Quoting: denyasisI've thought about getting my child an Rpi4, but the more I think about it, I think I'd rather get a NUC

You should get three. You could call them Larry, Curly and Moe. NUC NUC NUC!

View this comment - View article - View full comments

Quoting: STiAT

Quoting: pleasereadthemanualThis would be particularly unappealing for Adobe, for example.

Most of Adobes software is actually web-based now.

Hmmm . . . I just had a tech guy come down to help me be able to use Acrobat again, and that didn't seem to be quite the case.
I work in a university library, fair number of employees across three locations. For all the other software, what happens as far as I can tell is the IT people buy a bunch of licenses and then the software is available on your desktop. And that's what old Adobe Acrobat Pro was like. But with the new "Creative Cloud" thingie, the way it worked was they bought a bunch of licenses subscriptions, and then I had to click on this Adobe Creative Cloud thing on the Windows software menu, and then that brought me to a web portal, and I had to make a password and stuff--that's me, individually, but only after I was pre-authenticated as being someone who was allowed to in the first place, right.

But then it wasn't like anything was a web application. What the cloudy portal let me do was download the application (and in theory other Adobe applications). As far as I can tell, the only thing "cloud" about it is that it phones home to the "cloud" to remind the mother ship of its existence and if the library's subscription runs out Adobe will nuke it from my desktop directly. Assholes.

View this comment - View article - View full comments

Quoting: pleasereadthemanual

Quoting: Purple Library GuyWhat interests me more is the closed source, commercial software side. This could set up Flathub as a distribution source for Linux software from closed source software publishers. That could be good for Linux. Not in the short term good for open source, but I think that on Linux, open source software tends in the long term to win out (less so in games, which have various weird characteristics such as being ephemeral and art-heavy). What one could see is an increased ecosystem of closed source (mostly non-game) Linux software distributed via Flathub, complementing the ecosystem of mostly closed source game software from Steam, and combining to make Linux a viable desktop for more and more use cases. But the open source software continues to be distributed on it mostly via more traditional distro repositories and package management. As the Linux desktop grows, with more user and developer base I would expect more and more of the open source alternatives to these Flathub-distributed closed source applications would reach critical mass and gradually supplant the closed offerings.

I would like to agree with you, but I can't see Adobe ever writing software for GNU/Linux. Affinity Serif maybe, though unlikely.

True enough. Adobe, fuggedaboutit. Mind you, Adobe are turning so evil these days that forget Linux, software users in general are going to have to find some way to dump them. What the fuck is this "Adobe Creative Cloud" subscription bullshit?

Quoting: pleasereadthemanualI don't think Flatpak, regardless of how polished it is, will get software publishers caring about GNU/Linux, but it does solve a pain point for many small developers who want to support the platform but are put off by the horde of distribution methods. I think the market share needs to come before publishers will consider distributing their software on GNU/Linux, and the only tool we really have to help with that is WINE.

Makes me wonder how workable it would be to build Flatpaks with (Windows application + Wine), and whether Codeweavers could make a few pennies setting up such packages on Flathub for Windows software sellers?

View this comment - View article - View full comments

Quoting: Guest

Quoting: Klaas

Quoting: CyborgZetaNothing wrong with supporting the people who make, or package, the software I use.

I wonder how they prevent someone from lazily throwing a package together and selling it on flathub while the original authors get nothing. Similar to what happens on Steam, e.g. https://www.gamingonlinux.com/2022/10/someone-released-the-foss-rts-0-ad-on-steam-without-speaking-to-the-developers/.

Nothing, anyone can sell free software.

Contrariwise, though, if it's FLOSS there's also nothing to stop anyone from putting it up for free. Charging money for Free Software is perfectly fine, but somewhat self-limiting.

Personally, I think that vis-a-vis Free Software this is not an important move--it might help a few projects get some contributions, and that's fine, but it's not going to damage the open source model; anyone who thinks it could is not remembering their open source history.

