Yes it is possible, and I ask the same question I ask every Linux themed poster who has come before you.
WHY????
Name one single good reason why it is a good idea.
Please none the dumb reasons about Linux developers - they will not come - they like Linux.
Please nothing about using the Linux kernel, you probably have no idea how much of a rewrite that will be or how much of the present Haiku code will need to be dumped first.
To have a distribution already configured with applications installed.
How about that for one good reason? It would be called "Haiku Application Distribution". It would have minimal configuration changes. It would not have forked code.
Even if you create a distribution with pre-selected software (as opposed to 'vanilla' Haiku), you still can't call it 'Haiku'.
Such an attempt has already been done, BTW, by Haikuware's very own Karl (and maybe others as well).
The distro used to be called 'Senryu', IIRC.
Problem is the changes in Haiku are not locked in stone yet.
You can deliver such a distro and the following week it could end up missing major improvements such as the recent webkit and videos updates.
Let me make a suggestion. Instead of making an entire distro, why don't you consider a package of .HPKGs for certain areas of interest. Even better yet a zip file of .HPKG of certain types of programs.
IE. A zip file full of game .HPKGs, one for programming tools, another for graphic tools, etc.
So instead of a person downloading dozens of packages the user only needs to get 3-6 zips files and once downloaded it will be easy to re-install the .HPKGs to a new nightly.
Alpha-4 is nothing like the latest versions of Haiku, so even thinking of doing such before Alpha-5 seems hasty to me.
However, setting up collections of programs and documentations seems like something you can do now.
Where did you say so before? We are not mind readers. Alpha5 and Beta1 do not presently exist and you gave no hint that you meant to use either. The logical conclusion is you are going to use Alpha4 or a nightly, with both approaches having lots of problems.
Stop trying to push an idea where there is no support.
Just realize YOU will have to do all the work needed.
Please understand. Haiku is *NOT* Linux.
And what is wrong with my suggestion? Your claimed goal can be reached just as easily with a set of zip files of collected .HPKGs, so are you doing this to help people or not?
Some responses come across quite aggressive. Could we all calm down a bit please?
Going through the thread, you'll notice that threadstarter Guiseppe didn't participate in this shouting match. Maybe the useful link posted by vooshy was all he needed. So there's no sense to argue against imagined reasons he might be interested in a (totally legal) Haiku based distribution.
Conceptually, don't you think the first distro of Haiku should be one without any forked code and lots of applications for new users to try?
It seems to me that this would be a really good introduction for new users and evaluators. I think the new package manager is just awesome. But I also think that not having all the best applications easily available and already installed presents a barrier of sorts.
I think including utilities, games, screensavers, wallpaper, productivity apps, etc will provide a better 'out of the box' experience for new users.
Conceptually, don't you think the first distro of Haiku should be one without any forked code and lots of applications for new users to try?
Not exactly sure what you mean with a distro "without any forked code", but I don't see how we have any influence on the nature of a Haiku based distro someone has determined he should put out.
Quote:
I think including utilities, games, screensavers, wallpaper, productivity apps, etc will provide a better 'out of the box' experience for new users.
Guess it's a matter of philosophy. For me, Haiku stands for an OS that doesn't get in your way and doesn't assume to know what the user wants. Why should I have apps in my system someone else thinks I want? We have (or soon will have) a repository where packages are sorted by category, have screenshots and user ratings. If I need some app, I'll have a look at what's available and try out the candidates I like best.
I don't think we should weigh down the Haiku image with loads of arbitrarily chosen apps. People that want to provide that could offer scripts with pkgman downloading a whole bunch of apps. "Meta-package-scripts" for audio/video/tools/network/whatever...
Where did you say so before? We are not mind readers. Alpha5 and Beta1 do not presently exist and you gave no hint that you meant to use either. The logical conclusion is you are going to use Alpha4 or a nightly, with both approaches having lots of problems.
Stop trying to push an idea where there is no support.
Just realize YOU will have to do all the work needed.
Please understand. Haiku is *NOT* Linux.
And what is wrong with my suggestion? Your claimed goal can be reached just as easily with a set of zip files of collected .HPKGs, so are you doing this to help people or not?
I think you confused AndrewZ as having initiated this discussion? (it was not actually him that proposed the idea).
I believe the OP meant well, however, until the first release, there would be absolutely no guarantee that some thing that was being relied upon didn't change.
In the Linux world, where a lot of people like to try out lots of different distros to find the one they like, there are lightweight, midweight, and heavyweight distros available.
(I would think this was the concept the OP was thinking of.)
When you look at the BSD's, they only come with the basic neccessities, you download and install the software that you want, and nothing else.
My personal thoughts are that a modern OS should connect to the internet flawlessly & come with a browser ready installed. Any other software should be available for easy download & be installed automatically for the user.
@Earl: No problem, Earl. You're passionate about Haiku and act mature when you realize you might've been trodding down the wrong track. A rare and commendable trait. Kudos!
@Earl: No problem, Earl. You're passionate about Haiku and act mature when you realize you might've been trodding down the wrong track. A rare and commendable trait. Kudos!
Seriously I see two causes one would want to create a distribution forking Haiku:
He doesn't like the Package Manager and simply prefers the simple old way of installing applications or wants to make simply something similar to an Apple bundle (yes I know that .hpkg could work as a bundle). This is the path of the devil!
Suppose I'm Dream Multimedia (the famous creator of the Linux based Dreambox) and I decide that for Enigma 3 Linux is totally sh*i not only as software but I don't like its license too (indeed Dreambox was badly cloned and as Enigma 2 was open source the cloner modified it for their clones!)... Haiku with his BSD license permits to derive work from it and then close the sources! Now I have Haiku that runs on a MIPS processor in a 32 MB flash memory but my App Server is really different from the Haiku vanilla one: it has only 65535 colors, my Windows are bi-dimensional and cannot ever cover the background (no always on top!) and so on. This is a "good" fork and I hope if it should happens Haiku will support them fervently!
That one lost me too. In the bad old days of hard drives being less that 5 GBytes in size bundles could be considered a problem for the amount of space they take up but on today's drives I don't see the problem.
Even if you are using smaller SSDs compared to HDDs, the space taken up with bundles is not that big a deal.
PS. The last two SSDs I bought were 240GB and costed less than $200US each.
Distros are certainly a possibility should Haiku gain a decent amount of interest, however there is certainly a less likelyhood of them emerging due to the nature of Haiku compared to say the Linux situation due to Haiku being a complete OS rather than a OS component (which is what Linux is, albeit a rather large one).
Furthermore, one of the most likely reasons for the existance of Haiku distros, as well as the only distro I know of now (Senryu), was/is to package Haiku with a specific set of software 'out-of-the-box'.
This reason is no longer really exist in my opinion as the package manager (once it is finished) will allow users easy install of any software they want with a click of a button more or less.
Certainly you could create some specifically themed version of Haiku, but given how Haiku is a complete integrated system there's little room beyond that to differentiate a 'distro' enough for it to be of any real interest.
To be honest, I kind of disagree with the idea of forking for the sake of it. A few years back, I posted several LONG threads about starting a fork/distro. I quit the idea and left, going instead to start a Gnu/Linux distribution.
That whole experience taught me a LOT, including why the free (a.k.a. Gnu/Linux) desktop is failing. A package manager is heavenly and I'm really thankful it's coming, but we need more than that--I think we need a Web-centric system. After 3 years, I know in my heart I'm ready to start working a distro for real this time and not quit on it.
(For the interested in my reply), this is what I'd add when I finish my distro that I think we're missing:
- EFI(64) support (there's several ways to do this)
- Modern look and feel
- Web shell and next-gen applications
- Accessibility (including for modern displays); if I was blind or deaf, I couldn't comfortably use Haiku right now.
- And other, smaller tidbits...
But we can't become like Gnu/Linux, either. If I may quote, "Certainly you could create some specifically themed version of Haiku" -- I respectfully digress. This is what I honestly think destroyed the Linux desktop. They have 278 contenders/distros out with 10 big ones, over 5 official DEs, and 900+ distros that have been/are inactive; thankfully, we have less than half a dozen. I think we ought to work to keep it simple like it is, and only fork if there's real improvements and the developer(s) are willing to commit.
As I said, an "app store" or package manager of sorts is great, but I think crafting a well-designed distro would help to boost Haiku popularity, and I wouldn't (and haven't over the years) dismissed it.
To be honest, I kind of disagree with the idea of forking for the sake of it.
I wouldn't call a distro a 'fork', atleast not in terms of Linux as there is no 'official' Linux distro from which to fork, Linux distros all have one common denominator, which is the Linux kernel, 99.9% also use the GNU coreutils/toolchain, but the Linux kernel is always the same, give or take a few patches.
Also I'm not sure what you mean with 'forking for the sake of it'?
agreimann wrote:
That whole experience taught me a LOT, including why the free (a.k.a. Gnu/Linux) desktop is failing.
Failing? If Haiku would ever reach anything close to the amount of users the Linux desktop has I would consider it a massive success and be super happy, we are talking millions of users here.
agreimann wrote:
- EFI(64) support (there's several ways to do this)
- Modern look and feel
- Web shell and next-gen applications
- Accessibility (including for modern displays); if I was blind or deaf, I couldn't comfortably use Haiku right now.
- And other, smaller tidbits...
Could you elaborate on Web shell and next-gen applications? Apart from that, I think the other features are things that the official Haiku will improve on/add as it continues to develop, not sure I see any reason for a fork to achieve this.
agreimann wrote:
If I may quote, "Certainly you could create some specifically themed version of Haiku" -- I respectfully digress.
