's metal and really tightly well built, I often will look over the MacBooks and go, man, if those could flawlessly run Linux, I would consider it.
But until that day, it's not for me.
So I always felt like I've kind of been in this zone of searching and never really found the perfect solution for me personally.
So is that what sort of motivated you to start this project?
Actually, it's interesting. I mean, I might as well have recorded that and played it back. That's that is one of the driving forces of why I formed Purism is for for decades. Well,
maybe not decades, but over a decade, I've been hunting for, you know, every year, right, or
every year and a half or so getting a new laptop, where I bought from
Think Penguin, I bought from System76.
And, you know, I've always liked the fact that what they believe in and what they promote
has been great.
The laptops themselves arrive, and, you know, they are off-the-shelf products that can be
purchased in ones or quantities of 10, usually, and then stripped, installed Linux, and delivered it.
So I've never really been, you know, really satisfied.
And I think to your point is, you know, I have for one consulting gig,
they shipped me a MacBook Pro, and I really like the hardware.
I thought
it was well-built. It clearly would last.
And so for me, I wanted to get something
that was custom-built
specifically for GNU Linux, where all the chips
would work without any binaries,
and then just decided to
if no one else had done it,
that I might as well be the one.
Yeah, I can understand that.
So, in my room, if you guys have any questions, go ahead and speak up. But Todd, I'll have a link in
the show notes, and I would encourage people to go check it out. So one last question. Say this
makes its funding. Can I put more than eight gigabytes of RAM? What's the limitation there?
Because I see right now in the dropdown, it's eight gigs. Is that a physical limitation,
or is that just the specs you've offered? Can I can I put up to 32 gigs or give me an idea of like the expandability in this
thing? That's, that is our number one question. So, um, so right now we actually are going back
to our reference design to actually increase that to 32. It is a physical limitation right now in
our design. However, that is something we can change the issue. Actually, it comes down to,
um, the space within it.
So what we're doing is we might end up having to do something where you can go with a secondary solid state drive in the CD-ROM, DVD-ROM drive bay and then be able to expand your memory.
So we're basically looking within, you know, trying not to because I can't change the case.
Right. So I'm looking at ways in which we can fit more memory within the same footprint.
Sure, and you know what?
As somebody who, I bought a few CD-ROMs recently,
on my more recent purchases, and haven't used them like once.
I used them for the first time this last weekend,
so I would totally take that offer.
Yes, we end up, when we have that offer now,
you can swap the CD DVD for a second
bay of solid state drive, or hard drive if you want. But the, so the issue ends up being clearly
a drive is smaller in footprint, so we can end up, it would be almost a, it would be a requirement,
potentially a requirement, to drop the DVD and go with an extra RAM. But we're still toying with that, and we'll update the project page
once we get the confirmation
to be able to go up as much.
But a lot of people, like on the Cubes OS,
they really want to have more than 8 gig
because for running VM software,
you need to have a lot more RAM.
All right, that was all of my questions.
Wimpy, I know you had one
that will probably be of interest to the audience.
Go ahead.
Yeah, so you already spoke about the FSF certification,
and I wondered if you might try and find out if they're going to be flexible on that,
if you can run the laptop in, say, an Intel GPU-only mode,
and therefore you're unlocking the full potential of what that configuration is capable of,
if they may endorse or certify the laptop.
Yeah, actually, that would happen.
The hang-up with the Free Software Foundation certification is actually the BIOS binary,
the Intel FSP or Intel ME.
That binary is the one that we have to free that, which we're working towards.
We have a couple of avenues to produce a free, completely free BIOS. So we use core boot,
but we'd have to still use this Intel FSP binary inside of core boot to actually boot the machine.
And it's a long sorted story of as,orted story as it's moved out of the kernel
and more and more into the BIOS level for these controls.
