Douglas Crockford revealed that JSON had been partly influenced by Rebol:
Another influence was Rebol. Rebol’s a more modern language, but with some very similar ideas to Lisp, in that it’s all built upon a representation of data which is then executable as programs. But it’s a much richer thing syntactically. Rebol is a brilliant language, and it’s a shame it’s not more popular, because it deserves to be.
JSON would not have emerged without the communication talent and obstination of Douglas Crockford. He also focused on making his concept usable by current mainstream technologies like Ajax. The Saga of JSON he told in this Yahoo Webminar should be an IT marketing lesson for Rebol.
Douglas Crockford said “Rebol is a brilliant language, and it’s a shame it’s not more popular, because it deserves to be”.
But how can people be attracted by Rebol when:
- the company’s website looks like it was my neighbour’s kid who would have webdesigned it in the eighties :).
- instead of creating libraries for or in PHP, Java, Ruby, .NET, javascript, xml, … to handle Rebol’s Data Format or Interoperability, they hang on their own Rebol Server IOS technology which you cannot even download to look at the wonder (at least I have never found any link).
- nearly none Hosting Company has ever heard of Rebol whereas they should have been targeted as partners and natural prescriptors, that’s how I guess PHP has really taken off.
- and where are all the Rebol’s Community ? They are hidden in a world called Altme, isolated from the Internet, the last contact was the Rebolask forum which since was deserted due to lack of moderation.
It’s not such a shame that Rebol is not popular it is a logical consequence. It isn’t a fatality, things can still change.
With Cloud Computing in the corner, let’s hope Rebol will think about how it will integrate with other platforms like .NET and Java. Integration is the very reason Scala Language has such a huge success, that was the reason also for JSON as it integrates nicely with Ajax even when Ajax was originally meant for XML.
admin:
Hostile Fork:
Ladislav:
Let me address your analysis. Your analysis is at least some 40% off imo. I am with REBOL since 1996 and I know all the good and bad things, which happened to it.
Most things, are becoming popular because of hype - not everybody has the same luck. There would be no Ruby if Rails would not be hyped upon its real importance and you could find another examples of good technologies becoming popular, or not …
- IOS - noone hangs on anything - IOS link is not even available from the main site. I think that IOS is not supported nor sold anymore. We are well aware of integration issues and mistakes, and I think it can change with R3, which is going to be highly modularised, extensible, and some 80% open-sourced.
- websites - we are very well aware of our situation and I agree current look&feel is, well, suboptimal. But new websites will come, after R3 is in better shape. Hopefully we are in beta rather soon and can get back to websites redo. There is internal group, which analysed the topic.
- REBOL community. You talk about the “dark net”. We are not visible on the web, right? But such opinion is just typical mental block of those who are ignorant enough
You either get your support, or not. So let’s look, what have we got in last few years - prominent AltME channels (you can get your answer in seconds, not like with web forum), it is much better channel, believe me. 2 blogs from language author, public wiki, public bug base, devbase (R3 Chat) implemented directly in R3 console, detailed release notes. You can talk to Carl (language author) on almost daiyl basis. Dunno how you, but I can see it as very well maintained communication channels.)
- Hosting companies - just ask them to place one exe for you to work in CGI mode, that is all
Believe me, we have rather good plans, and we are getting there, slowly, but still … just stay tuned!
Hello Petr,
Glad you reacted and I will answer below:
Luck is part of life but you must also do things to attract luck when it presents itself. Currently both Java VM and .NET have opened their platforms or Virtual OS to multiple languages, every dynamic language from PHP to Python has jumped on the bandwagon and will be ready for the cloud which should be widespread in ten years, what about Rebol ? Will you still say we didn’t have luck because you refused to foresee the coming future
About the website, it isn’t so important, I just took it as symbol of lack of presentation. Same for Rebol GUI standard look and feel. In the end I would rather use Rebol as back end with a .NET Front End. Hope in Rebol 3 that will change (and hope will be able to customize the not so nice Rebol icon). Going back to webdesign why do you want to do it yourself ? A professional webdesign is not so costly 1000$ should be enough to rewamp the logo and site’s look.