What interests me more is the closed source, commercial software side. This could set up Flathub as a distribution source for Linux software from closed source software publishers. That could be good for Linux. Not in the short term good for open source, but I think that on Linux, open source software tends in the long term to win out (less so in games, which have various weird characteristics such as being ephemeral and art-heavy). What one could see is an increased ecosystem of closed source (mostly non-game) Linux software distributed via Flathub, complementing the ecosystem of mostly closed source game software from Steam, and combining to make Linux a viable desktop for more and more use cases. But the open source software continues to be distributed on it mostly via more traditional distro repositories and package management. As the Linux desktop grows, with more user and developer base I would expect more and more of the open source alternatives to these Flathub-distributed closed source applications would reach critical mass and gradually supplant the closed offerings.

So. Summing up, I think:
--Short term, for open source, some people get a bit of money to help development, impact minor
--Short term, for closed source, could help distribution of closed software to the Linux desktop, creating more interest in providing same
--Medium term, could help grow Linux desktop but somewhat crowd out open source software on Linux
--Longer term, I'd expect on a bigger-share Linux desktop, open source alternatives would grow again and displace the closed offerings. In the vague, nebulous future a good thing.

View this comment - View article - View full comments

Quoting: sarmad

Quoting: Purple Library GuyThe only way Snaps become dominant is if Ubuntu grows and displaces the use of most other distros.

This is not necessarily true. Snaps are independent from Ubuntu.

No they are not.

Quoting: sarmadIf you go to the snap store you'll see no Ubuntu branding there. Canonical is marketing snap as a Linux store, not as an Ubuntu store, much like Steam is a Linux gaming store regardless of which distro you're using.

That is an amazingly irrelevant argument. How Canonical markets a thing has no bearing on whether Canonical controls a thing. Except in the sense that you rarely market a thing at all unless you control it.

Quoting: sarmadSo, if people have no problem installing Steam, which is proprietary and tightly controlled by Valve, why would they have problem installing snap, which at least has an open source client?

First, I don't think people should have a problem installing Snaps. Canonical tightly controlling Snaps is a reason for other distros not to adopt them, but it's not in any major way a reason users should have a problem with them. If you use Ubuntu, use Snaps, there's no problem with it. Just don't expect they're going to spread much beyond Ubuntu, because they won't.
As to Steam . . . well, if there were an open source store controlled by a non-profit foundation on which I could find all the games and so on and so forth, I'm sure I, and many Linux distributions, would use it. But there isn't.

View this comment - View article - View full comments

Quoting: pleasereadthemanual

Quoting: Purple Library GuyIs it really a "competitor option"? I don't think Flathub is maintained by some specific other distro company. My understanding of Flatpaks is that they are just an independent piece of software, like LibreOffice or something. If Canonical were to start considering the whole open source ecosystem to be "competition" they would have problems.

On the other hand, for every other distro, Snaps really are a "competitor option" since they're created and maintained by not just a particular distro, but by a for-profit corporation, and Snap as currently written works with just one repository, Canonical's.

I think there is a disconnect between your two arguments here.

I don't think there is. At most, I think you're greatly overstating it.

Quoting: pleasereadthemanualLibreoffice is shepherded by Collabora; they sell support/development services for it, so I don't think that's a great example. But also, businesses can sell services for free software (and sell free software on its own as Ardour does, though this is rarer), so doesn't it make sense that anybody who offers a service which competes for their market is a competitor, regardless of whether they're earning money from it? I mean, Amazon lost money for years so they could service more of the market than competing companies.

And businesses can also sell competing services for the same software. Mumble is a great example of this. You'll see a lot of businesses offering Mumble servers.

All that is all very fine, but the motivation you're sketching is for most software pretty vague and attenuated. You could say that for any distro, or at least any distro with a for-profit company at its centre, they have a motivation to never use any software they didn't write in-house. But for most things it's a very small motivation compared to the motivation of not wanting to write the damn software.

(As to LibreOffice, I didn't remember it being run by a distro, and I just looked it up, and it is not. They've got this "The Document Foundation" thing which as far as I can tell isn't even associated with any distro in particular. I dunno, maybe Collabora contribute more code than others, but that is not the same thing. And let's not forget that Libreoffice is Libreoffice precisely because people got sick of OpenOffice being tightly controlled by a corporation. So my example was fine, even excellent)

But packaging is infrastructure at the core of what makes a distro, a distro. So, I'm running a distribution. I have a choice: A package system which is not controlled by any particular other distro, and which has a significant "reference" hub for packages, but not an exclusive one--I could do my own packages, host a hub of my own and so on. Or, a package system which is controlled by a large distro which is fairly explicitly a competitor for my "market share", and which does not allow me to make or host any of my own packages unless I rewrite it first--I would have to use my competitor's hub and go with their choice of what should be available. Who on earth would make the second choice?