Just to make it clear, this was not something I proposed, it was instead the only thing that came to my mind in terms of what reason 'someone' would have to create a distro, I don't personally think this is a good reason to create a Haiku distro.
agreimann wrote:
This is what I honestly think destroyed the Linux desktop. They have 278 contenders/distros out with 10 big ones, over 5 official DEs, and 900+ distros that have been/are inactive;
I don't see how the Linux desktop is in any way 'destroyed', if so then Haiku is absolutely dead, which of course it is not.
Now the Linux and Haiku situation is entirely different, again Linux is just a kernel, and as such it is not an operating system, but is instead combined with a lot of other components to make an operating system.
This both means a lot of flexibility but it also means a less integrated operating system. A Linux distro will of course use Linux, and most likely the GNU toolchain, but apart from that it's very much up for grabs, you can ship with a complete desktop environment (Gnome, KDE, XFCE, Enlightenment, etc) or window managers (Openbox, BlackBox, insert tiling WM here), or without any preinstalled graphical environment at all etc. We are seeing movement towards consolidating functionality outside of the kernel with projects like systemd, but there will never be a standard Linux desktop, of that we can be quite sure.
Haiku by comparison, comes as one tightly integrated operating system, furthermore it is directly targeted at desktop use, which means that it comes with a default GUI and desktop environment.
As such there is a lot less (I'd say practically nothing) with which to differentiate a potential Haiku distro, which is also why I doubt we will see many. Particularly since the most likely reason for a distro would be packaging it with preselected software, however with a simple to use package manager that allows users to easily and quickly install software themselves, such a distro will not be very interesting to end users.
agreimann wrote:
As I said, an "app store" or package manager of sorts is great, but I think crafting a well-designed distro would help to boost Haiku popularity, and I wouldn't (and haven't over the years) dismissed it.
While I personally don't see any reason for a Haiku distro, even with the bullet points you listed, you are of course free to give it a shot.
Sure there could be a market for a distro. TuneTracker is developing one as we speak. They've identified a very specific niche market and are creating a product for it.
Or do this: write a bunch of hardware drivers for oddball devices that the official release does not support, then put a restrictive license on those drivers so that they can't be folded back into the main project. Create the office suite we've been desperately looking for for over ten years now. People will grumble, but they will pay. (Actually, they would pay you just for a good office suite, no need to fold a distro around it, but I'm trying to make a general point here.)
But nobody will thank you for changing the title and a few graphics. There have to be real advantages. And you'll have to keep it in sync with the nightlies. This community seems to be split between those still running Alpha 4 or even alpha 3 and those picking up a new nightly every week. Which of those groups are you aiming for?
I know in a sense I've taken the focus of the post away from the OP, but I think this is an important thing to discuss for everyone interested in doing a distribution (or even a fork). After carefully reading the whole thread and the last few posts, I wanted to respond best I can in regards to what I wrote earlier if I may:
Rox wrote:
I wouldn't call a distro a 'fork', atleast not in terms of Linux as there is no 'official' Linux distro from which to fork, Linux distros all have one common denominator, which is the Linux kernel, 99.9% also use the GNU coreutils/toolchain, but the Linux kernel is always the same, give or take a few patches.
Also I'm not sure what you mean with 'forking for the sake of it'?
I wouldn't either. Notice that I have always called Gnu/Linux exactly that, not Linux, as I think that it's the proper title for it as an OS. Linux is only a kernel, but at the same time, a very powerful one that comprises a micro OS of sorts as a sole entity even without the GNU tools. I've looked at the Linux source myself; it's quite impressive, actually. However, as a direct example of how diverse Gnu/Linux can be, I consider Slackware and Ubuntu to both be an OS because the thinking is different, though internally, they share the same "roots". But Gnu/Linux (or just "Linux" if you prefer) is a controversial topic that I'm going to avoid so I'm not responsible for a flame war.
I'd be glad to answer your second question; sorry I wasn't concise on it before. What I meant by "forking for the sake of it" are people who make a distribution of some sort for pleasure, etc. and just quickly change the logo(s), change the look to their favorite colors and theme, add a few packages, and paste it on the Web. And, yes, I do honestly believe this destroyed the Linux desktop, as now there are hundreds of dead and active distros, with no hope of uniting with one dominant leader. We can never do this.
Rox wrote:
Failing? If Haiku would ever reach anything close to the amount of users the Linux desktop has I would consider it a massive success and be super happy, we are talking millions of users here.
I wouldn't dare post something without doing some research first to back it; Miguel de Icaza, one of the original developers for the Gnome project, noted what "killed" Linux as a desktop in 2012. The group behind the AbiWord project wrote a paper on why they felt Linux was failing a while back. And lastly, please see Net Applications or other firms' web and market share for Linux as a system group on the desktop--it does not hold much share in any view.
Undoubtedly, Linux runs the majority of supercomputers and mobile devices by market share, which can as easily be proven. Linux also doesn't lack people, either, as you mentioned. Ubuntu, Linux Mint, OpenSUSE, and Fedora, as well as others in the top ten have quite a few followers. But, compared to a world market, desktop Linux has been soundly beaten... for now. There's still hope. Whether the situation with Windows 8.x causes a major flux in this has yet to be seen, but I know that as of now, the statistics show the situation for themselves, though their true accuracy due to Gnu/Linux's open nature is under debate by some (i.e. are there more or less installs and users than polled?)
Rox wrote:
Just to make it clear, this was not something I proposed, it was instead the only thing that came to my mind in terms of what reason 'someone' would have to create a distro, I don't personally think this is a good reason to create a Haiku distro.
My apologies over my quick reply. I now understand you were just looking for a point. I also agree that simply "theming" Haiku wouldn't be a good idea, either.
Rox wrote:
This both means a lot of flexibility but it also means a less integrated operating system. A Linux distro will of course use Linux, and most likely the GNU toolchain, but apart from that it's very much up for grabs, you can ship with a complete desktop environment (...) or window managers (...) or without any preinstalled graphical environment at all etc.
I think these are good points any dev knows well. I like Haiku because it has the "look and feel" of Apple Platinum (UI style for MacOS 8/9) with *nix tools and a slick, integrated, and compact framework. Unlike Gnu/Linux, there's not a ton of layers, multiple window servers as of 2013, compositors, etc. Furthermore, I've authored a small desktop environment as a proof of concept, and I'm aware of the "layering" of Gnu/Linux. To be honest, I'll be thankful to leave that mess for something that I believe is cleaner and better.
Michel wrote:
Or do this: write a bunch of hardware drivers for oddball devices that the official release does not support, then put a restrictive license on those drivers so that they can't be folded back into the main project. Create the office suite we've been desperately looking for for over ten years now. People will grumble, but they will pay.
This is a commendable idea, but personally, it'd be against my ideals. We need a free environment, and unlike some of the Linux distros I've seen that do everything for themselves, I believe in community. If Newleaf improves on Haiku, I'd be very grateful if someone reused the code, etc. as long as they gave Haiku or any new distribution credit, as I think would be ethical and honorable to do. I'd like to be able to give back by allowing others to see and use it openly, and for anyone else's endeavors to do the same.
Michel wrote:
But nobody will thank you for changing the title and a few graphics. There have to be real advantages. And you'll have to keep it in sync with the nightlies.
I wouldn't thank myself for doing that! I believe that there must be real advantages for anything to be worthy of calling itself a distribution. Of course, there would be a new name and logo, but there would be several huge improvements for ANY free desktop--not just Haiku!
1. A web shell An app store/package manager is a GREAT start and I'm thankful for those inside and outside the Haiku team focused on it, but it's only the beginning! Clearly, I think that a dynamic web-based application framework is the future of the open (and closed) desktop. Firefox OS, Chrome OS, Android, iOS and Windows 8.x I think help prove this somewhat. Haiku doesn't have this... if it did, we would attract developers to a live environment fast, because there would be something they'd actually want to build their modern apps on. I'd also include tools to build simple apps of this class into the distro. Haiku simply does not have any of these capabilities in it at the moment. Without it/them, an OS is dead already. I'm shocked how much this is being dismissed as another feature! It's the whole face of the OS that will define the future... if anyone can see it. I'm highly optimistic about it and can't make this point any stronger.
2. Accessibility Can the system speak by default or read contents of applications? Can we easily control the zoom onscreen or scale the display for large displays like I can with my Mac? Do we have a way to easily use a virtual keyboard or easy-use controls? If the answer is no, then I think we need to get Haiku there to be a truly modern operating system. This would be a second area I'd seriously work on. Accessibility IS the heart of ANY system, and if we ignore these details, we've ignored how to make really awesome software.
3. The tidbits I think new colors and icons are a definite improvement for the modern crowd. Also, I would include a few packages; I think Firefox, VLC, and Qemu with a ReactOS image would be helpful. As an extra, I think using Grub or a base Debian image could allow EFI64 users to boot their machines; it's worth a try for now, as implementing a new way would quite honestly waste time and take away from working on user-related areas.
And finally, I would follow the nightlies, because I would want to keep up with any bug fixes and additions. I hope this answers everyone's concerns and questions. I really do want to say thanks to the Haiku team, and after reading the guidelines on considering a distro, I think I'm doing the right thing.
By nature, as open source users and developers, we're rebels, because we desire something better while preserving our rights. I think we also offer hope. And I hope I can get such a distribution out by Christmas... and, as a life goal, put a dent in the open source arena before I'm ever gone.
My apologies that this was so long; thank you all for reading what I've had to write! Please let me know what you think! :-)
Miguel de Icaza, one of the original developers for the Gnome project, noted what "killed" Linux as a desktop in 2012.