As for Community, I don’t mean support, generally I get great answers when I used to ask them on Rebolask forum and now on Stackoverflow. I mean being part of. When I am on a forum and there are plenty I can switch from C# to PHP to javascript, when it comes to Rebol, I can only go to Altme, oops bad luck my Vista refuses to connect to it. I’m not a great geek as for firewall. But above all, the Rebol Community of Altme is mostly a community of a few Rebol Gurus, I’m not a Guru.
When it come to hosting, sure you can ask to place the exe, except the hosting service doesn’t want to because they don’t know rebol and so fear for security. More and more they use CentOS and they don’t see CentOS listed on Rebol’s site either when I point the page to them.
Well, working as CTO for one large company, I face the schizophreny of what is the reality, where is IT heading, and what I can see as optimal solution, on a daily basis
I don’t know. We have SharePoint here, SAP (ABAP), some JAVA stuff, yet I think that JAVA is at its ends
Just my internal feeling. It is not absolutly interesting technology. Scripting and more dynamic solutions will replace it. Or can you tell me, why JAVA FX almost copies REBOL/View syntax? Well, even mighty corporate programmers are tired of constant OOP aproach, so that they need some rest with more declarative syntax? 
If you really think, that JAVA or .NET = cloud, then you as well might be mistaken
The problem you are trying to describe is - problem of “deployment”, and here you are right. If you look at various libraries, you can see plenty of langs listed, but no REBOL. Hopefully it changes with R3. Last month we got proper integration Extension API. .NET integration will come, not so sure about JAVA. Don’t worry, first we replace JAVA, then .NET
As for community, you are not all that correct here imo - there is web forum, although not much visited by community - reboltalk. There is also ML, where you get your answer in minutes. On altme - in seconds. AltME is social channel. There are not just gurus. REBOL3 AltME world has over 300 registered users, something like 10 always available. You have to get feeling for it to fully embrace the difference to web forum. I am on many web forums, maybe over 20, but sometimes I don’t get any answer. REBOL community support in that regard is excelent in comparison to even more popular systems. With R3 Chat system, we will get to web later, it is going to be Altme NG - both rich and web clients will be available …
Any hosting site which tells you they have problem placing REBOL on their server is run by incompetent ppl
They fear the security from false reasons - becaue REBOL is not part of default distro, and not being open-source, it might have problems to be included by default. But is any language delivered with OS as part of default install more secure and bug-free by default? As for CentOS - this is mostly Fedora/RedHat, no? It should work for CentOS imo.
Professionally I also work in multiple Big Corporates Environments. So we can understand each other. As a CTO you are the first to know that decision is based on Risk Management and ROI. If Rebol refuses to work on Java VM or .NET Runtime which are deployed everywhere how do you expect Rebol to be able to enter the arena except for a few exception (that is limited to a small service) ? Infrastructure is a huge cost and people are less and less akeen to be flexible. You talk about incompentencies, well may be, but it’s the way it is, if you expect the world to change for you instead of adapting to it, you’re doomed to fail in the Business World.
I agree JavaFX may have copied VID syntax, but the stake is no more about language but about Virtual Platforms that’s why I talk about Java VM and .NET CLR not about Java Language or .NET language. So you pretend to go with another virtual OS that would compete with these big twos ? Well this ressemble much the IOS which you admitted had been a failure. So why don’t you learn from past errors ?
Programming Languages debate are much like Religion’s Debate. As for me, I am agnostic. Rebol is great language but not great platform as long as it refuses to integrate .NET and or Java VM. Why don’t you want to create a REBNET for .NET ? Is there a technical difficulty ? Or is it in the tube ?
As long as Rebol refuses to do so, it will stay a hobbyst or admin tool for most users and the winner will be Scala. Scala users clearly claim that the reason for switching to scala was not only the language itself but above all the support for Java VM which allow to reuse the huge java libraries available and interoperate with existing java applications as well as benefiting from an existing infrastructure. Where’s the incentive for switching to Rebol ? Rebol is no more the single dynamic language on the planet, which can do dsl: Scala can also and since I am agnostic I intend to launch another blog scalatutorial.com and I will follow both languages and see who the winner will be I hope both :).
Geeks are not Business Men but if you do Business whereas you like it or not you must take into account the Market Environment. I hope R3 will more listen to it.
As you probably know, the REBOL language is the work of a single visionary, Carl Sassenrath. Carl is a strict adherent to a set of design principles that have served him well in his career. Two examples of the principles are: Avoid software bloat. Shun complexity.
