Guest post by BeOS may be dead, but over a decade after its lamentable demise the open source Haiku project keeps its legacy alive. Haiku is an attempt to build a drop-in, binary compatible replacement for BeOS, as well as extending the defunct OS’s functionality and support for modern hardware. At least, that’s the short-term goal – eventually, Haiku is intended significantly enhance BeOS while maintaining the same philosophy of simplicity and transparency, and without being weighed down with the legacy code of many other contemporary operating systems. Computerworld Australia recently caught up with Stephan Assmus, who has been a key contributor to the project for seven years for a lengthy chat about BeOS, the current state of Haiku and the project’s future plans.
Modern OSes are able to display Aßmus just fine.
Please correct it!
Gracias.
Ringel-s may also be written as a double-s (I speak German, you know
). Since OSNews has issues with weird characters, I have to do it like this.
Edited 2012-08-08 16:04 UTC
actually it’s sz, but everyone uses ss (including me)
nonetheless it’s a useless character and should have been abolished many decades ago…
Unless you happen to live in Switzerland…
it’s pronounciation differs dramatically from what you would expect from ss, namely it’s the complete opposite (long instead of short).
If you want to get rid of it, please don’t try to make things more wrong.
Disagree here; those special characters shape a language. I’m a native Spanish speaker and ‘á’, ‘ñ’, ‘¿’ or ‘ü’ are part of its history and personality. Currently a lot of people write without using tildes, and, though their texts are perfectly understandable, their texts show their poor level of language handling too.
Especially compount words in the german language profit from ligatures like the Eszett. (Yes, it’s not a letter, it’s a ligature consisting of two lowercase s, precisely long-s plus round-s, forming a unit that’s not hyphenated, dissolving to SS when capitalized, and has nothing to do with “vowel length”).
Example: Meßstrecke vs. Messstrecke (measuring track)
Just count the consonants!
It also improves reading:
Bambuseßstäbchen vs. Bambussessstäbchen (today’s common “short vowel” nonsense spelling error included) (bamboo chopstick)
The Eszett is a typographical aspect of the language that helps to differentiate words. In Switzerland, those two are only distinguishable from context, not from written representation:
Buße (penalty fee) vs. Busse (buses)
Maße (measures) vs. Masse (mass)
If Eszett cannot be typed, it’s common to replace it by ss (which is a valid and common replacement). If it’s neccessary to “preserve” the Eszett (e. g. to write a name of a person), instead of ss also sz may be used (because it appears “validly” only in easily recognizable exceptions, so you can be sure sz means Eszett). This concept is nearly exclusively used in teletype and data transmission via basic ASCII.
Except OSNews headlines
I submitted it with an ß, though I have to admit I didn’t notice in the preview if it was displaying correctly…
Modern OS’s yes, OSNews headlines no.
It’s quite appalling to think about it, but Haiku, as incomplete and visually aged as it is today, still manages to be a better desktop OS than any Linux distro with a modern DE.
It’s a shame that it never got serious attention of any large company.
I can understand ‘incomplete’, but ‘visually aged’? There’s Linux DEs and window managers even more minimalistic, but no one calls them ‘old’ or ‘aged’. It’s just classic kind of style, bro.
Maybe back in 1999 that was the case, but now I would take a Fluxbox desktop over a BeOS/ Haiku one any day of the week.
The BeOS/Haiku widget set and interface are more cartoonish than Windows XP, and feel more like a mock up than a real UI for a so called media OS.
The issue here is not only the look (imho, Haiku indeed looks dated), but their overall architecture: Haiku is more integrated, more standardized and is a overall cleaner design for a desktop OS (not server) than Linux.
Despite dozens of efforts by Linux community to integrate the DE with the core OS, it still fells like a gigantic wrapper on top of a mess, who gives a overall fragility to it.
For a power user who likes total control and has patience to keep track of every single application (and their versions) installed, this is not a issue, they don’t need a wrapper. Just a window manager who can put a terminal anywhere in the screen is enough.
But, for a naive but curious and tech-savvy user, who likes to mess with his computer hardware but don’t care that much about the OS, he is forced so many times to use a terminal, so he can edit entirely non intuitive configurations, that it gives the feeling that a single mistake can bring the whole world down. (while on Windows, it would mean, more often than less, a box about “where is the driver?”.)
