World: r3wp

My opinion is, that it is not me, who was off base, but it was Reichart's 
comment, which was simply inappropriate imo. But I talked to Reichart 
later privately, and he talked normally to me. It was also me, who 
reacted with explanation of my points. Well, while I used "Could 
someone please ask Carl", I used bold style, which could sound like 
I ordered RT to post a blog. I decided to use bold letters because 
sometimes the discussion here scrolls so fast, that I want it to 
be noticed by folks like Gabriele, who might connect later. btw - 
I still can see Reichart with his name here ....

ah, I can see it too now ... in user list, he is still listed as 
Reichart, his messages are listed under "none" name.

Well, historically, there is one thing which is interesting. RT's 
communication nowadays is 100% better, than it was during IOS days. 
During all that long years, I received various messages from various 
community members, supporting my point of view. I also sometimes 
asked, why that person does not state it publicly, and the answer 
often was the fear, that it is inappropriate and that such persons 
could be kind of persecuded, or that their reputation in the eyes 
of RT could be lowered. I of course remember very well my overreactions, 
and there is not excuse for such.


But - I like things being communicated. I don't like ppl passing 
by one each other, with heads down, eyes not meeting. I can criticise, 
and I can accept critique. I like the fresh air, not hiding of problems.


I am just like that - "I don't fear dragons." As for me, I even don't 
feel offended by Terry's views. I can understand both sides pov, 
and mostly what we are spending hours with here in such regard, is 
already being a wrong kind of communication. Instead of someone asking 
privately - hey, Terry, dude, what is problem in your opinion, we 
call for official user.r agendas non english man can't even properly 
follow :-) Well, whatever, I am staying ;-)

A new blog entry -- and related to R3 too :-)
http://www.rebol.net/r3blogs/0096.html

And: http://www.rebol.net/r3blogs/0095.html

I was leaving thart one as a pleasant surprise for Petr to discover 
for himself :-)

I know about it for quite some hours prior to that post :-)

OT: so Qt has resizing too :-) http://labs.trolltech.com/blogs/2007/08/09/qt-invaded-by-aliens-the-end-of-all-flicker/

Yup, communication is almost everything. I'm working at a small company 
right now, about 10 people, and even here most problems can be tracked 
down to mis(sing) communications. 
THat should tell you something

latest R3 alpha, which came a few hours ago has:

- built in VID
- built in HTTP
- more undocumented features
- about the same amount of documented features as last alpha
- dll file is 262 kb
- exe file is 336 kb


There are 225 bugreports in our database, of which perhaps half are 
fixed. Two reports were added today.

That's my status update for now. :-)

Thanks Henrik. It's good to know that REBOL 3.0 will remain small.

So what is the story with Alpha?  Are we getting a beta anytime soon. 
 I haven't heard the latest eta on some of these distros.

Thanks Henrik!

I hope VID is shaping well :-) I was quite worried, when I read your 
"quirks document" :-)

That seems to indicate that there still get more bugs detected than 
fixed compared with http://www.rebol.net/r3blogs/0094.html

btw - I had an idea, nothing concrete, but - could it be possible 
to build styles in complete visual style, almost without coding? 
Imagine visual style builder. You basically would have container 
and you would be able to place gobs there. You would have also available 
palette of handlers, which you could kind of plug into your container. 
Then you would have available testing methods for dragging, mouse 
moving, and test how your style reacts. It would kind of remind Photoshop 
work with layers. It would create tree structure and then it would 
compose a style for you as a resul, with minimal coding effort. Just 
not sure it would be possible. But - for new VID, you guys should 
thought in such an abstract way to make it possible :-)

sqlab - yes, but they also plug-in new subsystems in there, which 
were not there previously, or so I think. That can mean bug reports 
are rising, but those are new bugs due to extended codebase ... just 
my guess ...

ah, it's so nice to be designing a progress bar with sub-pixel precision. 
:-)

ah, my last message did not pop up? Well, subpixel accuracy? What 
to say? Cool :-)

sub-pixel Precision
?? New Vid?