Based on this, I’m confident that you will never see REBOL integrated into an enterprise platform such as Java or .NET. I don’t think I’m going out on a limb in saying that. To Carl and the community, these languages represent almost everything REBOL opposes. I think that’s an objectively true statement.
This means that REBOL won’t ever be a mainstream corporate language. Although it may sound corny, I think REBOL has loftier aspirations (e.g., changing the face of personal computing). I would be happy to see REBOL be recognized as the best personal programming language available– that would be a satisfying accomplishment.
Changing the face of personal computing ? How can you change anything when nobody knows
I’m nearly pretty sure JavaFX and Scala have “borrowed” to Carl’s ideas, isn’t it a pity that they won’t even make a public declaration to recognize it like Douglas Crockford has the honesty to do ?
Creating a Java or .NET CLR version of Rebol on TOP of Java VM or .NET CRL has nothing to do with bloating the language since the language will stay the same except it will be able to run within the Java VM and .NET CRL.
For Java VM and .NET CRL are Virtual OS, since Rebol claims to be multi-platform it should be able to run on these OS otherwise it isn’t really Multiplatform
Look, I think you look way too much into corporate sphere, like the computing world would be only JAVA and .NET. And you are you are imo wrong even for corporate sphere. I know at least 10 companies, which do use SAP for e.g. But none of them uses JAVA container, but ABAP one. There is mostly no JAVA without WebSphere in the enterprise, as there is nearly no .NET without the SharePoint.
Another point where we differ is - no - I am not doomed by not trying to integrate REBOL into .NET. 1) There are some users, who will bring it to .NET, noone ever said, that we don’t want to do it, it was just not possible with R2 architecture, but - 2) No matter what you think, as a CTO, I will not let my folks to use Python, Ruby, or even REBOL for .NET work, even if available. This is my risk management. Because coding in Python means Python code, and I would have to have at least two devs to support the code anyway. Those decisions are rather difficult to take and I will think twice before I enable another technology to our IT infrastructure, being it REBOL or anything else.
And yes - we will replace JAVA in the future, don’t worry. We let the bloat for others
I’m not sure to understand: you said no I won’t take the risk to use Python or even Rebol and then you said, we’ll replace java ?
I had been working in the past on an application with SAPBW (which uses a mix of ABAP and Java) myself and currently I’m working in Banking Sphere which uses Cobol, so sure enough I do know some companies still use old tech like ABAP or COBOL but is it because of EXISTING old code. When they have the choice they use Java or .NET. In the case of SAPBW the web server is in Java only the back end is in ABAP.
My current client in Banking looks for getting rid off their Cobol stuff which cost them a leg and an arm and precisely because less and less people know Cobol. As for SAP they still maintain business because a LOT of people including NON-IT especially BUSINESS-ORIENTED people know SAP. SAP is a Business Application that is the reason for ABAP to be maintained otherwise it would be like Cobol.
Rebol is NEW (well 10 years old is not so new but it is so unknown) and is not BUSINESS-ORIENTED. Do you really think seriously that I’m gonna tell my client to use Rebol ? I won’t even try
Rebol could have some serious BUSINESS NICHES, for example, since I mentionned SAPBW, for Business Intelligence Front-End. That was a big starting niche a few years ago, now it’s well taken. It’s not too late though but how can you use Rebol when the plugin is in eternal beta, has no connector to Oracle, SQL Server, Web Service, SAPBW, Business Object etc ?
Rebol as front end on this kind of niche is acceptable because it doesn’t induce such big risk for back-end infrastructure. The problem is Rebol is not yet very sexy for front-end or rather it is hugely difficult to build such application when you have an ugly rectangular window you cannot even customize, and there’s no component ide to help you build them. It is as if you had to build Microsoft WPF/XAML with Notepad only.
For me the fundamental reason of all this current state is, as you said Rebol is a ONE-MAN company whose vision is essentially driven by PURE TECHNICAL background. And of course he only gather around him with people of the same feather. I guess and understand Carl has some fear of Marketing (I do myself) and has no time to take some marketing coach but at least he should think about taking advice from some marketing consultant.
You said I look too much in corporate sphere ? It is the corporate sphere that drives the job and people learn a language mainly IF IT CAN BRING JOB ! Where’s the job for Rebol ?