So to answer your question directly, the Free Software Foundation,
the main blocker for us getting the Respects Your Freedom certification
is the Intel BIOS to free the BIOS. It is not having a secondary video card that is
available to use as a binary or with a binary. That makes sense. So I think the other,
this is a common question. I see it right now in the chat room and I've seen it when I covered it
before. I don't actually hold this opinion, but I think a lot of people think it's expensive for
the specs. And I'm thinking that might be due to the quantities or the scale of the build.
Do you want to touch on the pricing and why it is the price that it is?
Yeah, clearly I'd love to sell it for $200 if I could.
But the issue is exactly as you described.
When we're talking about manufacturing a motherboard,
we have a minimum order quantity to be able to place that order
to actually have the motherboard manufactured.
It's all about the tooling.
So we paid an NRE fee to get it down from a 5,000-unit order
to really in the few hundreds.
So what, of course, that means is that our per-unit cost has to go up.
And so we're taking a – we have the belief that people will back the business because of the belief
system and what we're trying to do and understand that it's a short term,
larger price tag to be able to get to a completely free manufactured device
later with a lower price tag when we can actually order it in larger volumes.
So if we were to be able to place an order for 5,000 units, then we could actually have
that price probably cut at least, it would be close to a half the price.
Wow.
Well, that makes sense, too.
I mean, that's sort of just the dynamics of the market.
So, Todd, I really do wish you the best luck.
And you guys can go to crowdsupply.com and look for the Librem 15 from Purism.
And we'll have a link in the show notes.
You can go directly to it.
And if you all go back, then I'll probably get mine.
And then I'll give you a review on the Linux Action Show.
But, Todd, you are welcome to stick around.
We're going to continue on with the show.
And I wish you the best of luck.
And I hope I get mine in around March or April time frame.
That would be really cool.
Hey, thanks so much for having me.
So we're going to – I want to talk about something the CoreOS project is doing,
and I want to break it down because I think it's huge and I think it's super necessary.
And it's already been piled on by the free software press – that's not even a thing –
by the people who pretend to be press of the free software movement.
I'm sticking myself in a bigger hole, but it sounds closer.
They've already called it the container war.
Docker goes to battle.
I'm going to tell you what's really going on.
We're going to break it down for you guys.
And it's actually a really, really good thing.
But first, I've got to tell you about a really great thing.
That's DigitalOcean.
Head over to DigitalOcean.com right now
and go get yourself your own simple cloud server.
DigitalOcean is dedicated to offering the most intuitive and easy way for you to spin up your own server.
And you can probably do so in less than 55 seconds.
Pricing plans start at only $5.
That'll get you 512 megabytes of RAM, a 20 gigabyte SSD, one CPU, and a terabyte of transfer.
It's awesome because DigitalOcean has data center locations in New York, San Francisco, Singapore, Amsterdam, and London.
I got an email from listener Patrick today.
And he's like, Chris, you've talked a lot about things you use your DigitalOcean droplet for.
But have you ever considered the fact that $5 a month is cheaper than most any really good Minecraft hosting solution out there?
And with DigitalOcean, not only do you get the ability to host your own Minecraft server, but you get the snapshots and the backup and the templates, right?
And, of course, Minecraft's one example.
It could be WordPress, GitLab.
It doesn't matter.
BitTorrent, Sync, SyncThing.
Why not roll your own SyncThing solution,
and the DigitalOcean droplet could be the off-site cloud storage?
You don't need anything else.
But it's really about the interface.
This is kind of why I think DigitalOcean, they sort of broke through the barrier when it came to making this something that's accessible to anybody.
Like, you know, you can be a 15-year server admin, you could be a 10-minute server admin, and you're going to find the DigitalOcean dashboard is amazing.
It's intuitive and power users can replicate it on a larger scale with a straightforward API.
And for me, I can't use that API.
I couldn't use an API to save my life.
I could look at it and be like, oh, so that's how you do that. But to actually use it? No.
Now, I know for a lot of you out there, you're rolling your eyes and like, oh, Chris,
you're such an old man. And that's great because then you can take advantage of it today. But for
me, I take advantage of all the awesome community resources built around that API. There's applets
for your Linux desktop. There's ways to snap it in with your existing management infrastructure, command lines.
