http://m.slashdot.org/story/213469

I also think if you read the rest of what they're saying, it's probably is true that we will only get Linux GUIs which are competitive with Windows or Mac when people are willing to fund their professional development. OSS works exceptionally well for lower level components, or for components targeted at developers, because everyone involved benefits. For a professional developer a contribution acts as a good mechanism to raise your profile within the community and improve your ability to earn a decent salary elsewhere. For a company, running open source projects means they can solicit contributions and improve their profile; they contribute to other projects because it gives them a base to be able to build their products. We don't expect or find that OSS developers of high profile projects are starving in garrets, they are in fact leaders of their communities, highly prized and in-demand.

But these mechanisms do not have the same force when the consumer of the software is a general user, not a developer in the same field. i.e. when you're writing software in C++ which is then going to be used by graphic designers or artists, or by a casual user. In those case, the direct reputational benefits just aren't there, and that's why those sort of projects often struggle to take off. And on top of that those projects also involve a lot of painstaking and arguably boring work to get everything polished to a fine sheen. Krita for instance is excellent, but it has found a way to fund development through yearly crowdfunding. I think the same applies to Linux GUI's, that ultimately you need some pot of money able to support developers, which moves around according to the desires of informed consumers. In my opinion, that's what it will take for desktop Linux to become a serious mass market alternative to Mac and Windows.

It's not rational for me as a consumer to fund such an unsure bet, especially when changing an operating system has lots of friction. I assume that the proposed talent of an individual investor is in the ability to sieve good from bad in a sea of maybe. But I also don't understand the market well enough to see how such an unproven bet might succeed in a very stale market dominated by arguably one entity (you can't really buy macOS) -- maybe two.

I'd also say that while software work is expensive and requires talented labor, and while operating system work takes an intense amount of effort, there's something to be said about the fact that Elementary OS could even build their OS to begin with. How did they do it? On the backs of billions of dollars of free labor. I can't imagine how it would be if the whole world hid their source code with proprietary-only software, and Elementary had to start from ground 0.

What rational model governs this situation?

You mention that investors have lots of money and are therefore able to take on risks. I suspect that nearly everyone who would consider downloading this operating system has a lot of money relative to the $5 suggested donation, and certainly relative to the $1 you could choose instead. And if you truly don't have that, then you can do $0 instead.

You can make the risk to yourself arbitrarily low, selecting whatever you think fits your profile of cost vs potential reward. That is their rational model.

http://blog.elementary.io/post/110645528530/payments

But this time I regret every cent. Coming from Mac OS, or even Fedora, Elementary seems unpolished and unstable.

As it is, my Macbook Pro Retina has been crashing a lot and doing weird things frequently since i upgraded to Sierra. Between that and the latest Macbook Pro notebooks, i just don't like Apple anymore. I'm pretty set on buying a highend laptop _(similar build quality to Apple)_ and putting Linux on it... but that's where the fun stops. Linux UIs tend to lag behind significantly.

I want a for-profit UI company working on a Linux UI. OSS just can't compete with Apple on design it seems.. and in my opinion, users shouldn't have to choose between OSS and pretty design.

It's bad enough that if/when i leave OSX, i lose a lot of my apps due to them not supporting (or not supporting competently) Linux, but losing apps and UX.. well, it's a tough pill to swallow.

I think you're massively underestimating the magnitude and difficulty of what you're seeking. elementary has been at this for 5 years and made an immense amount of progress, though. But it's definitely still not at the level of Apple, which has had decades and billions of dollars poured into this.

It's not a "paid OS" in that it doesn't force you to pay a particular dollar amount, though. (People tend to lose their minds when you charge them for open-source software.)

https://elementary.io/docs/human-interface-guidelines

f/d: I was on the elementary core team

People complain about iOS' race to the bottom, but in the OSS community, I frel like this has been there for a long time. "You want to _charge me for it_ ?! Pffft!"

We may complain about macOS now with Sierra but for thr most part, macOS has been stable for a long time. I haven't seen a kernel panic in probably a decade. Not saying it's perfect, but there's a point you hit when people are paid to make things happen and macOS hit that point yeats ago.

I would argue Elementary provides a lot of value and should therefore weed out the complainers and worst of the bunch by charging. I would remove the 'pay what you want' and just charge $30 or $40.

We don't live in the Star Trek universe yet, so people still need money to live. I fully believe in and support companies, individuals, and groups that provide a good value for what they charge.

I _want_ those people making great things and doing it full time. Otherwise, you just get abandonware or crappy products where people don't fix bugs because they have a day job to deal with. That's the reality in my mind.

