Studio33 has released part II [part I] in its series of articles looking at the current state of Haiku. “In the previous part I talked about the achievements of the Haiku Team since the project was first started, this time I will go deeper into the work that has been done lately and which parts need serious attention in the coming months.” Screenshots o’plenty, boys and girls.
both links go to the old article…
Fixed already.
The article states that there is no audio playback in Haiku because there is no audio_server.
Audio works, his soundcard probably isn’t supported.
The audio_server in R5 was just for backwards compatability with previous BeOS releases. There will be no audio_server in Haiku.
He also says that the current network stack doesn’t do anything. While it is correct that the current netstack will go away and be replaced with a BSD stack, it currently works fairly well. Browing in NetPositive works.
You are correct of course – my SB Live card is not supported by Haiku currently.
The netstack in Haiku is functional, but there is no preferences dialog to configure it, and only a couple network cards supported (SiS900 and RTL8139/69 I believe – maybe more?).
The netstack is still missing a DHCP client, however, and there are a few other issues reported in the forums and mailing lists.
Edited 2006-03-21 01:46
What is the status of the low-levels of the OS? I think there is still some confusion about if Haiku is still running inside or on the old BeOS in some form.
This is just a whine, but I wish that Haiku had used the Linux kernel. In the future it will be an uphill battle to get drivers created… when Linux already has so many! Oh well…
Keep up the good work, however!
The Haiku install shown by Studio 33 is completely compiled from source (except for some of the programs being run such as VLC, CL-Amp. etc – those are actual binaries copied from a BeOS R5 system run natively in Haiku)
You obviously do not understand the fundamentals of OS design if you think Haiku could have (let alone should have) used the Linux kernel.
Put simply, it’s not feasible.
Put slightly less simply, Haiku is an OS where the GUI is integrated tightly (e.g. Windows). Linux architecture does not fit Haiku paradigms. Linux itself is a huge codebase and designed to be a general solution. Haiku (and BeOS) is a high performance desktop OS. They have very different goals.
Agreed, but a “wrapper” sub-system for Linux ethernet and PCI drivers (or other drivers, for that matter) would be a nice short-term solution. In fact, this has already been implemented for some 3COM cards — check the lib directory:
http://www.augusta.de/~michael/BeOS/3c509/3c509.r5.zip
Isn’t this how Syllable’s drivers work?
Sort of. For network drivers and (currently) soundcard drivers, we generally port from Linux. For that small subset of devices the kernel APIs are close enough that it isn’t too much work; you generally only need to tweak the way ethernet frames are pushed up and down the stack and re-write the PCI and logical device management. We do have a linux_compat.h system header for drivers, which is a small set of macros and typedefs to make life a little easier.
All of this ignores the fact that the Haiku/NewOS kernel isn’t GPL though, which would make it tricky to use GPLd drivers as Syllable does.
“All of this ignores the fact that the Haiku/NewOS kernel isn’t GPL though, which would make it tricky to use GPLd drivers as Syllable does.”
Sorry, this doesn’t make sense. We already have some GPLd drivers in our tree.
I didn’t say it was impossible, just tricky. It’s a gray area of the GPL, and if the Syllable kernel was not under the GPL it isn’t an area I’d want to deal with myself, personally. There isn’t even a clear-cut consensus on which way around linking between the kernel and a driver occurs: does the driver link against the kernel, or does the kernel link against the driver?
Provided you’re happy with your answers to those questions then that’s fine. Linux drivers are an amazingly useful resource, so it’s great that you’re making the most of them.
“does the driver link against the kernel, or does the kernel link against the driver? ”
I’d say the latter, because the kernel can live without the driver but the opposite isn’t true.
“Provided you’re happy with your answers to those questions then that’s fine. Linux drivers are an amazingly useful resource, so it’s great that you’re making the most of them.”
That’s true, although I definitely prefer *bsd drivers, for the license and also because linux drivers’s code (at least, the ones I looked at) is a bit more messy.
Sorry, this “gray area of the GPL” just doesn’t exist for open source kernels with a compatible license. Even though GPL v3 is a lot more specific about this, there is nothing that would prevent anyone from relicensing the Haiku kernel under the GPL.
IOW even if shipping a kernel with GPL drivers makes the kernel GPL (in case of GPL v2, this is a possible understanding), this wouldn’t be a problem; ongoing development would still be put under the MIT license, of course.
Oh sure, as a developer it’s not much of an issue. A distributor would probably want to know for certain what the licence conditions are though.
I’m not saying that is an issue, just that it is possible and not something I would personally wish to deal with.
