[other] Haiku Project Announces Availability of Haiku R1/Alpha 1 (THE multimedia-OS resurrection)

Community Ideas for Ardour
#11
if it has a functioning POSIX interface and working X11 implementation (or something else which GDK supports)
Check this out: http://tiltos.com/drupal/node/16
I wanted to have a GIMP on haiku, so I decided to do something with that. There were two paths to achieve this goal. The first one was to port a GTK+ and then replace X11 calls from a GIMP with the Haiku's ones (complex task). The second one was to port X11 to the Haiku (easy task). I chose the second path for now. The GIMP on haiku is quite stable except the bug #283 , which also affects Abiword and Gnumeric.
;)
Better people than us have produced better OS's than Haiku, only to watch them fall by the wayside. NeXTstep
Once again, Wikipedia is your friend: http://en.wikipedia.org/wiki/BeOS
Apple CEO Gil Amelio started negotiations to buy Be Inc., but negotiations stalled when Be CEO Jean-Louis Gassée wanted $200 million; Apple was unwilling to offer any more than $125 million. Apple’s board of directors decided NeXTSTEP was a better choice and purchased NeXT in 1996 for $429 million, bringing back Apple co-founder Steve Jobs.
:|
I've managed to port Ardour to Cygwin and it was a relatively painless task - but only because someone else had already ported many of the technologies that Ardour relies on (such as gtk-x11, libsndfile, libfftw3 and various other things). If I'd had to port all those things myself (as would probably have been the case with Haiku) I'd have had a much bigger mountain to climb.
That's why platform indipendency is a really important thing (which Ardour, unfortunally, don't have).

I have to say it again: Haiku port it’s not that important, but “getting rid” of 3rd party technologies would result in “portability easyness”.

"the beautiful chaos that is linux"
The beautiful of linux is the openness, not the chaos. Haiku team has introduced a kind of hierarchy in the open source scene (for example you can't make a distro without permission), it's not the cathedral nor baazar, but "kibbutz" approach: http://homes.dico.unimi.it/~monga/lib/oss-icse04.pdf
#12

I somewhat tend to agree with you (not on apples niche but on the rest of it). Haiku’s spiritual ancestor is BeOS and well I remember back into the BeOS 5 days that some of the larger DAW makers were looking at the system with great interest. Unfortunately Be went over to embedded systems and was bought out by Palm not long after.

Basically Hardware compatibility is going to suck, software compatibility is going to suck. The clean design might be nice (compared to the beautiful chaos that is linux). But the bottom line is - if Haiku is able to extract say 40% extra processing power from a quad core processor then it might be worth the effort. Of course Linux, Windows and Mac and associated software are probably a lot better on multiprocessor/multicore machines then they were when BeOS was around and the benchmarks I remember were being done.

#13

Forart.it, you are making yourself look ignorant, perhaps even stupid. You keep repeating stuff like this: “That’s why platform indipendency is a really important thing (which Ardour, unfortunally, don’t have).”. By the definitions you are apparently working with, there is no such thing as a platform independent program of any complexity. Its been patiently explained to you that Ardour relies on POSIX, the C++ standard library, GTK and Glib all of which are HIGHLY portable APIs. They exist on every operating system that has sold more than 100 copies. If Haiku doesn’t support these APIs, then it is the one that is deficient. If it does, then stop bickering on this forum and go find some people that actually want to port Ardour to your preferred OS.

#14

and in addition… isn’t there always going to be a risk with any open source OS that it will go the same way as Linux and end up with a disparate bunch of forked or specialized distros?

#15

@forart.it

I stand corrected, BeOs and therefore Haiku was not UNIX based as I had thought, I realize Haiku isn’t “Linux” based, my point still remains that the type of high level developers Linux in general could lose to this project would be a shame since Linux has been continuously making such good progress especially in the multimedia field. Just an opinion.

Moving from Windows to Linux as a multimedia platform for many is equal parts labour of love/nightmare so I wonder how many people will have the sustained interest to make the leap again starting from the ground floor.

