Can a distro of Haiku be made in the future?

Forum thread started by apgreimann on Wed, 2010-09-08 03:56

Hello, all. Earlier, I had posted on allowing WPA Wi-fi to work on an Aspire One(tm) netbook, and found that Ethernet is much more stable--the netbook's now got Haiku dual-booting on it happily--with Internet. :)

I absolutely like the Haiku, and the simple UI it presents. The developers did a fantastic job, and hours of work were clearly put into it. However, could I be brave enough to put forth a simple question? Could Haiku ever be "re-spinned" or forked into a distro by *anyone*, like how Linux is? (I know Haiku's MIT, and Linux is GPL.) I think if re-spinning was allowed, it'd increase Haiku's usage share, as it'd allow more users to get aquainted with it quickly. It'd be as popular or even more popular than Linux, if so. :)

(And, if I may, can I add a sub-question to that? Does anyone know what causes the PC to not turn off after shutting down? Again, I'm on a netbook--an Acer(tm) AOA150. Maybe others are having this issue, too, so I thought I'd post that. :D)

Personally, I've dreamed of re-spinning like in Linux, but it'd be cool if anyone could do it. Is this something that can be done in a Linux-like fashion? If not or so, can you please post back?

Thanks for reading! :)

Comments

Re: Can a distro of Haiku be made in the future?

It is certainly possible to make a distro, as there are no restrictions about it in the license. But I fail to see any point in this. Most Linux distros aren't that unique, they're all pretty much different combinations of the same components and packages. What is the difference between Ubuntu and Fedora? Different bugs, different default apps, different themes and different package managers, mainly.

How would creating distros increase Haiku's usage? I don't think it helped Linux. Everything about Linux is scattered, there are no standards, they don't even have a unified sound API that works Ok and is used by most applications. Every year is apparently "The Year Of The Linux Desktop" but yet Linux isn't really getting anywhere. The amount of users are still very few.

Oh and, I will look forward to the day when the Linux desktop has 32bpp support. :P

Re: Can a distro of Haiku be made in the future?

I hear ya, sparklewind. :) That's so true. However, as much I do not like Linux, Ubuntu has allowed more non-programmers to use it and has stepped up Linux's market share by a few points. It'd be *really* cool to see Haiku race right past Linux.

I might add a comment. Why can't the Linux community figure out that people might just want it to be one OS in looks and performance? You do have a very good second point. :) A Fluxbox WM on one, a KDE-based desktop on another, and no true GUI intregration. *However*, Haiku's got a good UI, a good base, and is truly an OS that presents a spun-together front to it--it stands a much better chance than Linux does, in my opinion. ;) That's why I think building deriatives of Haiku would be a cool idea, and I think it's cool to hear it's totally legal, too. :)

Let's use something that ain't GPLed and show the world we can! (Yeah, I know, Haiku's got a little in it anyway.) :D

Thanks for replying!

(P.S. There have been posts and tickets on PCs not turning off--I'm not the only guy that has a shutdown problem on his PC. How can I determine ACPI/APM status or toggle it on/off in Haiku?)

Re: Can a distro of Haiku be made in the future?

Well, I just don't see what any distro would be for. Could you elaborate a bit on this? Just like with Windows, we might one day see people bundling Haiku with a bunch of applications in one ISO and putting out a torrent somewhere. Then there's TiltOS, which seems to be a project to make Haiku more Linux-ish in several ways. The former seems too insignificant to make much mention of, the latter is a bit too much. So I am not really sure what kind of distros should be encouraged. I would like to hear more from you about this.

I agree that Ubuntu has helped Linux to some extent, and it's really popular among Linux users. When I use Linux, I stick to Ubuntu too. But is this because of the distro idea? They could've improved Linux directly, if Linux were a complete OS. :)

I have no idea about your problems with shutting down your PC, sorry.

Re: Can a distro of Haiku be made in the future?

Yes, distros are allowed. Nothing to stop them from coming around.

I don't mind a couple of distros but hope Haiku does not become flavor of the week with lots & lots of distros.

I believe that when Haiku does proper package management the need for distros will decrease.

The only reasons for a distro are: a) programs installed b) package management c) artwork differences d) compiler optimizations (ie: use O2, remove debug, etc.)

I would hope that any code fixes or optimizations were put back into Haiku.

Shutdown issue like this one I'm guessing (ACPI related):
http://dev.haiku-os.org/ticket/5880

Look through here for others:
http://dev.haiku-os.org/search?q=shutdown

Re: Can a distro of Haiku be made in the future?

"If Linux were a complete OS. :)" :D

There's a *really* good point, and one I agree with. Linux is NOT complete, and is not stable for the end user. All it is is thrown together parts, borrowed from every software that's called "free", but I don't think Linux gets it. For instance, on a basic level, gnu programs are used, Fluxbox is used, and the X Window System is used--all in the same system--and we're only skimming Linux here! I think that the *Linux* programmer, who thinks Linux is the best OS that ever came out, is really Linux's audience. And if they keep insulting Windows(r) users, and not respecting them as human beings who are using a grossly overpaid buggy S/W, then they're not going to be very successful.