That’s correct, but it will be a problem for companies wanting to create a proprietary distribution based on Haiku. And since that is one of the main reasons for using the MIT license accepting GPL’ed code is problematic (and yes, there is GPL’ed code in Haiku. It’s not a problem unless you want to create a proprietary solution).
I wonder if the submit button will work in aBrowse? The login button didn’t work
(Reading this, Vanders?
The BeOS GUI is no more integrated with the kernel than it is in Linux. The BeOS GUI is just a userspace server, running a userspace toolkit (technically, no different than X11 + GTK). The difference between the two is not technical, but organizational. BeOS *seems* more tightly integrated because there was one entity in charge of it, which could set standards about the design.
You obviously do not understand the fundamentals of OS design if you think that a BeOS replacement could not be built on the Linux kernel. BeOS itself is an OS of separate components outside of its kernel, including the GUI processes.
A high performance desktop kernel can be built from the Linux kernel codebase. It all depends on how you build it and what features from the codebase that you want in your compiled version of the kernel.
“A high performance desktop kernel can be built from the Linux kernel codebase.”
Yes yes yes, we’ve heard this before and we will hear it again.
The whole point is, Haiku has made decisions (well thought through) about how they want to handle everything from the Kernel to the logo on deskbar. The “one size fits all” rule doesn’t apply to Haiku.
In case you missed out on that, B.E OS (Blue Eyed) already tried the path with the Linux kernel and surely they got results, just that Haikus was more impressive. Now maybe they implemented something wrong, I don’t really care, the thing is, Haiku has shown it’s the way to go.
Besides, there are already a 100 Linux distros, and if it’s suitable for “High performing desktop” I still don’t understand why I constantly have the feeling it sucks (for my purposes)?
Not to mention, that the license philosophy makes the Linux choice even more difficult.
I’m glad Haiku has come the way it has, I hope the remaining journey will be fast and prosperous!
In case you missed it: Linux Kernel != Linux Distro. Not even close.
I hope Haiku succeeds. But software needs hardware to work, and those drivers are crucial. Who knows, maybe people will build computers targeted for Haiku, enough to satisfy the niche.
As far as I know, there are already companies building computers for Haiku. Pegasos is one of them aren’t they? Well, they don’t build only for Haiku, but as far as I know they’ve given Axel a computer to make Haiku run on it. And also Pegasos seems to target a lot of platforms.
In the end, it’s all about preference. Do you want a system that can do everything and buy that 2000$ computer. Does it matter to you that it supports all the latest and the greatest? Or is it just important that you can get some HW to work properly and buy those pieces.
For the upcoming years, Haiku will fight an uphill battle with drivers and such. But this fight has been fought before, and usually the result has been pretty ok. Linux is one example, BSDs another. I think there is a big pool of users who are shaking with the urge to try out this system, and if we can only get 1% of the *Nix userbase to use Haiku, the amount of drivers will explode! Not only drivers I’m sure, but alot of other things as well.
The real big obstacle ahead (in my point of view) is to make people develop stuff in the Haiku spirit, not just develop stuff. The last thing I wanna see is “CLI only apps”. I wanna see people think in lateral, make things like OOo, but fast, efficient and beautiful (like GoBe). Things like Firefox, but faster, more efficient and more beautiful (like Net+). If the spirit can be spread, and devs drop by, general computing will be in for a fantastic ride, which I will enjoy with all of my heart!
I wanna see people think in lateral, make things like OOo, but fast, efficient and beautiful (like GoBe). Things like Firefox, but faster, more efficient and more beautiful (like Net+). If the spirit can be spread, and devs drop by, general computing will be in for a fantastic ride, which I will enjoy with all of my heart!
You’re dreaming, and if there is one thing the computing world is good at, it is crushing the hopes of dreamers…
You’re line of reasoning is fallacious. Just because B.E.O.S failed doesn’t imply that their technical decision to use the Linux kernel was at fault. Haiku had much more mindshare and developer support than B.E.O.S did. Now, perhaps you can indirectly relate the magnitude of developer support for Haiku to the decision to use a different kernel, and you might be onto something, but that’s a secondary issue entirely.
Besides, the point was not that Haiku should have used the Linux kernel. The original poster said that Linux would’ve been unsuitable for Haiku’s purposes. Now it’s fine that Haiku went with the kernel they wanted, but that doesn’t mean that Linux would’ve been unsuitable.
The point about the “Linux distros” just shows how little you understand the underlying situation. Linux distros look the way they do because they all use the same software — the GNU utilities, X.org, GNOME, etc. A Linux-based Haiku wouldn’t use these things, and wouldn’t look or feel anything like a Linux distro. Utlimately, Apple has shown that the underlying kernel, as long as it supports certain features, is mostly irrelevent to the desktop experience. OS X is a very fine desktop OS built on a kernel cruftier than Linux has ever been.
