The Road to Luna | Hacker News Reader

The Road to Luna

elementaryos.org

68 pointsmacco68 comments

As much as it brightens my day to see a project I worked on (I'm actually the guy who filmed that video in the article) make it to HN's front page, it also makes me sad a bit.

It makes me sad because of the usual troupe of folks who complain that we don't have the right to go around calling ourselves an "OS", we don't bring anything original or innovative into the world, or that we are plainly just OS X clones.

I was extremely active on the thread[0] posted to HN when Luna launched (which was at #1 for the better part of that evening), but I don't think I'm going to do that same level of in depth responding here.

Instead, I'm simply going to leave you a link to the below article[0]. Please read my comment if you want to understand how we as elementary evaluate and understand our design process and the criticism which seems to come with it.

Thanks for listening.

[0] https://news.ycombinator.com/item?id=6193148

I can't imagine that people making those silly comments have 1) RTFA, or 2) done anything remotely like what you guys are doing. You guys are doing some amazing things with the Linux desktop, don't let the peanut gallery get you down.

PS - I fervently hate the word "exciting" and think it is by far the most overused and diluted word in contemporary US English these days, at least in the business and startup realm. Too many people seem to think that calling something "exciting" actually makes it so, and call everything they want investment or traction for "exciting". Almost none of it actually is.

eOS is one of the few exceptions. The first project I've seen where reading your blog posts about it feels like reading folklore.org, and for a Linux fan at least, truly is exciting. So please keep doing what you're doing, ignore the negative nancies.

I have yet to test it, but having seen the video, "haters" probably lack knowledge about desktops, user interface or simply history (how many desktop systems have been made that lacked the integration, speed and polish ?). Hoping you don't take it too hard and keep on experimenting and growing.

You should simply be proud, to have this endurance, to bring Luna to the masses.

I don't agree on everything you did, but I really appreciate what you did.

And you design consideration about how applications should interact with the users are ground breaking - at least a bit.

Sending minimize to hell is a smart move.

Kudos

Be proud. I have been showing Luna around and it has been met with nearly universal praise. One of my friends switched his Ubuntu desktop within 24 hours and won't stop going on about how amazed he is how everything just works, how clean it is, etc. I'd also Luna can absolutely be called an OS, using the current vernacular.

I think the nature of innovation is to have heavy criticism. You guys are doing an amazing job. You have reached the level of design apple has, and now the only way is forward.

I expect great things from you as a team. You guys have brought back the passion to linux and you have pulled me back to linux from Mac and Windows.

You guys have an amazing product. Keep moving forward. Keep being hungry and bring that passion back to the open source community.

Don't be sad. Be proud of what you guys have accomplished so far.

I've never spent a few days thinking about a Linux distro except for ElementaryOS: Luna. I spent 3 days including burning multiple CD of the 64-bit image (I don't know why but it seems to me that your 32-bit image works better) and finally got mine setup a few hours ago today.

I've imported my pictures. Installed a few other software (VLC, Chrome) and even thinking of learning Vala! (got MonoDevelop with Vala plugin installed). It's very unusual of me to jump into an obscure language (mind you I do Java for work and I'm spoiled with the tools, the best practices, the mature 3rd-party libraries).

Mind to share some tips of Vala development (project structures, IDE/editor, unit-testing, continuous-integration, etc)?

PS: I might be the minority but I never like launchpad and secretly wish you guys move out of it. Their UI is very confusing and in most cases I can't even see a screenshot for some of the projects that link back to Launchpad.

A lot of people here fervently complaining about eOS again.

The most important thing is that there's finally a linux distro with a team behind it that really cares about design. I think it's a huge step forward, and when eOS gains in popularity, I hope other teams follow in their footsteps.

No, distros like Linux Mint aren't the right way forward. Stripping off everything until what you have left is a taskbar isn't exactly my idea of good design.

Completely agree. Their focus on design as all of how-it-looks, how-it-works, and performance-is-a-feature is very promising for a Linux Desktop, as are the hard work and sacrifices they're willing to go to get that - rewriting everything in Vala and GTK3 namely.

