AltME groups: search

Davide: there is new trace mode which prints out a lot of informations.. 
for example:
>> trace on
<-- trace => unset!
>> print 1 + 2
 1: print : native! [value]
 2: 1
 3: + : op! [value1 value2]
 4: 2
<-- + => 3
3
<-- print => unset!
>>

Judging by myself, I think a lot of REBOL projects are postponed 
waiting for stuff that doesn't come

Kaj, there is a lot of work being done. A lot needs to be finished 
before R3 can be useful for anything that R2 can't do outside benchmarking 
and a few demos. I wouldn't start planning thing until the first 
beta comes out.

Of course not. I think the problem is that people expect R3 to be 
done in a hurry, since we already have R2, so what's the big hold 
up? (and personally I think the beta date of 1st August caused more 
negative talk and damage than I think Carl expected.)

I don't need a perfect R3. I just need to make some forms and other 
simple things, but I don't want to have to redo them a few months 
later

I don't need a perfect R3. I just need to make some forms... 

 <-- For that you need VID3 to be complete and the view system as 
 well. You need skins and you need a functioning event system. You 
 need keyboard input and a proper way to direct form text data to 
 objects or file storage. Furthermore since you don't want to change 
 it, we have to have the layout dialect in feature freeze, which it 
 isn't. It's currently undergoing design review, before we add more 
 features. Not simple. :-)

Than even a lot less was done so far than I thought until now. Why 
did we hear that implementing basic View would "take a day" after 
Core?

ok. that's not a good thing to say.

I understand that geeks tend to brag about their abilities, but this 
is a business that has been running for a decade, for heaven's sake

I'm sorry, but there is a breaking point for everyone

Kaj, I don't know what to say, other than a lot of work is being 
done. Please don't resort to planning around R3 yet. Use R2.

It's far too late for that. The promises started almost two years 
ago, as an extension of much earlier promises. You shouldn't be the 
one to field this criticism, but I am putting my reputation on the 
line for this in a number of places, and I hate to be pushed around

I can understand your situation and it's an unfortunate thing. The 
only thing I can say that might help, is that there have been suggestions 
about a plan to extend the alpha to the users in this world, so experienced 
REBOLers can take a look at it and "kick the tires", so to speak.


The problem with doing that is that it creates talk. Lots of talk. 
"why is this there? why does this function do that? I don't like 
this!" We have had a TON of discussions over the design of VID3 on 
how to do this and that, and we're not done with that yet. It's very 
time consuming to do that, when one man (Gabriele) wants to sit quietly 
and work out the design on his own until it's ready. It's just faster 
that way in the long run. Every time a new guy comes in, 500 questions 
need to be answered and it's usually the same 500 questions as the 
last new guy. :-)


Ideally, no questions should be asked until after about a week of 
use and start testing it right away. If there is a problem or a bug, 
consult the bug tracker or the documentation database, look at the 
discussions and the design documents and keep out of particularly 
Carl's, Gabriele's and Cyphre's hair until they crawl out from their 
holes on their own.

I'm in a big project myself, and there's no way you can ignore these 
issues. You have to deal with them as best you can. The world isn't 
perfect, and neither will R3 ever be

I use R2 in all my products. Sure you can use it. I have no plan 
to move to R3 until a good time after R3.0 is done, perhaps for R3.1.

Well, then you will have to use something else. I also have products 
in the pipeline that are impossible to do in R2. They will be done 
in R3 in about a year.

So we just have to increase all given estimates by a year or so?

You can try to make a joke out of it, but I'm not amused

You're not amused, because you made a plan around what Carl said 
(and again, I think he really shouldn't have said). I agree it's 
not funny, and it may put your reputation on the line, but use it 
as an experience in when to plan around an alpha product the next 
time, no matter what the manufacturer says.

You really want me to conclude that RT is placing REBOL outside a 
real market?

Since we're still in the alpha state, it'll probably be ignored. 
RT has no time to talk about R3 right now, so as a countermeasure, 
I and a few others do that.