…great news!
Any vmware player images for this?
Yes. One made automatically every six hours.
http://www.schmidp.com/index.php?option=com_files&path=/haiku/image…
I have vmware player, it wants to open a .vmx file..that site has vmdk images.
A .vmx file is just a text file that tells VMWare how to load the image.
see the following links to the Haiku-os forums for more info on how to get/create one:
http://www.haiku-os.org/forums/viewtopic.php?t=575&highlight=vmx
http://www.haiku-os.org/forums/viewtopic.php?t=655&highlight=vmx
http://www.haiku-os.org/forums/viewtopic.php?t=678&highlight=vmx
thx, I realized just after I posted and looked at a sample from google and made 1.
W O W : ))!
In the future, Haiku really needs to implement some sort of compatability port of X11 (similar to what Wine does for Win32). There’s no compelling reason for me to use Haiku R1 over linux at the moment. It’s bad enough I can’t run Windows apps–now I can’t use linux apps as well?
I realize what the goals of Haiku R1 are, nobody needs to remind me. But let’s face it, if Haiku is planning on becoming mainstream it must have a decent app collection. In my opinion, the easiest way to do this is to be compatible (source-compatible at least) with linux apps. I think I’d be more interested in an R2 release which looks outside the box, then a boring R1 release which is basically a clone of an extinct BeOS. Just my $0.02.
In the future, Haiku really needs to implement some sort of compatability port of X11 (similar to what Wine does for Win32). There’s no compelling reason for me to use Haiku R1 over linux at the moment. It’s bad enough I can’t run Windows apps–now I can’t use linux apps as well?
There’s an X11 server for Beos to facilitate porting graphical unix applications you know.
http://www.bebits.com/search?search=x11
There’s an X11 server for Beos to facilitate porting graphical unix applications you know.
Yes, but by most accounts it’s dated, feature-incomplete and crappy.
There’s an X11 server for Beos to facilitate porting graphical unix applications you know.
Yes, but by most accounts it’s dated, feature-incomplete and crappy.
Yes but then again a lot of software for BeOS is getting dated. It apparently is open source though, so people can work on that and the other 2 problems (though lord knows being open source isn’t a guarantee against being crappy).
So it’s definitely X11 then?
Yes, but by most accounts it’s dated, feature-incomplete and crappy
Yeah, that sounds like X11 to me.
Why not run OS/2 then? You can run Windows apps, Linux apps, DOS apps, Java apps – heck, they even have Virtual PC and Virtual Station to run the whole other needed OS if need be. We do have X11 as well, along with a posix layer. 😉 Somehow I doubt this comment will sway you or anyone else. hehe…
Edited 2006-03-21 17:47
Obviously, you have no idea about the BeOS/Haiku desktop. It’s all about the experience of NATIVE applications, written for BeOS and INTEGRATED in BeOS. We don’t want bloatware like X on Haiku. If we’d had it, Haiku would be just another crappy linux distro. So if you want to use your X applications, please stick to Linux and leave Haiku alone.
Obviously, you have no idea about the BeOS/Haiku desktop. It’s all about the experience of NATIVE applications, written for BeOS and INTEGRATED in BeOS. We don’t want bloatware like X on Haiku. If we’d had it, Haiku would be just another crappy linux distro. So if you want to use your X applications, please stick to Linux and leave Haiku alone.
Welcome to knee-jerk city, population: You. For your info I’m a mac user and though I try to use X11 apps as little as possible it is a good stop-gap measure. Eg. for apps like the Gimp on OSX. Haiku would do good to provide a decent X11 server as an add-on if they’re smart.
To alexorizor:
1. An X compatability layer running on top of Haiku would not bloat the OS. Think of it as an emulation layer (like Wine is). I agree that an actual X server running would slow things down, and nobody wants that.
2. If Haiku is to get more popular than BeOS was back in the day, it needs to appeal to a wider audience. Where will new users *most likely* come from? Hint – it’s not the windows crowd.
3. If you so vehemently hate X compatability then feel free to rewrite the 1000’s of apps in the linux library for Haiku. Sure, you can experience NATIVE BeOS applications — or the 10 of them that are actually worth a crap.
It surprises me how some people are so unopen to change. If you want only the features of BeOS, stick to BeOS! If you’re like the rest of us and want the best of both worlds — a simple, modern, efficient OS that runs OpenOffice, Evolution, etc. — then hope that such features are on Haiku’s roadmap.
or the 10 of them that are actually worth a crap.
*cough* http://www.bebits.com
There is absoutely no need AT ALL for an X server on BeOS. There are more than enough perfectly functional applications for BeOS. X is for UNIX and UNIX-like systems: keep it there.