My first idea was to build a small shell, and build a Linux distro. After seeing how unstable it was, and especially seeing the fact I'm *NOT* a die-hard GPL fan, I began to look into Haiku. I'd already begun using it, and really like it. It's as stable and simple as you'd find on a Palm(tm) or a Mac. I thought, combined with it's liquid performance and stability, it would be possible to use it as a base in the future, per-say. Of course, I understand anyone else could as well. I didn't just post this for myself, but hopefully for others to get the concept. Open source is about helping others with what you know, and with what others know. It's not about one OS, one license, like the Linux community thinks. :)

With that in mind, I was thinking to make things plug-and-playable, and build a very simple HTML shell people could use like they use webpages. Why HTML, and not C? I'd like to go for simplicity, so that anyone could pick up how to contribute to it without crashes, dependencies, bugs, etc. There are more complicated programs, but as programmers, we're missing the fact that any open OS must cater to the user, and not us. Haiku's got that vision, and I believe it's because they have great programmers who understand how an OS is done. :) If only Linux'd catch on and take a look at Haiku, instead of being arrogant about Linux....

(As for the PC, I'll try to find out why it just blanks the screen at shutdown. It's a bug somewhere--but it's the only major bug I've found in Haiku so far. :) And I know the real release will have it fixed. :D)

Thanks for replying, sparklewind. That's my thoughts.

Re: Can a distro of Haiku be made in the future?

sparklewind wrote:

It is certainly possible to make a distro, as there are no restrictions about it in the license. But I fail to see any point in this. Most Linux distros aren't that unique, they're all pretty much different combinations of the same components and packages. What is the difference between Ubuntu and Fedora? Different bugs, different default apps, different themes and different package managers, mainly.

In most cases Linux distributions have a clear focus which makes one distro more suitable for a particular purpose or user than another. Fedora has a short life cycle and concentrates on "leading the consensus", showcasing new Free Software that often goes on to be universal. It acts to some extent as a preview of what's coming for RHEL, which is important for the large number of people who administrate or develop for Red Hat's enterprise products. Ubuntu has a mix of long-lived and short-lived releases, and concentrates on the desktop experience almost to the exclusion of other features.

Quote:

How would creating distros increase Haiku's usage? I don't think it helped Linux. Everything about Linux is scattered, there are no standards, they don't even have a unified sound API that works Ok and is used by most applications. Every year is apparently "The Year Of The Linux Desktop" but yet Linux isn't really getting anywhere. The amount of users are still very few.

Realistically distros would indeed fragment Haiku. But with Linux such a big success across so many fields obviously there is room for more than one distribution. The people doing hearing aid research had a custom Studio64 variant made for them, such software won't even run on a 32-bit CPU, while something like Puppy Linux runs nicely on a Pentium II.

I think "third place" (Linux distros are used far more "on the desktop" than anything except Windows and OS X) in a globally competitive market is rather good. Millions of users, "on the desktop", and great success above and beyond that, in everything from smart phones to supercomputers.

Quote:

Oh and, I will look forward to the day when the Linux desktop has 32bpp support. :P

I'm not really sure what this means. Maybe you're confused by the 24bpp consolidation? This seems very long ago now, but perhaps it's fresh in your mind. X.org got rid of the 24/32 distinction which had caused confusion to users. All 8-bit per channel video modes were treated identically, and pixmaps with either 3 or 4 byte pixel strides were automatically handled for either kind of mode. Where enough VRAM was available for a 4 byte pixel, it was generally used because it was faster.

Re: Can a distro of Haiku be made in the future?

Thank you, tonestone57! I had located this ticket earlier. I always run a search engine before asking questions, as I think it's more polite, and saves time on the forum. :) This certainly helps :), however, I'd like to find the terminal command to toggle acpi support on or off, and see if that works.

I also agree here too, tonestone57--Haiku cannot become 400+ distros like Linux is--that's just ridiculous. :) We need more like, 2-10, or somewhere around that value. That way Haiku remains the flagship OS that'd start it off.

Thanks, all, for your replies. It's appreciated.

Re: Can a distro of Haiku be made in the future?

No please no DISTRO!
This is, indeed, the cause why Linux is not "ready for the Desktop"... if all the resource was used to create ONLY one OS what they could have in all this years years?

Now we've a no complete OS, 10 ways to have a GUI (and is a software a-la windows 98), when you have to do some configuration we have to use the shell (yes sometimes there are GUI applets, but simple not works), Kernel incompatibility, binary incompatibility!

They use the name "distro", but they are, in reality totally different OS...

Ubuntu has tried to go the Desktop way, but sometimes the GUI apps not work and you should use the terminal... and the very Linux haker hate it! If you want do a more professional work they force you to use CentOS (and break it! Forcing installing library as it is 10 year old... to do more modern things! At the end the same thing Ununtu guys already done... well but CentOS is more professional (?))

No Haiku should be ONE no forks: they are lost resources!

Can be acceptable to have different version for specify usage (that you could compile yourself):

  1. Embedded edition (<64 MB usable on set-top-box or smartphones!)
  2. Kiosk/Mediacenter (Only an application in foreground, no tracker or desktop... can not killed in any way! It's a kiosk)
  3. Desktop (The classic OS release, tracker, web+, future office, Mediaplayer etc...)
  4. Developer edition (What all is needed to create application or compile Haiku itself)

Obviusly all this "distro" (if you want call in this way are in reality "edition"!) are binary compatible, and certain ways upgradable (if you have the desktop and to upgrade to developer you can download the "developer Bundle", double click on it you are set!).

