Comments on: Status: R3 A30 released to larger group
Today R3-A30 was released to a larger group of developers for testing. If this test goes well, we will make R3 available to even more developers in coming releases.
Included within this release are mechanisms for:
- Communicating about it - the CHAT function
- Browsing the main document wiki - the DOCS function
- Keeping it updated - the UPGRADE function
- Browsing the bug database - the BUGS function
- All developers should carefully read the banner that comes up when R3 is run.
- To test, you must have direct TCP/IP access to HTTP (not via proxy).
- There is no sandbox in this release, so...
- Use only at your own risk. This is a test prototype, not yet for building your applications.
- Please read: R3 Alpha - contains important notes.
With feedback and help from developers, we hope to advance R3 more quickly now.
This is an excellent news.I have waited for it long time.|
But Where is URL?
The URL is only available to those on the R3 alpha world.|
Congrats - very exciting!|
I do not see url at R3 alpha world .|
Micha, there is Release group on r3-alpha AltME world. Go and look in there, the download link is there. The group is public, so you should see it.|
It would be great to see some application examples and not just some screenshots.
I understand that R3 alpha is not yet available for everybody. But is it possible to publish encaped version of demo/test applications with commented source code ?
It could be interesting for people like me that want to see how Rebol language evolved but do not have to many time to spent.
I hope that a public beta will be released soon.
when will we see an osx version?
when will host code be released?|
Goldevil, there is no Encap application yet. But - last year, there was public release, which contained lot's of demo scripts. But - it used Gab's VID 3.3, whereas new releases will use Carl's VID 3.4 (which is mostly a rewrite). If you want to see how the language evolves, watch REBOL DocBase, which is public - there are many cute things in R3 when compared to R2. I think that fully public release will be out in few weeks, once we proceed with new VID. But - I can't speak for Carl of course ...|
"The group is public"
But how can we get there, into secure AltMe world?
It can not be a public until requering
the approval for registration
and the registration itself.
Why are you ignoring common practices of software development and distribution totally?
VS, the message you quote was in reply to a question asked by someone with an account on that world. The group is public in that world.
Later, there will be a public release. A common practice in software development: Release to a select group at first, so as to test the release process and shake out a few bugs before a wider release.
For instance, R3 comes with a chat client that allows private chat. The limited release allowed us to fix a few bugs that affected who could see the private messages. There should also be some tweaks to the moderation functions before a public release, or we'll be buried in spam.
These are all reasons that many companies start out with a limited release.
Brian Hawley -- a good and common sense way to proceed, even though I have to wait ; )
Mind you waiting is going to be damn hard.
I can understand VS, I'm exactly in the same situation just because it's not clear enough (sorry, I'm French so maybe that's why I'm a bit confused).
It is said R3.0 is released for those on the R3 alpha world, but when you're new, you just can't know that this group is private!(and you don't even see it on AltMe).
Moreover, when I go to the R3 wiki it is clearly said (just a copy/paste) "View.exe 2.100.30 12-Jan-2009" : No longer private. A URL will be provided.
For me, and I suppose for most of you, "No longer private" means "public". So my question is when R3Alpha will be a *real* public one?
Software releases aren't simple, I guess, huh? :-)
Long story short: There are many private (REBOL related) conversations on r3-alpha, which could cause speculations or be released by a user. The people that reside in that world are assuming that all conversations are private. All of them are long term users of AltME.
Since r3-alpha is for early R3 developers and to keep discussion relevant to development instead of for beginners or the curious, I'm not sure it will ever become public.
There has been quite some time spent on managing communication, with the RebDev system, and in the end RebDev will prove a better solution for you than letting tens of folks into the r3-alpha AltME world. It's not practical.
So... releasing R3 publicly? We'll see in a few releases. When a public release is made, you should have access to RebDev via the new CHAT command and can speak directly to Carl that way or most of the people who already reside in r3-alpha.
Feel free to join the REBOL3 AltME world though. :-) We'll answer as many R3 questions as we can there.
Henrik's long reply does not answer kib2's simple question - will it be real public one? Yes, I believe so, it will be so quite shortly.|
I had a small discussion with Henrik and a few other users on AltMe after his answer here, and I'm quiet pessimistic about R3 coming public soon.[that's the third time I hear such things in 3 discussions].