Miguel de Icaza has held a grudge against the Linux desktop ever since 'it' (and the rest of the *nix ecosystem) turned down his beloved C#/Net port Mono, sorry, but he is hardly what I would call 'objective'. Also as I recall it he tried to put the blame on the Linux kernel devs, which was pure nonsense as they, unlike Miguel and company, are deeply against breaking user space functionality.
agreimann wrote:
But, compared to a world market, desktop Linux has been soundly beaten... for now.
Actually I've never seen how the Linux desktop (or any other desktop, Haiku included) could ever be anything but 'soundly beaten', Microsoft owns the PC desktop through the dominance they laid the foundation for back in the days of their anti-competitive OEM deals.
Not even Apple while riding their massive wave of 'coolness', with their product placements in everything on television and box office, was able to put a dent in Microsoft's PC desktop monopoly.
Of course from now on the battle for the 'computing masses' won't be on the PC desktop, but on mobile (phone/tablet/whatever comes next, glasses?) so there is a chance that we will see a larger incentive among those who still use the desktop PC to widen their horizon as the makers of their OS of choice starts focusing heavily on the mobile sphere.
Still the PC desktop market is by all accounts rapidly shrinking, so there will be less users to fight for.
agreimann wrote:
To be honest, I'll be thankful to leave that mess for something that I believe is cleaner and better.
Certainly there are drawbacks to the flexibility that loosely coupled components as those of the 'Linux desktop', such as the pieces often not 'fitting' very well, and of course that an application running on the Linux desktop can't expect that the end user will have desktop feature/functionality X.
Again Haiku, being written not only with the desktop as it's primary target, but also as a complete system where all pieces 'fit', does not suffer from those drawbacks.
And while I prefer Linux for many things (and I don't hate it as a desktop which is what I use every day), I would really like to use Haiku for my desktop needs, with the main barrier as of now being lack of software, but also system maintenance functionality which will eventually be solved as part of fully working package management.
As for your distro ideas, I don't feel the need to 'judge' them, if you want to pursue this then of course go for it, I would comment on the 'web shell' though, one of the areas where Haiku could become a 1st class citizen is of course on the 'web', as it is in itself platform agnostic as long as you can support the browser technology needed to display/run the served web content, is this what you are referring to (as in supporting existing/upcoming web frameworks) or are you suggesting the creation of some home-grown framework?
Agreimann, if you can get us up-to-date ports of Firefox, VLC, and Qemu, you will be the hero of this community. No need to worry about a distro. Just get those .hpkgs out there. What you call the tidbits is what we need the most. The versions we have are ten years out of date. The people with the skills to port these monsters are too busy hacking on the OS itself.
I think you are underestimating just how much work we're talking about here. Porting FireFox is a major job. But hey, dreaming big is good. Go for it.
This is really becoming an engaging discussion! Thanks for your comments! I'll see if I can respond best I can if time allows...
Rox wrote:
Miguel de Icaza has held a grudge against the Linux desktop ever since 'it' (and the rest of the *nix ecosystem) turned down his beloved C#/Net port Mono
Absolutely true. But De Icaza should earn respect at least as the original Gnome co-head with Mena nonetheless. There's no doubt Mono is a bad framework and that his argument was rather unfounded toward the kernel devs, but I wouldn't quit listening to what he had/has to write/say because of an idea. Current designers and developers for Gnome are rather brilliant, though Gnome Shell was a mess when it started.
Nevertheless, the Linux kernel can be a mess at times, when major transitions call for it or some new idea is proposed. I run Debian Gnu/Linux on my workstation, but even so, I think the direction Gnu/Linux is going is truly a mess. Wayland, Mir, and a host of other changes may be good for users, but they do nothing but keep software developers dependent on those changes constantly anxious and having to memorize new ways of writing (or rewriting) the code and doing things. Proprietary environments are stable, and while there are changes, they don't break modern software. After years of running and coding various projects on Gnu/Linux, I'm close to done with it. I'm sorry if that offends anyone; I genuinely don't mean to, but I think it's true.
Rox wrote:
Microsoft owns the PC desktop through the dominance they laid the foundation for back in the days of their anti-competitive OEM deals.
Not even Apple while riding their massive wave of 'coolness', with their product placements in everything on television and box office, was able to put a dent in Microsoft's PC desktop monopoly.
Of course from now on the battle for the 'computing masses' won't be on the PC desktop, but on mobile (phone/tablet/whatever comes next, glasses?) so there is a chance that we will see a larger incentive among those who still use the desktop PC to widen their horizon as the makers of their OS of choice starts focusing heavily on the mobile sphere.
I couldn't have addressed this better! Honestly--this about describes the situation. Remember the DOS wars, or Netscape? Worse yet, I do remember Dell couldn't offer an alternative OS outside of FreeDOS without losing benefits and being penalized. M$ is a monster, but the fact remains... they're here and that's the past.
Now, Apple actually did begin to gain more market share after Vista, but recently, their mobile division (the iDevices) is now what's the large area of sales for them.
I do think the future is a mobile one as well... I think we should look to the Raspberry Pi or other SoC alternatives and netbooks as the next step... for now. That's one reason why I think we need a lot of improvement in how the system appeals to people. Are we aiming for modern Web 2.0 developers, or the languages from the 90s? Viewpoint is everything that defines the "life" of an OS. And I do see the desktop shrinking quickly. Overall, we agree here completely as well.
Rox wrote:
As for your distro ideas, I don't feel the need to 'judge' them, if you want to pursue this then of course go for it, I would comment on the 'web shell' though (...) is this what you are referring to (as in supporting existing/upcoming web frameworks) or are you suggesting the creation of some home-grown framework?
I appreciate your consideration... although feel free to judge/critique my ideas and I'll listen best I can. Feedback is always important. My earlier post I think explains the web shell clearly, however. In short, it's not for supporting anything, but rather, to build something new. I think I've reiterated this point enough.
Michel wrote:
I think you are underestimating just how much work we're talking about here. Porting FireFox is a major job.
I've thought a long time about starting anything. I even discussed doing a distribution for a while back in 2011, but quit on it, as I thought quitting was the best option then. I think that now, with all the changes taking place in the world today and a less unified Gnu/Linux, there's no better opportunity to make things better and I'm not quitting on it this time. I don't think I've underestimated anything, but I do know it's going to be hard work.
Also, about Firefox... I was actually going to simply include the 3.0a release, as we'd need to port Cairo libraries (and more) for anything higher. That doesn't mean I've given up on such a task... it's that I've got to be open, honest and realistic about the situation. I can't promise impossible things like a politician, and porting a NEW version of Fx may or may not happen. But I definitely am going to try to get Qemu and VLC up-to-date for sure, as these are key for what I'd like to do. All I can write in this case is the usual "we'll see how it goes". I'm pretty sure, as I said before. that I probably won't have anything until Christmas. And about what you wrote about being a hero... I don't know, and frankly, don't care about that.
I just want to change the world. :-) Thanks all for the replies, and again, for reading...
I have only see tiltos personaly. I think if haiku becumes more populer it might have "distros for differnt arketheures that the devlopers don't use for example a dreamcast or ps2 haiku distrobution. Also stop using the packagemanger as an exucse or defend your statements as they don't count as no one uses nightly images.
Please... check your spelling if English is your native language, as your post was hard to read without saying it aloud. I don't mean to be a jerk... I promise I'm not a grammar nut and can live with mistakes, as no one's perfect... but several words were badly mangled.
In response to your post, one point at a time:
- I understand TiltOS is a good idea--but I do see a lot more needs to be done than simply porting X.org & apps. It has been successful, and if you're using it, wonderful. I can't argue my ideas will be better or worse, honestly.
- Haiku can already run on several architectures--please check HaikuPorts or the nightlies, or check out & compile a build yourself through subversion and try booting one. Be warned it probably won't boot without modifying it yourself.
- I'm not using the Package Manager as an excuse by any means. I've looked at it... it's a great concept that I think Haiku needs, and brings it nearer to post-2000 usage, but we don't have modern Web-style or next-gen applications. My stating that I believe the package manager wasn't the end of things is not an excuse... it's the truth. It's a brilliant start... and I definitely want to say thank you to all the developers that made HaikuDepot possible!
- Some people do surprisingly use Nightly images. I, for one, prefer Haiku nightly and like being able to use new (but unstable) features and capabilities.
[quote=agreimann]I understand TiltOS is a good idea--but I do see a lot more needs to be done than simply porting X.org & apps.[/quote]
Porting X.org is never the solution. Ever. And TiltOS didn't do that, they basically just recompiled a package manager and stuck it on Haiku.
[quote=agreimann]Haiku can already run on several architectures--please check HaikuPorts or the nightlies, or check out & compile a build yourself through subversion and try booting one. Be warned it probably won't boot without modifying it yourself.[/quote]
Subversion is dead; Haiku has not used it for a *very* long time. And HaikuPorts simply serves to build packages for Haiku, and in building the base system packages during the bootstrapping process -- AFAIK Haiku's configure and other tools use it automatically without user input (unless it breaks, which happens quite often because bootstrapping is rocket science...)
[quote=agreimann]I'm not using the Package Manager as an excuse by any means. I've looked at it... it's a great concept that I think Haiku needs, and brings it nearer to post-2000 usage, but we don't have modern Web-style or next-gen applications.[/quote]
Having "whatever's hip right now" is never Haiku's goal. If you want to port something, that's great, but don't expect that it will ever become a permanent part of the OS. It likely won't. But WebPositive on the nightlies is much improved, and HTML5 games are playable -- to some extent.
Also: Haiku's goal never was and never will be to attract developers. For R1 and R2 we have the goal of a BeOS clone with modern features; after that in R3, we'll break binary compatibility and do things "the right way" and not just because "everyone else is doing it and if we do it, we'll get more devs".