I can throw $XX at a linux distro, pray that it goes to the right things and gets improved proper, and then wait until it becomes good, or I could throw $XX at a closed source solution, and have something that works good right now.

Its all fine to donate $XX, but buying for $XX to get something significantly worse than its competitors is a bit weird, and can only really be justified through "I support free software even if it has binary blobs", which is still a bit weird

Allows students, poor etc, still manages to weed out the most entitled ones it seems ;-)

Windows almost has me sold with their new Ubuntu subsystem.

I've been orphaned from several platforms. OSX surprised me. It appeared to be the dream commodity unix. But then a release broke my workspace usage.

Eventually, I completely switched to a stripped-down unix. I use a bash menu for controlling things like wireless and screen lock. Tmux/dwm are my window managers, I use whatever the latest browser is.

All this required effort to learn and set up. People who look to the big companies for fashion would sneer at this. But it's a sharp tool, and it'll never be made obsolete by shifting trends.

[1] https://www.darlinghq.org/project-status/

Anything in particular you are missing?

I use Docker for Windows.

It works tolerably well, but the hyper-v thing that it runs on is currently my #1 suspect when Windows takes > 30 seconds to connect to any network :-/

> Is that not what Red Hat is? I thought that's what the Red Hat license provided?

I won't say anything bad about Red Hat, but calling the UX awesome isn't something I would do based on my experience. Then again, they might have changed but I haven't heard anybody mentioning it.

(This used to be Canonicals niche before they picked up the assumption that Mac-like == Good ; )

I'll agree most UX on Linux isn't great. But GUI is such a pain in the ass that most people really don't want to work on it. Plus, plenty of Linux users aren't casual users and will accept or prefer function over form.

I myself am a heavy terminal user and pretty much limit my GUI use to the occasional image editor and the browser.

RedHat makes their money on support licensing for the server. Employees are paid to work on Fedora, but it is not the RedHat product.

Which is really too bad. It's the only RPM distro I've seen that's reasonably well put together.

Edit: this -

> Which is really too bad. It's the only RPM distro I've seen that's reasonably well put together.

Anything specific I should be aware of wrt OpenSUSE?

It is definitely a case of "to each their own", but the only issue with KDE right now is developer churn. As the writers and more experienced developers in some applications reduce their participation due to time constraints, it is incredibly hard for anyone to step into those shoes in often millions of LOCs code bases. A lot of UI rot isn't because we don't have technologies (Kirigami, Qt Quick Controls 2, Plasmoids, etc) that enable fantastic UI and UX experiences, it is because porting forward hundreds of applications comes down to who has the time and willpower to either continue maintaining their 10 year old project or someone having the willpower to learn what can often be very ugly C++ or ancient PyQt codebases.

I also have no problem with what Gnome is doing. Their applications may not be for me, but I can recognize the beauty in what they are trying to accomplish, and still think they have a much better design language than anything MS / Apple is putting out. Whenever I sit down at OSX / Windows I feel like both are stuck in the 90s, often because half their applications (especially system management) were written then and never updated since, and because they just make random UI splits every release to look shiny and new by just changing the shell look and keeping the rest jarringly legacy.

GNOME sits on the other side of the spectrum on these issues but I personally find it and in fact most GTK+ apps more aesthetically pleasing and more well organized even if all the buttons are too huge and its white space is like a football field.

The problem is that many people have differing ideas of what the UI/UX would look like.

If you asked me to provide such a thing I'd just use Debian. Others might prefer RedHat, or similar. But copying an existing distribution, and keeping up to date with new versions would be a full-time job in itself, and that wouldn't leave any room for actual development.

The GNOME project, and KDE project for that matter, have spent years creating a unified environment, and even with their funds and developers the project is never-ending.

I wouldn't say a minimal & unified distribution is impossible, but it would require a lot more users to pay, up-front.

or, you can get the Dell XPS Developer Edition with Ubuntu pre-installed.

Good luck with that. The problem is that no one will make a UI/UX that's both awesome and stable. It's not like the resources aren't there: there's a ton of DEs for Linux, and they've been working on them for 20 years now. But the problem is they can't ever agree on anything, and they can't ever just leave well enough alone: every time they get something stable, they abandon it and make up something completely new and unstable. It's easily the 2nd biggest hindrance to Linux-on-the-desktop adoption (the biggest of course being application compatibility/inertia with Windows).

>but that's where the fun stops. Linux UIs tend to lag behind significantly.