Well, apparently nobody because I don't know of any non-Ubuntu distro that has adopted Snaps, not even distros that are derivatives of Ubuntu. So I'd say there is some empirical confirmation of my theory.

Now, again, I don't think this is actually some kind of mistake on Canonical's part. I don't think they're doing Snaps for the purpose of having anyone outside Ubuntu + "flavours" use them. They're doing Snaps so they can control the user experience, and they would totally understand if other distros, also wanting some control over the user experience, wouldn't want to give that control to Canonical. But I think if other people believe there's a real chance of Snaps spreading outside Ubuntu, they are mistaken.

View this comment - View article - View full comments

Quoting: sarmad

Quoting: whizseI can't say I have a stake in this, but it's interesting to ponder that Ubuntu seems to have a habit of betting on the wrong horse, Mir and upstart, for example.

Except that this time around they are betting on the right horse. snap is more mature than flatpak (gives the user more granular control over permissions, can be used for cli apps, can be used for non-sandboxed apps, can be used with system services, etc), snapcraft is more mature than flathub (it takes me 10 minutes to update flatpak vs the less than a minute snap takes to update), and snap has much wider adoption from 3rd party app vendors.

If they are betting on a horse at all, then I think it's the wrong one, not because of anything about the technology itself, but precisely because it is theirs. Their control, as I said in another comment, is too tight for it to get more widely adopted by other distros. And it seems they're quite open and clear about the importance to them of close control over the technology and packages built by it, as noted in the long developer quote posted by dziadulewicz. The only way Snaps become dominant is if Ubuntu grows and displaces the use of most other distros. And while there was a time when that seemed like it might happen, I feel that time has passed.

But I'm not convinced Canonical are "betting on a horse" at all. Apparently, they want a technology they can tightly control so they can release tightly controlled packages that will allow them to maintain very high stability (again, as quoted in dziadulewicz's post on p 3 of this discussion). As long as they get to use it to do that, I'm not sure Canonical care if Snaps are adopted anywhere else. It's fair enough, although I don't entirely approve of the attitude, which might be OK for Ubuntu but seems like it will promote fragmentation of Linux and not strengthen the overall platform in the long term.

View this comment - View article - View full comments

Quoting: IzaicMy concern is that this tactic may not work again because now it's public.

Might have become known to the cheaters whether they went public with what they did or not. But it might be difficult to avoid, too--sounds like it works based off one of the fundamental "cheaty" things to do, and not even the mechanics of it but just the raw ability to see things you're not supposed to. So it doesn't matter what method cheaters are using, if they can see things others can't and you put some stuff others can't see, they'll react to it differently unless they spend so much time second-guessing everything they see that they'd be better off not cheating . . .

View this comment - View article - View full comments

Quoting: PhiladelphusI see it's not mentioned in the feature notes here, but the Dev Diaries they've basically confirmed that cloaking will, in fact, let you send cloaked ships into empires with closed borders, so that sounds fun. :happy: I mean, on top of all the pre-FTL interaction stuff which is what I'm really here for.

I don't think I'd use that for spying very much--when it comes to wars I just rely on big fleets crushing stuff. But I'd love it for those times when I've got some kind of research project chain, like finding all those critters, and it turns out they're in places that have closed their borders to me and I'm thinking "OK, should I start a war just to get my hands on the Elusive Zoolantha or whatever?" A nice cloaked science ship would be just the ticket.

View this comment - View article - View full comments

Quoting: sprocketThe one argument I have for Snap over Flatpak is that Snap works for CLI software and system software. As of now, Flatpak is only Userland and GUI-based software.

Um . . . I can kind of see how Flatpak might only work for userland software. And I can see how Flatpak itself might have only GUI-based installation software, and no CLI-based tool for installing Flatpaks. But I don't see how, technically, it would be possible for Flatpak to even be able to tell if the software being packaged was CLI software. I suppose it would feel weird and backwards to use a GUI to install CLI software, but I don't get how they could make that not work.

View this comment - View article - View full comments