Re: Can a distro of Haiku be made in the future?

you can use safe-mode options to test with (it allows turning off ACPI) - while booting hold down SHIFT key or keep pressing SPACEBAR key. ACPI is on by default.

You can also edit kernel config file found under:
/boot/home/config/settings/kernel/drivers
Use text editor (StyledEdit)

Re: Can a distro of Haiku be made in the future?

Thanks, tonestone57. This is what I was looking for. I'll give it a shot. :)

Re: Can a distro of Haiku be made in the future?

@ fano: Thanks for your input. I honestly do appreciate your post, and I like looking at several different points of view on computing and operating systems. :) However, may I reply to a few comments?

Firstly, Linux couldn't have done this (that is, build a strong, full OS)--it needed a kernel and GUI shell, and more. Linux started as Minix as a command-line environment, and for the most part, still is :), underneath all the eye candy. Taking into consideration how Linux is built would better help you understand why. The core of Linux rests on top of the kernel, and from there, a command-line or the "shell" awaits it's user. To fix this, the X window system is thrown on top, along with a login manager, along with a GUI shell, and along with applications all requiring hundreds of dependencies. Your average Ubuntu installation with KDE or Gnome requires at minimum, 800 packages installed--most dependencies. This is due to applications being written in at least 5 different languages, and once that's done, on top of that, they are compiled as Gtk+, QT, or another application type. This is what makes Linux so unstable, and therefore, this both answers your question on why applets don't work, for instance, and destroys the previous argument that hackers prefer the GUI over the CLI. The CLI is how the base of Linux can be accessed. Furthermore, the programmer does use and enjoy the CLI in *any* OS, including Windows(r), as it gives them full power over their system in situations where the GUI was not designed to. Personally, I think an OS is only half an OS without a text interaction of some kind. :) I strongly disagree that the programmer prefers all GUI--but I *really* think the GUI is the future. Users will not use a CLI. They want a simple, point-and-click interface that they can understand without hassle. Programmers should work to make the shell integrate for the user.

Secondly, Linux is and isn't professional. Linux runs on the world's most powerful supercomputers, and on most of the smartphones today, thereby making it professional in that field. On i386 or powerpc, however, as mentioned, Linux is *not* professional by any means--I do not consider Ubuntu or CentOS to be professional. Linux users need to be honest, and realize this is the case.

Thirdly, you had mentioned editions, which *would* partition the OS. Two computing history examples serve to prove this, respectively. Apple(r) made the Centris(tm), Quadra(tm), Performa(tm), Power Macintosh(tm), and other desktop computer products. Until Apple(r) brought the computer lines into four main product lines, and three classes of computer (notebooks, desktops, mobile) users were confused and sales were down until the introduction of the iMac(tm). Recently, Windows(r) has made the same mistake, splitting itself into Starter (that is, the netbook edition), Home Basic, Home Premium, Professional, Ultimate, x64, Embedded, Server (which I agree with--that port makes sense), and of course on the side, Phone. Vista(tm) did poorly because of this, and Apple(r) gained a huge market share in 2008 and into 2009.

Fourthly, why not make a distro? OpenSolaris(tm) has about 8-10 distributions, and all it's users recognize the fact that it is the same OS with different aspects to it. This would help the OS by opening new ways of thinking how to build it and fix bugs. That way, everyone'd benefit.

In conclusion, I think a distro would be a great idea, but that's just my opinion. I appreciate the post and I honestly do like hearing different points of view--I just wanted to clear up a few points here. I apologize if I was rude.

Thanks. :)

Re: Can a distro of Haiku be made in the future?

Yes I know the Linux history... as you have described for obtain a complete OS you have to mix all the pieces: a Kernel (you hav 1000s to choose you can personalize itself, too!), a shell (there are 1000 of these, too bash, tcsh, ksh...), X (there was at least 2 version... nowadays only one is largely used), a Desktop Enviroment (1000's of this too... KDE, Gnome, FCE...) and in the and a package system (rpm, deb...), it's pretty obvious in this way the different factor is huge: Fedora and Ubuntu has so little in common that can be considered 2 totally different OSses!!!
Neither binary compatibily is garantueed...

One can think well this a freedom it's beautiful... I can create my own OS! Yes but in this case a think as this freedom is bad: it is for this Linux is "not ready for the Desktop" first what Linux? There is no "Linux"! There is Ubuntu, Fedora, CentOS...

Why have 2,4 o 1000 Haiku distros can be good for us?
If the difference are so little: a simple name change and some package installed instead of another can be OK, but well why I should use ArcaneOS instead of the original Haiku?
Why not add that functionality at the real Haiku codebase instead of do a fork?

But if the distro take the Linux way? They modify HUGE the kernel, leading to binary incompatibly, use a different system to install packages (I don't know Haiku use "bundle", ArcaneOS uses RPM!), they choose to substitute Traker with KDE (I know, I know... they should be masochistic)...