Also: Haiku's goal never was and never will be to attract developers.
Huh. You might need to tell the Haiku project about their goals:
Haiku's own Welcome page wrote:
We hope to attract new developers to our project
This sort of goal has also been expressed by key Haiku developers at various times, should I dig all those examples out too ?
waddlesplash wrote:
For R1 and R2 we have the goal of a BeOS clone with modern features; after that in R3, we'll break binary compatibility and do things "the right way" and not just because "everyone else is doing it and if we do it, we'll get more devs".
The Haiku R1 "BeOS clone" goal has been the same for almost 13 years now. Meanwhile not only have the goal posts shifted, the team moved to a different stadium and is playing a different sport. In 2016 announcing a new 32-bit desktop OS will be like announcing a new CRT television. A few fans will be excited, and everybody else will be momentarily puzzled and then move on.
Having "whatever's hip right now" is never Haiku's goal.
We're "hip" since 1998!
NoHaikuForMe wrote:
The Haiku R1 "BeOS clone" goal has been the same for almost 13 years now. Meanwhile not only have the goal posts shifted, the team moved to a different stadium and is playing a different sport. In 2016 announcing a new 32-bit desktop OS will be like announcing a new CRT television. A few fans will be excited, and everybody else will be momentarily puzzled and then move on.
No User cares about 32/64Bit, if everything is working (Webbrowser, e-Mail, Mediaplayer, Solitaire and/or Minesweeper!!). ^^
Haiku R1 has been decided to be a replacement for BeOS R5. If you don't like that, go use a Linux distribution. In R2, we'll switch to GCC4Hybrid (plus some bugfixes), which will happen a few months or so after R1.
Meanwhile, I don't see you helping towards development.
As for developers -- if any come along that want to help the project, that's great! But we aren't going to change our goals just to attract users or developers.
Haiku R1 has been decided to be a replacement for BeOS R5. If you don't like that, go use a Linux distribution. In R2, we'll switch to GCC4Hybrid (plus some bugfixes), which will happen a few months or so after R1.
We? Last I checked, you don't speak for the project, nor do you represent everyone else's opinion.
Quote:
Meanwhile, I don't see you helping towards development.
I personally think a non-forked distro would be a great way to attract new users, reviewers, and developers. I also think a distro could easily co-exist, even 'sit on the shoulders' of the package manager.
Here's what a distro would have in addition to the 'clean install' of R1:
A good selection of games (selected from Haikuware)
A good selection of wallpaper (we should hire a graphic artist for some Haiku stuff)
A good selection of screen savers (our current selection is Lame-O, we desperately need better!)
A selection of the best utilities (we desperately need office productivity tools)
A selection of developer tools (Paladin, gvim, Go, etc)
This way when someone installs Haiku for the first time they will have a better out-of-box experience, they won't have to poke around and do research to find the good stuff.
A distro like this could be implemented using the package manager. All this stuff could exist as uninstalled packages that get installed by the package manager at start-up. We could even make some cool graphics to distract people while the packages install.
This methodology would be flexible and maintainable. When Haiku gets a new release, only affected packages have to change.
"Meanwhile, I don't see you helping towards development."
Waddlesplash, please meet NoHaikuForMe. NoHaikuForMe, please meet Waddlesplash. NoHaikuForMe is a long time critic of the Haiku project. He is a very bright contrarian. I find that even though he occasionally gets his facts wrong :-), he does have a keen understanding of OS fundamentals. Sadly, he's not a very productive developer and seldom if ever submits code :-)
Waddlesplash is a junior Haiku contributor and is release coordinator for A5.
The Haiku R1 "BeOS clone" goal has been the same for almost 13 years now. Meanwhile not only have the goal posts shifted, the team moved to a different stadium and is playing a different sport.
This is as it should be. This project is not done with static 'waterfall' requirements and design. It's done through learning and adaptation and limited resources.
NoHaikuForMe wrote:
In 2016 announcing a new 32-bit desktop OS will be like announcing a new CRT television. A few fans will be excited, and everybody else will be momentarily puzzled and then move on.
Since Haiku is not a server oriented project there is little need for the benefits of 64-bit architecture, which as you know are primarily more RAM and security. Once Haiku is stable and has more apps that need more RAM, and has more users with more security issues, then emphasis and resources can shift to 64-bit concerns. The new package manager can actually be adapted to manage 32-bit and 64-bit 'fat' executables. Much like Mac PowerPC/68000 executables. There is even the possibility of a thunking layer.
I have been contributing, although mostly in HaikuPorts, which is where most of the action for R1a5 will happen (as already stated by Ingo and Oliver on the ML).
I'm also manager of HaikuArchives -- otherwise you captured me pretty well :-) and pleased to meet you, NoHaikuForMe!
As for my R2 statement -- true, nothing has been set in stone. However, I'm not the first one to suggest something like that, I recall Adrien (PulkoMandy) saying something similar to me in IRC.
No User cares about 32/64Bit, if everything is working (Webbrowser, e-Mail, Mediaplayer, Solitaire and/or Minesweeper!!). ^^
Chrome, Windows, Mac OS X, and OpenIndiana/Solaris ALL support 64-bit and transistioning to it completely will be as inevitable as the obsolescence of 16-bit. The total shift from 32-bit has not happened now, and probably won't happen for another 10+ years as mobile/embedded devices are still heavily using 32-bit architectures, but it IS occurring and it is important to look at the future, not the past.
Intel already offers the x64 Minnowboard Max, and mainstream computers are shifting now. As for the mobile or future market, the iPhone 5s introduced the collaborative 64-bit A7 and the upcoming Octa-core is 64-bit; it is safe to not only assume, but state, that ARM processors/architecture as a collective whole will eventually follow the transistion to 64-bit as well.
What is my point? x86 is still here, but it won't be forever. AMD64 and x86_64 on the desktop, notebook, and even on the high-end tablets (such as an EliteBook convertible), are becoming common. ARM will shift to 64-bit. What will that leave? In my opinion, it will leave an outdated 32-bit group of hobbyists with a cool system that only they want.
PLEASE NOTE: I do not wish to contribute to, nor start a flame war; I'm just stating a conclusion as it appears. In summary, Haiku and any derivatives MUST support more than x86 if they are to survive and be used by the general market.
1. We have x86_64 nightly builds that work for the most part (they even have WebPositive!) but they aren't our highest priority. Not until after R1 at least.
Since tilt os was mentioned... does anyone know where the source for the tilt OS ports went? Even though it's a bit messy, an X windows port would be handy!
That would be restarting Senryu rather than continuing work on it...
I would suggest you wait for alpha5 and use that as your base. But yes, I think there is room for a DVD-sized "distro" with more software on it than will fit on a CD. Look at Ubuntu: you can get it in versions ranging from 40 MB command-line only up to the "several kitchen sinks" version.
I would love to have a multimedia production distro (something like Dyne:bolic - http://www.dynebolic.org/), instead...
same (r.i.p. dyne:bolic). a media production os can still work in 32-but, though, especially considering must digital media recorders (cameras, audio recorders etc) are 32-bit and anyone not working in 4k raw video wouldn't need more. with the way the media kit works, efficient workflows can be built with minimal overhead.
I don't mind work restarting/continuing on Senryu. That was quite a while ago, so it's probably not relevant anymore with all the changes made to Haiku. I lost all the sources and changes that made the distro compatible with Haiku's 'distro rules'. I do think it's a good idea to demonstrate what Haiku can do and am all for it, just don't have the energy/will to commit or help in such a project. I'd also be in favour adding a Haikuware repository to the package manager; restricted to the Senryu distro (certain Haiku devs consider software at Haikuware 'crap'). I just consider the software available at Haikuware 'use at your own risk'. If it works great! If it doesn't fine, who cares. It's better having it there than not - especially for research purposes to read what the application was capable of and to contact authors for sources. At one time most of the software at Haikuware worked on Haiku releases. That's why I am not sorting what works, what doesn't on Haikuware until a final R1 comes around. Even current package management software breaks all the time because of changes to Haiku, so it's hard to define what 'crap' is until R1 comes around (when and if that that comes)
I find that even though he occasionally gets his facts wrong :-), he does have a keen understanding of OS fundamentals. Sadly, he's not a very productive developer and seldom if ever submits code :-)
Bane's voice:For you
Quote:
Waddlesplash is a junior Haiku contributor and is release coordinator for A5.
By the way it seems a bit unfair to let a kid who is new to the project take the bullet for Haiku's ongoing logistical problems.
After the R1 alpha 4 mess a post-mortem was conducted. That should have directed, or at the very least influenced, the planning for R1 alpha 5. Most obviously it was clear that R1 alpha 5 needed to be out the door in 2013, but that's not possible now. Still there are other lessons and it's unfortunate that so far there's no sign that anybody actually learned anything.
Comments
Re: Distro based on Haiku
Yes it is possible, and I ask the same question I ask every Linux themed poster who has come before you.
WHY????
Name one single good reason why it is a good idea.
Please none the dumb reasons about Linux developers - they will not come - they like Linux.
Please nothing about using the Linux kernel, you probably have no idea how much of a rewrite that will be or how much of the present Haiku code will need to be dumped first.
AGAIN WHY?????
Then go read http://www.haiku-os.org/community/forum/haiku_fork
And answer with a worthwhile statement WHY?
Re: Distro based on Haiku
https://www.haiku-os.org/community/guidelines_creating_haiku_distribution
Re: Distro based on Haiku
Here's one good reason to have a Haiku distro:
To have a distribution already configured with applications installed.
