Common fonts as primitives.

[Orion Pobursky]

Why do we care about this again? I came across this comment from Chris and it got me thinking:
https://www.ldraw.org/cgi-bin/ptdetail.c...04589j.dat
I think consistency within the Architecture tiles alone would make letter prims useful but I forgot why we cared enough to stop.

[Travis Cobbs]

Why do we care about this again? I came across this comment from Chris and it got me thinking:
https://www.ldraw.org/cgi-bin/ptdetail.c...04589j.dat
I think consistency within the Architecture tiles alone would make letter prims useful but I forgot why we cared enough to stop.

I think it was because it is easier to use txt2dat than to manually position primitives. Manually positioned ones might also have incorrect kerning (although I guess there's no guarantee that the kerning from txt2dat matches the original parts).

[Willy Tschager]

Why do we care about this again? I came across this comment from Chris and it got me thinking:
https://www.ldraw.org/cgi-bin/ptdetail.c...04589j.dat
I think consistency within the Architecture tiles alone would make letter prims useful but I forgot why we cared enough to stop.

We tried to squeezes the fonts into standardized quads (A REALLY STUPID IDEA) which affected kerning especially on lower case letters instead of designing the room around the font to its need.

w.

[Orion Pobursky]

We tried to squeezes the fonts into standardized quads (A REALLY STUPID IDEA) which affected kerning especially on lower case letters instead of designing the room around the font to its need.

w.

I still think these prims are a Good Idea ™ but how do we do it right?

[Willy Tschager]

I still think these prims are a Good Idea ™ but how do we do it right?

You misunderstud my comment. The prims are fine, trying to squeeze them into a quad is stupid. The current helvetica should be obsoleted the new fonts should allow kerning on A, P, l, i getting away from the blocky background (and using a higher resolution as the current fonts).

w.

[Magnus Forsberg]

IMO,
They should be treated as templates, to be used by an author as he/she sees fit. A quick way to get a correct letter.
They are not mandatory to use. Nor is it mandatory, but recommended, to use them inlined.


Completing the set/font we have. Good idea.
Adding more fonts as templates. Not so good, unless we can prove it is very commonly used.

[N. W. Perry]

IMO,
They should be treated as templates, to be used by an author as he/she sees fit. A quick way to get a correct letter.

And in addition, a way to get a scalable letter. Glyphs should be made of curved prims where appropriate, rather than tris & quads, so they can render at any size with prim substitution. They should also (and maybe they already do) use the text baseline as an origin.

[N. W. Perry]

I think consistency within the Architecture tiles alone would make letter prims useful but I forgot why we cared enough to stop.

It looks like the Architecture tiles have converted to a new typeface. Is it more about compiling a set in this new font, or retrofitting the existing tiles (and thus, also any other existing parts using that same font)?

[Orion Pobursky]

It looks like the Architecture tiles have converted to a new typeface. Is it more about compiling a set in this new font, or retrofitting the existing tiles (and thus, also any other existing parts using that same font)?

Considering that most of the tile don't use the correct font and the conversion is only 2 sets in, it's mostly for correctness and consistency.

[Orion Pobursky]

My brain isn't working today apparently. Can you submit a proposal in simple term with examples? Nothing fancy just get me to understand.

[Willy Tschager]

   

Hope this helps.

w.

[Orion Pobursky]

I get it now, thanks.
Can we think of "good" way to make the text color 16 without having to have 2 different prims for each letter? I'm thinking it's unavoidable unless these are going to be "inline intended" as mentioned elsewhere in this topic.

[Willy Tschager]

I get it now, thanks.
Can we think of "good" way to make the text color 16 without having to have 2 different prims for each letter? I'm thinking it's unavoidable unless these are going to be "inline intended" as mentioned elsewhere in this topic.

We should go the way the logos with a separeted background are done.

w.

[Orion Pobursky]

We should go the way the logos with a separeted background are done.

w.

Ok.  I like that. I think this'll be the next project once the npegholes are done.

[Magnus Forsberg]

There are 2 sets of pattern subfiles only used in the RCX 1.0 module

letaa
letab
letba
letbb
letca
letcb
letra
letrb
letxa
letxb

numb0a
numb0b
numb1a
numb1b
numb2a
numb2b
numb3a
numb3b
numb4a
numb4b
numb5a
numb5b
numb7a
numb7b

I want to make them obsolete.
Mainly because they are made using the wrong font, but also impossible to use in the RCX modules.

Or, is there anything else we could do with these?

[Vincent Messenet]

I agree with inlining+obsoleting these if they are not used anywhere else

[Vincent Messenet]

I have started working on patterns using the Cera Pro font with Medium style now commonly used by Lego and I was planning to create the characters as new primitives using txt2dat and following recommendations discussed above.

However I have further questions/points requiring consensus

1. Where should be the origin of such primitives? I don't like the origin from the current type-swiss721bt-bold prims. From my point of view the origin should be set on the vertex where:
x = 0 = x from the leftmost vertex
y = 0
z = 0 = the baseline of the font

2. How should we name the files? I started with few characters with following suggestion, having one file with the character surfaces (inside) and one file with the outside:
Lowercase e: font-cera-medium-lc-ei.dat + font-cera-medium-lc-eo.dat
Uppercase e: font-cera-medium-uc-ei.dat + font-cera-medium-uc-eo.dat
8: type-cera-medium-8i.dat + type-cera-medium-8o.dat
Copyright: font-cera-medium-copyrighti.dat + font-cera-medium-copyrighto.dat

3. What should be the dimensions of these prims? For now I produced them using Txt2Dat available in LDPE using following parameters:
Mode: Characters only, without background
Flatness: 0.2
Margin Percentage: 1.0
Font Height: 100 LDU
And then I kept them as is

Attaching samples, Please, let me know your thoughts

[Orion Pobursky]

I think the existence of Txt2Dat makes letter prims pointless. I could (if I had the time) make a web form that would return a file with Txt2Dat output. Maybe I will when I finish this absolutely massive PT update that I somehow unwittingly roped myself into.

[Magnus Forsberg]

I have started working on patterns using the Cera Pro font with Medium style now commonly used by Lego and I was planning to create the characters as new primitives using txt2dat and following recommendations discussed above.

Don't spend time creating something no one wants, since we have already decided we don't want to have more font primitives.

[Vincent Messenet]

Don't spend time creating something no one wants, since we have already decided we don't want to have more font primitives.

Ok clear, sorry if I missed that.