I don't believe nothing can be done about this. As I suggested on 
Carl's blog, a decent stepwise release plan would already help a 
lot

I can take a more relaxed view and pick whatever I want for my first 
R3 project.

For me, I have been rescheduling to other things for a long time. 
If I can't start working on R3, it makes most sense to work on non-REBOL 
projects

I've become a Ruby programmer over the past five years, for example, 
while I wanted to be a REBOL programmer

It's good, but not inspiring such as REBOL. Code-wise, it leads to 
very similar results, but usually not as elegant (although there 
are a few exceptions). But it lacks all the data capabilities

wait, 1st august was not "beta", it was "getting the alpha to a wider 
audience". it has been suggested that it would have been bad to do 
so, so in the end it wasn't done. the dates were not unrealisting, 
they were based on the idea of releasing "early". that idea was later 
on opposed and so it wasn't done. also, the august date was based 
on the fact that me and cyphre had to be away from r3 development 
for august and september, and so we thought it was better to release 
rather than delay two months.

An early alpha followed by a beta a month later was promised.

i think it was stated clearly that the "july release" if you want 
to call it that way would not have been complete. there has been 
a delay in that it went from 15 july to basically mid august. but 
at that point, Carl has been convinced that it was not a good idea 
to release an uncomplete product. so that's what it is - saying that 
there's nothing does not help. what was supposed to be released on 
july is there already. it was just been asked not to release it because 
it is not complete.

damn, already a month ago? :-)

If I may offer some experience from our release-early, release-often 
Syllable project: we make semi-weekly development builds, even between 
our alpha releases every three or four months. I have been amazed 
at the download peaks we get when we announce a release, while the 
weekly builds were almost it already

People understand it very well when you say: this is an official 
release, and that is a development release

Yeah. I understand all the arguments given. The problem is, when 
you open yourself for feedback, you get a lot of idiots, if I may 
say so. But this obscures the fact that the silent majority are reasonable 
humans

For our download numbers, I'm also amazed at how little feedback 
we actually get, even about things that we know are broken. Often, 
we try to push people to report stuff we don't know yet, but they 
assume that they shouldn't speak up about a non-finished release

such a depressing conversation

Kaj, if you just want to create simple forms, I don't know, why you 
cannot use R2. I'm still developing for R2 and don't have any problem 
if the R3 will come a few months later. And I think, you can simply 
ask Gabriele to get access to R3-alpha world and help if you can.

R2 is a monolithic, closed product that doesn't integrate on a platform 
level. I need the R3 libraries, even if it does a lot less than R2

Kaj - you should know that I strongly disagree with Henrik's aproach. 
Although I have total trust to his developer's skills, I would probably 
fire him as a marketing responsible person :-)

I don't want RT becoming another Amiga Inc. And I will not move my 
opinion position a milimetter. Maybe I even started to being new 
marketing guy for RT! For me it is like the last call. If things 
don't work out, I don't trust RT, REBOL, whatever related anymore.

I do see the other side. The other side caused me quite a lot of 
trouble on some projects. I'm not going to talk about it, but suffice 
it to say, it had to do with prematurely released deadlines.

Last time I strictly expressed my opinion on RT's not sufficient 
communication skills, it nearly felt as I ordered Carl to update 
his blog. Reichart jumped in with disagreement, only stating the 
group is doing good progress, and then he left this group. But once 
again - I will not move a millimeter!

Kaj, it's ugly with A. Inc. I'm not even sure that real tech people 
are involved. They have ridiculed themselves beyond comprehension.

if a company produces something interesting, people will come, even 
if they've left it before. that's my theory.