That's the Linux community's fault. It's not the features that lag, either; KDE, feature-wise, has always been far ahead of the commercial UIs. For one simple example, I can't middle-click and vertically maximize a window on any other DE that I know of, but in KDE that's been a standard feature for at least 15 years. For another simple example, multiple desktops have been standard on Unix UIs for over 30 years, but never in Windows or Mac. The problem with Linux UIs isn't design, it's stability: everyone's constantly revamping stuff (Gnome3, KDE4, KDE5, etc.) and never spending any time fixing bugs. For instance, I have the Dvorak keyboard set on my LM17.3/KDE desktop, but several KDE apps don't respect that and go back to Qwerty. (To be fair, Windows is even worse: I have my work Win7 desktop set so I can switch between Qwerty and Dvorak, and the stupid thing is constantly switching back and forth randomly.)

But to your point of bug fixing, I absolutely agree. macOS is plagued by this as well, but mot to the degree Linux DE's are.

I've used multiple desktops since the 90's and loved it. macOS has had support for a while, but the animation switching from desktop to desktop was horrendous. I got a utility to fix that, but it's a joke they don't just include a checkbox to toggle the animations. This is one huge UX fail in my mind.

I really wonder sometimes why distros don't use it more, and make their own custom themes for it, instead of just leaving it at the default. It wouldn't take that much work to make themes to make it look almost like anything you want. Or you could even make a different version of Plasma which emulates some other OS/DE if you needed to, leaving all the underpinnings intact.

Talk more than others about usability - happily gloss over all kinds of weirdness[0] in your expensive setups.[1]

Happily pays quite a lot for your Macs - "regret every cent" after donating peanuts for the development of a Linux distro.

Just saying.

Edit: downvotes welcome! I cannot buy anything for my Internet points here in Western Europe.

A textual description of my wrongdoings would be even more awesome!

Edit2: you don't need to upvote me! Instead just tell me what I did wrong!

[0]: Yes, there is. If you cannot spot it I won't recommend you for any lead UX role.

[1]: Not saying it is bad or not worth the price though. Just expensive (given that you have to have a Mac to run it legally) and far from perfect.

I hackintosh so I'm less well-placed to answer your second point, but I'll try: A Mac is expensive, but you know exactly what you're getting and you know it will work well (recent dongle madness notwithstanding).

Meanwhile, eOS markets itself (previously implicitly and now apparently explicitly) as a potential replacement for OSX. But it's nowhere near. It's an outdated version of Ubuntu with some basic custom apps on top (a calculator, another basic wrapper around webkit, a slightly cleaned up fork of a file browser, etc).

And then you stray to the third party apps, and as always with desktop Linux, they're awful. Everything uses a slightly different toolkit, the UI elements are ugly, UX seems to be devs' lowest priority item, and of course you can't completely interact with industry standard formats (DOCX, PSD). The Intel Mesa driver still tears and/or stutters (actually all drivers do except Nvidia's) so watching videos is an awful experience.

Most of those items above are not fatal by themselves, but together they make for a miserable experience. I used desktop Linux (mainly Ubuntu) for over a year since I needed to do some dev and couldn't run OSX at the time. I won't try again unless there's a revolution in the desktop FOSS world. (Good video drivers, standard well-designed toolkit, UX as a priority, deeper integration of things like WINE.)

I do have some family members who use ancient core2duo laptops and do absolutely everything in Chrome. I pop Ubuntu MATE on their machines so they can avoid malware. I also always keep a live USB stick handy for flexible troubleshooting. As far as I'm concerned, this is basically the limit of usefulness for desktop Linux.

I understand some people value the flexibility above everything else and thrive under desktop Linux. I'm not trying to devalue their views or say that their setup is wrong. Use whatever works for you- but if you haven't gave OSX a serious shot yet, do try it.

I don't think it's fair to compare OS X (with runs on a very limited set of hardware) with a Linux distribution running on a random not-well-supported system. If you want an apples-to-apples comparison I'd suggest to test Ubuntu on one of the officially supported systems - see https://certification.ubuntu.com/desktop/ for a list.

But you are right, some people mix these different issues up. My next laptop will come with Linux pre-installed - so I expect to report that all the bits work correctly (except Linux will be running on top of Qubes rather than on the metal, so...)

"How much do you want to pay?"

"0 Dollars."

"Great!"

Why is that so horrible?

Really great line, i liked that a lot.

"We want users to understand that they’re pretty much cheating the system when they choose not to pay for software. We didn’t exclude a $0 button to deceive you; we believe our software really is worth something."

That's why you let users decide after they've used the software. You don't tell someone that up front and expect a payment when you haven't been able to try and test the software for yourself and your team.

If it's worthy, then users will come back and offer a payment or contact you to find out how they can contribute if they feel its really worth it.

