After Axel Dorfler and Robert Szeleney, it is Kristian ‘Vanders’ van der Vliet’s turn to answer the Five Questions. Vanders is one of the primary developers behind Syllable, the fork of (the now dead) AtheOS which saw the light of day July, 2002, because several AtheOS developers were concerned about the project’s long-term goals. Syllable is free/open source software under the GPL license.1. What are Syllable’s strong points?
It is a complete system where all the parts have been designed to fit together well: Kaj calls this “holistically designed”, which is a good description. We take the best design features from past operating systems and implementing them with modern, mainstream, open-source parts. We are also not afraid to cut out the bloat and implement our own solution if we think we can do it in a way that fits the rest of the system better. This has resulted in a very fast, extremely efficient, remarkably easy to use system. We think we’ve hit the sweet spot between elegance and practicality.
Syllable is fairly complete and has good hardware support, which makes it usable right now. It’s also heavily multi-threaded [ed. note: think BeOS] and supports multiple processors, so it’s better suited to modern multi-core CPUs than a lot of other OSes right now.
2. What are Syllable’s weak points?
We’re the new kid on the block. Although Syllable gains a lot of stability from the mainstream parts we use – only a few percent of Syllable’s total source code is new – not all of our own subsystems are fully stable yet. Because we have our own graphical environment with our own APIs, few native applications have been written so far. This means we have a limited number of users and developers at this point, although attracting more is just a matter of time. Part of the problem may be that Syllable currently only has limited support for languages other than C++ and C.
Hardware support is a perennial problem, even though we do pretty well compared to most other OSes in our class. Getting hardware support to a point that is even as good as Linux will take time. If someone tries
Syllable and it fails to boot, that can irritate them and we may lose a potential new user. We always try our best to fix any bugs and add more drivers, but it can be tough with such limited resources.
3. What applications are sorely lacking from Syllable?
All but a few. We have two decent editors, a web browser and a mail program that are shaping up. We have an integrated development environment that fits Syllable very nicely and is easy to use. We are introducing an address book application and will be polishing up the CD burning application. There are the usual small tools that one expects with a system, and some games. We basically have all of the well-known command-line applications, and SDL and Curses programs are easy to port. Other than that, we have a lot of opportunities for developers to start writing applications. There are the obvious ones that everyone mentions like an office suite or Flash. What a healthy OS really needs is a large number of small applications. As Syllable matures and people pick up on our development environment & layout designer that should begin to change. I’m sure Java and .Net support wouldn’t hurt, either.
4. If there were two features you could magically get from other operating systems, what would they be and why?
Do developers and users count as features? If not, a complete version of REBOL with REBOL/View would be nice, and maybe a working, fully supported port of OpenOffice.org
5. What project, feature, or application currently in development for Syllable excites you the most, and why?
Only one? Then it would have to be Syllable Server. Kaj has been planning it for a long time and it’s exciting that it’s really happening now. We have already been noticing that it opens doors for us. On the other hand, it’s just a side project to help Syllable Desktop out. Using the Linux kernel for Syllable Server gives us several things: stability, REBOL/Core, and ReiserFS 3. Putting our graphical environment on top of that will make it immediately more usable. People will be able to start writing applications for our APIs without fear for instability and limited hardware support and those applications will be almost trivial to port to Syllable Desktop. We’ll be able to do some exciting things as we integrate Syllable Desktop and Syllable Server.
There’s a lot of other things that excite me, though. Dee Sharpe has been working on OpenGL support and how to integrate it properly. Jonas Jarvoll has been working on oCADis, which is a large scale application that proves Syllable can handle those sort of real-world applications. We’ve just released a new version of ABrowse, which gives us a very fast and well supported web browser.
Thanks to Vanders for taking the time to answer the Five Questions.
If you would like to see your thoughts or experiences with technology published, please consider writing an article for OSAlert.
I really like Syllable, but hate GPL, so I’m torn.
Keep up the good work, though.