It's so cool.
And one of the things that we love about DigitalOcean is you can go over there and you can apply
our discount to your balance.
This is a great way if you forgot to use it when you signed up.
Now, we're doing the transition between months right now, so we have two promo codes.
Right now, I know that Unplug November works for you, so you can go over there and use
the promo code UnplugNove November and get a $10 credit. Then you can try out the $5 rig
for two months for absolutely nothing, for free. But it turns out it's December. So Unplug December
is probably going to start working later this afternoon. And you can go over there and use that
as well and get the $10 credit. Go check it out. Really, if you're even a casual Linux user,
this is such a fun experience. And since you get it for
free, why not just try it for a little month
for a couple of months and just play
with it. It's so neat. I have
right now, I have my Quasals core up there.
And one of the things that turns out
didn't know this about Quasal, but
when you're in all these chat rooms logging all the time
that database gets
big. Like 8 gigabytes big.
And if I didn't have that sucker sitting on an SSD drive, I would be in a world of hurt every time I go to open up my IRC client.
But because DigitalOcean has incredible bandwidth speeds, has those SSD drives, when I open up my IRC clients, boom, it all gets pushed down to me.
It's amazing the amount of transfer I get.
And I get up to a terabyte for free. And then DigitalOcean has easy step plans to upgrade to the pricing. In fact, they have it all listed out down to me. It's amazing the amount of transfer I get. I get up to a terabyte for free. And then DigitalOcean
has easy step plans to upgrade to the pricing.
In fact, they have it all listed out on their site.
You just click that pricing tab and it's really
straightforward. And the other thing that's cool,
hourly pricing. You know like
your favorite hotel? They have hourly
pricing over at DigitalOcean. So if you need to scale for a little
while, you're doing a big deployment, something like that,
you can spin up a server and then shut it down when you're done.
And DigitalOcean's up in the game on the tutorials for the community.
In fact, Patrick, who sent in the email about his Minecraft server,
used a DigitalOcean community tutorial to set up that Minecraft server.
And this is what's neat is they realize this is a massive advantage
that they have over their competition.
So instead of just sitting there and just kind of sitting back and be like,
oh, yeah, everybody, write it in.
No, they're actually going to pay you,
and they have editing staff that's going to work with you.
You can earn up to $200 to write a tutorial for DigitalOcean.
Now, if you are a Linux geek and you know how to set something up
like a frickin' Minecraft server, why not go make $200?
Seriously.
Now, somebody already took that one,
but there's hundreds of other topics you could write
and work with their staff. DigitalOcean.com. Unplug November to get that $10
credit. Try out that $5 rig for two months for absolutely free. So freaking cool. It is the
coolest thing ever. Seriously. As somebody who's worked in IT for 15 years, when I see this,
I'm sorry, I cannot help but geek out.
Today, after all of this time of using
DigitalOcean for months and months and months,
I still sit down and I think,
holy smokes, this is crazy cool.
The fact that they saw the potential
behind SSD when nobody would touch
SSD because of the cost. The fact that they
saw that KVM was an awesome virtualization
platform before anybody else saw it,
and that they build it all on top of Linux,
and they integrate with CoreOS.
They're all about that Docker.
One-click deployments.
It's so slick.
It's such a good system.
And on top of all of that,
you get the satisfaction of supporting
your favorite podcast network, DigitalOcean.com.
Use the promo code UNPLUGGEDNOVEMBER.
Seriously, go see what I've been talking about.
And a huge thank you to DigitalOcean
for sponsoring Linux Unplugged.
All right, Matt, I don't know if you saw the big hoopla,
but you know Docker's all the rage these days.
Everybody's talking about Docker, right?
Well, CoreOS wants to talk about Rocket.
Wimpy, did you see the announcement about Rocket?
I have not, no.
All right, all right, good. I'm going to teach you something today, Wimpy, so buckle up. Here we go. I am super excited about Rocket? I have not, no. Alright, alright, good. I'm going to teach you something
today, Wimpy, so buckle up. Here we go.