Edited 2012-08-08 18:58 UTC
Quite true.
Back in BeOS days, it was for me the right successor to Amiga, sadly never took off.
If Apple had taken BeOS instead of NeXStep, while managing to arrive at the same state the company has now, no one would be getting Mac OS X (former BeOS) because of its UNIX roots, funnily enough.
On the other hand we would have a great multimedia OS.
No, it would entail messing around with the even more fragile registry.
Yes this is the strength of Haiku, as it’s totally focused on providing a full desktop experience out-of-the-box, which means it comes with all necessary components tightly integrated.
Linux by contrast is just a kernel to which you add whatever components you need/prefer, as such Linux strenght is flexibility/customization.
For a desktop environment I prefer the Haiku approach (although I also like the maximal flexibility I get from Linux and the tailoring of environments it offers), and as such it is what I really want to run as my day-to-day desktop OS.
Currently the greatest hinderance for Haiku adoption is hardware/software support, and the short-term solution for software in my opinion is to get more ports for Haiku, native applications are of course much preferred but they will only come once people actually use the system, and before they do that there must be software available to make it useable.
Well, they pretty much are ‘wrappers’ by comparison as there can never be the type of integration between the kernel and outside components as that which Haiku enjoys due to Haiku being developed as a whole OS with a specific purpose, from kernel to gui.
On the other hand, the vast majority likely don’t notice this in any really practical way and although I love Haiku with all my heart I’d have to say that the vast majority of people out there get’s a much better ‘experience’ out of something like Ubuntu then out of Haiku at it’s current and foreseeable future state.
Loved the article, it’s always interesting hearing about a project from the viewpoint of one of it’s developers. Also Stephan’s comments seems very down to earth rather than trying to drum up some hype which lends credibility to his views IMO.
From what I gather the big stumbling point towards a R1 release is the package management, so here’s hoping someone steps up to the plate and finishes this so that Haiku can finally take that last huge step.
I think R1 would do wonders for the sense of achievement and overall interest in Haiku, of course package management in itself would do wonders for day-to-day use of Haiku, particularly updates.
What the UI looks like is probably the least important factor IMHO. It’s like the walls in your home, after a short while you forget what color they are. Of course, for marketing purposes it’s vital, but Haiku isn’t designed to attract customers, it’s designed to be usable.
Really? I think its icons are nicer looking than XP’s myself, in fact I used to use icon sets like XPIze when i was on XP just to get rid of those fugly default icons whereas when i ran BeOS back in the day and looking at Haiku now i think it looks quite nice.
Of course the problem with both BeOS and Amiga clones is what made those systems cool was NOT the software, it was the hardware. Back then those OSes were written for the chips which also couldn’t really be had on any other machine and that tight connection between hardware and software was what made them multimedia monsters. Back then those machines could do things that other machines simply couldn’t touch, serious multitasking, media playback while doing other tasks, it was ahead of its time.
Today frankly any bottom of the line AMD or Intel multicore can do everything you could do then while having so many cycles to spare its just not funny so there really isn’t a point in having the OS and hardware tied tightly together like that. the MHz wars left us with such superpowered hardware that even MSFT can’t pile enough bloat and bling to slow down even a low end quad and the GPUs that Intel and AMD integrate now can just tear through multimedia.
So while I can see why some would want to play with them, I ran both BeOS and OS/2 Warp back in the day and both kicked the snot out of Win9X and System 7….sigh, maybe we’re just spoiled. i know sitting here typing this while transcoding a video and listening to music if you’d have told me then i could do all this without swapping or slowdown I’d have laughed you out the building, now i just take it for granted.
BeOS wasn’t really tied that much, architecturally, to any particular hardware. Sure, Be at first made the NeXt-style error of limiting BeOS at its launch (which happened a bit too late vs Windows entrenchment…) only to the buyers of the expensive & uberniche BeBox, and afterwards (definitely too late…) to the “intermediate” like that Powermacs. But, some time later, they didn’t seem to have issues with porting it to the PC (waaay too late…), where it ran just as well.
