In the last weeks, Jérôme Duval has worked on completing the input_server. He is happy to announce that the input_server is in a working state now, and can even be compiled and used as a drop-in replacement for the R5 input_server. Read more. Elsewhere, YellowTAB released a Zeta magazine.
Haiku: Input Server Up and Running
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Beyond that, the input_server addons for mice/keyboards/usb HID are done, as are the actual device drivers for ps/2 mouse and atps/2 keyboards. The entire path from devices to app_server unless you use USB
Oh, and the keymap chooser is done too, which isn’t as important but still nice to have working.
Looks very promising. Only the application and media kits look like they need more developers (according to status on haiku website)
i would love to give haiku a try as soon as the first test iso is out, and so are many other here i guess. a good free desktop OS is something we all need, and kde/gnome is nice but not as cool and responsive as beos was.
only too bad many applications need to be ported – that will prove te be difficult. does the haiku team any efforts to make sure win32 / linux applications can easily be ported to haiku?
and i hope i could use it in my life…
Haiku will be highly POSIX compliant, more so than BeOS R5, but porting stuff is not the point – if you want Linux apps, use Linux.
any binary reduction?
any speed up? (even if it’s one hard to check test case)
any known bug of the older input server fixed? (i don’t know any myself but some stuff caused it to hang from time to time).
can haiku really be more posix compliant than the original beos? i mean with the completely diferent architecture of beos can stuff like threading or asynchronous i/o be posix compliant? and even so is that even something you should want? you’ll end up like windows where 99% of the programmers use select() instead of windows native completion ports.
..or FreeBSD, or Solaris x86. Both run (most) linux apps natively. Sometimes better than linux itself, in fact.
Go Haiku Team Go! I can’t wait for a beta release of the whole OS, even tho I realize its a few years off. But damn I’m impatient
And with that I say keep on truckin’ because your progress is going well.
Not sure if this is all POSIX or not, but its going to have mmap() and pthreads anyway, as well as better networking support (AF_UNIX/PF_UNIX). This would allow MySQL, etc to work, as well as better, cleaner ports of Apache, PHP, PostgreSQL, etc.
But wheter it’ll be stable/good/whatever I don’t know. Ported apps are far from the be all and end all of an OS – BeOS’s architechture means native will always be better.
of course, native apps are better, but haike needs apps to gain momentum and porting apps is still easier then writing new ones. when haiku becomes really big, only then it’s interesting to build native apps IMHO
…..wake me when the app_server works. As if.
I would just like to try to boot thing thing and see how it looks. For fun
“Looks very promising. Only the application and media kits look like they need more developers (according to status on haiku website)”
AFAIK, the status isn’t really updated very often. The Media kit is actually pretty far along. However the key pieces still nowhere near completion (once again, AFAIK) are the kernel and the app server, which means no Haiku for quite awhile yet.
pthreads == PosixThreads
I will say two years from now they still will not have a beta version of haiku for people to try.
Thats the key get a beta version going.
One by one the parts are completed. So what if it takes another year? It’s our code, and beyond the BeOS R5 clone objective (aka Release 1), we can take this project in any direction we please.
In case anyone wonders the license is MIT: http://www.haiku-os.org/learn.php?mode=nsl_view&id=23#74
They have Beta versions of virtually every single part of the OS for people to try. Just because “people” in general are too damn lazy to test them doesn’t mean they’re not there.
Nice to see progress. Both on Haiku and Zeta development. Still out there are people who claim it’s dead OS. For me its not dead anymore! I bealive that Haiku reach R3 before linux will be desktop ready.
Some programs are so complex to port or are directly tried to a OS market that it is very unlikely to be ported to a new OS – examples AutoCad, OpCode, MicroSoft’s Development tools, Apple Ixxxxx series of programs.
However there is also a danger is your OS needs to little porting effort or even none. The C128 developed very few programs for itself because C64 programs would run fine on it, so why develop for the smaller market. We also saw the same thing happen to the OS/2 market. Most Windows programs (Win32 spec?) would work fine on it so why develop an OS/2 version.
Make porting/writting programs for BeOS too hard and very few will be written, make it too easy and we see either the original code recompiled or just a few simple mods being made, result a lot of me-too programs instead of BeOS programs.
That is a hard balance.
Right, and R5+BONE had a good balance – it was easier to write a native graphical app than port it, but there was still enough compliance that even major UNIX serving programmes (Apache, UnrealIRCD, CommuniGate) were made available for it.
Having no X11 layer will help Haiku more than it could hinder it.
I bealive that Haiku reach R3 before linux will be desktop ready.
Which appears to be mainly because of your prejudice against Linux.
I’m using Linux for 5 years on my desktop (just as with BeOS), and i’ve “switched” computer-illiterate people who are very happy with it (GNOME).
However, Linux will become my secondary OS once Haiku reaches R1. I had the opportunity to try Zeta “RC3”, and if they dare to call this a Release Candidate, I cannot suggest buying Zeta. Even Dano was more stable, and they Windows-ified a lot of the interface (Preferences Apps, Be Menu => “Software” submenu).
I doubt it is a non-Linux user’s prejudice against Linux. Personally I have found the people who consider themselves the Linux elite to be the worse people against making Linux more easy to use. They seem to like the complex parts and resist changes to make the system easyier to use.
Well I’m familiar with linux (running gentoo), and i have found it is very powerfull, very user unfriendly, and quite inconsistent OS. I can handle this, but average Joe can’t.
Joes computer is fast, equiped with color monitor, 2d/3d accelerated graphics cards, and… YES! He has a mouse! So why bother native terminal and many text config files edited manualy on desktop system? Maybe it is good for server like OS. I’m sure it’s not good for Joes desktop OS. Joe just want to surf the net, hear some music, watch a movie, contact with friends etc.
It’s just a mine opinion. I do not want start flame war here.
Personally I have found the people who consider themselves the Linux elite to be the worse people against making Linux more easy to use. They seem to like the complex parts and resist changes to make the system easyier to use.
You have probably dealt with KDE Users…
I guess that at this point I should quote past0R:
I do not want start flame war here.
I can’t wait for someone to put together just a basic iso so we regular joes can test the work, right now it’s just announcements for most people with no relevance to our every day lives.
Well, for the average Be user this news is relevant. This means that you can get to testing the input server, as it is feature equivalent to R5 and can be used as a drop-in replacement. It also means there’s progress as now there’s one more kit that has a checkmark next to it.
I am not elite, I can only code a tiny bit, etc. I like the complexity, within reason, if it makes my OS more efficient to use.
>> Personally I have found the people who consider themselves the Linux elite to be the worse people against making Linux more easy to use. They seem to like the complex parts and resist changes to make the system easyier to use.
> You have probably dealt with KDE Users…
> I guess that at this point I should quote past0R:
> I do not want start flame war here.
Sorry to say while the number of people I have meet who consider themselves to be elite were few, they seem to cover the entire range of UNIX and Linux systems.