“We will replace JAVA” - you should read between the lines
It was meant as an ambitious wish …
I thought you come from banking sphere. It is typical scheme - JAVA, Oracle. Medical systems - FireBird used very often - don’t ask me why
If you think that COBOL or ABAP are going to be prelaced anytime soon, you migt as well be very mistaken
It would already happen in the past, but did not happen. SAP is never going to redo old modules in ABAP. New stuff, which is not part of SAP/R3, so mostly their middleware infrastructure, might be JAVA, yes. As for me, as for CTO, I see no difference in price of Cobol, ABAP, JAVA programmer - we are charded the same tax per day from company, which we selected to SLA our SAP.
While I agree that I would not tell anyone from my programmers to use REBOL, I would not tell them even to use Python, Scala, anything. We are fully MS house here, and we will not let even Linux, PostGress, iven if it would work for us. You know why? Because of cost of support. I have so-so staff members, and I need at least 2 guys to support one solution, I need continuity.
As for Ugly rectangular REBOL windows - as for R2 - try to use RebGUI - still not there, but better than VID. Look at its sources. You don’t need visual editor to build GUIs in such a simple way. R3 is coming with VIVID (Vastly Improved VID
But even then - most companies move their stuff into browser mostly.
There’s very little job oportunities for REBOL, but there are some. Some REBOL folks work as freelancers too.
As I said - we did not start to implement the marketing plan yet. Without a product, what is cool new website good for? So - R3 first, then the rest
We are patient, because as a programmer, you either like REBOL, or you ignore it. And if you like it and stick with us, the oportunities will come, we just need a little bit more time.
I don’t think this analysis is very far off the mark. I think every point mentioned has been voiced at some point during the past decade.
As a new language, it is incumbent on REBOL’s proponents to make the case as to why REBOL should be a more widely adopted language. REBOL Technologies (aka Carl Sassenrath) is a strong tech outfit, but this is a marketing challenge. The marketing challenge was made much more difficult by some of the tech and business decisions.
- betting on the vision of an X-Internet future (Never heard of X-Internet? There’s part of the problem.).
- focusing on providing an end-to-end, single-vendor solution (IOS) instead of integrating with or leveraging established open-source or commercial technologies.
- creating a desktop language which functions outside the browser — features which overlap with or compete with browsers.
- not having a robust CGI or web framework as the world transitioned almost entirely to the web, instead devoting significant energy to a non-standard GUI.
- hoping that a grass-roots community would popularize a closed, proprietary language. licensing for any of the products was never clear and out in the open.
- the belief that a simple, consistent syntax packed in a small exectuable would woo developers from the complexity of web markups, protocols and platform stacks.
- the expectation that language speed and benchmarks are not especially important– i.e. Moore’s Law would make speed a non-issue.
- Mac marginalized as a target platform, while virtually extinct platforms were elevated.
- lack of out-of-the-box data storage — i.e., connection to standard databases. lack of support for Unicod chars, lack of print solution.
- reducing developer choice by forced platform independence– any distinguishing features provided by a specific brand of OS were abstracted from the language.
- lack of open-ness and/or weak communication plan. sketchy near, medium and long term goals. no language specs, lack of licensing details, out-of-date docs, silos of information and a community split across different channels and time-zones.
Despite some of these sins of the past, REBOL could still enjoy a nice future. A few of the above issues are being addressed, and that may be enough to help pull REBOL out of its slump. I think a lot of people would benefit from the promise of R3.
Thanks for this lengthy and interesting comment.
Edoc - there were many problems with R2, RT’s aproach. What is important though, is that nearly everything has changed with upcoming R3. Architecture which resembles more an OS than a language. Communications channels being MUCH more open and accessible. RT, a company which listens to developer needs. REBOL gets more fixes on daily basis than some other langs between major versions
(well, not fair, we are in alpha stage 
As for the past - few comments:
- Betting on x-internet - REBOL tends to be ahead, like Amiga was - or can you explain me, how is that even Flash and mighty browsers try to mimick the ability to run “outside the browser”?
- what is bad upon stand-alone IOS? IOS was one of the coolest things in groupware like products at that time. Don’t you remember it was also nominated for Webby awards together with Google? The problem of IOS was that it was not better supported and pushed into more enterprise companies.