In this way we should be in the same situation of Linux: Haiku is 1000 different OS, UserBase and worst coders are fractionated no chance to substitute (seriously) Windows... will be a "toy OS" as Linux, Aros, React OS etc...

No I boldly hope there is never an Haiku distribution... "edition" can have sense , indeed (endorsed by the real Haiku Inc... if in the future I want to install Haiku in a smartphone or in set-top-box instead of Linux I should have something different from the "Desktop Edition" for start little size ideally 64 o 128 MB to put it in a flash drive, a minimal tracker, only a small number of application, busybox...)

Re: Can a distro of Haiku be made in the future?

Linux is prely a kernel. Which means everything else required for linux to function must be built on top of it, and because there's so many alternative packages, distros were born to create a complete OS.

Haiku IS a complete OS. There's no reason to add packages on top of it, as the default system is feature-complete. There's no reason to have many distro's, as Haiku works out of the box. Because distro's arn't required, there will be very few, if any, big Haiku distributions outside Haiku Inc's official one.

Re: Can a distro of Haiku be made in the future?

One reason to have distros for Haiku is fat versus thin. Many folks want only the basic applications, nothing more. Truly, there are heated discussions on the developer mailing list about whether or not to include vim, which isn't that big. Other folks want apps and libs, the more the better. This article describes some very nice apps that would go into the 'fat' distro:
http://www.osnews.com/story/23445/The_State_of_Applications_On_Haiku

And there are more libraries here:
http://haikuware.com/directory/view-details/development/class-libraries/...

There are more apps coming each week, and this will accelerate as be bring in more new Haiku developers.

What's really different about fat and thin Haiku distro idea is that each has the same kernel, same window system, same demos, etc. It's really not a 'split' or schism like most traditional Linux distros.

Hell, we have so many games now we could have a Haiku Gaming distro.

Re: Can a distro of Haiku be made in the future?

The questions was really about if distros are allowed. Yes they are. Anyone can make a distro if they want to but must follow the guidelines and avoid using Haiku trademarks. MIT-BSD license allow it.

Some may like distros and others will not! That is life.

Personally, I can see a few distros being a good thing but too many a bad thing. No way to stop them from coming about.

Re: Can a distro of Haiku be made in the future?

As a end user I would prefer "as would most" to have the option of what to install. I don't want to have to track all over the internet to find it and the dependnacys. Make the disk fat and allow the user to select software groups or applications to install.

Just a user chiming in.

Re: Can a distro of Haiku be made in the future?

I have read the posts so far carefully here, because I certainly think it is wise to. :) I think listening allows learning.

@fano: I'm well aware of the Linux history, but like I have mentioned earlier, it's audience will never really be the end user, unless it is on a cell phone or generic device because Linux is not a complete OS. I've said before, and I can state again that Haiku *is* a complete operating system, which comprises it's pieces similar in fashion to the MacOS(tm). Linux started as a kernel cloned from Minix, and it really, in 18 years or so, hasn't gotten much further in terms of use, other than a bunch of eye candy in Gnome and KDE. True, there are very different ports of it, but underneath, Linux is just Linux. And again, I believe that editions would cause the end user to get confused as in what to choose. 2 or 3 might be fine--like if Windows(r) were just Corporate, Home User, and Mobile, it'd be easy to figure it out. *However*, NT did start with two editions, "Workstation" and "Server", and the home user had 98. This fanned out to the editions that they have today. Simply put, I think we can learn from this.

@thatguy: That'd be a great idea to implement in the software. Making things plug-and-playable would make the experience as simple as installing add-ons in a browser. :)

As for the whole distro idea, I can see points where it could be great, and also where it could not be so great. It sounds like there's different opinions on it. :)

And, once again, as I've repeated before, I think that a distro could have some use. However, this is only if it'd help out the Haiku in expanding. Bringing great concepts to life, along with God on their side, is what's made inventors like Edison innovative.

Re: Can a distro of Haiku be made in the future?

I believe that the term distro only really fits that well when talking about Linux. I just can't imagine a distro in that kind of sense for Haiku. However, it might be fitting to talk about editions and forks.

I don't think there's anything wrong with editions. The difference between third-party editions of Windows and distributions of Linux, is that Windows is still the same system in just about every way, they just add their favourite applications, wallpapers and settings to the ISO and sometimes there's a new installer. There's often not even any official website, you just see a new torrent once in a while on The Pirate Bay. Provided that you don't violate the terms of any license or law (Obviously the case with Windows, but not necessarily with Haiku), I think this is completely fine. This wouldn't get much attention, but it would still be out there for those who seek it out and are interested.

Then there are forks. Syllable was forked from AtheOS, Haiku's kernel from the NewOS kernel if I remember correctly. Basically you take source code with potential and make it into a new OS. This is also fine, and they have no real affiliation with the OSes and kernels they were originally forked from now. In the same way, one could, say, take Haiku and Étoilé and make a new Mac OS X-like system. You'd have a new OS.

No matter if you make an edition or a fork, the point is that Haiku is Haiku, and not Haiku Hotsprings Distribution, Haiku Cat Distribution, Haiku Coffee Distribution, and so on. I hope I was clear and my point got through and made sense. ;)

Re: Can a distro of Haiku be made in the future?

sparklewind, once again, I agree with you on this. :)

I was using the term 'distro' so that any Linux programmers or users could catch my meaning. But, yes, I was referring more to a 'fork' of Haiku.