I am super excited about Rocket.
You put your pants on the outside
of your trousers now.
My pajamas, sir. The fact he's wearing pants
at all is just like... No, I'm just kidding.
Keep in mind it's pajamas. Let's not get crazy.
Alright, so
if you listen, depending on who
you read, like the register, apparently a new Linux container war has begun.
That's crap, right?
That's total crap.
First of all, I want to set something up right up front.
Core OS is not moving away from Docker.
Docker will continue to ship in Core OS and be well supported.
And I want to give you the recap because the Core OS guys, which who we've interviewed on Linux Action Show, I tried to talk to Mark Shuttleworth
about it last week, and he very quickly changed the topic, which to me was maybe an indicator
that he's aware of the heat that CoreOS is bringing. There's already major clients, enterprise
grade clients that use CoreOS. The whole idea behind CoreOS, right, is you isolate out a
lot of the applications and the server services into containers, and then you just roll that
base, baby. You get that OpenSSL patch. You get that Shellshock patch. You patch, patch, patch,
mother, because you know what? All the maps are in containers and it's safe. Well, that's the idea.
But the problem is, and this is something that's been bothering me for a while,
Docker has gotten big. And I don't mean like big code-wise. I mean big like publicity-wise, big company-wise, big contracts-wise.
And Docker is expanding out way, way beyond the original vision, right?
It's a whole platform. Docker is not a container anymore.
Docker is a platform of services in which containerization is one of those things.
So that's where CoreOS picks up. They say, when Docker was first introduced in early 2013,
the idea of a standard container was striking and immediately attractive.
A simple component, a composable unit that could be used on a variety of systems.
The Docker repository included a manifesto of what a standard container should be.
This was a rally cry to the industry, and we quickly followed.
Brandon Phillips, the co-founder and CTO of CoreOS, became a top cry to the industry, and we quickly followed. Brandon Phillips, the co-founder
and CTO of CoreOS, became a top contributor to the Docker project. He now serves on the Docker board,
the governance board. We thought Docker would become a simple unit that we could all agree on.
Unfortunately, a simple reusable component is not how things played out. Docker is now building
tools for launching cloud services, systems for clustering,
and a wide range of functions from building images
to running images, uploading, downloading,
and eventually even overlaying networking,
all compiled into one monolithic binary
running as root on your server.
The standard container manifesto was removed.
We should really stop talking about Docker containers
and start talking about the
Docker platform. It's not becoming a simple composable building block we had envisioned.
So they're introducing Rocket. We still believe in the original premise of containers that Docker
introduced. So we're going to do something about it. While we're at it, we're cleaning up and
fixing a few things we'd like to see in a production ready container. And the things
that they lay out here are things that
have fundamentally bothered me about Docker. I am a huge supporter of Docker. I think it's great. I
think it continues to be great. But a few things have bothered me. Isolation should be pluggable,
and the crypto primitives for trust should be strong. Image auditing and application identity
need to exist from day one, i.e. signed containers, signed applications, so that way when you get them distributed to you,
you know they are legitimate from the original source that you expect them.
Image distribution.
Now, the Docker Hub is neat because you can check in something like Smokeping,
check it up to the Docker Hub, and anybody can check it out.
Boy, that's great.
But the problem is it's centralized, right?
It's an open source thing that's centralized.
Well, we know what happens when you centralize things.
CoreOS thinks that discovery of container images should be simple
and they should have a facilitated by a federated namespace.
So you can do distributed retrieval.
This opens up the possibility for alternative protocols to distribute these images,
potentially like BitTorrent,
and deployments inside private environments
without the need of an external registry
like the Docker Hub.
That's a huge one for me.
Also, they believe the format and runtime
should be well-specified
and developed by an open community
with implementations and tools
that can run the same container
consistently across different platforms.
Rocket is what they're introducing.
It's a command line tool for running app containers.
An app container is a specification of an image format, container runtime, and discovery mechanism.
So an app container tells you what it is, what the runtime stuff is you need, and how you get this container.