Their app-data-sharing component, Contractor, is also interesting. I've had at least one idea I wanted to try in the past but couldn't figure out how to implement it b/c apps are not aware of other running apps. I wonder if Contractor will make that possible.

These guys have created a really strong base to build upon, and appear to be innovative problem solvers. They're also tuned into the problems of power users, referencing the whole Gnome2/3 brouhaha. Really looking forward to seeing how this develops, and to it getting to a point I can completely replace Ubuntu desktop with it.

I'm glad to see that the top comment is a positive one. I haven't personally used eOS, but I'm also glad to see they are focused on usability and design. I switched from a decade of Linux to OSX a few years ago and have been pretty happy with my choice - OSX is unix-y enough, and tends to just get out of the way when necessary (some of Apple's more recent Gatekeeper changes withstanding).

In the meantime, I'm disappointed that Hacker News commenters continue to bring out constant negativity against new projects. What happened to original HN ethos that celebrated people _doing things_? Seriously: if you feel like you have to complain about somebody's hard work, try being constructive and giving feedback, or not commenting and just doing it better instead.

"Cares about design" implies that there is one optimal design for everything (perhaps Apple's). But a lot of design is very subjective.

Actually there are multiple distros which care about design, but they have different ideas about what to try to optimize for.

They don't care about design. If they did they would have created something original. Or at least they would have more than one source of inspiration to pick from. If they wanted to take the whole shape of OSX applications, at least they should have picked a different color theme. Here they stole the functionalities, the shapes, the colors and the metal textures that Apple's designers chose. eOS is designed by Apple, there's no one who care about design in the eOS team, at least not for the apps shown in the video. They're receiving bad critics because thieves receive bad critics. Hopefully they'll learn and their next version will have its own identity.

Apple doesn't care about design. If they did they would have created something original. Here they stole their entire windowing interface from Xerox and their kernel from FreeBSD and Mach. OS X is designed by other people, Apple only cares about metallic textures. They're receiving bad critics because ... oh wait, they're not b/c Steve's said "great artists steal" and made it all ok.

What's good for the goose is good for the gander.

Why do you guys make a big deal of Contractor in the article when actual development of the service is rather neglected looking at your milestones for your next release's beta[0] and Contractor's revision history[1]?

On your mailing list you guys are already getting ready to update your Ubuntu Precise base to a more recent release[2], but why? Precise is stable and will be continue to be so long enough for you guys to spend time developing the developer-facing treasures of your platform, and yet it seems you're a little too focused on user-facing niceties. Or maybe the better question is when will you guys sit down and start making the stuff that'll entice third-party developers?

Let me know if I'm getting things confused, I'm actually a big fan of the project.

[0]: https://launchpad.net/elementaryos/+milestone/isis-beta1

[1]: http://bazaar.launchpad.net/~contractor-dev/contractor/trunk...

[2]: https://lists.launchpad.net/elementary-dev-community/msg0257...

Hi, I'm one of the devs.

Regarding Contractor, its big proponent is on vacations and we still need to move some of the blueprints about it to the appropriated milestones. In fact, we still have to report many bugs about it, but in time I'm sure it'll be done.

Regarding the new release, we get a lot of criticism from all kinds of users saying they want us to be based on a more modern version of Ubuntu and as so, we plan to be based on 14.04 for next release. We aim to release shortly after Ubuntu 14.04 comes out and that's the plan we're sticking to. I'm sure 14.04 will be just as stable, we can't be stuck in the past and moving is a lot of work - we need to start working on it as soon as possible!

Besides, we also need many new GTK+, Vala and other libraries' features.

I like the idea of a new competitor in this space, however I was a bit disappointed in their direction. Not sure of others but I don't need another gnome3-like (dumbed-down, grandma-friendly, immature paradigm, rather be a tablet) interface. These linux-based desktops will never be popular; that ship sailed long ago.

However, I desperately need a professional, stable, feature-complete, mostly bug-free workstation GUI with modern capabilities at least as good as Win2000/XP, Irix, NeXT like I had 10+ years ago.

Something like Xfce polished to OSX-level quality.

If one of the design-goals is that "menus are too difficult" there is no market available for that product. The people that would benefit use Windows, OSX, iOS, and Android already.