Amiga Inc. generated quite a lot of interest before the first release 
of AmigaAnywhere, because of the Tao Group technology. The technology 
was interesting, but it was an "R2"; closed and hard to develop for, 
sparse documentation, etc. but impressive to look at. Today they 
have squandered their opportunity window, because their technology 
was really depending on such a window. But if they produce something 
new, something incredible, of course I'd be interested, and so will 
many other people.

it's the same thing with REBOL. "oh, I liked the language a lot, 
but I couldn't do XYZ." well, if you can do XYZ now, maybe you'll 
be interested again. I think most people who have left REBOL saw 
it as something with great potential and great fun to use. They won't 
forget that. It'll be easy to regain momentum with these people.

Henrik; I have to disagree again.  I had a friend that saved up for 
years to buy a really nice BMW motorcycle.  He was so excited.  Then 
come delivery day he was told he had to wait.  For two months he 
was told "maybe tomorrow".  Then when it was delivered and I helped 
him load it on my truck, I could see in his eyes that he just didn't 
care anymore.  He road that bike like it was something to be tolerated, 
not enjoyed.

Amiga Inc. will never do such a thing any more. I'm not exactly sure 
what its current purpose is, but it has one and it's not technology 
development

People come back in droves all the time. Look at Apple how they lost 
themselves in the 90's and now are back again, more popular than 
ever. A Danish IT company here was widely criticized for extremely 
poor customer support and lost a lot of customers. They turned around 
and people came back.

Look at Linux. Maybe you won't install Ubuntu as your main desktop 
this time around, but maybe in a year. It's lurking in the background 
and you'll make the jump to it when you want to.

I'm not sure if the REBOL community has contracted, but I suspect 
so. When I joined five years ago, there was a lot of activity, but 
it has mostly gone downhill since then

There' s no publicity any more, no books for example. And it's a 
different world now: open source languages have taken over

I can't find myself in such a defaitistic outlook. REBOL is a great, 
fundamental advancement in technology that would be awful to loose. 
Like that BMW

the only time I would really worry, would be if Carl just stopped 
developing R3. that would be a terrible thing, but that would be 
the only risk of losing it. it won't get cut by a clueless executive 
or if there is no money (because there appearently is).

matters to us in here, of course, but is not even a blip on the radar 
anywhere, never was. perhaps in that, we have the greatest strength. 
like the stealth bomber going to war. :-)

Right. And GNU/Linux doesn't even belong in that list. RT does not 
exist in a vacuum, and Carl is not an island. His funding is coming 
from somewhere, and he has a responsibility over it, and a dependency

now if REBOL was developed by Commodore, it would have been truly 
lost in a patent/license/bankrupcy issue.

I added Linux as an after thought...something that matters but doesn't 
really have the mainstream in tow  ...yet.  Like most worthwhile 
things, the mainstream doesn't really seem to have a clue but carries 
an immense momentum, for good or ill ... usually ill but everyone 
goes home paid ... paid just enough to need to stay in the mainstream 
and quietly desire something better.  :)

This is a small community ... and to break it into two with the haves 
and have nots is very bad.

We could tolerate a few weeks ... but 6 months??  That's really pushing 
it.

Now that's depressing... Ithink that Pekr's solution makes a lot 
of  sense: Use R2-Core's feature set as a first beta goal so that 
people can start using R3 for new projects without having to recode 
later. I do not know how practical or possible that would be interms 
of development though.

note about asking me to join: it's not my decision who joins and 
who does not. so you are free to ask but i need to get the ok to 
add more people. as i said, i was for just releasing to the people 
here no matter what the status. you can either fix the feature list, 
or fix the date, you can't fix both. my assumption was that for r3, 
we had fixed the date, and not the feature list (except for a couple 
things we were ok to move the date for). don't blame me if it wasn't 
released in august.

petr, if it's Core only as you say, what would be different from 
the current status of R3? just bug fixes? so why should people here 
wait for xmas for? i really don't understand. if we want release 
early, forget features, just release what you have and make it clear 
that there is no support. if you want a final product you can use 
in production, just wait until all bugs are fixed and everything 
is settled on, which could be three months or three years depending 
on what the community asks for and which bugs they find.

And R3/Core 3.1 - add rebcode as a base. Ppl already found its uses. 
IMO it was already usefull and it is a pity it was later excluded 
from the Core ...