I think it is more View kernel related than VID related?

it's simple: since the progress bar is drawn with DRAW, the horizontal 
resolution can be scaled up 10 times, so you get theoretically 10 
times better precision. this way you can still see progress on progress 
bars that update very slowly.

ah, so it is not subpixel accuracy we talked about earlier? Does 
pair allow decimals already?

no, it's handled with scaling pairs up 10x. So a 500 px progressbar 
has a 5000 pixel resolution.

hmm, but that is something different, no? It is not subpixel accuracy 
AGG allows?

you can do it in R2 draw as well:

draw [scale 0.1 0.1 box 100x100]

draws a box that appears to be 10x10

draw [scale 0.1 0.1 box 99x99]

draws a box that appears to be 9.9x9.9

So, Henrik - are you familiar with new VID to that extent, so you 
can start porting your grid? :-)

No, there are some things still missing. I'm not even sure that a 
port can be warranted as the system is very different, so it would 
take shorter time to rewrite it. I have long wanted to rewrite it 
to separate list operations from the display more.

so when approximatelly starts it be about styles only? Are other 
subsystems ready already?

we're doing both subsystems and the styles at the same time. neither 
are finished at this point.

will your Quirks doc "worries" be adressed? :-)

I'm sure they will.

hopefully decimal pair! will be supported in R3.x

Carl started his active blogging once again. Thaks for that very 
much!

http://www.rebol.net/r3blogs/0098.html-  drag&drop in R3!

A little update from alpha testing, which I joined about a week ago:

- Carl managed two releases today, the last one added a request-file.

- I've been working on some very thorough 64-bit decimal testing, 
and I found a couple of bugs. Better to find them now than after 
a public release.

- I made a little program to help getting a view on the alpha situation. 
(tests, bugs, etc.)

- I just had a talk with Carl, and he has some more to say about 
testing. He has a test script, that was used for VID2, that he wants 
to get running under R3. It should make it easier for all the testers 
to get more done faster. Testing this is big project, but there is 
progress all the time. Just 4 updates to the bug tracker yesterday, 
but more than 20 updates the day before. (Both new reports and solving/testing 
old bugs.)

When in R3 doing:
? action! reveals 129 actions.
? native! reveals 144 natives.
? logic! reveals 8 logics.
? op! reveals 18 operators.
? function! reveals 168 functions.

i repost a comment on the last Carl's blog post:
I noticed the use of the new Rich Text Format. Bravo !!! 

will have us the possibility of including sub-blocks of text with 
inheritance of the style properties to the sub-blocks ? 

something like: [bold “bla bla” [red “i' m red”]]

some infos ?

I haven't looked much into the graphics yet, so maybe someone else 
can answer that!?

that would be yet another Cyphre question :-)

Some more output from the R3 console:

>> length? system/words
== 1296

>> ? system/datatypes

SYSTEM/DATATYPES is a block of value: [end! unset! none! logic! integer! 
decimal

! percent! money! char! pair! tuple! time! date! string! binary! 
file! email! ur

l! tag! issue! bitset! unicode! vector! image! block! paren! path! 
set-path! get

-path! lit-path! dictionary! datatype! typeset! word! set-word! get-word! 
lit-wo

rd! refinement! native! action! routine! rebcode! op! closure! function! 
frame!

object! module! error! task! port! gob! event! handle! struct! library! 
utype!]

>> ? system/view/event-types
SYSTEM/VIEW/EVENT-TYPES is a block of value: [

    ignore interrupt device custom error init open close connect accept 
    read write wrote lookup ready done time show hide offset resize active 
    inactive minimize maximize restore move down up alt-down alt-up aux-down 
    aux-up key key-up scroll-line scroll-page drop-file
]

Wasn't the 14th the day R3 was going beta ??

Graham - not sure - it was first supposed to be out on 15th July, 
then Gabriele was talking about 1st of August, but that was not official 
statement. IIRC no further dates were mentioned ...

Geomol: good work! I am glad they extended the group! btw - where 
is Volker, has anyone seen him recently?

graham, that was Petr's release date. you should get an R3 exe from 
Petr... i would not attempt to run it though.

Steeve: sub blocks are possible in DRAW but not with TEXT (it complicates 
a bit caret handling; we may add that in the future but so far we 
have not needed them).

also, i think you want something like DRAW's PUSH command actually, 
I don't know if that is planned or not.

Ah, 14.8.? That was btiffin's birthday IIRC, it was posted on announce 
channel as a joke ... well, call it kind of a black humour :-)