It seems like KDE already handles your needs: it's well-integrated, stable and has all the modern capabilities you could want. Certainly far better than XP! I've used KDE a fair bit, and the recent versions have been superb. Sure, it's heavyweight compared to some of its bare-bones competitors, but that's the only way to fulfill all your conditions.

Of course, the truly "professional" choice would be XMonad layered over something minimal. But that's a different discussion for a different time :P.

It would seem so. Unfortunately, I've given it a try every other year and have been disgusted every time. The technology is nice, but it feels like it was designed by 90's IT guy. Clutter and programmer-art is not what I am looking for either.

KDE at least the versions I have used have a bunch of little annoyances that just break usage for me.

Mostly having to do with things like KWallet integration or other little pieces not working properly. I've used workstations that still run Windows 95/2000 for machines and they work great but KDE 4 leaves me missing its 3.x variant.

XFCE could probably be a viable replacement though, it is familiar and easy to use it just has its little gotchas as well (which are more annoying on a laptop than a desktop though).

>there is no market available for that product. The people that would benefit use Windows, OSX, iOS, and Android already.

There's a huge market for that product - everyone already using Windows, OSX, iOS, and Android. Talk about validation (minus Windows at least, where it's not clear what portion use it b/c they have to vs want to).

I remember at the time of the Vista launch spending a quiet afternoon going in to donate blood at the Red Cross. They had quite some more donors than they had expected, so there was lots of time to wait. While I waited, I discussed Vista with someone who was clearly a Windows enthusiast. It was also fairly clear that they weren't much of a technologist. I sort of spent the time trying to find the words to encourage him to give it a little time. I was and still am using XP. I have Linux in some VMs. I can see transitioning, but haven't bitten the bullet.

More succinctly, I wonder if you are projecting your doubts a little bit.

>More succinctly, I wonder if you are projecting your doubts a little bit.

Not sure I understand you correctly. To clarify, what I'm saying is, any OS that people use because they want to is an opportunity to do it better and sway them. They may be personally invested, but they're invested because it's a good product, so if you can one-up that you have a chance of winning them over, and it's actually possible they could swap to your OS.

If they're using an OS b/c corporate IT mandates it, fat chance they'll ever reinstall their work computer's desktop.

I just don't have any good guess as to what portion of Windows users are the former or latter.

Not sure about that, but we know approximately how many people have changed to a Linux desktop, and it has always been near 1%.

And that's all it may ever be, and that's fine as long as the reason isn't b/c Linux is self-sabotaging by not having a highly polished, competitive desktop experience.

It's getting close, Ubuntu is arguably there, but obviously alternatives are needed since not everyone likes Unity. Kudos to eOS for shaping up to be another viable contender.

Ok, a huge market that has 0% interest in a new Linux desktop.

IMHO there is a huge market for replacing menus with somthing "simpler" and faster, look at the "ribbon" in MS Office.

ps: not trolling but I gave to my grandma a linux laptop (ubuntu and unity) and I think it's always great to have alternatives

There is no market for desktop linux. The only people there in substantial quantities are technical workers, where menus have worked just fine for thirty years.

sickening to see the negativity here, I just bought a copy because this thing looks great and because I want to support the team.

Even though the comparisions to OS X seem inevitable due to the launch bar, I like the snappiness of the OS, and one of the few distros I felt playing around. I am a primary windows user, who tinkers and plays around with Linux and OS X, I definitely like the OS. My wireless drivers being proprietary could not work with Luna, so as much as I want to have Luna - I could not use on the old laptop, but I will definitely consider installing it on my VM.

Crunchbang and Luna are the distros I am liking.

It's a really cool OS. The only thing I don't like is that they force their own applications. I always install Firefox and Thunderbird, and then it's a real thing.

So installing something by default is forcing it on people now? It's still Linux.

> So installing something by default is forcing it on people now?

Only for Microsoft.