How about that for one good reason? It would be called "Haiku Application Distribution". It would have minimal configuration changes. It would not have forked code.
Re: Distro based on Haiku
Even if you create a distribution with pre-selected software (as opposed to 'vanilla' Haiku), you still can't call it 'Haiku'.
Such an attempt has already been done, BTW, by Haikuware's very own Karl (and maybe others as well).
The distro used to be called 'Senryu', IIRC.
-- louisdem
Re: Distro based on Haiku
Unless of course it was done with permission of Haiku Inc.
Re: Distro based on Haiku
Problem is the changes in Haiku are not locked in stone yet.
You can deliver such a distro and the following week it could end up missing major improvements such as the recent webkit and videos updates.
Let me make a suggestion. Instead of making an entire distro, why don't you consider a package of .HPKGs for certain areas of interest. Even better yet a zip file of .HPKG of certain types of programs.
IE. A zip file full of game .HPKGs, one for programming tools, another for graphic tools, etc.
So instead of a person downloading dozens of packages the user only needs to get 3-6 zips files and once downloaded it will be easy to re-install the .HPKGs to a new nightly.
Alpha-4 is nothing like the latest versions of Haiku, so even thinking of doing such before Alpha-5 seems hasty to me.
However, setting up collections of programs and documentations seems like something you can do now.
Re: Distro based on Haiku
Unless of course it was done with permission of Haiku Inc.
I don't think so (but I could be wrong).
I also agree with Earl's last comment.
-- louisdem
Re: Distro based on Haiku
Alpha-4 is nothing like the latest versions of Haiku, so even thinking of doing such before Alpha-5 seems hasty to me.
This would be Alpha5 or Beta1.
Re: Distro based on Haiku
Where did you say so before? We are not mind readers. Alpha5 and Beta1 do not presently exist and you gave no hint that you meant to use either. The logical conclusion is you are going to use Alpha4 or a nightly, with both approaches having lots of problems.
Stop trying to push an idea where there is no support.
Just realize YOU will have to do all the work needed.
Please understand. Haiku is *NOT* Linux.
And what is wrong with my suggestion? Your claimed goal can be reached just as easily with a set of zip files of collected .HPKGs, so are you doing this to help people or not?
Re: Distro based on Haiku
Hey guys!
Some responses come across quite aggressive. Could we all calm down a bit please?
Going through the thread, you'll notice that threadstarter Guiseppe didn't participate in this shouting match. Maybe the useful link posted by vooshy was all he needed. So there's no sense to argue against imagined reasons he might be interested in a (totally legal) Haiku based distribution.
Regards,
Humdinger
Re: Distro based on Haiku
Okay, you are right. Sorry to all, I was getting carried away again.
Once again sorry to all.
Re: Distro based on Haiku
:-)
Conceptually, don't you think the first distro of Haiku should be one without any forked code and lots of applications for new users to try?
It seems to me that this would be a really good introduction for new users and evaluators. I think the new package manager is just awesome. But I also think that not having all the best applications easily available and already installed presents a barrier of sorts.
I think including utilities, games, screensavers, wallpaper, productivity apps, etc will provide a better 'out of the box' experience for new users.
Re: Distro based on Haiku
Conceptually, don't you think the first distro of Haiku should be one without any forked code and lots of applications for new users to try?
Not exactly sure what you mean with a distro "without any forked code", but I don't see how we have any influence on the nature of a Haiku based distro someone has determined he should put out.
I think including utilities, games, screensavers, wallpaper, productivity apps, etc will provide a better 'out of the box' experience for new users.
Guess it's a matter of philosophy. For me, Haiku stands for an OS that doesn't get in your way and doesn't assume to know what the user wants. Why should I have apps in my system someone else thinks I want? We have (or soon will have) a repository where packages are sorted by category, have screenshots and user ratings. If I need some app, I'll have a look at what's available and try out the candidates I like best.
I don't think we should weigh down the Haiku image with loads of arbitrarily chosen apps. People that want to provide that could offer scripts with pkgman downloading a whole bunch of apps. "Meta-package-scripts" for audio/video/tools/network/whatever...
Regards,
Humdinger
Re: Distro based on Haiku
Where did you say so before? We are not mind readers. Alpha5 and Beta1 do not presently exist and you gave no hint that you meant to use either. The logical conclusion is you are going to use Alpha4 or a nightly, with both approaches having lots of problems.
Stop trying to push an idea where there is no support.
Just realize YOU will have to do all the work needed.
Please understand. Haiku is *NOT* Linux.
And what is wrong with my suggestion? Your claimed goal can be reached just as easily with a set of zip files of collected .HPKGs, so are you doing this to help people or not?
I think you confused AndrewZ as having initiated this discussion? (it was not actually him that proposed the idea).
Re: Distro based on Haiku
You are right, that is another reason to say sorry as I did not follow the thread properly.
Again to all sorry for my haste.
Re: Distro based on Haiku
I believe the OP meant well, however, until the first release, there would be absolutely no guarantee that some thing that was being relied upon didn't change.
In the Linux world, where a lot of people like to try out lots of different distros to find the one they like, there are lightweight, midweight, and heavyweight distros available.
(I would think this was the concept the OP was thinking of.)
When you look at the BSD's, they only come with the basic neccessities, you download and install the software that you want, and nothing else.
My personal thoughts are that a modern OS should connect to the internet flawlessly & come with a browser ready installed. Any other software should be available for easy download & be installed automatically for the user.
Re: Distro based on Haiku
Hi!
@Earl: No problem, Earl. You're passionate about Haiku and act mature when you realize you might've been trodding down the wrong track. A rare and commendable trait. Kudos!
Regards,
Humdinger
Re: Distro based on Haiku
Hi!
@Earl: No problem, Earl. You're passionate about Haiku and act mature when you realize you might've been trodding down the wrong track. A rare and commendable trait. Kudos!
Regards,
Humdinger
+1
Re: Distro based on Haiku
Seriously I see two causes one would want to create a distribution forking Haiku:
Re: Distro based on Haiku
Bundles are the "path of the devil"? Please explain... as I think OS X bundles work quite well.
Re: Distro based on Haiku
That one lost me too. In the bad old days of hard drives being less that 5 GBytes in size bundles could be considered a problem for the amount of space they take up but on today's drives I don't see the problem.
Even if you are using smaller SSDs compared to HDDs, the space taken up with bundles is not that big a deal.
PS. The last two SSDs I bought were 240GB and costed less than $200US each.
Re: Distro based on Haiku
Distros are certainly a possibility should Haiku gain a decent amount of interest, however there is certainly a less likelyhood of them emerging due to the nature of Haiku compared to say the Linux situation due to Haiku being a complete OS rather than a OS component (which is what Linux is, albeit a rather large one).
Furthermore, one of the most likely reasons for the existance of Haiku distros, as well as the only distro I know of now (Senryu), was/is to package Haiku with a specific set of software 'out-of-the-box'.
This reason is no longer really exist in my opinion as the package manager (once it is finished) will allow users easy install of any software they want with a click of a button more or less.
Certainly you could create some specifically themed version of Haiku, but given how Haiku is a complete integrated system there's little room beyond that to differentiate a 'distro' enough for it to be of any real interest.
Re: Distro based on Haiku
To be honest, I kind of disagree with the idea of forking for the sake of it. A few years back, I posted several LONG threads about starting a fork/distro. I quit the idea and left, going instead to start a Gnu/Linux distribution.
That whole experience taught me a LOT, including why the free (a.k.a. Gnu/Linux) desktop is failing. A package manager is heavenly and I'm really thankful it's coming, but we need more than that--I think we need a Web-centric system. After 3 years, I know in my heart I'm ready to start working a distro for real this time and not quit on it.
(For the interested in my reply), this is what I'd add when I finish my distro that I think we're missing:
- EFI(64) support (there's several ways to do this)
- Modern look and feel
- Web shell and next-gen applications
- Accessibility (including for modern displays); if I was blind or deaf, I couldn't comfortably use Haiku right now.
- And other, smaller tidbits...
But we can't become like Gnu/Linux, either. If I may quote, "Certainly you could create some specifically themed version of Haiku" -- I respectfully digress. This is what I honestly think destroyed the Linux desktop. They have 278 contenders/distros out with 10 big ones, over 5 official DEs, and 900+ distros that have been/are inactive; thankfully, we have less than half a dozen. I think we ought to work to keep it simple like it is, and only fork if there's real improvements and the developer(s) are willing to commit.
As I said, an "app store" or package manager of sorts is great, but I think crafting a well-designed distro would help to boost Haiku popularity, and I wouldn't (and haven't over the years) dismissed it.
Re: Distro based on Haiku
To be honest, I kind of disagree with the idea of forking for the sake of it.
I wouldn't call a distro a 'fork', atleast not in terms of Linux as there is no 'official' Linux distro from which to fork, Linux distros all have one common denominator, which is the Linux kernel, 99.9% also use the GNU coreutils/toolchain, but the Linux kernel is always the same, give or take a few patches.
Also I'm not sure what you mean with 'forking for the sake of it'?
That whole experience taught me a LOT, including why the free (a.k.a. Gnu/Linux) desktop is failing.
Failing? If Haiku would ever reach anything close to the amount of users the Linux desktop has I would consider it a massive success and be super happy, we are talking millions of users here.
- EFI(64) support (there's several ways to do this)
- Modern look and feel
- Web shell and next-gen applications
- Accessibility (including for modern displays); if I was blind or deaf, I couldn't comfortably use Haiku right now.
- And other, smaller tidbits...
Could you elaborate on Web shell and next-gen applications? Apart from that, I think the other features are things that the official Haiku will improve on/add as it continues to develop, not sure I see any reason for a fork to achieve this.