But - if I should choose between the project staying closed, or release 
in current state, then let's release and not frustrate ppl further. 
I was trying to offer some intermediate solution, a little bit better 
organised effort ...

If a release happens to this group, people will have to understand 
that it's a work in progress and commenting on design decisions and 
general behavior after limited use is not recommended. Spend time 
in the docbase and on the bug tracker to see if an issue has been 
discussed. Current VID3 bugs are not in the tracker, due to its incomplete 
state, so there are probably 50 of those lurking.


If you find a bug that causes a crash, i.e. WinXP's bug report window 
and it isn't in the tracker, then it should of course be placed there.

Petr's proposal sounds very good to me. Except that RebCode is probably 
a luxury at this moment

Reading the R3 alpha changelog, it is clear that Core and View are 
developed together now. While I know that this leads to a balanced, 
integrated product, it is an enormous blocker to release anything. 
This is redoing a decade of existing work. There is a point that 
you must decide that the Core/View integration issues are worked 
out far enough to focus on the first development release, and that 
must naturally be Core

petr, what would having pop or imap change? (btw, ftp... that's a 
mess of a protocol. find someone wanting to write that one :P) i 
really don't understand the point. pop and imap are "trivial". VID 
is the focus. i can't wast time on pop now. that can be done later 
on. i must spend my time on what's important - VID. having an R2-like 
R3 is just crazy. R2 is already here.

Yes, at some time the decision has to be done, but I don't think 
we're talking about a developers release now. A developers release 
won't happen in the timeframe you are wishing for. So what we're 
talking is trash release (or pre-developer - by pure conincidence 
some things might work ...) so there's no need to add anything to 
this release just to give it the look of a developer release.

It looks like from what Gabriele is saying, it's at a state where 
one can write their own protocols such as pop, and imap

Petr;  I'll defend Gabriele on this.  Yes I would say most REBOL 
developers can code schemes given the motivation and due diligence 
reading the IETF docs.  I count myself amongst the clowns and after 
a few days I had a dict: scheme converted from old Jeff Kries code 
that I'm pretty proud of.  It would be a little different if that 
work had been in a vacuum, but it's not.  There are templates already 
and the RFCs are well established professional specs.  Am I dutifully 
impressed with some rebol's code, absolutely, but don't underestimate 
the power of "average" REBOL coders.

Pekr, you can say that about many things. There are many protocols. 
Which one should go in first? HTTP was logical here. Now which one 
goes next? If Gabriele spends time on FTP and not on VID, I can't 
work on skinning. It could be other things, but VID is far more important 
right now than FTP, because FTP is probably fairly trivial to do, 
but still a one-man project that should be done when Gabriele can 
be free to do that.

Henrik - as a developer, you should understand priorities.

yes, that would be a good argument, but Carl wants VID out as quickly 
as possible.

gotta go to the other world. Carl is delivering a lot of info right 
now :-)

I think there is strategy...we may not be in that loop...yet.  And 
I'll agree it's a little frustrating, mainly because we all feel 
and care deeply.  We all want "better" for REBOL.

I agree completely.  Let's push to get more people included.  There 
are people here who have put in the time and just deserve an open 
door, or at the very very least, a window seat.

yeah, that was a quick turnaround there :-)

petr: no matter how much time does vid take, either we release in 
whatever state it is, or we wait until it's finished. inventing something 
unrelated like adding pop just takes time from the end goal of R3, 
and does not help anyone. people want a toy to play it? so, let's 
just release early and often, no matter how badly it crashes or how 
much it destroys your hd - we put a warning in there. people want 
a finished product? then, just wait until it is there, and don't 
tell me you want a date for it, because noone in the world is able 
to give you one.

petr... yes... anyone is able to write imap or pop. i don't think 
i'm a god or something like that, you know.

I would prefer R3 now, even if it is unfinished, but with a clear 
plan and roadmap what it should support and what not.