Elementary OS looks like a OS X knock off. Which is a shame, in my opinion. They say they reached the same conclusion as Apple's designers, but I think they just have a bias. They should be taking inspiration from apps like Fantastical, Soulver, etc. and building in those concepts to the O.S. They should try and set themselves apart from OS X. It just looks too similar. Remember what Picasa said: "Good artist copy. Great artist steal." Elementary OS is just copying. They should be stealing. I'd love to see something along the lines of this Dribbble shot, honestly.

http://dribbble.com/shots/576250-Windows-UI-Concept/attachme...

While I agree with your premise that Elementary is nothing more than a repetitive clone of a well trodden trope, I could not disagree more with the "steal" aspect.

Instead, I would suggest that any further designs should drop the privileged terms of "intuitive" and other such rubbish[1] and instead focus on designing a system that specifically meets the needs of a very particular demographic / audience.

Finally, there is little to no evidence to suggest that Picasso said that infamous line. It was likely misattributed by Mr. Jobs from a biography.

It is instead a mangling of the words of T.S. Eliot[2] that seriously detracts from Eliot's original (and relevant to this discourse) intention.

[1] http://www.uigarden.net/english/easy-intuitive-and-metaphor [2] http://nancyprager.wordpress.com/2007/05/08/good-poets-borro...

For those saying Elementary OS isn't a blatant copy of OS X. Just look at those icons in the dock.

"Good artists copy, great artists steal." So by Steve Jobs (and Picasso's) own definition, the eOS guys are merely good, not yet great. They've still got some catching up to do.

https://www.youtube.com/watch?v=CW0DUg63lqU

That quote isn't actually Picasso's. It's T.S. Eliot's, and it goes, "Good poets borrow, great poets steal."

That's not quite what Eliot said either. From elsewhere in this thread, 'foolrush points to this:

http://nancyprager.wordpress.com/2007/05/08/good-poets-borro...

mere copying: i called theft.

I emailed support a question and never received a reply back. Using Geary, once deleted/archived, emails almost always come back. Sometimes, they come back after several attempts.

Hi, I'm on the elementary team.

I'm not sure how exactly you "emailed support" because we don't have a support email address, but I would encourage you to check out our support landing page here: http://elementaryos.org/support and specifically search/ask a question here: http://elementaryos.org/answers/+/all/solved/highest

If nobody seems to have a conclusive answer for you (mind you, Geary is not an app we wrote in house, so there's liable to be less Geary developers hanging around our support channels), I'd ask you (on behalf of the community) to report a bug here: https://bugs.launchpad.net/geary

Thanks :)

How can they name their mostly visual customizations of the existing projects "OS" and be accepted as serious? Such things were called skins and themes.

What is really original there apart from the visual design elements and the UI customizations to allow them calling their contribution an "OS"? Just that they actually write some code? Where are their own kernel and their own drivers? Aren't they just one more derivative GNU/Linux distro?

So by that characterization, virtally no GNU/Linux distro is actually worthy of being called an OS.

That's fine if you want to say that, but you have to understand what that actually means.

As for differences beyond the "visual", we wrote our entire desktop environment and a number of our apps from scratch. We did't just skin things and make opinionated decisions about design. We actually write code (apps, libraries, our DE), as this article expresses.

Full disclosure: I am an elementary team member and I helped edit the linked article.

By that characterization, virtally no GNU/Linux distro is actually worthy of being called an OS.

Exactly, it isn't. Fedora, Debian, Ubuntu, they are all distros. The real OSes are Unix, VMS, Plan 9, BeOS, Minix, Windows, GNU/Linux, BSD, OSX and Android.

Making some user space applications or libraries don't make you OS authors. And I don't understand why you guys are hurt when I say that. And why you are not open about what you're actually doing.

Maintaining a distro is not a small thing though, so I'd like to see how serious you are in that aspects. What are your upstreams? What are your release policies and schedules? Update schedules? Verifications? Quality assurance? How many people do you have in charge of all that?

I see you've had one release which you call yourself "Ubuntu 10.10 remaster," and now you prepare for second based on... what? I see you make your own patches of Nautilus. Once you accept you're a distribution you can be realistically compared to others.

"Ubuntu: The world's most popular free OS" http://www.ubuntu.com/

"Debian is a free operating system (OS) for your computer." http://www.debian.com/

It's like calling the specific desktop computer "a CPU." Some people do this too. It's just wrong. If I plug the pieces together in a workable machine it doesn't mean I've made a CPU, even if I've painted the box myself.