I decided to switch from Python a month ago after 5 years using it (and a few Common Lisp too, but I keep it). So making my choice was difficult, because there are plenty of programming langages out there (ok, not so much are interesting). I was not expecting to wait for an undetermined time for R3 to come.
Maybe I will come back to Rebol later, when R3 is in *real* alpha, ie for everybody and not a handful of lucky developpers. Don't blame me, I'm just giving my opinion from the "outside" non-private world.
Untill then, I have to find another candidate.
... and I simply don't believe you. You either choose tools upon their quality and potential, or you are just lurking around. ALL of us do wait. Even if I send you r3 alpha in email today, what does it change, if it is not completed yet? It is about ppl willing to cooperate to reach some target in the future. So - you are tool user, while we are tool developers and followers. Good luck ...|
Kib2, I hope that you realize that a public alpha will still be an alpha, and thus not completed yet. An R3 alpha is not a good target for any short-term projects unless those projects can afford to be rewritten later. Some stuff doesn't work yet, some will change. This is what alphas are.|
I totally understand that, but I really don't care leaving on the edge if it suits my needs. And they are simple: working on Unicode text and with antialiased fonts in my GUI, so what else but R3a ?
First, thanks for answering with my name; that's a bit weird don't you think so ? First you misunderstood something : I don't care if you to send me R3a, I was just asking myself why it was so closed, and why Carl is writting so much about it if it serves only a few of us? Take ie Python : it has produced a lot of alphas before the latest 3.0, and it was not a problem. I think it helped a lot in the developpment process. Remember that even newbies may help, because they don't have your viewpoint.
I'm a tool user, but also a tool developper, and I was pleased when others (whatever their level was) helped me debugging some things or pointing me ideas.
>>So - you are tool user, while we are tool
>>developers and followers.
I can be either user/developper, certainly not the follower : the world is not a one man show.
Kib2, I understand your frustration.
What you may not realize is that when you are in the early alpha stages, adding developers can slow down development. This goes doubly so during the design stage. You can count on doubling your development time for every one developer added, just because of management and communications overhead.
We have a wide variety of expert and newbie perspectives more than covered already. As it is, we are just getting to the point where we can handle the bugs and ideas we have been generating so far. We have been trying to get the management and communications infrastructure in place to be able to handle more developers, and we are almost there. You don't have long to wait.
Python 3.0 is not a fair comparison:
- Python already had the communication and management infrastructure in place for large-scale open source development - REBOL is in a much earlier stage of that process.
- The Python 2 to 3 change was much less ambitious than REBOL 3.
- REBOL has higher standards for language design than Python. It takes a lot of careful choice to be as clean and simple as REBOL.
By the way, Unicode text and antialiased fonts don't work yet in the R3 GUI, so R3 might be more on the potential side of the equation as far as your project is concerned. At least we have the quality.
kib2, sorry, that was my reply. I did it already once in the past. Living so much on AltME or IRC, I am used to fast "kib2: here is my answer". So instead of putting my name there, I posted by your name.
Anyway - Python3, for me, is like some R2.x to R2.x+1 changes. R3 is total rewrite of REBOL. We are adding things (as Unicode for e.g.), which we lacked, while Python had them.
As for full alpha release, I AM the one, who suggests to release fully public, but I have to respect other's decisions too. As for your "few of us". It is not just "few" - first alphas were released to something like 10 ppl, now another 20 might be in question? (r3-alpha). However - general public release might be out sooner than you might think ...
I reckon it's not just the fact that it's closed to a few testers but that we can't 'see' what's going on. Yes, there is AltMe, but I don't want to go there to get blog+1.
Inasmuch as it states on other pages that it's cross/multi platform, there's no 'clear' indication that this is being addressed. Not just for Core, but for GUI and for application deployment (way back I was just about to purchase when at the last page there was no dev version for OS X. Note: I'm not tied to any specific OS).
If a similar parallel blog, or even comments by Mr. Sassenrath, stating development status on the other platforms would make it clearer for all to see how this is progressing.