However... I still think chopping any software up into editions may not be the smartest idea, unless Haiku in the future just stuck with Desktop, Server, Mobile, to keep things simple--if nothing else, for users. Or, 32 and 64 bit? The names here, for instance, would be self-explanatory. And, I know OS X has Server and the retail version--two editions, and I agree with the idea there--*much* simpler/better than the half a dozen Windows(r) editions. :) But, isn't Haiku already at where it's powerpc, i386, mips, and arm? Maybe I'm wrong on this.

I don't know... what do you think?

Re: Can a distro of Haiku be made in the future?

I think more distros are a good thing. Not more for the Desktop, but different distros for different needs. Like one for smartphones (I think it would be a great smartphone os), one with an interface adapted for netbooks and tablets, one for game consoles, one for digital media receivers (like Google TV and Apple TV), one for digital video recorders, one for big-screen tv's, one for set-top boxes etc.

Re: Can a distro of Haiku be made in the future?

My dream phone would be a Droid X running a mobile version of Haiku.

Imagine the apps that could be made for a mobile Haiku considering it's multitasking and multimedia capabilities.

Re: Can a distro of Haiku be made in the future?

thatguy wrote:

As a end user I would prefer "as would most" to have the option of what to install. I don't want to have to track all over the internet to find it and the dependnacys. Make the disk fat and allow the user to select software groups or applications to install.

Exactly. The user should be able to decide at install time how fat his system will be, and be able to add/remove parts at any time later.

Maybe even provide a download service where you can cherry-pick which packages should be included on your install media, and the anyboot image is assembled on the fly?

By making Haiku flexible in this regard, and also treating bundled and 3rd party software very similarly, the need for "distros" and "editions" will be almost non-existent.

Re: Can a distro of Haiku be made in the future?

agreimann wrote:

sparklewind, once again, I agree with you on this. :)

I was using the term 'distro' so that any Linux programmers or users could catch my meaning. But, yes, I was referring more to a 'fork' of Haiku.

However... I still think chopping any software up into editions may not be the smartest idea, unless Haiku in the future just stuck with Desktop, Server, Mobile, to keep things simple--if nothing else, for users. Or, 32 and 64 bit? The names here, for instance, would be self-explanatory. And, I know OS X has Server and the retail version--two editions, and I agree with the idea there--*much* simpler/better than the half a dozen Windows(r) editions. :) But, isn't Haiku already at where it's powerpc, i386, mips, and arm? Maybe I'm wrong on this.

I don't know... what do you think?

The many official editions might indeed become a concern. Right now we have 4 editions for only one kind of system on one platform (x86 32-bit); gcc2, gcc2hybrid, gcc4, hcc4hybrid. This will most likely get very frustrating once we throw 64-bit, the Clang compiler and different hardware platforms into the mix. Personally, I believe we should settle on one compiler and add optional compatibility libraries, and then make 32/64-bit hybrid builds. This would leave us with one build for each hardware platform (Maybe two, Desktop and Mobile, for ARM), which shouldn't be too confusing for users. I can't say how the developers will handle this though, as there seems to be a lot of disagreement about this issue.

When I said I find it Ok with alternative editions, I didn't mean that we should officially make editions with extra apps and stuff, but rather that if a somewhat anonymous person decides to make something without bringing too much attention, then it wouldn't affect us who just download the official builds. You'd have to seek the unofficial editions out if you're interested in them, and if you're not then they're easy to ignore. ;)

Re: Can a distro of Haiku be made in the future?

sparklewind wrote:

The many official editions might indeed become a concern. Right now we have 4 editions for only one kind of system on one platform (x86 32-bit); gcc2, gcc2hybrid, gcc4, hcc4hybrid.

The nightly images are purely for development related purposes and are not official releases of Haiku (eg, R1/Alpha 1 and R1/Alpha 2 are released only as a gcc2 hybrid as that is the official type).Granted the nightly images still bear the HAIKU development logo -- perhaps that should be changed so that only gcc2hybrids contain it.

As for the 'editions' or 'distributions', there's still some ambiguity on which graphics can and should be used for them. According to the current Haiku Distribution Guidelines, the HAIKU logo (and other trademarks) cannot be used. Instead a HAIKU Compatible Logo may be possible to use.

Basically, before creating any 'editions', 'distributions', 'remixes', ...., it'd be best to discuss it on at least the General Mailing List.

Re: Can a distro of Haiku be made in the future?

mmadia wrote:

The nightly images are purely for development related purposes and are not official releases of Haiku (eg, R1/Alpha 1 and R1/Alpha 2 are released only as a gcc2 hybrid as that is the official type).Granted the nightly images still bear the HAIKU development logo -- perhaps that should be changed so that only gcc2hybrids contain it.

Wow, I didn't even think about this! I guess I am just too used to using the nightly builds. Well then it's not quite as complicated as I thought.

Re: Can a distro of Haiku be made in the future?

I'm optimistic about the Haiku. :) And, I think there's a lot of good ideas here, too.