That's what an app container is.
Rocket is the first implementation of an app container.
And they expect there will be other ones. And all of this information is linked up in the show notes, but they have a fact that I think I want to cover. They say here, why are you doing this now?
enterprise environments and we cannot in good faith continue to support docker's broken security model without addressing these issues this is something we've talked about on tech snap on
several episodes of tech snap there's fundamentally some issues with the docker security model
and rocket plans to address those they go on to say will core west continue to ship docker yes
period we will continue to make sure core west is the best place to run Docker. So this is not exactly a war,
so much as a fundamental divergence of how container technology on Linux should work.
Now, the Docker project very quickly posted a follow-up.
There's a lot in this, but I highlighted a few important areas.
They say, while Docker continues to define a single container format,
it's clear that our users and the vast majority of contributors and vendors want Docker
to enable the distributed applications consisting of multiple discrete containers running across multiple hosts.
In other words, some of the things we're doing is because people want us to do them.
We think it would be a shame if the clean open interfaces and anywhere portability
and robust set of ecosystem tools that exist for a single Docker container applications were lost if we went to a world of multiple containers and distributed applications.
They say we'll provide more detail at DockerCon conference this week in Amsterdam.
vendors, and contributors, whether people add value in the form of contributions to Docker as independent projects that build on the Docker container format or as plug-ins
to the Docker orchestration APIs or otherwise.
We hope that the open, layered approach provides the best options for all.
So Docker says we are an open, layered approach.
Yes, we have some bits and components here that we keep secret sauce, but we have open
governance.
They say in some cases, of course, there are technical or philosophical differences.
That appears to be the case with the recent announcement of Rocket.
We hope to address some of these technical arguments posed by the Rocket project in a subsequent post.
So more to come.
Now, I thought it was interesting.
Some of you might not care at all, but what I found to be interesting is two things. Containers are obviously a huge deal. Not new to Linux, not new to operating systems in general, but there's a lot of heat around them right now. I think that's undeniable. you can look at the massive pickup that a freshman project like CoreOS has seen
because it's based around containers.
Red Hat Enterprise 7, kind of, I think some could argue,
rushed to implement Docker support
because it was such a talked-about feature.
The largest enterprise Linux distribution in the world,
the top granddad of enterprise Linux sisters
that everybody recognized,
oh, it's Red Hat, oh, it's Red Hat, it's Red Hat.
They rushed to ship Docker support
because they felt the pressure to be hip and modern
and have Docker in there, right?
So to have CoreOS, one of the biggest contributors to Docker,
come along and announce something else,
people have reacted kind of strongly to this.
And Matt, you can probably guess,
some people have thrown their arms up in the air and said,
oh, this is the problem with Linux.
Everything's always forking.
Yeah, it's like fork this, fork that, fork everyone.
And I think that's kind of the attitude that tends to kind of
dwell out of this whole thing.
But I think they're actually, I think Rocket is actually going
to address some serious competitive, I think it's going to bring
some serious competitiveness to Docker.
I think it's like it's addressing issues that I've watchedcker go from this is a really great way to make sure this application runs the same
exact way on any distribution to the docker platform which is this is huge thing and it's
almost too big now i i think so i think you know of course there's the old marketing law first but
as we saw with google they can you can obviously trump that but i think the real key is what they're
pushing forward is their big claim to fame is security.
That will resonate with enterprise folks.
So that may actually work out in their favor.
And they may get the traction they need.
Yeah, and I like this federated discovery system they're talking about where the protocol to distribute the container is pluggable.
It could be BitTorrent.
There's obviously some advantages to that, especially for smaller projects that don't have huge bandwidth capabilities.
That's kind of a neat way to distribute things.
I like, too, that I maybe could have a private, quote-unquote, hub of these things.
Maybe we could have an internal JB server where we check things in and out.
It doesn't need to be up on the centralized Docker hub.
So I think they're actually doing – oh, and also they're doing some actual real good serious crypto signing.
They had a Google Hangout today where they talked about some of the technology they're using and why they chose it.