Hey, maybe even a bar graph/pie chart for the platforms and modules status would help (could this be tied in to DevBase and syphon off status data via a Rebol app?)...Hmmm...A Rebol dev status widgety type thing...Like a star shaped flowering lollipop. Or maybe I need more coffee...
The comparison with Python 3.0 was not the best I recognize, but I was just using the langage I know the best. Pekr is right here, it seems like some R2.x to R2.x+1 changes for me too (does this really deserve a new version number?). It has partial unicode support before, now they rewrote it from scratch wich is certainly some work. Also it is more consistent because Python was (and still is) stealing a lot from the others.
So take Ruby 1.9 : it is a major rewrite, and it has public alphas.
If R3a does not have antialiasing yet, what are those screenshots/videos ? http://rebol.hmkdesign.dk/files/r3/gui/ (sorry, sometimes the server seems to have problems)
I predicted a year ago that there would be people turning up who had tried R2 once, had a great experience, but found parts that would not satisfy their needs and then leave again. They would however remember the experience of using REBOL. If only R2 would do XYZ, it would be wonderful!
Now that R3 is moving closer, it will solve an enormous amount of the issues that R2 has. And those people are popping up again asking for an alpha. But we are not there yet.
I've encountered kib2's opinion before, which basically boils down to "please release an R3 alpha, or I'll use some competing language."
Well, please use a competing language, then.
I proposed at the time that you should study other languages in the meantime if R2 is not good enough for you. I still hold that opinion.
There is no way that we can adjust for opportunities that turn up, since there is so much left to do.
R3 development is steadily moving forward, now with a good built-in upgrade scheme. But as BrianH says, R3 is still Alpha software. The graphics core has some nasty unsolved bugs that stem from bugs in the C compiler used by Carl, a very tricky problem to solve. There are no tasks and no modules, two very essential parts of R3, that could redefine how many scripts are written. Large parts of the GUI system have yet to be decided how they will be done. Those decisions will also affect how the GUI is used.
I would therefore never consider it in a production environment until it's polished and shown to be hard as a diamond with an insanely great design.
But! You are in luck that Carl has decided to release a public alpha soon. This is time consuming work, that I personally would have found unnecessary at this point, although it's a good test for R3 to find glaring omissions in its design. It could probably have been done in other, more time efficient ways.
Way I see it (here I go again...:P ) is those that want the alpha, want R3 now (moi?!...) _want_ Rebol. We/They've seen the potential: We're showing our colours!
It's a bit like getting the toy in the cereal box makes it taste better, so you want it first.
Luis, I can understand you. In fact I proposed system which would show something like feature graphs and their level of completion. It could be attached to bug-base too, showing which bugs are marked to be fixed for the next release. But - we are not there. R3 so far exists for Windows only, and it will be so imo untill Carl decides to open-source host code. That will not happen sooner after GUI is in certain shape, so it will take at least few months imo.
As for RT's communication, it imo changed. We have two blogs, and also R3 section, where Carl reshapes his 1 - 2 months plans, so you at least know, what is planned to happen. I do remember few months information outages, so just let's be happy for what we have :-)
Perhaps the constant tease of R3-a is "almost here" is a little too much for us. I don't want to use R2 because I do not want to have to port any large development effort to R3.
As a result, I am stuck in limbo and have been holding off using rebol altogether. (Not because I want to)
It may have been wiser for RT to NOT talk about R3 publicly until there was at least a beta. It would have prevented people turning away from RT due to the incorrect timing of the marketing of R3.
Give a date and stick with it.
Len - when Carl goes into cage, not informing community, then we complain. If Carl communicates, then we complain. First - new alpha wes released to something like 10 ppl. Now it was extended to r3-alpha AltME group, which is something like 40 ppl. There are being made some changes/additions, to better handle higher user load, and once ready, it will go full public alpha. What is bad about such an aproach?
I am strictly agains not talking about R3 - this rewrite si complicated and long term development effort, which needs to be managed some way. I think that we got to the stage, where community is well informed. What I would like to improve in the future is - do some feature sets and milestones, displaying the level of completion etc.
We are talking about a product which is desperately trying to penetrate into commerical applications. It shouldn't be handled in such a "Hang on, it's almost here" approach. (again, and again, and again)
How can RT be taken seriously by the commercial sect?
