(Note for Antoine: Thanks for sending me the link. I got confused and thought it was your blog, but apparently it is someone else’s…just mentally mixed up Albert/Antoine when I clicked through. Left a comment there.)

I’ve said repeatedly that where Rebol has most clearly failed is not an issue of technology so much as properly managing expectations. A case of consistently terrible marketing, at almost every level.

But Rebol is Carl’s “grand experiment” in language design. It is driven mostly by his curiosity and not practical concerns. That leaves a lot of people in the lurch because they encounter Rebol the way you might come to a language like Python or Ruby or C++… expecting a well-documented way of getting from point A to point B.

It’s not that at all. But it’s tried to wear the costume, and it’s clever enough to get people’s hopes up. Those hopes are quickly dashed when they interact with the self-selected community. Some nice people, some less nice people, but everyone who’s stuck with the language over the past decade is missing a lot of the energy, savvy, and charisma you’ll find with Ruby on Rails or Python etc.

Yet as a project Rebol is profound and interesting. So all the haters make me sad. If Carl was in academia, and presented it as his evolving work that grad students were hacking on, I think it would be quite beloved. But he’s been in industry for a long time and sees the world through that lens, and I think he has made Rebol look like a “shrinkwrapped product” and it’s… not.

What I wonder is–how can those who find the project interesting come together to support its unique ideas while giving meaningful feedback to steer it on course? I’ve used Rebol enough to really like it and I’d like to be able to in good conscience suggest others use it… but there are some barriers to that right now. Can we take the barriers down?

–Brian

P.S. Yep, touting number of downloads is on par with having a hit counter on your site. Very 90s. You can get away with it these days if you’re Firefox but most people can’t. Still.. rather than tear Rebol down can’t we build it up? If you use it you have to see the potential… so why not see the lousy marketing as something to improve?

Scalability :-
One thing that I consistently see missing about rebol is : How scalable is Rebol ? Has anyone done any test?

Is it scalable and how fast is it? again no answers.

can the Rebol fans send their stats ?

The documentation is really pathetic.
One thing I don’t get is that they brag about 3,000,000 downloads. If they have that many people, how come there is barely anyone contributing to it. Is it 3,000, 000 downloads by 100 people who download daily to pump up their numbers ?

I personally am convinced that rebol is not going to become a mainstream language. It will just be for some “hobbyists” rather than used by the general programming community.

The world has changed. It’s no longer the 1980s, It looks to me that Carl Sassenrath is still living in the 80s.

We are interested in free or open languages. Yes he has to make money, but his chances are very low with the model he is using. He can still have it fully open so that there is higher dissemination of the language, and charge only those who use it in a business setting. Just like Resin (caucho.com)

Why will someone use his language when they have the likes of PHP, SCALA, CLOJURE, RUBY, PERL …. and so on free and that can do everything Rebol does ?

actually Rebol is still missing a lot of features e.g multi threading and nobody knows how it scales. There is no MAJOR enterprise application developed with this language. Yes they say there is Alt me and cheyenne, but who is using them in real life under a load of over a thousand users or more ?

Its rate of adoption is so small as to be considered insignificant and will be so for ever.

cheers

The REBOL/open-source discussion strays a bit from the main topic of the article (which is about complexity). But I think most of the difference in our opinion is probably covered by the comment I posted immediately above.

I’ll send you an email to follow up further and we can chat there…

I already wrote you about the way you went around the initial limit of REBOL you saw in the product, that is not supporting an enumaretd type. Nice work and there are now some similar thinking done from other sources too. See the Reboltutorial website for yourself.

All in all I agree with you that it even if it would be nice if every product was free for us developers, it is not the way things are normally done. And they don’t have to be so either. Look around. Almost every customer product available is made and delivered on a use it as is basis. Even then some users care to own their own and use it as advertised without wanting to modify the way things are done. It’s only in the CS domain that we see things like that. And even if it is interesting to be able to modify concepts, nothing prevents us to redo similar things from scratch either.

I think Carl did and now is refactoring a great work by otself from its own experience and this is OK. He is humble enough to offer his product for free to use it and keep exploring with, while also hearing from many people that take time to suggest him some niceties to add, modify or remove from his original design. I don’t know how he will do with his future licensing but even if it keeps them as they are now, many people can benefit from his work for “some” free part that would be hard to recreate for most of us…

So Carl’s pov is debatable at any level but I can’t see why his way of doing things is not at least as good as the way others do similar software tools development.

After all, when someone needs and wants a special tool to help him getting the job done, generally he has to decide for himself if he will buy or not the product. The alternative is to recreate a similar product, find and/or get an alternative product (may be cheaper but not free either) or not use the product at all.

Sure it would be simpler to get everything for free in our short life but it’s an utopy and it is even a rarity to find many persons that offer much use of their original thinking for free, may be except in the educational domain. And Carl is not in this domain while it offers his tool for free use.

For me it’s having this ultimate choice to get for free the product for use as a personal tool, that represents the TRUE freedom, as I envisage it.

To go further and giving for free or selling his work to others is a commercial issue and I don’t want to judge anybody on this basis. I consciously thanks everybody that works this way but I can’t seriously ask everyone to agree to this way of doing things.

As a child, I used to wait anxiously from one Christmas or birthday to the next for a LEGO set. This meant I only could have one or two a year, despite dreaming of LEGO cities. Certainly my creativity was stifled by spending more time wishing than learning the lessons that come from building.

