r3wp [groups: 83 posts: 189283]
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search
 

World: r3wp

[!REBOL3-OLD1]

Steeve
21-Mar-2009
[12139x5]
another one bug, if the size of the gob is previously set to 0x0, 
size-text is entering in an eternal loop.
gob: make gob! [size: 0x0 text: ["test"]]
size-text gob
i wonder how th
i wonder how the VID can work currently
size-text is used by some styles
Actually, not.

In VID, Size-text is only used to change the y coordinate of a gob, 
which means the bug has never been traced.
Henrik
21-Mar-2009
[12144x2]
yes, I've not noticed the problem before. Now that I look at it, 
it seems to be constantly about 2 chars too narrow and a tad too 
short in height.
actually less than 2 chars.
[unknown: 5]
21-Mar-2009
[12146]
That  sounds like a very high priority bug to me.
Steeve
21-Mar-2009
[12147]
I agree
Henrik
21-Mar-2009
[12148]
probably something Cyphre needs to fix.
Steeve
21-Mar-2009
[12149]
Could you post the 2 bugs ?
Henrik
21-Mar-2009
[12150]
Yes, however I think they need to be more exact. It looks like it 
could be related to the old rich text bug I saw a long time ago.
Steeve
21-Mar-2009
[12151]
because of that i can't progress in my rich-text editor ;-)
Henrik
21-Mar-2009
[12152]
exact = test with different font sizes and string lengths to see 
if the error is constant.
Steeve
21-Mar-2009
[12153]
the error is constant, i tried several font and several para
Henrik
21-Mar-2009
[12154x3]
http://curecode.org/rebol3/ticket.rsp?id=345&cursor=3

The original report.
OTOH it may not be related.
ok, thanks. I'll write up the report.
Steeve
21-Mar-2009
[12157]
thanks
Henrik
21-Mar-2009
[12158]
btw. are there any margins to adjust? could it be that?
Steeve
21-Mar-2009
[12159x2]
i don't think so, the margin and origin are too low to explain the 
gap
i tried with a margin of 2x2 and an origin of 2x2, but what is missing 
is 30-50 pixels
Henrik
21-Mar-2009
[12161]
then the error is not constant. I experienced an error of about 2-4 
pixels here.
Steeve
21-Mar-2009
[12162x3]
try with a big font
hmmmm.... seems you're right. By removing origin and margin, i got 
the correct size.
Arghhhh, it was not a bug actually
Forget the bug Henrik, thanks for your help.

A good evolution would have size-text taking in account origin and 
margin
Henrik
21-Mar-2009
[12165]
no problem and yes, that could be a good idea.
Ammon
22-Mar-2009
[12166]
Heh, Steeve.  You are working on a rich text editor too, huh? 


Are you using a single face, gob and draw, faces and gobs, just gobs 
or something entirely different?
Steeve
22-Mar-2009
[12167]
1/ I mix several gobs of all types (effect draw text color). Basicaly, 
drawings can be mixed with texts.

2/ only gobs which are modified are reconstructed, it should be fast 
to insert modify text.

3/ The structure of gobs is build using an external  grammar (with 
parse), depending of the source file type. The parser convert the 
source text into a DOM structure,  which is converted into a gobs 
structure further usings a style grammar (external too).

4/ It will be a verstatil richtext editor. By changing of external 
grammars (basically a set of parse rules), we change the rendering.

5/ The grammar for a specific markup langage will be tiny. Currently, 
the grammar of makedoc2 is less than 2 Kb.

6/ In theory, easy conversion between different makup language could 
be done by the engine (makedoc <-> postscript <-> pdf ...)

7/ After makedoc2 rules will be ok, I'll build a grammar for rebol 
sources of course.
Ammon
22-Mar-2009
[12168]
Sounds like a pretty awesome engine.  Can't wait to play with it. 
=D
Steeve
23-Mar-2009
[12169]
Is there a way to complement bitsets in R3 ?
I liked the NOT proposal in parse, but it's not done yet.
Ammon
23-Mar-2009
[12170]
Theoretically you could build a complement set in R2 and just copy 
the bitset to your R3 script but I think that will cause you some 
problems with unicode, which, as I understand it, is why complement 
has been removed in R3.
Steeve
23-Mar-2009
[12171x2]
Currently, i do this, but...

no-chars: make bitset! 256
foreach c "^/^-" [append no-chars c]
no-chars: complement no-chars
geez...
drop and off don't work in rich-text
Henrik
23-Mar-2009
[12173]
known bug
Steeve
23-Mar-2009
[12174x2]
Fuck !
sorry
Henrik
23-Mar-2009
[12176x2]
:-) I'm not sure if there is a workaround.
Generally we'll have to talk to Cyphre one day. His bug list is getting 
quite large.
Steeve
23-Mar-2009
[12178x3]
workaround...
bye bye my rich-text engine
another one project fail, my speciality....
Ammon
23-Mar-2009
[12181]
Drop and Off?  What are you talking about?
Steeve
23-Mar-2009
[12182]
used to discard previous commands in rich-text blocks (like bold, 
underline, italic)
Ammon
23-Mar-2009
[12183]
Ah...  I'm sure they'll fix that eventually...
Steeve
23-Mar-2009
[12184]
eventually, mwahahahaha
Ammon
23-Mar-2009
[12185]
I'm building my rich text editor on the gob/face level so that won't 
be bothering me. Getting this parser right is proving to be tricky 
though!
Steeve
23-Mar-2009
[12186x3]
eh ? i don't understand why you are not impacted...
when i talk about rich-text, i mean gob/text, so you are impacted
me too, i create my own gobs