RT should be thinking in a business manner in order to get business adoption. Be serious, commit to dates, have press releases.
Keeping it private is totally acceptable. Just keep it private. When you have a product to release, just release it.
What we have is a public altme world which is private.
We have a super private alpha world which is private in
the private altme world.
And to add insult to injury, we have RT promoting a product to the WWW that no one has access to.
Pick a date and stick with it.
Make announcements in a professional way.
Keep the development private within the elect.
When the beta is release to the general public, solicit feedback and factor it into V1.
I believe in Carl's vision, but RT really needs to work on the commercial appeal and you don't do it this way.
Len, I think you are just wrong, and maybe a bit upset, because of not being part of initial selected group of developers. I can understand it, but - if you really want serious commercial aproach, then RT would not let community to be involved at all. Product would be announced, once ready. But then - you are also not right further thru your post. RT did NOT announce release data, no commercial company does so, so how could business partners not take RT seriously?.
All information comes thru RT's blogs. Come one - even big Microsoft blogs about Windows 7 - it is simply a communication channel to those who do care, who want to eventually test early product. And there were closed betas, as well as there is a public beta of Windows 7 now. Exactly the same aproach as RT took.
Rebol.net is developer's page, REBOL.com is official one. There you have info about current stage of R3's development. The product is NOT ready yet. What does it change, if it is fully publicly released? It will be so probably soon, but you would not bet your business upon some alpha stage product, would you?
Gee, I did it once again - instead of putting my name in there, I put there name of person I reply to. So - the previous post is mine, sorry for that ...|
...maybe a bit upset, because of not being part of initial selected group of developers...
this is a 'kindergarden' argument if you don't use rebol as a toy or just as entertainment pekr.
len has made a clear statement based on commercial thoughs like I do. rebol gives me/us a
huge development advantage so I stick with it but everything has it's costs/usage relation.
I know rebol is not vaporware but it is close to be one if every commercial oriented
user is handled like a disturbing nerd...
so lerning erlang now is my second choice until rebol3 will be ready, hopefully this year.
and by the way, I interpret people heavily shouting for rebol3 at this stage (like me) as a
positive sign. you wouldn't do that if you really don't care about it or do you?
the rebol addictive factor is still high. we will see how resilient this factor will be.
I do not have the time to become part of the rebol development crew. I am not upset with the fact that rebol3 is not distributed. I am upset with the fact that the community is being strung along with "It's almost here" all the time.
If there was true honesty here, RT would have posted the REAL date along time ago and accepted the fact that there would be a potential loss of user base. As long as this game of "It's almost here" we are hesitant to move on because of such an awesome technology.
Blog.. Ok, I can wait another month, Blog.. ok I can wait another week, Blog.. ok another month is so bad. Add it all up and I believe most people who are thinking of using rebol as part of a significant development effort within their company would have cut bait a long time ago.
Have your blogs, have your alphas, solicit feed back, but don' set expectations unless you can stand behind them. And why alienate the community by this Private-> Ultra Private community that has public reference and blogs.
Do us a favor. If it's not ready yet, don't keep saying it almost is. Have an alpha, pick a date for beta, and publish the beta ready or not. Solicit feedback in beta and apply do V1. You will do far more for loyalty then what's happening now.
I agree with you completely. It's the addictiveness, efficiency, and cost reduction that I can get by using R3 that is making me wait.
I am looking at this from purely a cost and risk factor but I am now at a tipping point.
I have held of developing on a project for more then a month because of a pretty promising blog or post I read. I need to move on our initatives and want to use R3.
I have 5 developers I need to train. The project could span 4 to 5 months and I don't want to take on the cost of porting as I would not be able to justify it to our client.
If it were "TRUELY" almost here, I would wait, evaluate to see how stable and take on a calculated risk with the technology. (Hoping the announcement games would stop and it would be solid by our roll out)
I don't want to use NewLisp, python, erlang either, but given the past, I can't see anyone can justify using any version of Rebol in a commercial effort.
I do feel that we are being treated like hacks and hobbiest and not real potential commercial partners.
