Guest post by Red/System, the new programming language that is used in the Syllable project, has reached its next milestone: an ARM code generator backend for its compiler. It supports Android (screenshot) and generic ARM Linux (screenshot on Debian). Earlier, the backend for generating Mac OS X executables was already completed (for x86 CPUs so far).
Red now has two complete CPU code generator backends: one for x86 and one for ARM. The new ARM backend supports the ARMv5 architecture and up. This means that all current ARM CPUs are supported except the smallest embedded versions, which use a distinct Thumbs instruction set that is not yet supported by Red.
Apart from these 32 bits CPUs, work has also been done on a code generator for AVR Atmel 8 bits CPUs. This has been demonstrated running on an Arduino board at Software Freedom Day 2011 in the Netherlands (the Arduino demonstration is towards the end of the English spoken video).
Executable file formats now supported are ELF, Mach-O for OS X, PE for Windows and Intel HEX format for embedded devices such as Arduino. ELF support has variations for x86 Linux, ARM Linux, Android and Syllable Desktop. FreeBSD support is under development.
Cross-compiling is trivially easy with Red. For example, the canonical demo program can be compiled to target Android ARM from Syllable Server or any of the other supported host platforms like this:
rebol -s rsc.r -t Android /path/to/tests/hello.reds
Thanks for the update. I love cross-platform support, especially for alternative processors! Keep up the good work.
The first thing I expect to see out of a new programming language is tutorials, which they truly give you the necessary glimpse of what the language is about. I searched online, but I couldn’t find RED tutorials, only Rebol tutorials. Does anyone know a link to some RED tutorials?
Isn’t RED a compiled variant of REBOL? I thought the tutorials for REBOL would apply in RED as well.
It seems RED is a subset of Rebol, so Rebol tutorials are not exactly RED tutorials. But I could be wrong…anyone more knowledgeable to help us with this issue?
The name implies that it’s a Red(uced) subset of REBOL, but it’s more intricate than that. It’s more a pun on the important REDUCE function in REBOL, which could be thought of as the REBOL interpreter’s way of compiling pieces of code. Red is a full compiler, which is innovative, since during most of REBOL’s lifetime it was commonly thought that a compiler was almost impossible due to REBOL’s highly dynamic nature. Red reduces that level a bit to the point where compilation is possible.
So Red will not be entirely REBOL compatible, for example because it uses static type inference instead of dynamic typing. Further, Red doesn’t exist yet. The current language level is Red/System, which could be shortly described as REBOL syntax with C semantics. Code that is fairly static looks remarkably like REBOL, but code that needs more dynamic data management will look more like C until Red is implemented.
Therefore, REBOL tutorials can give you an impression of the look and feel of Red/System, but they won’t work as is. Also, the language is less than a year old, so it’s very early days for tutorials, and there’s a huge amount of other work to do. The reference manual is of high quality, though:
[ http://static.red-lang.org/red-system-specs-light.html ]
I guess closest to tutorials are the examples in the bindings that are currently available:
[ http://www.red-lang.org/p/contributions_21.html ]
Ok, thank you for the information. Perhaps it is too early to be talking about it, isn’t it?
If there is nothing to talk about, anything is good enough.
I just wondered, where’s our Syllable troll, Ciphri, and here you come It’s more about RED,than Syllable. It is mostly a starting project, so why not let ppl to know about it? Ppl talked RaspberryPi even before anyone had any HW in their hands, hence preliminary. So what?
Red/System is a complete language which can now target most hardware currently on the market with the x86 backend and this new ARM backend. The reference documentation is complete and there are enough bindings to write full programs.
What it’s too early for is Red, because it’s another language that is currently being built on top of Red/System. It’s also too early to expect mature support such as extensive tutorials and third-party websites.
Kaj, there’s one 3rd party site to RED, it’s called Red Chronicle
http://www.red-chronicle.com/
Thanks for reminding me. I should start one myself, too.
Can’t get excited about Red. Nor Syllable. Wait… I’m the Syllable troll? Oops. Syllable went down the toilet when REBOL came in to view, and the death was sealed when Kai ousted Vanders. Sad, but very true.
To answer Kai’s observation made back on a story we last both commented – no, after you hit a certain threshold, OS News auto promotes your comments to +2. I guess you don’t post here often enough.
Henderson – you should either know what you are talking about, or better stay silent, if you can’t make a logical argument. Cipri is Syllable troll, period. You can go into ANY Syllable past related article here, and see his posts, very often addressing his own agenda and being topic unrelated.
So Syllable failed, because REBOL came into town? Great to know, that language can destroy OS development And as for Vanders, he said his own words to the topic too.
Try better next time. There’s many topics here at OSAlert, and if I am not interested, I either don’t read it, or I stay away from commenting. So Syllable got early RED support – what’s wrong with that? That news is not a rocket science to understand …
To be honest, it was more a message to Kai regarding his former accusation. He implied I was shilling my comment score, when it *just* posta as +2… I do nothing.
As for Syllable – you assume I’m not in the loop? Maybe – I stopped any Syllable interest 3 odd years ago. But, I was there at the start. It floundered with the advent of Syllable Server, got overly bogged down in REBOL, then Vanders left. Syllable now is a former shade of what it might have been that much is clear.
Look at a successful project like Haiku. Started around the same time. Which is more mature? Which has traction?
Edited 2012-01-05 23:43 UTC
Haiku is a bit different league, no? Well, just my feeling about the project. Apart from technical things, it also seems the project is rather well organised. I wonder, in that regard, what internal organisation was there for Syllable? I mean – e.g. some council, voting?