Edited 2006-03-21 12:14
There is absoutely no need AT ALL for an X server on BeOS. There are more than enough perfectly functional applications for BeOS. X is for UNIX and UNIX-like systems: keep it there.
There is absolutely no need AT ALL for a win32api compatibility layer on Unix. There are more than enough perfectly functional applicaions for Unix. win32api is for windows systems: keep it there.
Except, that argument is weak. Weaker in the case of X, where some run it even on windows in order to allow for forwarding.
Isn’t interoperability a good thing? Are BeOS users that exclusive? If this attitude was put forward with Linux, people would rightly cry zealot.
Is it a compatibility layer your after, or is it X you’re after?
Sure, if someone wanna make a compatibility layer for it, I guess that’s good, but please understand, at the current stage this is not a priority. Maybe it’s a priority for people running Linux/Unix, but for those developing Haiku and using BeOS (ATM) obviously don’t think of it that way.
IMO, and please note IMO, Gnome/KDE has put a lot of effort earlier on being a lot like Windows, and as of lately a lot of effort on imitating OSX. Haiku wanna live it’s own life, being a model, not modelling someone else.
Interoperability is more about fileformats and protocols than apps and compatibility layers in the API. So that being said, I’d prefer GoBe productive being able to handle OASIS than having OOo dominate the scene. I’d prefer having Refraction handle Photoshop/Gimp files than having GIMP onboard.
If someone dislikes that and feel the urge to make a compatibility layer etc (similar to CygWin I suppose, maybe CygHaiku), then go for it. Just don’t expect it to fall into the main tree of development.
the Haiku devs naturally want it to have as widespread usage as possible, but on the merits of Haiku/Beos itself. and the only way gain users based on the merits of the system is through native apps that fully takes advantage of it.
compromises will have to be made, ports of stuff like Firefox and OpenOffice for example, they don’t integrate into the system the way native applications do, but they are critical applications of which there is no native equivalent, nor likely to be for quite some time.
but again, if anything is ever going to ‘sell’ the system, it’s native applications. taking advantage of the filesystem metadata, translators, replicants and the slew of other ingenious components is what will set Haiku apart from the rest, just as it did with Beos. and yes, Beos failed mainly due to lack software support, but back then there was no portable open source alternatives for those aforementioned critical applications. add to that the fact that you had to shell out for the operating system, now anyone can download Haiku for free and try it out, there is no risk involved for the curious user.
finally, though I hope otherwise, I doubt Haiku will ever become anything even remotely mainstream, but hopefully it will carve itself a niche amongst tinkerers, developers and general computer entusiasts, much like the current non-corporate linux userbase. I for one will be there.
peace and prosperity to all!
“There is absoutely no need AT ALL for an X server on BeOS. There are more than enough perfectly functional applications for BeOS. X is for UNIX and UNIX-like systems: keep it there. ”
X is a protocol with an implementation available for almost all modern operating systems, including GNU/Linux, FreeBSD, OpenVMS, Windows and Mac OS.
Edited 2006-03-21 17:40
@underthebridge
User-inteface consistency is the key for Haiku. It’s designed by and for people who value that consistency, just like BeOS or MacOS.
I can’t see it being any more than a side-project, just as it is on MacOS X. After all, one of the aims is not to get into a Gnome/KDE/X11 situation, so why slip into bad habits right at the beginning?
BeOS is not completely starved of apps, and it has all the webby apps that people use on a daily basis.
In the long-term, if Haiku is a success the apps will come with native UIs. It looks like Haiku it’s in it for the long-haul and won’t cut corners for a quick result against Linux.
“User-inteface consistency is the key for Haiku. It’s designed by and for people who value that consistency, just like BeOS or MacOS. ”
I don’t think MacOS values consistency that much these days. I’d even say GNOME beats it in that regard.
Where will new users *most likely* come from? Hint – it’s not the windows crowd.
I completely disagree. Linux users want … Linux.
They want to run a server, or they have an irrational hatred of Microsoft. They are almost never the type that want to create audio, video, and graphic designs.
Haiku/BeOS is essentially a very responive MacOS 8/9.x on x86 hardware. It is also incredibly user friendly, and appeals to a wonderful aesthetic. Linux is none of these things.
Please leave Haiku alone, OSNews-Linux-Regulars.
That’s a completely retarded position. What about people who want to use Haiku, but are tied to particular X11 apps? OS X is a comparatively large platform, with orders of magnitude more native software than will ever exist on Haiku, yet I still need to use X11 compatibility on a daily basis.