Did I mention it was just too late?*
At my place, one of the most popular PC magazines (at the time, most likely well above 100k in a <40M country) included once the BeOS PE on its CD. That didn’t result in much.
OS/2 was similarly essentially given away, back then, with a demo that could be easily (even accidentally) unlocked to be time-unlimited. That OS had an underlying goal of returning to IBM the control over the PC – this is why it failed (of course other OEMs wouldn’t play along, preferred Windows), also why it’s probably good that it failed.
*Plus, focusing on mostly useless demos (why should we be interested in several videos or GL teacups at the same time?), while Win was perfectly good enough (NVM its software library). And also largely stuck in the past, being single-user, with no security model to speak of, and poor internationalisation (while the favourite pet story of BeOS refugees is how a Japanese line of Win+BeOS PCs, from Hitachi, was meddled with by MS…).
[/miss_q]
It was a great desktop, stunningly better than Windows 9x. It is a shame that it never became a viable desktop OS mainly due to lack of applications. I still think its a great OS that in some ways compares favourably to any desktop Linux – Windows or whatever.
Just exactly what does Haiku or BeOS do better than Linux? My question is as genuine as it is rhetorical.
That’s easy. Two things…
Uncompromising responsiveness. It is a heavily threaded, tightly scheduled, single user OS with a UI that is designed to respond with absolute minimal latency, even when heavily taxed.
Simplicity. When you start digging into the guts of the file system and peel back the initial layers of the OS you find… Nothing. There are few if any layered abstractions. The same goes for the API, the number of moving parts is quite small.
Is either of these things the most important attributes of an OS? Not really, but that is kind of the point. The priorities of BeOS/Haiku are and have always been different from most other Operating Systems.
On the other hand, Haiku still has to improve a lot in terms of security.
Currently it has a security model similar to what Amiga, Atari ST, Windows 9x had. We all know what that meant in terms of virus and getting your data secure.
If the an application gets owned, it will have access to the complete filesystem.
This is very important issue to fix, even on a single user desktop OS.
You’ll have to wait for R2 for a fix for that. R1 is supposed to be a drop-in replacement for BeOS R5, and that means pure single-user. R2 is where Haiku will start to take off on its own direction.
Yeah, but all other systems keep moving on as well.
GNU/Linux only managed to achieve what it is today thanks to the support of companies that wanted to improve the ecosystem for their purposes, regardless of good or bad.
Haiku needs something similar, otherwise it will never be much more than a hobby OS.
Nice to cure the nostalgia kick for a few hours and that is it.
but the haiku at the moment says: we finish R1, then we continue with R2.
But they could do something different. They could already now start working on R2 while also working on R1. And if something in R2 would be important to R1, it can be ported back to R1. For example while R1 is not finished, the R1 version could be the “official release”, and the R2 would be just something like a prototype which doesnt need to be backward compatible with each release. And just after R1 is finished, the R2 becomes the “official version”, which from there on needs to be backward compatible and not break api/abi. This kind of “freedom” perhaps could help to make greater progress and to come-up with new ideas.
There should something like an “experimental haiku”, in my vision.
Be, so sleek, sublime
Threads ev’rywhere, by design
Come alive Haiku
Haikus are easy
But sometimes they don’t make sense
Refrigerator.
If only BeOS had succeed and if only Haiku wasn’t so slow developing when it didn’t.
Of all the OS’s I’ve had anything to do with BeOS was the one & only OS that just seemed right, the real deal.
Sadly…if only.
Funny how Haiku is still able to overwrite other partitions.
It says Ass-mus. Haha. Oh the lack-o-i18n fun.
i must admit the development of haiku slowed down, just the gsoc could save haikus image. That’s one advantage of haiku, that even if the core developers of haiku slow down, or loose their interest, there are still new developers coming that are doing a great job.
Haiku has a lot of good stuff under the hood, but it’s user interface makes in my opinion a too old impression. I guess most haiku/beos subestimate the importance of looking pretty/fancy. Beos users are used to that current user interface, and they could swear how great it is, but some of them dont understand, that most other users that come from windows/linux perhpas prefere a more modern looking user interface.
I would prefer if haiku would already now start working on R2, and on R1 at the same time. And if there is something good/important in R2, it can be ported back to R1.