I am not sure the Syllable project collapse, just because one person, Kaj, wanted to have REBOL language on-board?
Sometimes simply project does not generate enough of community and volunteer following. This is e.g. case even for REBOL language, and in general almost anything Amiga related this days.
Well, while I think, that really – Syllable is getting nowhere nowadays, because the lack of man-power, I think that RED might get somewhere, because REBOL really has some qualities to follow. But – there’s so many languages out there, so let anyone follow anything he/she is interested into.
In the future I propose separate RED related news, with no link to Syllable, nor REBOL, as it is causing only a harm and generally negative reactions, because of such a relation. Pity for such a young project to carry that past heritage.
-pekr- , it’s true that sometimes I use to write also non-topic stuff, but at least syllable related. If you took a look at the major of comments are writing also non-topic stuff, even you.
Now you start attacking henderson just because he also says too that he lost interest in syllable because of kaj/rebol/server.
The low man power is a direct consequence of this.
Nearly all developers that left did it because of that.
I dont want to give more proofs for that assertion, because it’s private, if you dont believe it it’s fine. But you can also contact them yourself, and ask them, perhaps they will talk with you about that subject.
Another reason why I dont think that there will other programmers apear, which will follow kaj is based on the following idea:
Kaj has a very low c++ knowledge, and his system programming competence is also as good as non-existent. His competences are more in the domain of scripting. (If you dont trust me, just ask kaj a few details about the appserver)
If you want somebody to “follow” you, you need to be strong.
Vanders was a good leader, because people trusted in his programming capabilities. And that’s why haiku is a success, because they have a democratic voting system, and because they have exceptional programmers, which surpass even vanders by large margins. Their deep insight in nearly all parts of haiku (api,kernel,filesystem,..) give you a lot of trust.
And yes, the good organisation of haiku is also a reason for their success. It’s great to see how all the management of all donations and all money won by haiku is open for anybody to see. In the case of syllable you dont know what happens with your donation.
Can you be sure that Kaj never used donations for private use?
And a remark relate do Haiku/Syllable so that you can compare better their performance:
When Syllable started in 2002, they started with a fork of Atheos which has been by then already quite advanced. And syllable was free to break all binary/source code compatibily to atheos, and also to itself all the time.
While haiku started from scratch, with the goal to be even binary compatible to beos who’s source code they have never seen.
I guess you can see that, the task of haiku was one a lot more hard, and still they managed to surpass syllable already in as good as any domain. Even at scripting!
The last “development build” of syllable is now 2 years old, while haiku is compiled everyday.
I guess kaj is hoping to find new developers using propaganda, but i guess he doesnt understand that in the long-run propaganda is not enough.
It doesnt help to call everyone Troll when writing something bad related to syllable. This wont make syllable look better, because any serious developer takes first a look at the history of syllable before he invests his time in syllable. And the facts are very visible: The last release has been 2-3 years ago, the real improvements to the core system in the last years are nearly zero. In the last years most developers left, and the leader is a scripting-guy.
These facts are already enough for most serious developers to stay away from doing syllable-development.
I dont care much that you call me Troll, it’s not a problem for me, but I hope that perhaps one day kaj will come to the conclusion that i said a lot of true. I was “ringing the alarm clocks” many years before that something is wrong. And even by then I was nearly too late for saving the situation. It was perhaps the last train to reconciliation, but kaj was not wise enough to do a step back.
And since then the future of syllable was what I was supposing it to be. A big stagnation.
Ciphri – thanks for your post. Finally something reasonable. The reason I called you a troll is, because in the past, you jumped on Kaj almost at each Syllable mention.
Now to some points. From what you describe, I can’t easily imagine, how the language selection might ruin all the project, apart from wrong organisation model, competences, and ego being part of the game. So was there no formal structure, simply to say NO to REBOL related stuff/aproach? What difference would it make, if someone would choose Python, Lua, whatever? Imo Kaj has no problem to agree, that he is not a C++ guru. But that does not help the situation either.
Simply put – it is always about ppl. But – it is also about a momentum. You see – REBOL had some momentum xy years ago, and – it’s gone. We also have AmigaOS4.x, MorphOS 2.x, AROS – they might have wider userbase than even Haiku, but imo Haiku has the right momentum. Well, maybe it is just my feeling, I don’t know.
There is simply many factors, to create successfull product. And also – it does not matter, how cool the architecture is, the success comes from differn conditions. Even if you would put 10 C++ programmers to follow the Syllable, it would not necessarily help. Simply ppl are not interested anymore, and even Haiku might be in trouble. Recently, the momentum is mobile plus embedded, desktop PCs are going to die-off in a big picture imo …
That seems a bit backward to me (pretty much an outsider to those few projects; no side to take, I think …except, yeah, I can roughly see in what state Syllable is for some time now, compare it to some other projects; and by some almost local “patriotism” – lets not pretend it can’t influence overall trends in software – I should even stand more behind Syllable vs., say, Haiku).
It appears that cipri cares (or at least cared, which is for all intents and purposes just as good here) for Syllable, has quite a bit of inside experience with it and related – which places such somebody firmly in the group of interested & with valuable perspective to share, to comment.
PS. And you seem to be almost reaching also his assertions …but not quite / almost like being vary of that part of the conclusion. Yes, people are not really interested in Syllable for some time now – and cipri offers seemingly compelling answer to “why?” (partly at least)
Edited 2012-01-09 00:05 UTC