And I think some of the hostile attitude that people are taking away in the press is they're kind of calling Docker out for not totally being enterprise ready.
And not everybody really wants to hear that, but it's the truth.
I think you nailed it i also think the fact it seems like they did wait long enough for docker to kind of absorb major funds and say okay now what are you going to do about the security issues
and and then it was just dead air and so now i think they're kind of like okay you know what
fine we'll do it yeah yeah and i think you know core os obviously has a lot of skin in the game
because they're sure and they're this isn't core this is not going to be Core OS specific code.
This is going to be open code that's available to anybody to use just like they did with etcd.
And I'm thinking for us Linux users, this is a really good thing.
And every time we talk about this, of course, the FreeBSD guys always say,
Oh, we've had Jails forever and the Solaris guys, oh, we've had Zones.
And somebody's like, but I've been using LXC.
But the thing is, is this whole plug-in software
and pull-it-down thing is unique and special,
and I think it was sort of the tipping point
that made this sort of more of a common, popular concept.
I'm kind of excited about what the CoreOS guys are working on,
and so producer Q5Sys is working to line up a chat
with folks from CoreOS to talk to us about
it. So that way we can pick their brain and
kind of maybe get directly from them
where they feel like Docker was maybe slightly deficient
and needed to be replaced. And what
some of those serious issues they think they have
that don't make it ready for enterprise. I'd like to get their
answer on that too.
It's interesting times.
It just came out of nowhere.
Of course I don't follow CoreOS super closely, but I wasn't expecting this.
Yeah.
Mumble Room, any thoughts before we run on this topic?
Last chance or forever hold your peace.
Going once, going twice.
And no.
And sold.
So Q5 Sis in the production chat tells me right now that we might have Brandon on Sunday to do an interview.
I don't know if we'll air it in Sunday's show since we have like an hour
long chat with Lenart. But if you tune in
for the live show, you'll catch the live taping of it.
And I don't know when we'll air that.
But we'll work that out.
So we should have Brandon on
from CoreOS. Brandon is one of the
top contributors to Docker. He sits on
the governance board of Docker. He works on the
CoreOS project. He's one of the people behind Rocket.
And he'll be joining us on the Sunday live stream
to answer those questions,
if all the scheduling works out.
So that should be good.
Good stuff.
Okay.
Well, we just have a couple of things to cover before we run.
I'm still looking for the best of,
so if you go to the show notes,
you'll find a link for a form to fill out
for the best of moments of Jupyter Broadcasting.
We're going to try to take the holiday week off
and give people a recap of some of the best stuff.
You can find a link to that in the show notes. We need your help and your suggestions of moments that Jupiter Broadcasting. We're going to try to take the holiday week off and give people a recap of some of the best stuff. You can find a link to that in the show notes.
We need your help and your suggestions
of moments that you liked.
It doesn't have to be a lot, but you'll see the form.
We've got all the stuff in there.
And don't forget, we'd love to have you contact us.
Go to jupiterbroadcasting.com slash contact.
Choose Linux Unplugged from the dropdown
and send us in your feedback.
We love doing the feedback at the top of the show.
And you can always go to linuxactionshow.reddit.com.
We'll have threads for this specific episode,
but you can start your own thread.
Just comment on existing threads, vote things up or down.
Helps make a better show.
We've got lots of stuff, too.
I mean, more and more coming up.
It's funny, because I thought we'd have a big slowdown
with the holiday season, but our producers,
they're a crack team.
They're working on keeping content flowing.
And with all the news, and thanks to folks like Todd for coming on,
we've got lots of good stuff to talk about.
So I think it's going to be a couple of good weeks of shows.
So don't go anywhere.
Even if you get the time off, don't forget your Linux Action Show,
Linux Unplugged shows will be here.
And you can join us live.
We do this show Tuesday, 2 p.m. Pacific.
Go to jupyterbroadcasting.com slash calendar
to get that converted to your local time.