There is no getting away from X11 compat for an alternative OS that hopes to be useful to real users. The best you can do is emulate Apple and try to integrate the X server as cleanly as possible into the desktop.
rayiner:
tell me one thing: why the hell do you want to run Haiku? What’s the point? If you need your X11 apps so bad, why don’t you stick to Linux?
You guys just don’t understand. People are running Haiku/BeOS because they *don’t* want to have things like X11. Haiku is about innovation, about NEW concepts. I say it once again: if you need to use your old apps, stick to Linux.
alexorizor, like I said before, Haiku is not really doing anything different from the original BeOS. The only thing you are really gaining is support for modern hardware.
If you want your old BeOS apps, STICK TO BEOS. Leave Haiku alone.
it need first to catch up, then evolve.
At least those old BeOS apps are not so crappy : )
I don’t want to run Haiku. Between OS X on my desktop and GNOME on my laptop, I’m a happy camper. My point, however, is not just about myself. I find it hard to believe that the population of people who can be satisfied with an entirely Haiku-native set of apps is not an order of magnitude smaller than the population of people who could run Haiku if they could still access a few critical apps that have no Haiku equivalent.
Face the reality. There will never be a Haiku-native replacement for the GIMP. There will never be a Haiku-native replacement for Inkscake. There will never be a Haiku-native replacement for OpenOffice. There are barely enough developers to make Haiku, much less enough to make a plethora of native apps like these. Within that reality, there are only two options: embrace X11, and the negatives and positives that it brings you, or reject it and lose 90% of your potential userbase. If you have no need for X11 apps, then no harm, no foul. Don’t run it then! If you do, then at least you have the option to use Haiku.
Your entire line of reasoning is ridiculous. Say you have a prosperous Haiku desktop. You’ve got lots of native apps, etc. Then, sombody ports X11 to it. Does the entire thing come crashing down? Does stuff that worked yesterday suddenly stop working today? Does the Subversion repository dissapear and does the Fedora repository appear in its place? Clearly not! OS X has shown that the draw of native applications is enough that people will still write native apps even in the presence of X11 alternatives. The X11-alternative, however, allows a lot of people to use the platform that could not otherwise use it.
“I don’t want to run Haiku”
No comments… run vista then…
“There will never be a Haiku-native replacement for the GIMP”
Gosh! But we have Refraction & Wonderbrush! What we will do now? No GIMP??? heavens!
“There will never be a Haiku-native replacement for OpenOffice”
Of course! Zeta will never be a port too!
BeOS has Gobe, why Haiku will never be one?
“There are barely enough developers to make Haiku, much less enough to make a plethora of native apps like these”
Arram! The developers that don’t working in Haiku are making native apps…
“If you have no need for X11 apps, then no harm, no foul. Don’t run it then! If you do, then at least you have the option to use Haiku”
“The X11-alternative, however, allows a lot of people to use the platform that could not otherwise use it”
This is the same GPL vs MIT discussion in Haiku… bla bla bla… speak until parrot speaks… Time will tell…
Gosh! But we have Refraction & Wonderbrush! What we will do now? No GIMP??? heavens!
Neither of which have half the power of the GIMP or Inkscape.
BeOS has Gobe, why Haiku will never be one?
Gobe was written for BeOS back when BeOS still had a future. Haiku might very well run the old Gobe, but a piece of abandonware like that is not exactly something to pin your hopes on. Perhaps something like Gobe could be written again, but even if there was an Haiku-native Gobe equivalent, you still wouldn’t have an OpenOffice replacement. OpenOffice has an order of magnitude more features than Gobe did, and lots of people need access to it.
Arram! The developers that don’t working in Haiku are making native apps…
BeOS in its heyday didn’t have enough developers to make all the native apps people needed. That’s why the X11 port existed in the first place. Hell, OS X doesn’t have enough developers to make all the native apps people need, hence why it has an X11 port too! Haiku’s developer base is probably a fraction of the size of the original BeOS’s, and its ridiculous to expect that they’ll create all the native apps people need to use.
Don’t get me wrong, I was a BeOS user back in the day. I even have a copy of Corum III lying around here somewhere. However, I’m not delusional enough to expect that Haiku can try to go it entirely alone, creating a completely new platform from the ground up. History has shown that this is a fruitless endeavor.
I’m terribly confused – How is it that GIMP, OpenOffice.org, Inkscape, etc. require X11 in order to be ported to an OS? I’m pretty certain the Windows ports of these are not using X11 right?
I’m pretty certain that compatibility layers can be written for the porting of these apps. Mostly, POSIX compliance is the most important requirement – then most of the dependencies can be ported, and finally the apps that use them.