Len and Tom - I simply can't agree. OK, so I send you alpha by an email today - how does it change the fact, that it is still an alpha, hence non-finished product? How could you even imgine making your business dependant upon anything like non-finished product?
Len - can you please point me to blogs, which say something like "we are close"? You make it look like nearly each blog suggests something like that.
Besides that, believe me - I am your best fried here in trying to get full public release out.
you are missing the point here. I am absolutly not interested
in using a alpha or even beta. I am coding for money, this is
a different approach. Like len I am also shifted a project for
R3 release... and stop:.. before you say use R2 it's there you
have to know that R3 contains features wich are quiet important
for this project (e.g. tasks) a rewrite later is doable but
consider only the effort translating many GUI's from currently
view to gobs etc. it's better to invest this time with family.
So what I am talking here is the security of planning within a
commercial environment where the number of mistakes you can make
are small that's all. but anyway I don't want to put this further.
Once again, I absolutly agree with Len and yes pekr I appreciate
the advocacy work you do for rebol very much. At least I will wait
like Len for R3 even if it takes much longer because rebol is and
remains, even with this bad timing the most efficient language I
have ever coded. (so what does this telling you between the rows
From Len: "I do not have the time to become part of the rebol development crew."
Then the upcoming developers' release is not what you want. This is going to be an alpha release, to increase the development and testing crew. REBOL 3 will not be suitable for use in a large project for now. No blog post has claimed otherwise - quite the contrary.
It is inappropriate to make a release date for a REBOL 3 beta at this stage of development, so there is none. We are still in the alpha stage. Only when you near the end of the alpha stage is it possible to accurately estimate when a beta will be possible. Such is the nature of this kind of project: This is a programming platform, not a building.
I wish you luck in your endeavors, and look forward to working with you if you do decide to become part of the REBOL development crew.
Project started in 2006?
REBOL 3.0 Alpha Released
1-Jun-2007 2:41 GMT
"But, we need to keep it small and manageable for the next few weeks while we get it stable and polished up a bit more."
25-Oct-2007 21:53 GMT
"But, R3 is running well enough that someone contacted me recently wanting an "ok" to ship a product that uses R3. "
3-Nov-2008 21:22 GMT
"As we move toward the completion of REBOL 3.0 and it's related components, "
REBOL 3.0 Alpha Public Release
8-Jan-2008 20:54 GMT
So, now, the URL: http://www.rebol.com/rebol3/
Link does not allow download.
I took a few minutes to look through the blog. Maybe there are others. Perhaps it's wishful thinking on my part, but based on these dates and announcements doesn't it always seem like "It's almost here"?
Tom, tasks don't work yet and are due for a redesign, or at least a revisit. At the very least, its behavior will change, so any code written in the near future to take advantage of tasks will likely need some rewriting later.
I would recommend that you either write your system using the multiple process with interprocess communications for now, or use another platform in your project.
If your project is more near-term, you shouldn't use alpha software to implement it unless you are only using the functions that are already final. Tasks are not final yet.
I think the frustration speaks volumes for the work RT and crew have been doing. So honestly, hats off to you all. As you can see from my post regarding the blogs, it appears to always be right within grasp and this is the frustrating part.
Like many, I have searched for the "Secret Weapon" programming language for many years. Ureeka, I have found it. It is just a shame that there is no definative timeline. I can not use V3 nor can I make realistic plans on incorporating it in our projects because I don't know when it is coming out.
R2 is not an option because I don't want to have to port the work and what ever I choose today as an alternative will stay with our product offering for years. (That's what kills me)
As far as alpha is concerned, to me that is a label. I don't know what the current bug list looks like but it may be beta to some and a point release to others. I guess it is just based on need.
Why don't you let the community decide for themselves if they can use the release productively?
I don't know what Carl was expecting in writting his post, but there's a lot of blood and frustration now.
Maybe my fault; maybe I shouldn't have started writting my thoughts here ? I was expecting a reaction from Carl, but maybe we have to wait (an habit!) for the next post.
Len, the initial project was thought to be of lesser scope. It became clear very quickly that there was much more to do. Strangely enough, you caught the time period when the scope of the project changed: late 2007. That is when we realized that Unicode support would have major implications for the whole system, and made major redesign work necessary. There was a similar change in early 2008, when we realized that the GUI was going in the wrong direction. These things happen.