All right, Matt, well, I'll see you on Sunday with our chat with Leonard Pottering, all right? Sounds good.m. Pacific. Go to jupiterbroadcasting.com slash calendar to get that converted to your local time. All right, Matt, well, I'll see you on Sunday
with our chat with Leonard Pottering, all right?
Sounds good. See you then.
Okay, everybody.
Thanks so much for tuning in this week's episode
of Linux Action Show.
Nope, of Linux Unplugged.
We'll see you right back here next Tuesday. Thank you. Well, what show am I doing?
What day is it?
Linux Action Show.
I got too many Linux shows.
Who am I?
What am I doing?
Yeah, yeah.
I got too many.
Holy crap, I have no pants.
No, it's pajamas.
It's pajamas, I tell you.
All right, so go to jbtitles.com uh go pick a title
everybody go vote i got kind of those days where all the days tend to run together yeah yeah well
and you know what turns out uh pajama bottoms are way more uh like they keep the heat in a lot more
than jeans do so it is hot in here it's hot yeah I've been wearing my cell phone for years. What?
I'm sat here in my pants, Chris, because it's so humid.
Yeah, remind me where you're at again, Wimpy?
It's not our type of pants. Wimpy's on the road.
I'm in Kuala Lumpur.
Kuala Lumpur.
In Malaysia.
What are you doing?
What are you doing?
Well, that's a really good question.
I'd love to tell you all.
If you promise not to put it out in the show, I can explain what I'm doing.
You'll have to tell me later then because I'm still recording.
I don't want to get you in trouble.
Well, no, it doesn't matter if it goes out on the live stream.
Just don't put it in the edit.
All right.
Before you tell me, I want to give a plug to something that I meant to mention in the show.
Then I'll turn off the recording and you can tell it just to the live stream.
So before we go, I want to give a plug, and then we'll wrap up.
There is a QMU advent calendar, and this is the best thing I've seen all year.
So every day they're doing a QMU virtual image pick.
And today is day two, so we're just at the beginning of it,
and it's modern DOS.
The free DOS disk image contains the latest and greatest from the land of the disk operating system.
Besides checking out DOS with a TCP IP stack, he says you can relive id Software's early shareware hit, Commander Keen.
As a bonus, the free software edition of Jetpack is also included for your crazy jet-powered platformer action.
So qmu-advent-calendar.org.
I'll have a link in the show notes.
Day one was Slacker's time travel,
so you can go back and get the old ones.
This is when you dig up the earliest Slacker image to be had,
and you can go play around with it.
It ships with kernel 0.99.12.
Yeah, I said 0.99.12.
So isn't that a great idea, a QMU advent calendar?
That's crazy.
That's the best thing I've seen all week.
I just had a great experience
just a few minutes ago. For the first time, I needed
to actually print from my
desk workstation. I have printing set up on the machines
here in the studio for notes, but
today I was up in my office
and I realized I had CUPS installed, but I
didn't have the PPD
file for my brother laser printer. And I realized I had CUPS installed, but I didn't have the PPD file for my brother laser printer.
And I bought it because I had read that it was a fairly well Linux-supported laser printer.
And so I just did a Packer search for the model number of my brother printer.
And then the one result I got back was the PPD file for my exact brother printer.
I Packer-s that file,
that package name, and within 30 seconds, I had all of the cups printing set up for my brother laser printer, and I printed out the notes for today. It felt a little 2014
is all I'm saying. It just felt like the future a little bit. That was the easiest printer
setup I've ever had.
Yeah, there's lots of brother support in the AUR.
I think I maintain two of the brother printers there because I've got one at home.
And what I've done recently... The AUR actually has the exact same printer drivers that you'd get from Ubuntu in terms of the extras.
Yeah, they're usually dev files that are then extracted and sprayed around the appropriate places.
And what I've done recently at home to avoid that whole printer driver
sort of arrangement is there's somebody who's implemented
the Google Cloud Print API and services that integrates with CUPS.
So I've got a Debian Wheezy box that has got CUPS in it,
so it's actually running Open Media Vault.