If I may quote, "Certainly you could create some specifically themed version of Haiku" -- I respectfully digress.
Just to make it clear, this was not something I proposed, it was instead the only thing that came to my mind in terms of what reason 'someone' would have to create a distro, I don't personally think this is a good reason to create a Haiku distro.
This is what I honestly think destroyed the Linux desktop. They have 278 contenders/distros out with 10 big ones, over 5 official DEs, and 900+ distros that have been/are inactive;
I don't see how the Linux desktop is in any way 'destroyed', if so then Haiku is absolutely dead, which of course it is not.
Now the Linux and Haiku situation is entirely different, again Linux is just a kernel, and as such it is not an operating system, but is instead combined with a lot of other components to make an operating system.
This both means a lot of flexibility but it also means a less integrated operating system. A Linux distro will of course use Linux, and most likely the GNU toolchain, but apart from that it's very much up for grabs, you can ship with a complete desktop environment (Gnome, KDE, XFCE, Enlightenment, etc) or window managers (Openbox, BlackBox, insert tiling WM here), or without any preinstalled graphical environment at all etc. We are seeing movement towards consolidating functionality outside of the kernel with projects like systemd, but there will never be a standard Linux desktop, of that we can be quite sure.
Haiku by comparison, comes as one tightly integrated operating system, furthermore it is directly targeted at desktop use, which means that it comes with a default GUI and desktop environment.
As such there is a lot less (I'd say practically nothing) with which to differentiate a potential Haiku distro, which is also why I doubt we will see many. Particularly since the most likely reason for a distro would be packaging it with preselected software, however with a simple to use package manager that allows users to easily and quickly install software themselves, such a distro will not be very interesting to end users.
As I said, an "app store" or package manager of sorts is great, but I think crafting a well-designed distro would help to boost Haiku popularity, and I wouldn't (and haven't over the years) dismissed it.
While I personally don't see any reason for a Haiku distro, even with the bullet points you listed, you are of course free to give it a shot.
Re: Distro based on Haiku
Sure there could be a market for a distro. TuneTracker is developing one as we speak. They've identified a very specific niche market and are creating a product for it.
Or do this: write a bunch of hardware drivers for oddball devices that the official release does not support, then put a restrictive license on those drivers so that they can't be folded back into the main project. Create the office suite we've been desperately looking for for over ten years now. People will grumble, but they will pay. (Actually, they would pay you just for a good office suite, no need to fold a distro around it, but I'm trying to make a general point here.)
But nobody will thank you for changing the title and a few graphics. There have to be real advantages. And you'll have to keep it in sync with the nightlies. This community seems to be split between those still running Alpha 4 or even alpha 3 and those picking up a new nightly every week. Which of those groups are you aiming for?
Re: Distro based on Haiku
I know in a sense I've taken the focus of the post away from the OP, but I think this is an important thing to discuss for everyone interested in doing a distribution (or even a fork). After carefully reading the whole thread and the last few posts, I wanted to respond best I can in regards to what I wrote earlier if I may:
I wouldn't call a distro a 'fork', atleast not in terms of Linux as there is no 'official' Linux distro from which to fork, Linux distros all have one common denominator, which is the Linux kernel, 99.9% also use the GNU coreutils/toolchain, but the Linux kernel is always the same, give or take a few patches.
Also I'm not sure what you mean with 'forking for the sake of it'?
I wouldn't either. Notice that I have always called Gnu/Linux exactly that, not Linux, as I think that it's the proper title for it as an OS. Linux is only a kernel, but at the same time, a very powerful one that comprises a micro OS of sorts as a sole entity even without the GNU tools. I've looked at the Linux source myself; it's quite impressive, actually. However, as a direct example of how diverse Gnu/Linux can be, I consider Slackware and Ubuntu to both be an OS because the thinking is different, though internally, they share the same "roots". But Gnu/Linux (or just "Linux" if you prefer) is a controversial topic that I'm going to avoid so I'm not responsible for a flame war.
I'd be glad to answer your second question; sorry I wasn't concise on it before. What I meant by "forking for the sake of it" are people who make a distribution of some sort for pleasure, etc. and just quickly change the logo(s), change the look to their favorite colors and theme, add a few packages, and paste it on the Web. And, yes, I do honestly believe this destroyed the Linux desktop, as now there are hundreds of dead and active distros, with no hope of uniting with one dominant leader. We can never do this.
Failing? If Haiku would ever reach anything close to the amount of users the Linux desktop has I would consider it a massive success and be super happy, we are talking millions of users here.
I wouldn't dare post something without doing some research first to back it; Miguel de Icaza, one of the original developers for the Gnome project, noted what "killed" Linux as a desktop in 2012. The group behind the AbiWord project wrote a paper on why they felt Linux was failing a while back. And lastly, please see Net Applications or other firms' web and market share for Linux as a system group on the desktop--it does not hold much share in any view.
Undoubtedly, Linux runs the majority of supercomputers and mobile devices by market share, which can as easily be proven. Linux also doesn't lack people, either, as you mentioned. Ubuntu, Linux Mint, OpenSUSE, and Fedora, as well as others in the top ten have quite a few followers. But, compared to a world market, desktop Linux has been soundly beaten... for now. There's still hope. Whether the situation with Windows 8.x causes a major flux in this has yet to be seen, but I know that as of now, the statistics show the situation for themselves, though their true accuracy due to Gnu/Linux's open nature is under debate by some (i.e. are there more or less installs and users than polled?)
Just to make it clear, this was not something I proposed, it was instead the only thing that came to my mind in terms of what reason 'someone' would have to create a distro, I don't personally think this is a good reason to create a Haiku distro.
My apologies over my quick reply. I now understand you were just looking for a point. I also agree that simply "theming" Haiku wouldn't be a good idea, either.
This both means a lot of flexibility but it also means a less integrated operating system. A Linux distro will of course use Linux, and most likely the GNU toolchain, but apart from that it's very much up for grabs, you can ship with a complete desktop environment (...) or window managers (...) or without any preinstalled graphical environment at all etc.
I think these are good points any dev knows well. I like Haiku because it has the "look and feel" of Apple Platinum (UI style for MacOS 8/9) with *nix tools and a slick, integrated, and compact framework. Unlike Gnu/Linux, there's not a ton of layers, multiple window servers as of 2013, compositors, etc. Furthermore, I've authored a small desktop environment as a proof of concept, and I'm aware of the "layering" of Gnu/Linux. To be honest, I'll be thankful to leave that mess for something that I believe is cleaner and better.
Or do this: write a bunch of hardware drivers for oddball devices that the official release does not support, then put a restrictive license on those drivers so that they can't be folded back into the main project. Create the office suite we've been desperately looking for for over ten years now. People will grumble, but they will pay.
This is a commendable idea, but personally, it'd be against my ideals. We need a free environment, and unlike some of the Linux distros I've seen that do everything for themselves, I believe in community. If Newleaf improves on Haiku, I'd be very grateful if someone reused the code, etc. as long as they gave Haiku or any new distribution credit, as I think would be ethical and honorable to do. I'd like to be able to give back by allowing others to see and use it openly, and for anyone else's endeavors to do the same.
But nobody will thank you for changing the title and a few graphics. There have to be real advantages. And you'll have to keep it in sync with the nightlies.
I wouldn't thank myself for doing that! I believe that there must be real advantages for anything to be worthy of calling itself a distribution. Of course, there would be a new name and logo, but there would be several huge improvements for ANY free desktop--not just Haiku!
1. A web shell An app store/package manager is a GREAT start and I'm thankful for those inside and outside the Haiku team focused on it, but it's only the beginning! Clearly, I think that a dynamic web-based application framework is the future of the open (and closed) desktop. Firefox OS, Chrome OS, Android, iOS and Windows 8.x I think help prove this somewhat. Haiku doesn't have this... if it did, we would attract developers to a live environment fast, because there would be something they'd actually want to build their modern apps on. I'd also include tools to build simple apps of this class into the distro. Haiku simply does not have any of these capabilities in it at the moment. Without it/them, an OS is dead already. I'm shocked how much this is being dismissed as another feature! It's the whole face of the OS that will define the future... if anyone can see it. I'm highly optimistic about it and can't make this point any stronger.
2. Accessibility Can the system speak by default or read contents of applications? Can we easily control the zoom onscreen or scale the display for large displays like I can with my Mac? Do we have a way to easily use a virtual keyboard or easy-use controls? If the answer is no, then I think we need to get Haiku there to be a truly modern operating system. This would be a second area I'd seriously work on. Accessibility IS the heart of ANY system, and if we ignore these details, we've ignored how to make really awesome software.
3. The tidbits I think new colors and icons are a definite improvement for the modern crowd. Also, I would include a few packages; I think Firefox, VLC, and Qemu with a ReactOS image would be helpful. As an extra, I think using Grub or a base Debian image could allow EFI64 users to boot their machines; it's worth a try for now, as implementing a new way would quite honestly waste time and take away from working on user-related areas.
And finally, I would follow the nightlies, because I would want to keep up with any bug fixes and additions. I hope this answers everyone's concerns and questions. I really do want to say thanks to the Haiku team, and after reading the guidelines on considering a distro, I think I'm doing the right thing.
By nature, as open source users and developers, we're rebels, because we desire something better while preserving our rights. I think we also offer hope. And I hope I can get such a distribution out by Christmas... and, as a life goal, put a dent in the open source arena before I'm ever gone.
My apologies that this was so long; thank you all for reading what I've had to write! Please let me know what you think! :-)
Re: Distro based on Haiku
Miguel de Icaza, one of the original developers for the Gnome project, noted what "killed" Linux as a desktop in 2012.