Note the name is Ubuntu, not "Ubuntu OS." Debian and not "Debian OS."

I wouldn't have problem if the product is just named "Elementary" and described as a Linux based OS distribution, once they have a release policy (before it's just a Linux based release). Having OS in the name means a new OS, and it simply and plainly isn't.

I'm not sure I agree.

I could be said that "Windows", "OSX", "GNU/Linux" are not really "Operating Systems", intended as software products that you can install on a machine.

Windows XP, OSX Mountain Lion, Ubuntu 13.04 and yes, Elementary, are.

You don't download a specific "GNU/Linux" OS, you download an OS (intended as a software product made up of kernel+file system+desktop environment) that is based on that technology. The word "distribution" is something that was created for Linux given its specific nature.

Yes, if we stick to the classic CS definition of OS maybe you are right, but I believe that Ubuntu, Debian and Elementary can rightfully call themselves "OS"s without issue.

Windows XP, OSX Mountain Lion, Ubuntu 13.04 are all specific releases, not the new operating systems.

Elementary isn't, too. It is not even a distro until it has clear release and update policies.

Otherwise you'd consider any skinned ISO of any existing OS a new OS, and that's just wrong. "Hello all, I'm making one new OS in my basement every week!"

From Wikipedia[1]:

> a distribution is most simply described as a particular assortment of applications installed on top of a set of libraries married with a version of the kernel, such that its "out-of-the-box" capabilities meet most of the needs of its particular end-user base.

It seems you're just making up random criteria that elementary OS doesn't meet so you can dismiss it for what it is: an operating system.

[1] http://en.wikipedia.org/wiki/Linux_distribution

It's just wrong.

This is the wrong thread to kick up a fuss about this. If these guys are using "OS" the same way that Ubuntu and Debian do, that's just fine for a new project. It also means that your complaint has nothing to do with this new project in particular and is thus off topic. This matters because threads about new projects are the heart of HN and deserve special care.

They aren't using it "the same way Debian and Ubuntu do." Debian is named Debian not "Debian OS." Ubuntu is named Ubuntu not "Ubuntu OS." The releases have own names.

Elementary "OS" guys use it like OSX and iOS which both conform to the real view: Note "OSX Lion" or just Lion and definitely not "Lion OS."

Clarity is important.

I suppose that you had better go down and sort out those blatant liars CentOS, ClearOS, OS/4 OpenLinux, SolusOS, PureOS, Chrome OS, Joli OS, Peppermint OS, ZevenOS and PCLinuxOS.

How dare they try to advertise themselves as operating systems.

They should keep 'OS' in the name if for no other reason than to piss off annoying pedants on the internet.

Who the fuck gives a shit?

This sort of "elementary is not an OS but a desktop environment based on the Ubuntu distribution based on the GNU/Linux kernel..." nonsense that is so prevalent in the Linux community is probably part of the reason why elementary exists in the first place.

"Maintaining a distro is not a small thing though, so I'd like to see how serious you are in that aspects. What are your upstreams? What are your release policies and schedules? Update schedules?"

And will we find a note on the Web site one day saying 'Luna is closed now, we are working on Selene (or whatever)'?

Jupiter worked well on my Netbook. Luna works OKish (bit heavy). The UI is refreshing, and I appreciate the work going into this. But as acqq says, a roadmap might be important for wider adoption.

Throwing around "OS" used to bug me too. Even Android runs on top of Linux. So, is Android really an OS? A distribution? It seems like something more than a distribution--Android contributes drivers and patches to the kernel. But it's still Linux under there.

I finally stopped caring and accepted that "OS" has evolved for most end users to mean a user experience, an ecosystem of applications, a hardware support list, and a technical support strategy. For many developers, it's now the set of tools, services, and paradigms available for development along with an understanding of the number and types of users that can be reached by supporting it. These definitions span the classical distinctions between Operating System, distribution, and even sometimes application. In my line of work, I can communicate clearly with the majority of my customers by referring to things like Android and eOS as "OSes". It's what they expect and I'm okay with that.