I don’t see what the problem would be for you as an end user, unless your objection is purely ideological.
Not directly as an end user perhaps, but it will affect your selection of apps considering that unless there’s a license exception for syllable’s libs that I missed, it’s effectively impossible to write a non-GPL app for Syllable, which will turn off a lot of third party devs. Correct me if I’m wrong.
You are wrong.
Most of Syllable source code is not under the GPL but the LGPL.
It is perfectly possible to write proprietary applications for Syllable. But don’t expect to get a large following from that.
Ah, thanks. I couldn’t find anything on the site indicating that, so it seemed as if all of it was GPL only. Appreciate the clarification.
It is covered in the FAQ ( http://www.syllable.org/docs/developers/faq.html#1_2 ) but I’ll try to make that information more prominent.
You seem to be quite mistaken.
Linux, for example, also has many proprietary programs. If they aren’t as many as in Windows, the fault is to find in the companies you seem to favor – they just seem to be prefering supporting Windows.
Actually, with your reasonning, you should on the contrary be thankfull to all the GPL applications because they generously fill the void let by proprietary applications.
Just a question : where have you been the last 30 years?
ok here’s a question. Will it go under GPL 3?
Probably not. We’re happy enough with the GPL2 and the license retains the “or later” clause, so others are free to use the code under GPL3 if they wish.
How does Syllable compare to Haiku? Of course, the fact aside that Haiku is a remake of BeOS and Syllable isn’t.
Both projects seem to be based on the same premise: To get rid of bad legacy APIs, and to redesign the system API from scratch using the experience gained from previous approaches (read: Unix/POSIX).
Well, for one, Syllable will run all by itself. Haiku isn’t self hosting yet. You still have to use core parts from BeOS PE
“Well, for one, Syllable will run all by itself. Haiku isn’t self hosting yet. You still have to use core parts from BeOS PE”
It’s true that Haiku can’t yet compile on Haiku, but it does run all by itself.
Edited 2007-07-06 22:00
Yes, Haiku is not yet self-hosting, in the sense that building it is done primarily in BeOS or Linux. Haiku is not yet stable/complete enough to survive building itself. (You have I believe 150.000 files, post-build. Haiku itself is rather lean, currently only 1500 files.)
No, Haiku does not need anything from BeOS. Unless you count OpenTracker, the BeOS desktop which Be open-sourced.
Well that’s a lot of it right there. Haiku’s philosophy (from my perspective anyway) is that BeOS was awesome so we should revive it in open source form. It currently maintains both source and binary compatibility with BeOS and sticks closely to the BeOS look and feel. There are a few advantages to this approach: they don’t need to design an API (though internal design issues remain), existing BeOS applications will run on it without modification, and it has a built-in userbase of BeOS devotees.
Syllable has a somewhat different philosophy. It’s trying to be the best desktop OS it can be. It takes a lot of inspiration from BeOS because BeOS did a lot of things right, but it’s continually diverging away from BeOS and is not locked to any of BeOS’s design decisions. This is a blessing and a curse. It gives the developers a lot of freedom to do things better than they’ve been done before in other operating systems, but it means it doesn’t run existing BeOS applications. Lack of software is definately Syllable’s biggest weakness at the moment, but it is steadily improving as Vanders mentions in the interview.
From a practical perspective, Syllable is significantly more stable and has better hardware support than Haiku currently does, or at least that has been my impression (I have not personally used Haiku). Haiku also has a rather nasty choice ahead of them. The C++ ABI (or at least the name mangling scheme) used by GCC changed between version 2 (which BeOS uses) and all later versions starting with 3. So they either need to stick with old creaky GCC 2, break binary compatibility with BeOS or produce their own custom version of GCC 4 (or 3 I suppose) that is ABI compatible with 2.
It’s nice to see a positive mention of REBOL – it amazes me what you can do in that language. The next version will make porting much easier, so it’s a good thing to look forward to.
Syllable is one of those OSes that seems to prefer doing things the best way instead of the conventional way. I always like to hear more about it.