Miguel de Icaza has held a grudge against the Linux desktop ever since 'it' (and the rest of the *nix ecosystem) turned down his beloved C#/Net port Mono, sorry, but he is hardly what I would call 'objective'. Also as I recall it he tried to put the blame on the Linux kernel devs, which was pure nonsense as they, unlike Miguel and company, are deeply against breaking user space functionality.
But, compared to a world market, desktop Linux has been soundly beaten... for now.
Actually I've never seen how the Linux desktop (or any other desktop, Haiku included) could ever be anything but 'soundly beaten', Microsoft owns the PC desktop through the dominance they laid the foundation for back in the days of their anti-competitive OEM deals.
Not even Apple while riding their massive wave of 'coolness', with their product placements in everything on television and box office, was able to put a dent in Microsoft's PC desktop monopoly.
Of course from now on the battle for the 'computing masses' won't be on the PC desktop, but on mobile (phone/tablet/whatever comes next, glasses?) so there is a chance that we will see a larger incentive among those who still use the desktop PC to widen their horizon as the makers of their OS of choice starts focusing heavily on the mobile sphere.
Still the PC desktop market is by all accounts rapidly shrinking, so there will be less users to fight for.
To be honest, I'll be thankful to leave that mess for something that I believe is cleaner and better.
Certainly there are drawbacks to the flexibility that loosely coupled components as those of the 'Linux desktop', such as the pieces often not 'fitting' very well, and of course that an application running on the Linux desktop can't expect that the end user will have desktop feature/functionality X.
Again Haiku, being written not only with the desktop as it's primary target, but also as a complete system where all pieces 'fit', does not suffer from those drawbacks.
And while I prefer Linux for many things (and I don't hate it as a desktop which is what I use every day), I would really like to use Haiku for my desktop needs, with the main barrier as of now being lack of software, but also system maintenance functionality which will eventually be solved as part of fully working package management.
As for your distro ideas, I don't feel the need to 'judge' them, if you want to pursue this then of course go for it, I would comment on the 'web shell' though, one of the areas where Haiku could become a 1st class citizen is of course on the 'web', as it is in itself platform agnostic as long as you can support the browser technology needed to display/run the served web content, is this what you are referring to (as in supporting existing/upcoming web frameworks) or are you suggesting the creation of some home-grown framework?
Re: Distro based on Haiku
Agreimann, if you can get us up-to-date ports of Firefox, VLC, and Qemu, you will be the hero of this community. No need to worry about a distro. Just get those .hpkgs out there. What you call the tidbits is what we need the most. The versions we have are ten years out of date. The people with the skills to port these monsters are too busy hacking on the OS itself.
I think you are underestimating just how much work we're talking about here. Porting FireFox is a major job. But hey, dreaming big is good. Go for it.
Re: Distro based on Haiku
This is really becoming an engaging discussion! Thanks for your comments! I'll see if I can respond best I can if time allows...
Miguel de Icaza has held a grudge against the Linux desktop ever since 'it' (and the rest of the *nix ecosystem) turned down his beloved C#/Net port Mono
Absolutely true. But De Icaza should earn respect at least as the original Gnome co-head with Mena nonetheless. There's no doubt Mono is a bad framework and that his argument was rather unfounded toward the kernel devs, but I wouldn't quit listening to what he had/has to write/say because of an idea. Current designers and developers for Gnome are rather brilliant, though Gnome Shell was a mess when it started.
Nevertheless, the Linux kernel can be a mess at times, when major transitions call for it or some new idea is proposed. I run Debian Gnu/Linux on my workstation, but even so, I think the direction Gnu/Linux is going is truly a mess. Wayland, Mir, and a host of other changes may be good for users, but they do nothing but keep software developers dependent on those changes constantly anxious and having to memorize new ways of writing (or rewriting) the code and doing things. Proprietary environments are stable, and while there are changes, they don't break modern software. After years of running and coding various projects on Gnu/Linux, I'm close to done with it. I'm sorry if that offends anyone; I genuinely don't mean to, but I think it's true.
Microsoft owns the PC desktop through the dominance they laid the foundation for back in the days of their anti-competitive OEM deals.
Not even Apple while riding their massive wave of 'coolness', with their product placements in everything on television and box office, was able to put a dent in Microsoft's PC desktop monopoly.
Of course from now on the battle for the 'computing masses' won't be on the PC desktop, but on mobile (phone/tablet/whatever comes next, glasses?) so there is a chance that we will see a larger incentive among those who still use the desktop PC to widen their horizon as the makers of their OS of choice starts focusing heavily on the mobile sphere.
I couldn't have addressed this better! Honestly--this about describes the situation. Remember the DOS wars, or Netscape? Worse yet, I do remember Dell couldn't offer an alternative OS outside of FreeDOS without losing benefits and being penalized. M$ is a monster, but the fact remains... they're here and that's the past.
Now, Apple actually did begin to gain more market share after Vista, but recently, their mobile division (the iDevices) is now what's the large area of sales for them.
I do think the future is a mobile one as well... I think we should look to the Raspberry Pi or other SoC alternatives and netbooks as the next step... for now. That's one reason why I think we need a lot of improvement in how the system appeals to people. Are we aiming for modern Web 2.0 developers, or the languages from the 90s? Viewpoint is everything that defines the "life" of an OS. And I do see the desktop shrinking quickly. Overall, we agree here completely as well.
As for your distro ideas, I don't feel the need to 'judge' them, if you want to pursue this then of course go for it, I would comment on the 'web shell' though (...) is this what you are referring to (as in supporting existing/upcoming web frameworks) or are you suggesting the creation of some home-grown framework?
I appreciate your consideration... although feel free to judge/critique my ideas and I'll listen best I can. Feedback is always important. My earlier post I think explains the web shell clearly, however. In short, it's not for supporting anything, but rather, to build something new. I think I've reiterated this point enough.
I think you are underestimating just how much work we're talking about here. Porting FireFox is a major job.
I've thought a long time about starting anything. I even discussed doing a distribution for a while back in 2011, but quit on it, as I thought quitting was the best option then. I think that now, with all the changes taking place in the world today and a less unified Gnu/Linux, there's no better opportunity to make things better and I'm not quitting on it this time. I don't think I've underestimated anything, but I do know it's going to be hard work.
Also, about Firefox... I was actually going to simply include the 3.0a release, as we'd need to port Cairo libraries (and more) for anything higher. That doesn't mean I've given up on such a task... it's that I've got to be open, honest and realistic about the situation. I can't promise impossible things like a politician, and porting a NEW version of Fx may or may not happen. But I definitely am going to try to get Qemu and VLC up-to-date for sure, as these are key for what I'd like to do. All I can write in this case is the usual "we'll see how it goes". I'm pretty sure, as I said before. that I probably won't have anything until Christmas. And about what you wrote about being a hero... I don't know, and frankly, don't care about that.
I just want to change the world. :-) Thanks all for the replies, and again, for reading...
Re: Distro based on Haiku
I have only see tiltos personaly. I think if haiku becumes more populer it might have "distros for differnt arketheures that the devlopers don't use for example a dreamcast or ps2 haiku distrobution. Also stop using the packagemanger as an exucse or defend your statements as they don't count as no one uses nightly images.
Re: Distro based on Haiku
Please... check your spelling if English is your native language, as your post was hard to read without saying it aloud. I don't mean to be a jerk... I promise I'm not a grammar nut and can live with mistakes, as no one's perfect... but several words were badly mangled.
In response to your post, one point at a time:
- I understand TiltOS is a good idea--but I do see a lot more needs to be done than simply porting X.org & apps. It has been successful, and if you're using it, wonderful. I can't argue my ideas will be better or worse, honestly.
- Haiku can already run on several architectures--please check HaikuPorts or the nightlies, or check out & compile a build yourself through subversion and try booting one. Be warned it probably won't boot without modifying it yourself.
- I'm not using the Package Manager as an excuse by any means. I've looked at it... it's a great concept that I think Haiku needs, and brings it nearer to post-2000 usage, but we don't have modern Web-style or next-gen applications. My stating that I believe the package manager wasn't the end of things is not an excuse... it's the truth. It's a brilliant start... and I definitely want to say thank you to all the developers that made HaikuDepot possible!
- Some people do surprisingly use Nightly images. I, for one, prefer Haiku nightly and like being able to use new (but unstable) features and capabilities.
I hope this explains everything.....
Re: Distro based on Haiku
Porting X.org is never the solution. Ever. And TiltOS didn't do that, they basically just recompiled a package manager and stuck it on Haiku.
[quote=agreimann]Haiku can already run on several architectures--please check HaikuPorts or the nightlies, or check out & compile a build yourself through subversion and try booting one. Be warned it probably won't boot without modifying it yourself.[/quote]
Subversion is dead; Haiku has not used it for a *very* long time. And HaikuPorts simply serves to build packages for Haiku, and in building the base system packages during the bootstrapping process -- AFAIK Haiku's configure and other tools use it automatically without user input (unless it breaks, which happens quite often because bootstrapping is rocket science...)
[quote=agreimann]I'm not using the Package Manager as an excuse by any means. I've looked at it... it's a great concept that I think Haiku needs, and brings it nearer to post-2000 usage, but we don't have modern Web-style or next-gen applications.[/quote]
Having "whatever's hip right now" is never Haiku's goal. If you want to port something, that's great, but don't expect that it will ever become a permanent part of the OS. It likely won't. But WebPositive on the nightlies is much improved, and HTML5 games are playable -- to some extent.