- robust CGI framework? I studied tonnes of frameworks. And what are they? Just frameworks. Frameworks are nice, if they do exactly what they are supposed to do. They often tend to be rather complex, when you need to extend them - you need to adhere to them. Never heard of Cheyenne with its RSP? QM?
- licensing - problem only for die-hard open source fanatics, not a real problem if you know how to communicate with business entity (RT).
- yes, some of us believe that web full of markup mixture is - “overengineered”. Or do we all need to adhere to “status quo”? Where’s your web without JS today? Web has won - where’s the place for other languages inside the browser? Other languages face exactly the same problem here. The complexity is not in the exacutable size. It is in language design - one language for all.
- speed? JAVA is order of magnitude faster. Now how is that View’s UI might feel snappier? Apps are either fast enough for what you do, or not. What is your benchmark good for here?
- Mac marginalized? Ignoring facts? Just go to download section - Windows, Mac, Linux, then others …
- what does it mean “out-of-the-box” data storage? SQLite, mySQL, PostGress, ODBC, Oracle? Out-of-the-box, hmm, lazy to download “monstrose” 20KB mySQL protocol script?
- you are either platform independent, or not … platform specific things still possible.
- communication - big problem of the past.
The future of REBOL is R3. R3 architecture adresses almost everything. Still lots to do. I don’t know if e.g. View has any future, as nowadays IT lobotomised programmers to think that if you are not in the browser, you don’t exist
But - working in various IT fields, I know the world is more colorful than the browser - still some embedded stuff, platform native stuff. There’s a lot of possibilities where R3 could find its niche …
Other than that, your marketing analysis was always interesting and quite accurate. I just try to explain things from different perspective …
Best,
-pekr-
Hi Petr–
I think we have different perspectives, and as a CTO I’m sure you know firsthand why REBOL is not widely deployed in your enterprise. In any case, I think we both agree that the focus is on R3, and that productive/positive energy is best channeled in that direction. Developers “vote with their feet.” R3 will be our best shot at addressing the needs of all audiences.
REBOL is a very counter-culture language and has a strong independent streak. It is absurdly compact, yet has a lot of functionality and expressiveness. It’s written in C and is extremely dynamic and reflexive. It is designed to be more platform independent than Java or .NET, and it delivers higher levels of WORA. It is engineered to a degree that I doubt it could be replicated in Java or .NET without sacrificing significant characteristics of the language. Most likely, only a REBOL-like syntax would survive the port. So in many regards, porting REBOL to these languages would be a significant step down.
I acknowledge that ignoring those language platforms translates to missed opportunities. However, I credit Carl with sticking to his guns and devoting his energy and creativity to something more interesting than most of what passes for software inside the corporate firewall.
>It is engineered to a degree that I doubt it could be replicated in Java or .NET without sacrificing significant characteristics of the language.
I doubt, in fact someone had done the core of Rebol but of course such task should rather be done by RT.
Rebol was in advance 10 years ago, it risk to be retarded compared to a language like Scala which can now also build dialect while supporting Java VM.
R3 is only integrating what should have been done from the start: things like module which exist since long in other languages. And of course a more open architecture. Why didn’t he do it right from the start instead of waiting 10 years ? Will he wait and lose ten years more to realize it is mandatory to be compatible with Java VM and .NET to be able to pretend being MULTIPLATFORM ? Java is DECLINING as language but JAVA VM is becoming an OS, same for .NET now ported to Linux through MONO.
C/C++ is also declining because it is not compatible with managed platforms. Languages in the future will have nearly no chance to survive if they do not run on these OS, because OS is what it is about. Carl should not commit the same error as he did with Rebol as language, he should NOT try to transform Rebol into an OS but migrate Rebol LANGUAGE to these OS (the problem is he will tend to do so because of his past Amiga OS love-story). If he does so, I’m sure Rebol will jump into the sky as for popularity !
REBOL is a masterpiece. In many aspects. Those aspects were even strenghtened with R3. Internally, REBOL is becoming more an OS, than a simple language, and it can span world from embedded programming to other higher level areas - what makes you think, that just because Scala or other languages get some features, REBOL becomes “retarded”?