I think a lot of people have this misconception that everybody who uses open source software just assumes it should be free. In my own experience, it's quite the opposite. Most people who use open source are actually more willing to pay for something they feel is valuable to them without hesitation. They're also usually willing to pay a little more then perhaps is necessary in order to make sure its maintained and supported.

So apparently in the bait-and-switch mentality of the mac world, that you can pay '0 dollars' is not obvious.

[1]: http://blog.elementary.io/post/110645528530/payments

[2]: https://github.com/elementary/website/pull/655

It is a shame for Elementary OS Team, because this issue was many times discussed on GH issues / IRC about implementing just a "$0" button. As I remember in launchpad there is also a task or discussion about suggestion of the dialog which will remind about donation after certain time duration.

Edit:

Found https://blueprints.launchpad.net/elementaryos/+spec/donation...

An "I would like to try it first" button.

They should include more videos and screenshots so people have a better idea of what they are buying.

If you were to email Elementary and ask for support with a non-working 'sg' kernel module, I would be highly surprised if they offered you any support. More likely, they would tell you to contact one of the authors, who did not receive one dime for our work. He takes your money, and shirks off supporting you to people that _didn't_ take your money.

edit: typo

By voluntarily contributing to an OSS-licensed project, you don't have the moral right to begrudge him for it. What they are doing is completely permitted by the license; RMS has spelled it out over and over. So, no sympathy from me.

I would prefer to pay a reasonable subscription fee with a support contract which is also not illegal, but benefits all parties.

I run a small software company myself. We're selling our software on the internet. But our website focuses on getting people to download the free trial first. The "asking for money" is then done once the trial period ends.

It's already hard to get people to download your stuff for free. With obviously visible download buttons.

If we did what Elementary is doing we would be closing our shop rather soon.

Yet some projects still squeak out some money: Synergy, DOSBox Turbo, etc.

Freedom 0: "The freedom to run the program as you wish, for any purpose."

Does this line make a program un-GPL?

    exit(0) if Time.now > Time.new(2016,11,15,0,0,0)

_that's why it won't work_

From what I can remember, there was never a time where they disallowed a way of downloading the OS for free, they just made it more difficult to figure out over time.

[0] https://www.patreon.com/elementary

https://mpra.ub.uni-muenchen.de/53376/

First of all, it's a rather large download, so you're probably going to close the website as soon as you click on the download button. Second of all, to consume the product in this case, you have to install it. And yet again, after you do so, if you get greeted with the "donate" window right away, you're going to dismiss it because you still haven't tested out the product.

So, what you need is a timer, something that says like "after 30 minutes (or an hour, or two hours) of usage, display the window", but that would require the OS to make calls to external servers without the user consent, and, because we are talking about a Linux-based operating system, people would be complaining that this is a a privacy violation.

Not necessarily. It's not like you expect the buyer to do a careful analysis of what it's worth before they pay. You can usually decide this sort of thing based on very quick impressions. Also, there's the aspect of it just seeming nicer to ask after the download has happened. Then it's clear that you already have the product, now it's up to you to decide what to pay for it, if anything at all. Here people have been tricked into thinking they had to pay and begrudge the author.

Really? On most sites that I visit, I 'Save As' every link that I care to download, so that there is no separate tab to close. (But maybe you were referring only to this specific site.) Maybe the solution is just to make sure that there's more of interest on your download page than just the download button?

Stare into the abyss etc.

If I pay nothing, I don't pay.

I don't say "I paid zero." or "I paid nothing" unless someone asks the specific question "How much did you pay for that?"

It's obviously a way to nudge people into paying something.

edit: I made an aside about the use of the word 'donation' by cannabis collectives in the U.S. to denote 'payment' (as a way to operate as non-profits and avoid federal problems) as a reference to similarly weak wording used in a nefarious way, but I thought the relationship to the pattern here was weak, so I removed it.

They do serve a purpose though.

No thank you, I'd rather use Mint or Ubuntu if I want an easy distro.

Comparing that to GNOME, where everything is so damn big that it wastes half of my screen and KDE where the default apps are usually bloated and confusing and can do way too many things, I call Pantheon as a win for the developers.

EDIT: Also, one neat trick that I've never seen any other terminal emulator doing other than Pantheon Terminal, is that you get a desktop notification every time you're not focused on the terminal when it finishes its job. For example, let's say I do something like "sudo apt install netbeans". Since the download is large, I'm going to focus on something else in the meantime and I'll get a desktop notification when the installation is finished.

[1] https://launchpad.net/elementary

> Not now, take me to the download ›

Link.