So far the best ideas have been to not make the same mistakes as Linux (obviously) :), and to allow plug-and-playable apps in the system. I'm thinking to work with the gcc2 compiler (maybe we'll have an MIT C compiler someday.) ;)

Re: Can a distro of Haiku be made in the future?

agreimann wrote:

I'm optimistic about the Haiku. :) And, I think there's a lot of good ideas here, too.

So far the best ideas have been to not make the same mistakes as Linux (obviously) :), and to allow plug-and-playable apps in the system. I'm thinking to work with the gcc2 compiler (maybe we'll have an MIT C compiler someday.) ;)

I understand the feeling, I'm optimistic about it too. :) I just saw some of the work that has gone into MediaPlayer recently and it's looking good. Once we reach the final release, there probably won't be any basic functionality that isn't covered by the applications that are included with the OS! :D Not even Windows beats that. Most of the applications included with Windows are annoying and quickly get replaced.

Yeah, GCC2 is the best way to ensure binary compatibility for probably several years to come.

Clang uses a license similar to the MIT license: http://en.wikipedia.org/wiki/University_of_Illinois/NCSA_Open_Source_Lic...

Clang looks really interesting and hopefully it can replace GCC in the future. :)

Re: Can a distro of Haiku be made in the future?

Thanks for the link--Clang looks interesting. :)

As for the post, for everyone who has kindly taken time to reply, thank you. I've learned a lot from reading, and yes, for those wondering, I definitely admit I've commented a little too much. :)

Re: Can a distro of Haiku be made in the future?

Unfortunately, Linux can't be thought of as an single operating system, but more like a collection of operating systems. The fragmentation, the perpetually reinvention of the wheel, the poor programming and design decision, the poor integration, the poor implementation (eg. some gnome's features and components are written in python, for fu.ck's sake!) have plagued Linux and will continue to plague it until the end of time.

That's because there's no central authority, there's no team to coordinate and synchronize the different components of an OS. Unlike BSD's, Haiku, Reactos. SkyOS, Syllable, Qnx, Windows, Os X, QNX, you name it. More than that, Linux ecosystem isn't only a collection of software written with Linux in mind. Eg. Apache wasn't written specifically for Linux, MySQL, gcc, X etc.

Linux as a kernel is a wonderful piece of software. In fact, performance-wise has beaten all other known operating systems in benchmarks.

If "Linux" was led by a team instead of being an anarchy, if people worked on implementing new and better functionalities instead of reinventing the wheel, if "Linux" could have been an operating system instead of an "ecosystem", I think Linux would have been the world's most advanced operating system today. Server wise and desktop wise.

Unfortunately, most free operating systems, even if led by a team, have chosen "the linux way": instead of developing software tightly integrated with the OS, they chose to port software. I don't see a big difference between linux distros and freebsd, netbsd, open bsd. Beside the kernel and a few packages, all use the same software: gnu gcc, gnu binutils, X, Gnome, KDE, etc.

Maybe, this is the UNIX way...

The only UNIX variant that did a better job is Os X. But Os X isn't free. And although it did a better job at software integration, performance wise is a mess...

Re: Can a distro of Haiku be made in the future?

The problem with Linux is with so many distributions, you don't know what to go for. I don't really understand Linux frankly. With one specific version of Haiku, you know exactly what to expect. I'm sure it would even be possible to make a version of Haiku for a different class of PC hardware ... a PDA for example. But until then, Haiku is the only open source operating system that has more potential in actually overtaking Linux in market share, it's even very well publicised at the moment (I first discovered Haiku after seeing a multimedia OS called BeOS in a Mac OS History video, surfed it on the internet, then came across Haiku).

Re: Can a distro of Haiku be made in the future?

Quote: "The problem with Linux is with so many distributions, you don't know what to go for. I don't really understand Linux frankly."

@Papsie: Studying computer history would help out--here's a brief overview of Linux for you. :) Understanding operating systems is understanding the fundamentals that run the operating system--such as the kernel that the OS is built on, and the source code behind it--this is why open source software is better. It can be studied by the computer programmer, or even the end user, who can then study it to better understand computing, coupled with the study of computer science. Closed source hides the code, so only company employees know how it ticks. :)

Firstly, the world, regardless of the distribution name, generally calls 'distros' "Linux", and around this there is some controversy, because of two naming differences. Most will state that Linux is purely the kernel which allows different operating systems to emerge from building around the kernel. Collectively, it is called "Linux" or "gnu/linux" by different open source fans. This naming convention differentiation came from the building of linux and hurd--at the time, the hurd kernel was not mature enough to manage the OS, though the gnu community wished for all gnu software; as a result, the kernel for the OS became linux and the software was bundled out of gnu products. Some argue it is just "linux", whilst the gnu community argues "gnu/linux."