Congrats on the launch of Luna.

They use it like OSX and iOS which rightly have OS in their name: both are the new and original operating systems. Debian, Ubuntu rightly don't have OS in their own names. Elementary shouldn't too.

Poor choice of examples, mate. From the respective home pages:

1. "Ubuntu is the world’s favourite free operating system" - ubuntu.com

2. "Debian is a free operating system (OS) for your computer. An operating system is the set of basic programs and utilities that make your computer run." - debian.org

Then you have CentOS and their ilk. But that whole argument is irrelevant: take a step back and look at what you're arguing about. It is such a petty argument. I feel like I'm living in a crazy house where trivial things are important.

OS X is a distribution of Darwin. iOS is derived from it.

What was the motivation behind forking the distro rather than building packages and an apt repo for Debian or Ubuntu? This is certainly a neat desktop environment, but it always makes me sad to see things done in a manner that makes it difficult to use with other distros.

Money, time, and experience.

All of these things have changed since over the past year or so, and there's been a lot of internal discussion about moving to Debian. In fact, it's almost sure to happen for Isis+1 (Isis is the name of our next release).

Basically, setting up our own repo took really good packagers (we have them now, and you can also get pretty much all of our software from Gentoo and Arch), running our own repo required a solid and cost-effective server (we have this now), and forking let us inherit a LOT of stuff we didn't have the resources to customize back then.

As we grow, it seems only natural that we base less on Ubuntu. But a lot of us used to be Ubuntu developers/enthusiasts, and a bunch of us have good ties to the Ubuntu ecosystem (a number of us met up at UDS-O), etc.

So, tl;dr, it will happen soon enough :)

I must give you a heads up about moving over to Debian: currently, Debian installers do not have the option to format or install onto a BTRFS partition. (Nor does Fedora's installer, without difficulty. Ubuntu's does though.)

This matters to me because I was recently checking out an interesting installation method that allowed me to do away with Linux partitions by solely using BTRFS subvolumes. (If you look at the IRC chat logs of #btrfs, #archlinux-newbie, and #elementary, I was btrfs-newbie trying it a few days back.)

As of now, I am nearly done with my laptop set up, and I am happily running Arch Linux and elementaryOS with a shared /home subvolume all under the same BTRFS partition. This is a really neat feature that I could take advantage of in the Ubuntu installer (Ubiquity, was it?). I did not have this issue on Arch because Arch's installation method was practically a chroot and create the filesystem and package structure in the proper places.

Please keep this in mind when moving forward. It would be interesting if you can maintain that aspect in your next release, Isis. Thanks so much for your time, and keep up the amazing work!

Feature request - one of the few things I'd hate to give up by leaving Ubuntu is the PPA system, however the one thing I don't like about it is that it puts potentially untrusted (or at least not systematically vetted) PPA packages into the system folders.

An interesting improvement of that would be a PPA system that uses Debian's awesome Update Alternatives tool to install PPA's to, say, /opt, soft link them into the system folders, and then manage side-by-versions, rollbacks, etc. with update-alternatives.

By way of example, the Rbenv in Ubuntu repositories was rewritten to use Update Alternatives to manage Ruby version shims. It would be really interesting to have an entire PPA system similarly based on that.

What you saying is just. End users donwload operating systems and the know what a distro is.

Who are you, to define what a operating system is?

"Apple. How can they name their mostly API-level visual customizations of the existing projects "OS X" and be accepted as serious? Such things were called skins and themes.

If you ever looked at Darwin and compared it with any other kernel, you'd know it's a serious, big real OS work, far from any "UI" you mention.

Reading what you wrote, however, I don't believe you did something like that, or even that you'd be able to. I see there's no point arguing in such case.

Darwin was created using portions of FreeBSD's kernel and the work done for Mach at Carnegie Mellon. It's not Apple's sole creation in the slightest.

That's how OS X is now, after ~10 years of closed development. When it first started it was open source FreeBSD and Mach. I even remember the FOSS community's disappointment when Apple close sourced the project.

Who knows how eOS will look in ten years, and the state of OS X now is certainly no basis for criticism of brand new OS.

Reply on news.ycombinator.com