It was after seeing things like Pinball Construction Set that I began to get the inkling that virtual LEGOs could be infinite! I had the concept way before I’d ever seen anything like LDraw, and to the best of my knowledge no such programs existed at that time for the C64.

Yet a virtual LEGO is not infinite by definition (e.g. constrained only by the resource limits of your available hardware.) As we know with DRM, it’s possible to resurrect the familiar restrictions of matter into this new frontier. Then we are kind of back where we started.

Can you see the child with a virtual LEGO set having to beg their parents—who might or might not have a lot of money—to buy them more virtual LEGOs? I want the future to be guided by ethical principles that stop these situations from happening.

Also, I’d have liked to have written a LEGO CAD. I had the built-in BASIC interpreter and a machine language monitor, which I dabbled in. Little did I know most of the amazing C64 programs had been cross-compiled with proprietary assemblers and C compilers…no one told me, I was “just a kid”! So I blamed myself for not being a good enough programmer. Had I seen the source, I’d have realized how it was done—and I could have gotten started programming “for real” much earlier!

This is a piece of the historic lens through which I see this issue. How we’ve done it before doesn’t have to be the way we do it now. REBOL is mostly “gratis”, which is a nice change from when I grew up…and you can get it off the Internet, which didn’t exist for my C64. But the snowball is even bigger: programmers are now growing up with the expectation that if they receive something they can look at it to see how it works, improve it if they need to, and share it.

And to answer your question about can openers: I have a hand-operated one that I think is excellent. It’s actually one of my favorite designs…it cuts the lid of the can from the side, and has little pincers for picking off the lid built into it. I wish I could email it to you, but I can only send a picture

http://www.kuhnrikon.com/products/tools/tools.php3?id=34

Like Hostilefork like to do (as with car engine), just look at some others things arround you.

What are you eating ? What is in your plates ? Where does the food you buy in the supermarket come from ?

Is your food “free” as the software you only believe in ?

I don’t think so ! Ask the farmer. Many of them are doomed to the Monsanto world of contracts and layers (just to named it as M$ looks like to us), but it’s probably not the only one.

But I believe there is proprietary companies that make good seeds, not GMO. And they sold them to farmer who knows that the earth is the plant food, and that the plant will feed it of products that another plant will eat next year. So no need to put thousand of chemicals to have it alive. If earth is good, plants are strong. Simple enough and definitly knows by our ancestors.
But in the inductrial era, the biggest plan (was it planned?) is to make all day simple things any people knew how to do by himself in the past, become assisted by apparatus of any kind (they can sold), so complex and sophisticated.
So yes, having washing machine is great things for what it does, but just count how much electrical appliances of all kind stand in your kitchen (do you have an electrical one to open canned?).

Big software companies marketing like to say that novelties (more complex by before) is better than past. So in the same time, they said that what they have done before was not good!

If Rebol Technologies is a proprietary company, for sure, it’s a small one, and it is since the begining (even smaller today than 10 years ago). It will remain as is because it’s his philosophy “Keep it simple, not simpler”. There is no plan to make money (was Bill a good programmer? Hum bizness man for sure).
As Carl Sassenrath fight against software complexity, I could not believe he could let its company grows to an unmangeable big one. And even if its company failed (as that is the bigest fear of “free software” addicts) Rebol will be still available to us, maybe in a “free” and “open” way this time.

Just a last question “free” addicts : do you have an iPod, iPhone, Nike, KC underpants…???

To summarize : are you all so free?

I think the concept of a closed-source, ideologically pure RIA platform conceived primarily to live outside the browser is perhaps the antithesis of where the computing world is headed. To me, the road ahead looks multi-paradigm, multi-device, poly-language, and information-dense– neither simple nor pretty. REBOL is a unique and powerful tool. I hope it doesn’t become a museum piece like the Antikythera mechanism.

REBOL’s rejection of the multi-markup madness is a very good thing. I was working on a JavaScript program and trying to do CSS attributes a couple days ago, when I hit the issue that if you’re going to express something like “margin-right” you can’t use hyphens, and float is a reserved word. So to translate:

.style {float: left; margin-right: 80px; }

…this becomes in Javascript:

style = {floatCss: "left", marginRight: "80px"};

That’s to say nothing of the browser standards headache on top of that… it’s like nothing works! Web developers are basically doing a disservice by patching up their sites to work on broken browsers, they’ve allowed things like Internet Explorer to survive. Why is it so hard for people to stage revolts?

With compactness, platform-independence and design simplicity being paramount, there are compromises (depending on your pov), such as interoperability & native platform integration and raw speed. Also, partly because REBOL appeals to fervently independent “outsider” types, the community has remained quite small, and, with some exceptions, the available programs/scripts tend to reflect that. Add to that the fact that REBOL is not open-source and its closed development cycle seems to be sychronized to the Clock of the Long Now…

That is my perspective of the “dirt”. On the positive side, the people who create REBOL have bonafide principles and credentials. They’re on a mission and they refuse to release code before it’s ready. REBOL is already a joy to program now, and I expect that the next major release is going to reflect newly learned lessons as well as enough great features to remind people why they fell in love with it in the first place.

When played to its strengths (shell scripting, programming-in-the-small, network exploration), REBOL is a real joy to program with. I definitely hope REBOL is carefully domesticated from its current form in the wild. Either way, it’s definitely something every open-minded programmer should experience.