Yes, R3 is probably integrating features we wish they were available from the very beginning. On the other hand it integrates Amiga like asynchronous devices, with precisely defined isolation API. But in general - let’s not live the past mistakes. We learned from them and have R3 knocking the door of beta status - so - be happy
As for your notes of C/++, JAVA as “declining languages” vs JAVA/.NET VMs being more of hosting OS. Don’t think I don’t understand you point. In fact, it was me, who sometimes in 2000 pushed to port REBOL to AmigaDE (Tao Intent VM) and later JAVA. And why? Because of “deployment” and helping REBOL to reach new markets. Some folks were against, from various reasons, mostly because of the fear of lack of speed.
But now, as R3 is going to be 80% open source, and is very tiny package - what is the reason to port to JAVA VM, if we can port R3 itself to the target platform? Because of available libraries? Is that enough? You still ommit one aspect - as a CTO - would you let your guys to start using REBOL, Scala, Python together with your JAVA infrastructure? Because in fact - where’s your JAVA anymore? Nowhere - you have to support those new languages, no matter if they run OS-natively, or in terms of JAVA or .NET VMs.
I respect your opinion. A lot of people in the community share your passion. My opinion is that REBOL is a personal computing language– it’s very flexible and free-form, it allows a high degree of customization via DSLs, etc. Super-flexible languages like REBOL often don’t conform well in business environments– there’s little to no scaffolding, no methodology built-in to the language (OOP, frameworks), no IDE or library support.
I think shifting course to appease the enterprise market is a risky bet. I don’t think corporate environments place much value on the core principles of REBOL. The result of a porting effort might be similar to the lackluster adoption that many CLI and JVM languages face.
On the personal computing side, I’d be delighted to see great support for apache/lighttpd/cherokee/ningx, bindings for WebKit (E.g., Build a Desktop wysiwyg Editor with Webkit) and much better database support.
Thanks for the link on Webkit, seems interesting.
@Petr Krenzelok I’m not on altme (as I said my Vista refuses the connection) so I don’t know what’s happening exactly on R3 but wait for it.
As for Cobol and SAP we do agree that they won’t be replaced so sooner (though as for Cobol there are now tools to do so).
But for Rebol, it’s funny I am more optimistic than you for once
at least for front end IF ONLY Rebol was more sexy and has the available connectors to databases and webservices (those used in enterprises Oracle, SQL Server, SOAP). By sexy I also mean being able to customize the window like creating a round window for creating things like Dashboard that is mandatory. Tiny little details count, that makes a product finished, Rebol is a nearly finished product, it lacks the little polishment, I’m sure that’s not a lot to do technically that’s why it’s a pity.
Silverlight, Flash have no problem to penetrate front end in the enterprise since it doesn’t involve back-end infrastructure, so could Rebol with its plugin (as long as it isn’t in eternal beta
). It would be easier than for public website as people would not be akeen to download the plugin whereas in enterprise mass deployment can be done to unify all desktops.
As for AltME - running on Vista right now, for 2 years. What might be a problem is - are you behind a proxy? AltME can’t get thru proxy - old and bad defficiency.
As for other notes - I am rather optimistic, just in a little bit conservative way
Situation with R3 development is hundreds of percent better than with R2 in the past.
R3 GUI will adress most of R2 GUI defficiencies of R2 - the new GUI architecture/structure is based upon 10 years of experience of doing GUI in R2.
And yes - in the past I defined browser plugin as probably the most strategic product. And why? Because it significantly lowers end-user desktop deployment barrier. I hope plugin code is released, so we can plug R3 into the browser
Well great, I think everybody impatient for R3.
I also hope ADVANCED documentation will be available for example on how to do things in AGG. There are plenty of rebol’s official doc but there are mostly all hello world level so when you start on Rebol you can stay on the same level for years and years. That’s why I created this blog because I know how Rebol Newbies can feel frustrated as I am still myself in fact.
Another big frustration is that basic functions don’t work as promised. I tried esmtp, using the patch available, it doesn’t work, I can’t even make a tutorial on sending email or I would have to confess: “well, it should work but in truth it doesn’t ”
. FTP had the same issue in the past which was fixed since, I hope esmtp will be corrected in R3 or that someone posts something on rebol.org that works with free hotpop or yahoo.fr (yahoo.com not free) for example also example with stunnel because a lot of people uses Gmail which also allow sending email but through ssl.