Also: Haiku's goal never was and never will be to attract developers. For R1 and R2 we have the goal of a BeOS clone with modern features; after that in R3, we'll break binary compatibility and do things "the right way" and not just because "everyone else is doing it and if we do it, we'll get more devs".
Re: Distro based on Haiku
Also: Haiku's goal never was and never will be to attract developers.
Huh. You might need to tell the Haiku project about their goals:
We hope to attract new developers to our project
This sort of goal has also been expressed by key Haiku developers at various times, should I dig all those examples out too ?
For R1 and R2 we have the goal of a BeOS clone with modern features; after that in R3, we'll break binary compatibility and do things "the right way" and not just because "everyone else is doing it and if we do it, we'll get more devs".
The Haiku R1 "BeOS clone" goal has been the same for almost 13 years now. Meanwhile not only have the goal posts shifted, the team moved to a different stadium and is playing a different sport. In 2016 announcing a new 32-bit desktop OS will be like announcing a new CRT television. A few fans will be excited, and everybody else will be momentarily puzzled and then move on.
Re: Distro based on Haiku
Having "whatever's hip right now" is never Haiku's goal.
We're "hip" since 1998!
The Haiku R1 "BeOS clone" goal has been the same for almost 13 years now. Meanwhile not only have the goal posts shifted, the team moved to a different stadium and is playing a different sport. In 2016 announcing a new 32-bit desktop OS will be like announcing a new CRT television. A few fans will be excited, and everybody else will be momentarily puzzled and then move on.
No User cares about 32/64Bit, if everything is working (Webbrowser, e-Mail, Mediaplayer, Solitaire and/or Minesweeper!!). ^^
Re: Distro based on Haiku
Meanwhile, I don't see you helping towards development.
As for developers -- if any come along that want to help the project, that's great! But we aren't going to change our goals just to attract users or developers.
Re: Distro based on Haiku
Haiku R1 has been decided to be a replacement for BeOS R5. If you don't like that, go use a Linux distribution. In R2, we'll switch to GCC4Hybrid (plus some bugfixes), which will happen a few months or so after R1.
We? Last I checked, you don't speak for the project, nor do you represent everyone else's opinion.
Meanwhile, I don't see you helping towards development.
And you are?
Re: Distro based on Haiku
I personally think a non-forked distro would be a great way to attract new users, reviewers, and developers. I also think a distro could easily co-exist, even 'sit on the shoulders' of the package manager.
Here's what a distro would have in addition to the 'clean install' of R1:
A good selection of games (selected from Haikuware)
A good selection of wallpaper (we should hire a graphic artist for some Haiku stuff)
A good selection of screen savers (our current selection is Lame-O, we desperately need better!)
A selection of the best utilities (we desperately need office productivity tools)
A selection of developer tools (Paladin, gvim, Go, etc)
This way when someone installs Haiku for the first time they will have a better out-of-box experience, they won't have to poke around and do research to find the good stuff.
A distro like this could be implemented using the package manager. All this stuff could exist as uninstalled packages that get installed by the package manager at start-up. We could even make some cool graphics to distract people while the packages install.
This methodology would be flexible and maintainable. When Haiku gets a new release, only affected packages have to change.
I personally think this is the best way to go.
Re: Distro based on Haiku
"Meanwhile, I don't see you helping towards development."
Waddlesplash, please meet NoHaikuForMe. NoHaikuForMe, please meet Waddlesplash. NoHaikuForMe is a long time critic of the Haiku project. He is a very bright contrarian. I find that even though he occasionally gets his facts wrong :-), he does have a keen understanding of OS fundamentals. Sadly, he's not a very productive developer and seldom if ever submits code :-)
Waddlesplash is a junior Haiku contributor and is release coordinator for A5.
Re: Distro based on Haiku
The Haiku R1 "BeOS clone" goal has been the same for almost 13 years now. Meanwhile not only have the goal posts shifted, the team moved to a different stadium and is playing a different sport.
This is as it should be. This project is not done with static 'waterfall' requirements and design. It's done through learning and adaptation and limited resources.
In 2016 announcing a new 32-bit desktop OS will be like announcing a new CRT television. A few fans will be excited, and everybody else will be momentarily puzzled and then move on.
Since Haiku is not a server oriented project there is little need for the benefits of 64-bit architecture, which as you know are primarily more RAM and security. Once Haiku is stable and has more apps that need more RAM, and has more users with more security issues, then emphasis and resources can shift to 64-bit concerns. The new package manager can actually be adapted to manage 32-bit and 64-bit 'fat' executables. Much like Mac PowerPC/68000 executables. There is even the possibility of a thunking layer.
Re: Distro based on Haiku
I'm also manager of HaikuArchives -- otherwise you captured me pretty well :-) and pleased to meet you, NoHaikuForMe!
As for my R2 statement -- true, nothing has been set in stone. However, I'm not the first one to suggest something like that, I recall Adrien (PulkoMandy) saying something similar to me in IRC.
In defense of 64-bit
No User cares about 32/64Bit, if everything is working (Webbrowser, e-Mail, Mediaplayer, Solitaire and/or Minesweeper!!). ^^
Chrome, Windows, Mac OS X, and OpenIndiana/Solaris ALL support 64-bit and transistioning to it completely will be as inevitable as the obsolescence of 16-bit. The total shift from 32-bit has not happened now, and probably won't happen for another 10+ years as mobile/embedded devices are still heavily using 32-bit architectures, but it IS occurring and it is important to look at the future, not the past.
Intel already offers the x64 Minnowboard Max, and mainstream computers are shifting now. As for the mobile or future market, the iPhone 5s introduced the collaborative 64-bit A7 and the upcoming Octa-core is 64-bit; it is safe to not only assume, but state, that ARM processors/architecture as a collective whole will eventually follow the transistion to 64-bit as well.
What is my point? x86 is still here, but it won't be forever. AMD64 and x86_64 on the desktop, notebook, and even on the high-end tablets (such as an EliteBook convertible), are becoming common. ARM will shift to 64-bit. What will that leave? In my opinion, it will leave an outdated 32-bit group of hobbyists with a cool system that only they want.
PLEASE NOTE: I do not wish to contribute to, nor start a flame war; I'm just stating a conclusion as it appears. In summary, Haiku and any derivatives MUST support more than x86 if they are to survive and be used by the general market.
Re: Distro based on Haiku
2. ARM support is also coming via GSoC.
Re: Distro based on Haiku
I would like to continue work on SENRYU. If any of You are interested PM me on haikuware.com [ http://haikuware.com/streakx/ ]
Old senryu build:
http://haikuware.com/directory/view-details/utilities/system-system-info...
Re: Distro based on Haiku
Since tilt os was mentioned... does anyone know where the source for the tilt OS ports went? Even though it's a bit messy, an X windows port would be handy!
Re: Distro based on Haiku
No User cares about 32/64Bit, if everything is working (Webbrowser, e-Mail, Mediaplayer, Solitaire and/or Minesweeper!!). ^^
Well it depends on *what* type of users you target, IMHO.
Nowdays targeting "general purpose users" is a mistake, to me.
I would love to have a multimedia production distro (something like Dyne:bolic - http://www.dynebolic.org/), instead...
Re: Distro based on Haiku
That would be restarting Senryu rather than continuing work on it...
I would suggest you wait for alpha5 and use that as your base. But yes, I think there is room for a DVD-sized "distro" with more software on it than will fit on a CD. Look at Ubuntu: you can get it in versions ranging from 40 MB command-line only up to the "several kitchen sinks" version.
Re: Distro based on Haiku
I would love to have a multimedia production distro (something like Dyne:bolic - http://www.dynebolic.org/), instead...
same (r.i.p. dyne:bolic). a media production os can still work in 32-but, though, especially considering must digital media recorders (cameras, audio recorders etc) are 32-bit and anyone not working in 4k raw video wouldn't need more. with the way the media kit works, efficient workflows can be built with minimal overhead.
Re: Distro based on Haiku
I don't mind work restarting/continuing on Senryu. That was quite a while ago, so it's probably not relevant anymore with all the changes made to Haiku. I lost all the sources and changes that made the distro compatible with Haiku's 'distro rules'. I do think it's a good idea to demonstrate what Haiku can do and am all for it, just don't have the energy/will to commit or help in such a project. I'd also be in favour adding a Haikuware repository to the package manager; restricted to the Senryu distro (certain Haiku devs consider software at Haikuware 'crap'). I just consider the software available at Haikuware 'use at your own risk'. If it works great! If it doesn't fine, who cares. It's better having it there than not - especially for research purposes to read what the application was capable of and to contact authors for sources. At one time most of the software at Haikuware worked on Haiku releases. That's why I am not sorting what works, what doesn't on Haikuware until a final R1 comes around. Even current package management software breaks all the time because of changes to Haiku, so it's hard to define what 'crap' is until R1 comes around (when and if that that comes)
Re: Distro based on Haiku
How, exactly, does one go about setting up a folder full of hpkgs on a server as a repository? Anyone have documentation on that?
Re: Distro based on Haiku
I find that even though he occasionally gets his facts wrong :-), he does have a keen understanding of OS fundamentals. Sadly, he's not a very productive developer and seldom if ever submits code :-)
Bane's voice: For you
Waddlesplash is a junior Haiku contributor and is release coordinator for A5.
By the way it seems a bit unfair to let a kid who is new to the project take the bullet for Haiku's ongoing logistical problems.
After the R1 alpha 4 mess a post-mortem was conducted. That should have directed, or at the very least influenced, the planning for R1 alpha 5. Most obviously it was clear that R1 alpha 5 needed to be out the door in 2013, but that's not possible now. Still there are other lessons and it's unfortunate that so far there's no sign that anybody actually learned anything.