Release dates made before that time were based on insufficient information, and thus could only be speculative. This is acceptable for programmers because we are aware of the limits of knowledge, but is poor practice if you are talking to people who don't understand development processes. We have gotten Carl to stop answering that question when it can't be answered definitively (we hope).
I am sorry that you haven't been following the development process of REBOL 3 in the last year. As you say, you don't want to be part of the development crew.
Len, in this case "alpha" means that the system has planned features and subsystems that are not even through the design stage yet, but we have a lot there and mostly working already. Alpha means that code that is written for this platform must be expected to change in the future, except in specific cases. Ask first.
We have a bug tracking system up, and it contains a lot of the known bugs (all that are known to us, at least). It doesn't have much stuff related to systems that are still to come - that should be documented in the wiki (but in some cases is not yet).
The community-created portions of this project need more people working on them. This is the reason for the upcoming developers' release.
Len: "R2 is not an option because I don't want to have to port the work"
R2 isn't going away, and programs written in it will continue to run after R3 is released. We are also working hard to make the services model compatible between R2 and R3. You should be able to call R2 services from R3 and vice versa. You might not have to port as much as you think.
Not a recommendation, just a thought.
Kib2, Carl has been asked to keep the community informed of our progress, and this was the reason for this post. Don't be shy about expressing your opinion, but...
You keep comparing REBOL to programming languages where the situation is not really comparable. Python and Ruby have long-established open source communities and have already refined their community development processes, customized to their community culture. REBOL is only getting started with that and we have needed some refinement.
It is important to get the community infrastructure working properly or the development process falls apart. We have even had a few examples of this in the last year. I think we have a good model now, and our work of implementing it over the last few months will really pay off in the long run.
I have other things to say about the development models of Python and Ruby, and the effect it has on their language designs and implementations, but that is not a subject for polite company :(
For Python (I won't talk about Ruby, I'm using it sometimes only), I also have a lot to say about it, ie the fact that it has stolen a lot from the other really great langages like Haskell; but there are a lot of drawbacks in doing so, and maybe that was this point that causes the v3.0 to come, just to correct the faults.
Stealing ideas from other languages can be a good idea, but only as long as you really think about how it fits into the language. Most of the time the details have to change in the new context. Design-by-committee languages tend to do a poor job of this, and it shows. This is why I like REBOL.
Most major programming platforms are going through the same thing as REBOL, including Python, Perl, Ruby, Java, .NET and more. Between Unicode issues, parallelism, and the increasing relative cost of the programming portion of development, a lot of platforms are going for the redesign to varying extents. I've been watching the process for many of these platforms and it's been interesting to see what it says about the development philosophy of each community.
It's change-or-die time for us all.
Despite the frustrations expressed, it's great to see serious and long term interest in REBOL 3. Brian - you, Carl, and the other REBOL 3 developers have a tough job. I appreciate your comments and discussion :)|
Just my 2cents. I can understand the frustration some have with the implied release dates from blog comments when compared with other companies.
However, REBOL is not just another language. Its a new way of thinking and the most efficient programming language I've ever used.
I've too been waiting patiently for R3 b/c the program(s) I want to write will take me probably one or two years... even in REBOL. I've been looking for good cross-platform streamlined language to use for about ten years. Both scriptable and binary formats. Generally all of them are cumbersome to maintain. So I hide in a tech support role until something will enable me to do my ideas. REBOL may fit the bill and am working some with R2 just to get more familiar with it.
I am fortunate that I have the ability to wait and am not constrained by my environment. If money was burning away though on a daily basis, then another language might need to be chosen.
From Len: R2 is not an option because I don't want to have to port the work and what ever I choose today as an alternative will stay with our product offering for years...
If the REBOL company would commit to supporting REBOL2 for many years 10-15-20? after the R3 release or provide a porting mechanism then perhaps the barrier would be lowered for some to jump on board and use R2 today.
For me, I would rather have a completed R3 than an incomplete buggy version as some companies continue to thrust buggy versions into the mainstream dev world. So I will wait.
Post a Comment:
You can post a comment here. Keep it on-topic.