The software, however, I'll refer to as Linux--but you need to remember that Ubuntu, Fedora, and even Palm's webOS, Google's Chrome, and Android are all "Linux" but are all different operating systems by definition--some will use different package managers, computer languages, and ports to platforms. Obviously, anyone can choose and download a Linux distro--but because this is a brief overview, here's how it's built, and why I don't like Linux (I'll still use Linux a little, though.) :)

Firstly, you start with "Linux" itself by grabbing and compiling the Linux kernel in an active operating environment. You will then need to compile and add the coreutils, write in options and driver support into the kernel, add a package manager, and you will also need to choose what shell you wish to operate from (e.g. bash) and from here, if everything is successful, you can throw the small OS together and make the squashfs and image it into an .iso. Secondly, write it to CD or a USB drive and fire it up. Hopefully, you can at least connect to Ethernet, because you've only begun! You don't even have a GUI (graphical user interface) yet--unless you are a real geek, and honestly like the shell. :) To start, you'll need to use your new package manager and download the X Window System, along with fonts, and if you don't like the default, a cursor set. Of course there's a lot of options. Once that's done, what window manager do you want? Of course there's a lot of options--more than a dozen. What desktop environment do you want? Of course there's a lot of options. Then, you'll need to download apps. Of course there's a lot of options--up to 32,000. :)

Summing it up without going further writing what has already become a gigantic essay :D, Linux is not a complete operating system, like how arrogant Linux coders and fans look at Linux. And because of the vast array of choices at every corner, it has almost made Linux unusable for the average end user, who can't even figure out which distribution out of at least hundreds that have been developed to use! I would honestly dare the Linux fan to launch an application (written out of 1 or 2 of more than a dozen languages!) from the shell, and see if it opens cleanly--most will not--this is not a feature of good software.

However, for the entry-level user, most download Ubuntu or LinuxMint. Does this clarify your questions about Linux?

Haiku is a great operating system--I just can only hope that Haiku gears up for the end user, because it honestly needs to in order to be a great operating system people will want to use. Hope this was not too long. :D

Re: Can a distro of Haiku be made in the future?

What I would like to be done with Haiku in the future, is for all the necessary drivers for all the applications to be included in the OS, so that the end user doesn't have to do anything more then extract the zip, dc on the file, opens instantly, job done. I can't even get mplayer (with all the gcc4 plugins) to work, among the games. The trash is an absolute nightmare to operate especially on a USB thumb drive.

Re: Can a distro of Haiku be made in the future?

@Paspie: Could you expand on what you're saying here? It would help.

Re: Can a distro of Haiku be made in the future?

Right, when I unpack the gcc4 libraries to /boot/ it ends up sticking the files in a folder within the directory, and I'm not sure where. The app itself is just fine, with the app ready to use.

Re: Can a distro of Haiku be made in the future?

Here's what I meant when I said "Could you expand on what you're saying here? It would help." I was mentioning for all the details you had posted if you could explain:

a) "What [you] would like to be done with Haiku in the future, is for all the necessary drivers for all the applications to be included in the OS, so that the end user doesn't have to do anything more then extract the zip, dc on the file, opens instantly, job done."

b) Why "[you] can't even get mplayer (with all the gcc4 plugins) to work, among the games."

c) Why "the trash is an absolute nightmare to operate especially on a USB thumb drive."

I'm not quite sure why you're sticking the Media Player application in the /boot directory, instead of in apps. I'm only understanding part of your response to your Media Player issue. This is what I'm getting so far--thanks.

Re: Can a distro of Haiku be made in the future?

OK--I'll try to explain what I'm seeing for ya, and see if I'm guessing right. :) For a, I'm guessing this means a universal driver repository or something where a user can simply extract and open it? Or is this referring to the operating system? Haiku supports most drivers, including wireless Internet cards and touchpads out of the box. If you're having problems with a driver or your hardware, you're always welcome to write a new thread. :)

For b, Media Player should be in /system/apps. Media Player supports most common audio and video formats--what plugins are needed? Honestly asking to see what's up here, or really, why you need plugins.

For c, the trash operates differently than how the recycle bin does in Windows XP. The advantage is that files can be deleted on individual volumes, and that in different operating systems, such as Windows, the trash directory can be deleted and thereby 'emptied'. The disadvantage to this method is that it is slower and more tedious, since the trash in reality is a set of trash directories. However, I'm not stating that this is completely the case in Haiku--it was in Mac OS.

I apologize if I appeared rude or too straightforward earlier. :) Thanks for reading.

Re: Can a distro of Haiku be made in the future?

Sorry, I don't know how much more I can unpack. Mplayer was installed in either /boot/apps or /boot/. When I double-clicked on the application, nothing happened. Obviously the hard drive is now wiped (failed) so I'm not certain what went wrong when I did try to install it.

Re: Can a distro of Haiku be made in the future?

@Paspie: Yes--I had seen the 'trash' thread you started and that your HDD is in trouble here.

As far as making a fork, I'm looking through the source, but I'm still thinking whether it should be done... :)

Re: Can a distro of Haiku be made in the future?

To start this post off, I'm going to note that at some point, desktop Market share for Linux is going to collapse under too many distros (it won't die off completely thanks to the licence).

Tried to install FFplay today, another Libpak application, but as soon as it started extraction I entered kernel debugging land. Any way to get round this and get it working?

Re: Can a distro of Haiku be made in the future?

Paspie,

Is this the computer with the hard disk that is failing--more specifically, are you trying to install with the failed HDD? If so, the HDD is what is giving you trouble here, as far as you entering kernel debugging land. If not, if you are unzipping the file in alpha 2, the extraction process within Haiku appears to malfunction when unzipping certain files--and it is likely that your copy of the file is not working correctly with this service.