#16

No, Haiku IS NOT a Linux-based OS.

Do you know Wikipedia ?

Or, check out the official FAQs:

Is Haiku based on Linux?

Haiku is not a Linux distribution, nor does it use the Linux kernel.

#17

@forart.it: linux is not unix, read again :wink:

#18

For multimedia applications, moving away from Windows is good, but to Linux is not that good, IMHO.

Even Linus Torvalds has recently criticized the Linux growth, saying that the kernel has becomed “bloated and huge” replying to an Intel study which had found Linux performance dropping by two percent every release.
Check the article: http://www.h-online.com/open/Linux-Bloated-and-Huge-says-Linus--/news/114289

To understand even better what I mean, check out the funny Urias McCullough’s “attack” to Dell’s Ubuntu @ LinuxWorld ( http://www.haiku-os.org/blog/koki/2009-08-27/oscon_opensource_world_2009_double_report )

I don’t think that an Haiku port could “steal” developers from Linux; on contrary, good solutions from a real multimedia OS could be ported back to Linux, benefiting both worlds.

#19

IF Haiku really lacks a POSIX interface (which is all that “Unix” really means these days), then its absolutely true that any existing or potential Ardour developer spending time on a Haiku port will “steal” time away from both Linux and OS X and FEATURES IN GENERAL. This is also true if it lacks an X11 implementation or some other graphics API for which there is a GDK port. The amount of work involved in porting Ardour to a platform that lacks either one of these APIs is almost equivalent to completely rewriting the program. If you don’t understand why, you don’t understand the architecture of modern software. Just porting from GTK1 to GTK2 was a mammoth effort.

On the other hand, if it has a functioning POSIX interface and working X11 implementation (or something else which GDK supports), then porting Ardour to Haiku should be more or less trivial, and doable by any moderately skilled programmer, including someone who already uses Haiku and is enthusiastic about it.

#20

I don’t think there’s a “risk” (if that is the right word) that Haiku will fork in different distros, mainly because it is, in itself, quite monolithicaly developed.

A linux distro is a pick-n-choose; you take a kernel made by someone here, a windowmanager developed there, this IM-client, that image viewer and so on. Whereas the core and core applications of Haiku seems to be soley developed by the Haiku team.
I’d think you’d better compare it to the BSD camp, that eventually you might end up with two or three versions but nothing more.

I also hope someone (not necessarily an Ardour core developer) makes a port to Haiku, it’d be interresting to see if and how it would work.

#21

Mmh yeah Paul wrote ones in one of his posts that he started to understand the policy of ProTools in supporting only some hardware I believe.

The problem with Linux was in the last years mainly about ALSA/OSS/JACK and last but not least pulseaudio.
Also all the different distro’s are a problem for Ardour sometimes…

So this could be the future… but then you need people and devs who picks this up.
A nice OS with no good apps is not a good OS…

#22

From that I don’t see any indication that it will be a good multimedia OS. BeOS might have been in it’s time, I don’t know, but does that automatically mean that Haiku will be a good multimedia OS?
Why and which advantages should it have over linux in that respect?

And if at all I’m sure it’s a couple more years in the future.

#23

I think a lot will depend on the aims of the development community which don’t (to me) seem very encouraging. For example, one of the primary goals is to promote full compatibility between Haiku and BeOS (in the sense that applications written for one will run on the other). This has already caused problems because BeOS used an old version of gcc and didn’t support some of the newer features of versions 3 and 4. Some kind of hybrid build environment got rolled out to accommodate this but a similar problem awaits applications that use multimedia codecs. Many of today’s codecs weren’t invented (or weren’t in common use) when BeOS was in its heyday. If these differences can’t be reconciled, this is bound to create friction between developers and users. Haiku’s success will then depend upon how well or badly the development community can manage that friction.