And then put the Cloud Print API into that.
And in order to get it to work,
you actually have to put the printer drivers on the CUPS server.
So now that printer advertises itself through the Cloud Print API
for your Android phones and your Chromebooks.
Not that I've got a Chromebook.
And the Chrome browser.
Yes, but it also makes it publish it over multicast DNS on the LAN.
So you just get a Wi-Fi connection and that printer appears and you can do it without having to install anything on your machine.
So it works from live ISOs and stuff like that.
That's really – so you just set it up once on that dedicated box and then all –
You set it up once on that box and then it's available forever and always on everything.
Yeah, that Google Print thing is actually pretty handy, I got to admit.
It is kind of –
Yeah, it's a bit janky to set up, but it works really well once you go through the steps.
I've been thinking about doing the exact same thing using my server.
Only thing is it's running OpenSUSE, and it has been for several months now.
The problem with OpenSUSE is it's running an old version of CUPS
that's incompatible with that framework, the Google Print.
So I don't know.
Really? Which version of OpenSUSE?
13.1.
Okay.
But it's like running 1.5, and they haven't updated it even in 13.2.
Yeah. Okay.
Huh.
So my experience was sort of, it was just, it was so hassle-free that I thought to myself,
I'm pretty confident if you have the right conditions met, and I think that's always
a big if, but I think if you have the right conditions met, setting up printers is a lot
easier under Linux than it is under Windows.
Like, if you guys, I don't know when the last time you guys have set up a Windows printer,
but you still have to go through this arcane wizard,
and it still asks if you want to search a disk,
and all of this stuff that you would never expect a regular user to go through.
It makes me vomit every time.
Even for HP printers, it's worse,
because you have to go and get a 600 megabyte executable
and install boatloads of shit.
Last time I bought a printer, I unpacked it all, took the little strip off the printhead and plugged it into my laptop and sat it down, put some paper in.
And then I went and sat at my desk and there was a little pop-up that said, your printer is installed and ready to print on Linux.
And this was like four years ago. And I went to amazon to review it because that's where i bought it from and i reviewed it
and left a shining review and every single other review was really terrible panning it not because
of the printer but because of the windows drivers yeah yeah i'm i'm actually have a story as well
it's about printers and stuff because my uh had like constantly kept asking me like, will you fix this printer stuff?
And there's a ton of issues with his Windows thing and I showed it with my laptop.
I plugged it in, added it to my thing.
It took like 20 seconds.
I showed him.
It was already connected.
He's like, okay, give me that.
Yeah.
And now he's been using Linux for a couple months.
Yeah.
See, that's the thing about CUPS is it's just amazing.
Yeah. Yeah, when I started using CUPS,
to have all of your printers manageable through a web interface
was groundbreaking from a tech support standpoint,
just totally groundbreaking.
Not having to remote desktop into a Windows box
and go into the print queue.
And by the way, if the Windows box,
and this is still the same today,
if the Windows box is struggling with a print job for some reason,
sometimes the spooler is just totally nonresponsive. You can't go in there and delete
documents. It is so out of the 80s. It's almost like you go
into a time machine when you manage print jobs in Windows. And I became proficient at
net stop, quote, print spooler, close quote, enter,
net start to restart the spooler service because it constantly crashes under Windows.
And I'm talking in high load. For home printing, it's probably not so bad. But I just,
when cups came along and I could actually manage a failing print job and I could view all of the
printers in one window, in one screen, oh, it was so powerful. And then here I am thinking this is
going to be way overkill for just a home printer setup. Because honestly, my experience with cups
is multi-hundred printers. It's not one printer. So I'm thinking to myself,
oh, here I go. I got to go. This is the problem with Arch. I got to go install cups and I'm going
to spend 15 minutes configuring it. And then I got to figure out how to get my printer to work.
And I'm just going through this whole list. And then it ended up being like a 30 second process.
And I just thought back and good. Cups is amazing. It works for one printer all the way up to hundreds of printers,
and it's got awesome things about it in the whole range.
It's a really powerful printing system.