Gabriele, it is not going to be a toy. By playing with it, we can 
adjust our brains to it and be ready when R3 is ready, and we can 
even help. This is the very way a community works. Either you want 
a community or you don't

A community has to be developed together with the product

I'd rather have a polished R3/core sooner than later, same goes for 
R3/View. I don't have any inside into the current development state, 
so I don't know how much it will take to get a polished product ready.

On the other hand, I feel it is really important for the community 
to see _something_. To be able to adjust, learn, be ready.

Maybe some people will see the the current state, and think that 
this isn't worth their time, but nowadays people should know about 
alpha, beta, pre-alpha, development ... versions.

And Pekr, yes, everyone _can_ code a pop / imap / ... protocol in 
rebol. You may not yet be abel to, but you can. On the other hand, 
you can't code the port system, or view basics, because wou would 
need access to the c source code.

However - this discussion is probably a moot point, as Carl seems 
to be taking some contrete aproach to cure the situation.

Ingo - but Gabriele is coding VID, which is too only a rebol level 
code, not C one ...

VID3 is not as trivial to implement as protocols, so it's a higher 
priority task.

anyway - this discussion has nothing in common with initial ideas, 
so regard it just being a normal chat, not that I try to suggest 
what is more important. Once again - my mention of protocols was 
there only for the case, if Carl would not agree to "full" release, 
so I just mentioned it, what could be completed in some sane time-frame 
and released e.g. for Christmas, nothing more. That is no more valid 
or so it seems, but we will see what next week brings to the table 
:-)

Henrik - as for FTP - I think that Gabriele and Romano and Reichart 
might be actually right, that FTP is a real mess :-)

yes, it is unfortunately a mess. I'm not sure what should be done 
with it, other than be implemented by someone who is an expert on 
the FTP protocol, rather than just implement a rudimentary one that 
follows the RFC.

Ehm, I actually worked on the FTP scheme, testing it for Romano who 
did a LOT of bugfixing on it

I think we have two competing goals, quick path to real beta and 
community access and education.  Both are important but for different 
reasons.  It is self defeating for us the community to slow down 
the final goal but at the same time we can't plan or begin the ramp 
up while there is nothing available.  I am still willing to wait 
for the open beta but by the same hand I would also welcome an open 
alpha and understand that we have to take it as it is.  This means 
not pushing on the real team with "noise" of whatever type as to 
distract them.


I worry though that the ramp up is important - if you look at rebol.org 
and scripts it seems so much of it is old, years old, or ancient 
in internet time.  That has to change as well as the wealth of documentation 
and articles has to expand drastically.  The only way that can happen 
in a timely fashion is with community access.  I'd be very willing 
to help on those fronts as best I can.

Well, what I'm trying to say. If Carl is not capable or willing to 
keep up the communication channels while he is deep into coding, 
then let the code speak for itself.

I think all of us have seen to many great tools going down the drain, 
too many unkept promises (talking globally not about a specific product 
/ company). That's why I thiink that it's better to release whatever 
there is, now, than to wait for christmas to release a product which 
isn't meant to be perfect, either.

if you're cooking for a billion people .. one cook ain't enough

also a new word to learn: "Fireside". I hope the name won't change. 
:-) I'll let you guess for a bit what it is.

if I understood it correctly, it's a tool to let users edit all mezzanines 
of R3 directly and then submit those changes to RT, where they are 
filtered through qualified people into the source tree. it's also 
chained to the documentation of the function/protocol as well as 
test scenarios.

apparently there's already a backend for it and there has been a 
small test with a prototype of it, which was proven to work very 
welll, so a full tool will be built.

kaj: it's a toy because you can do whatever you want with it, but 
not put it in production, because you'll get no support at this point. 
well, of course, if you want to go into production with no support 
and knowing that the next release may break everything... of course 
you're free to do that, but don't complain to us then :)

a development tool is useless without developers

Our input is important but as a developer I know the best path to 
get to the finish of my project, not the users waiting for it.