So for the time being I’d say that I have to agree with Hollunder. There’s no reason (at this point in time) to believe that Haiku will develop into an impressive multimedia platform. Time will tell of course but unless I’ve missed something (which is entirely possible) that doesn’t seem to be one of Haiku’s goals.

#24

According to Zenja’s comment on http://www.haiku-os.org/community/forum/has_haiku_abandoned_beos_compatibility the goal for R1 (Haiku 1.0 so to speak) is to be BeOS compatible. This to ensure everyone works toward a common goal. From then on they’ll probably try to catch up with the rest of the world.

As a multimedia OS BeOS was really impressive for it’s time and possibly even compared with todays OS’s, so if the interest in developing Haiku further continues, and drivers and applications get written/ported, I think it could be a really interesting audio/video platform.
And since the point of Haiku is/was to have a free BeOS, why wouldn’t they have multimedia performance in mind?

#25

ardour is one thing, but jack is also needed. What jack backend would be required on HAIKU ?

#26

@John E… To be fair Apple Got Steve Jobs Back with next step. I think this was a major factor in the Deal… (either way it worked out well for apple).

The thing I recall about BeOS was that I was able to run 4 video streams simultaneously smoothly on a Pentium 166 with less than 100 MBs memory and no hardware acceleration. The system only lost responsiveness when I added the 5th.
How that translates to modern hardware and against modern OS’s I really don’t have a clue… Hopefully that will give you some idea as to where the niche may be.
a
@Paul – If my understanding is correct, Posix yes, Standard Libary yes, GTK & glib no…

@forart.it the good news is that (from listening to Pauls inexterviews today) that texshe Ardour source has a strong split between engine and UI. Should you feel the need it would be possible to use the Ardour backend and develop a native frontend. I think however that Paul and the existing devs have enough on their plates for the forceeable future

#27
That's why platform indipendency is a really important thing (which Ardour, unfortunally, don't have).

There are definite tradeoffs involved in platform (meaning API, really) independence.

First, there are performance tradeoffs. When building for a single platform, it is possible to leverage that platform’s performance to the best effect.

Secondly, when building to multiple platforms, that do things in many cases in radically different ways (GUI toolkits, for instance, have very very different paradigms; GTK and Qt work very differently under the hood, for instance, and one of the most popular cross-platform open source GUI Toolkits, WXWidgets (used by Audacity) is different yet again), the developer either has to have completely different libraries for basic things (like the GUI event loop) or the developer has to take a least common denominator approach. And while GTK has been successfully ported to more than one platform, GTK is but one API that Ardour needs.

Third, true cross-platform compatibility is very hard to retrofit. Cross-platform support really needs to be an early design goal, when the basic cross-platform infrastructure can be designed in and not bolted on. Ardour is complex enough that making it cross-platform would involve a vast amount of work that the developers have said they aren’t keen on doing.

So if you want a true cross-platform DAW, my recommendation is to go bug the Audacity developers about making their program more DAW-like. Or join up with one of the other cross-platform DAW groups.

And, having said all that, I don’t consider Ardour’s lack of full cross-platform support to be unfortunate in any way whatsoever. [EDIT] And I’m talking about API cross-platform-ness; as Paul said, most of the basic APIs are highly portable, but some are not (Windows, for instance, has some really different ideas API-wise even in their ‘POSIX’ layer…).

What is really important is getting the work done, not what tools you use to get the work done…l Although I must admit I wish that I hadn’t sold my old UREI Levelling Amplifier years ago…some tools just have class :slight_smile: Would love to see a good LA-2 sound-alike (don’t have to look like it, like the UAD plugins do) plugin for LADSPA/LV2

Although my RCA Audimax has much the same sound…

#28

let’s get ardour right on linux first, shall we ? :wink:

#29

Absolutely. Better one platform (or two) done well than many done not so well.

#30

forart.it - maybe another approach would be for you to lobby the Haiku devs to include a powerful VM feature so that Haiku could act as a host to an alternative OS or to apps written for an alternative OS (like Wine does, in Linux).