Edited 2006-03-22 02:57
umccullough wrote:
“I’m pretty certain the Windows ports of these are not using X11 right?”
shhhsh… don’t tell the x11 crowd. afaik gimp only requires the gtk+ runtime environment (go figure!). openoffice and inkscape has no x11 requirements when running under windows. however, iirc both gimp and openoffice on OSX does require x11, maybe thats what got rayiner so hung up about it, since he stated he was an OSX user?
rayiner wrote:
“OpenOffice has an order of magnitude more features than Gobe did, and lots of people need access to it.”
hence that is why Yellowtab is porting OpenOffice. and I pretty much doubt they’ll be running it under x.
rayiner wrote:
“However, I’m not delusional enough to expect that Haiku can try to go it entirely alone, creating a completely new platform from the ground up. History has shown that this is a fruitless endeavor.”
it’s not a ‘completely new platform’, it’s an open source clone of an existing platform named Beos, it’s also posix compatible.
and they’re not going at it alone, Yellowtab, a german vendor is selling a commercial version of Beos named Zeta, and are also porting the gcc 4.x toolchain, together with the aforementioned OpenOffice. also, they are developing commercial applications.
will Haiku be a runaway success? not likely, however, the existing Beos userbase will remain, ie migrate to Haiku or Zeta, and it will attract a bunch of new users with it’s elegance and some (I hope) osnews release hype ;D
world domination was never really on the table.
peace and prosperity!
Besides Haiku R1 actually being here and stable, the two major things for the Haiku community to tackle will be:
* Drivers
* Application availability
Without drivers for your machine, it will get you nowhere. Next on the need-list is the applications you use on a regular basis, and this list is diffrent for all of us.
Then why all this? To get as much people as possible to use Haiku on a daily basis. Lets not forget one important thing. Developers (like me) are users too, and with more developers comes more native apps and more innovation.
rayiner has a good point, in that more users, even though not using purely native apps, is a good thing. After all, Haiku will be better of with users running a great deal of native apps [1] and a few “Linux”-apps like GIMP, than users using only Linux apps.
[1] And thus contributing to them in various amounts, such as helping other users, giving bug reports or even doing actual development.
I for one belive Haiku has a good chance of becoming a success.
I’m terribly confused – How is it that GIMP, OpenOffice.org, Inkscape, etc. require X11 in order to be ported to an OS? I’m pretty certain the Windows ports of these are not using X11 right?
Little history lesson – OpenOffice started on Windows and OS/2 (back when it was still StarDivision’s product), the X11/Unix port came later.
Neither of which have half the power of the GIMP or Inkscape.
I wonder when You tried Wonderbrush last time?
Because i tried both of them and WonderBrush is just much better than Inkscape. I don’t know, maybe Inkscape has some features WonderBrush doesn’t have, but for me WonderBrush is just best.
As a kind of proof i can only add that currently i use Ubuntu most of the time (because of mentioned lack of software, and especially drivers for BeOS), but each time i have some gfx work to do i reboot to BeOS with WonderBrush.
Of course i don’t try to tell You that WonderBrush is ideal, end-of-all thing. It’s not. It has some things missing, but still it’s (one of) best application of that type IMHO. The only other i liked so much was Macromedia Fireworks, but i don’t use Windows anymore
.
As for GIMP… i understand and see it has a lot of features, but in terms of “user-friendliness” it’s a crap. I use it on rare occasions and it always is just a pain in the a** to use it. Yeah, i can get things done with it, but i waste a lot of time on it. And sanity
.
Forget it rayiner, it’s a lost cause talking to the BeOS fanboys. No matter how much you try to explain something, they just don’t get it. They think of X11 compatability as harm to the OS… yea sure.
Sheesh.. will you people stop being so defensive of your ever so precious Haiku! God forbid anything new is added from the original BeOS.
Alright then… live in your 1990’s BeOS world… the rest of us for the time being will be using linux/osx/windows until something better comes along.
And that is Haiku you smarta55, subscribe to the glass elevator list and you’ll get a lot of ‘new’ ideas for your reading pleasure. Calling BeOS old makes you an even bigger ignorant m0r0n. It were these new ideas which made BeOS so unique in the first place.
Why don’t you look at your stupid (legacy) linux view, linux by itself is pretty cool but too bad *it had to be based on ages old unix*. And I know that from my own experience when I had to learn that ‘in order to be compatible, it’s not recomended to change the resolution of the text terminal’ Uh oh some stupid text app might brake.
@rayiner
“Face the reality. There will never be a Haiku-native replacement for the GIMP. There will never be a Haiku-native replacement for Inkscake. There will never be a Haiku-native replacement for OpenOffice.”
People who throw the word “never” around usually end up looking stupid.