I'd love to further help you out here--however, may I kindly ask if you could please post a new thread? ;) It'd be appreciated, as this thread has gotten quite long. :)

Re: Can a distro of Haiku be made in the future?

It's with the replacement hard drive which I nuked successfully, so it's obviously the alpha 2 problem. Are there any nightly builds which fix this problem and run fairly stably? (not too concerned about stablity)

Re: Can a distro of Haiku be made in the future?

Paspie wrote:

To start this post off, I'm going to note that at some point, desktop Market share for Linux is going to collapse under too many distros (it won't die off completely thanks to the licence).

I don't quite follow this reasoning, Haiku can end up with just as many distributions. It doesn't matter though, just like with Linux distributions, Haiku distributions will be 'survival of the fittest'. The only difference is that there will be a Official distribution, but I guess Ubuntu can more or less be considered the 'official' desktop linux distribution these days aswell. In order for any distribution to attract usage it will have to offer something not offered by other distributions that users will want, in Haiku's case that could be a distribution optimized for a certain platform/use, future will tell.

Re: Can a distro of Haiku be made in the future?

Paspie wrote:

It's with the replacement hard drive which I nuked successfully, so it's obviously the alpha 2 problem. Are there any nightly builds which fix this problem and run fairly stably? (not too concerned about stablity)

Nightlys run fine for me. Pick the latest ones as the media player improvements have been huge. gcc2 hybrid is likely the best as I think it has the least trouble with gnash, but I haven't tested that in a good long while.

Re: Can a distro of Haiku be made in the future?

Going back to distros, why not allow them, but any extra features provided by these may be added to the official version if the main team is impressed. We can have distros, but purely for demonstrations.

Re: Can a distro of Haiku be made in the future?

Paspie wrote:

Going back to distros, why not allow them, but any extra features provided by these may be added to the official version if the main team is impressed. We can have distros, but purely for demonstrations.

How could you _not_ 'allow them'? Anyone can take the code and make a distro, anyone can take the code and make a proprietary version of Haiku aswell (although they must name it differently). The Haiku devs can not control the creation of distributions be them open or proprietary, however people can choose to use them or not.

Re: Can a distro of Haiku be made in the future?

Rox wrote:
Paspie wrote:

Going back to distros, why not allow them, but any extra features provided by these may be added to the official version if the main team is impressed. We can have distros, but purely for demonstrations.

How could you _not_ 'allow them'? Anyone can take the code and make a distro, anyone can take the code and make a proprietary version of Haiku aswell (although they must name it differently). The Haiku devs can not control the creation of distributions be them open or proprietary, however people can choose to use them or not.

Just to drive this point home ... practically anything can be done with the source code of Haiku, as it is MIT. However, the Haiku name and HAIKU logo are trademarks, which are seperate from the source code and do not share the same liberal usage.

As a working example, a few years ago Debian needed to rebrand their distribution of the Mozilla source code as 'Iceweasel', instead of the trademarked 'Firefox' : http://en.wikipedia.org/wiki/Mozilla_Corporation_software_rebranded_by_t...

Re: Can a distro of Haiku be made in the future?

I would love to use a mulimedia-specialized distro like dyne:bolic (or LMMS/ubuntustudio ?) but based on Haiku...

http://dynebolic.org/
http://lmms.sourceforge.net/
http://ubuntustudio.org/

Re: Can a distro of Haiku be made in the future?

The concept I had in mind, actually, was to make a netbook fork of Haiku. As someone who has enjoyed programming on Linux, (but doesn't like the GPL or Linux that much), I would look forward to writing new scripts, modifying existing source code, and again, writing a shell based on the hypertext markup language--the language of the Web--so that anyone can edit it without compilation or without having to know much about computers.

If anyone is interested in a studio-based distribution, I think they'd have to work on that--I'm a programmer who's focused on smaller computers. :)

Re: Can a distro of Haiku be made in the future?

No, please no FORKS they're evil!

As I already said is this the cause Linux has not success... 1000s of distribution all incompatible!
Distro divide the programmers (working togheter better result are accomplished and faster) and the userbase...

It is not better to do this as part of the real Haiku?

I understand usiing html can semplify the things for not real programmers (probabily there is necisity to write a new shell but to start WebPositive instead of tracker!)... but it is a pity to trash the beautiful Haiku API totally...

A think that I like to see is that I call Traker "personality" that is tracker can be configured to be:

  • Kiosk mode no tracker but your shell (webPositive or that you want)
  • Netbook personality: Meego Like
  • Mediacenter personality: big buttons all folder visible with icon (showing onlu media files), simply usable by a remote
  • Desktop personality: the actual (and only for now) Tracker personality

In this way we mantian an unique code base (not forks) but highly personalizable!

Re: Can a distro of Haiku be made in the future?

It sounds as if the Haiku community does not wish for a distribution... therefore, I'm thinking that I can work with a Linux-based OS, and somehow try to unify it into a desktop OS. Thank you all for your support, comments, and for ideas, but I can see that Haiku needs to remain Haiku.

I honestly do love the operating system--please do not misunderstand; I fully am grateful to the Haiku programmers and to those who have made this possible, but I can see the best way for the project to make it successfully is to have forks left alone. :)

Haiku team, thanks for making something great, and again, thank you!