You’ve got no idea whether there’ll be a port or not. You don’t know that the GIMP won’t be outclassed by another program. You don’t even know that Linux will be no.1 open-source OS on the desktop in years to come.
As for X11, the only questions are, should it be a core-component, and is it more important than the other features that Haiku needs? I’d say no to both. That’s different to saying it shouldn’t happen – I’ve got no problem with an X11 layer, as ugly as it will inevitably turn out to be.
@underthebridge
In the future, Haiku really needs to implement some sort of compatability port of X11 (similar to what Wine does for Win32). There’s no compelling reason for me to use Haiku R1 over linux at the moment. It’s bad enough I can’t run Windows apps–now I can’t use linux apps as well?
IIRC it wasn’t the X11 that was the problem but more of missing kernel instructions (NMAPsomethingamajig).
I realize what the goals of Haiku R1 are, nobody needs to remind me. But let’s face it, if Haiku is planning on becoming mainstream it must have a decent app collection. In my opinion, the easiest way to do this is to be compatible (source-compatible at least) with linux apps. I think I’d be more interested in an R2 release which looks outside the box, then a boring R1 release which is basically a clone of an extinct BeOS. Just my $0.02.
The original BeOS r5 was already more than 90% POSIX compatible. One of the aim of the rewrite was to fix that last problematic percentage.
A decent QT 3 toolkit port to BeOS/Haiku would be best.
here’s where I got this:
Michael Phipps: Our plan is that a few of the shortcomings will be fixed: the design issues. The 1GB RAM limit will become 2GB or possibly more. Sockets will be file descriptors. Select and mmap will work. HW OGL is a *HUGE* undertaking. It is a good thing and we would like to support it. But not for R1, unless we get a whole ton of very focused, very knowledgable people who want to work on it.
from http://www.osnews.com/story.php?news_id=5472&page=3
select: http://wiki.linux.net.nz/select(2)
mmap: http://wiki.linux.net.nz/mmap(2)
Edited 2006-03-21 16:21
There was already an X server for BeOS R5 available. It wasn’t that shiny, but it worked (I used it to have some ssh-forwarded X clients running).
Question only is, how useful a standard linux/unix app is on Be. Well, it could be an interims solution, but if you choose Be/Haiku, you would probably choose it because of the “whole desktop” experience..
…I need a multimedia OS (as BeOS was) for my activities (a/v multitrack recording/mixing & composing, graphic), so Haiku will probably be my next OS.
I hope to see mutch more multimedia apps for Haiku !
(note – i found this interesting article, check it out: http://arstechnica.com/reviews/01q1/be_qnx/be_qnx-1.html)
Quote from the article: “QNX’s software consists mostly of Linux ports, too. Why use ports on QNX if you could use a more mature native version on Linux anyway?” Hahaha!!
I rememeber running BeOS as my primary O/S for quite some time.. finally gave up due to lack of apps (even on Bebits).
It will be great if Haiku becomes viable ..
IMHO, it’s very nice have the access to some of nice applications developed for *nixes and all… but at same time, most of this applications would be VERY weak in a BeOS enviroment without a nice work porting them.
Of course there so many that don’t require so many changes of paradigm, like Firefox and so on… but take a look in Thunderbird, besides compatibility with your profile in other system, storing the e-mail in any way that is not simple folders and files doesn’t make ANY sense at all in BeOS.
BeOS isn’t about a kernel. It’s not just about a toolkit or a language too. It’s about a complete paradigm about how data is storaged and accessed. THEM BeOS make sense. with live queries, with tons of metadata, with translators for all applications and replicants for combining them… it’s all about reusability and simplifity. If you take that way, them it just don’t work in it’s complete form anymore, just some separeted pieces…
But still… without losting the BeOS spirit… there’s so many things that can be ported, so anyone can have the best of both worlds! ;]
axeld wrote:
> IOW even if shipping a kernel with GPL drivers makes
> the kernel GPL (in case of GPL v2, this is a possible
> understanding), this wouldn’t be a problem; ongoing
> development would still be put under the MIT license,
> of course.
Hmm… that sounds like you’re getting into pretty shaky ground there. Either you’re licensing your code MIT or you’re licensing it GPLv2. It will look pretty bad if you distribute Haiku with the MIT license, but with GPLv2 drivers thus being in violation of their license, while maybe saying on a mailing list, “sure, it’s GPLv2 then, fine.”.
The Haiku code is licensed MIT. The drivers are licensed GPLv2. They are separate works. Now, the derived work of the drivers linked to the kernel is GPL’ed (meaning that a statically-compiled kernel binary must be distributed under the GPL), but the original MIT source code can still be distributed under MIT.
As far as I can see, there is nothing wrong with this setup, since the MIT license is GPL-compatible. FreeBSD does the same thing too, btw. They have GPL’ed code in the tree, but isolate it so if you want a fully BSD-licensed source tree (for whatever reason), you can still get at it.
What axeld was stating was only that there were GPL drivers in the source tree. They are not “distributed” with Haiku by default.
When building Haiku, a developer must specify when running ./configure that they want to “include gpl addons” if they want them built with the OS. This means that a default distro of Haiku would not contain the GPL drivers and would therefore be distributed under the MIT license.
Since the MIT/BSD license is GPL compatible, someone distributing Haiku with the GPL addons can distribute the rest of the OS under that license as well and provide all source code as required by such.
Furthermore, drivers in BeOS/Haiku don’t necessarily need to be compiled into the kernel as they are in Linux – so someone building a driver for Haiku using GPL code would not be including their source as part of the kernel per-se. The drivers can be copy/pasted to the kernel’s add-ons directory where the kernel would have them loaded during bootup to see if they apply to any hardware. As such, I’m not sure that GPL drivers for Haiku would require the rest of the OS to be GPL’d.
Edited 2006-03-21 20:50
Hm. That sounds clumsy to me. So, your OS code is MIT-licensed, but every time you make a release containing GPLv2 drivers, since the kernel links to those drivers, that release as a whole is GPLv2? But everyone keeps working merrily along on Haiku using the MIT license?
It sounds like then, the de-facto license is actually the GPLv2… Anyhow, I’m just curious because I’ve never heard of such a thing.
“It sounds like then, the de-facto license is actually the GPLv2”
Not really for one reason: MIT licensed code can be directly copied and released under the GPL. You cannot however take code that is GPL and re-release it under the MIT license (unless of course you hold the copyright and then you can release it under anything you want).
underthebridge wrote:
> If you want your old BeOS apps, STICK TO BEOS. Leave Haiku alone.
Actually, the whole reason the Haiku team is sticking with their original goal of binary compatibility is so you can indeed run your old BeOS apps on Haiku.
And, BTW, I think what alexorizor is implying is correct: if I want to run an app on one machine but have it display on a different machine, I want a simple GUI dialog box with checkboxes and text fields specifying which machine the program will run on, which machine it will display on, which user it will be running as on the remote machine, which user it will be displaying as on the local machine (with a help box describing how the permission issues are handled), and big “run” and “stop” buttons at the bottom. And maybe color-coding so I can immediately tell when a window is from an app running on a remote machine. The worst thing about X11 is fiddling with its config files trying to get it to do what you want. And it looks like programming for it is no picnic either.
If you want X forwarding, use Unix.
Or, to put it another way: Unix (including Linux, BSD, etc.) is fundamentally designed to be network-transparent. This is good for a server operating system, in which it’s highly convenient to be able to have applications running on multiple different computers and displaying on the same display.
This is not good for a desktop operating system, because first of all, you can optimize local hardware access when you know that it’s actually local hardware access, and second, the last thing home users want (now we’re talking security) is to be accidentally running an application on a different computer entirely.
In the US, at least, it seems like “shared” devices (payphones, Internet terminals, etc.) are becoming less popular, and “personal” devices (cellphones, iPods, laptops, etc.) are becoming more popular. Since computers will only become more powerful, it stands to reason that most computing will continue to be done locally, instead of remotely.
So we have every reason to want a desktop system that makes local operations /fast/, and that somewhat sandboxes remote operations, so that people can use every Internet service they want without having to worry that they’re exposing their computer to ub3rh4xx0rs.
This is not good for a desktop operating system, because first of all, you can optimize local hardware access when you know that it’s actually local hardware access
This is an idiotic argument. BeOS doesn’t optimize for local hardware access in the GUI. It has APIs for direct hardware access (ala DirectDraw), but the regular GUI stuff runs over IPC. There is no reason, in principle, you couldn’t run the app_server on a seperate machine and send the BMessages over a socket. The only difference between how X11 works and BeOS works is the IPC mechanism in question — X11 uses sockets, BeOS uses ports.
rayiner said.
I don’t want to run Haiku. Between OS X on my desktop and GNOME on my laptop, I’m a happy camper.
So why post in a BeOs discussion then? Dont want to run it, dont comment on it and leave it to those of us that do!
Drivers are an issue for any OS except windows. It comes down to the companies releasing the specifications, which hardly ever happens. Even the drivers in Linux are pretty crappy, at least audio, supporting no real features. They do basic audio, and nothing more. Yippee. Might as well use a Soundblaster 16. Hows does your Creative XFi work in Linux???
I’m pretty amazed how well Beos/Haiku has kept up with driver development. The only area that I have trouble with is printer drivers.