Jim & all,

Please see:

http://wiki.povray.org/content/Reference_Talk:Tiling_Pattern

for the proposed update.

Bill P.

Post a reply to this message

On 9/23/2016 1:50 PM, William F Pokorny wrote:
> Jim & all,
>
> Please see:
>
> http://wiki.povray.org/content/Reference_Talk:Tiling_Pattern
>
> for the proposed update.
>
> Bill P.

not so lucky with this one ... i can't seem to get wiki to render this 
properly. works fine as standalone html page in Chrome, IE, and FF. will 
have to put this one on hold until i can find out whats happening ... i 
/really/ feel this is a wiki-ism ... stay tuned

here's a screenshot of what i going for:

Post a reply to this message

Attachments:
Download 'thinslines.png' (26 KB)

Preview of image 'thinslines.png'

On 09/24/2016 10:09 AM, Jim Holsenback wrote:
> On 9/23/2016 1:50 PM, William F Pokorny wrote:
>> Jim & all,
>>
>> Please see:
>>
>> http://wiki.povray.org/content/Reference_Talk:Tiling_Pattern
>>
>> for the proposed update.
>>
>> Bill P.
>
> not so lucky with this one ... i can't seem to get wiki to render this
> properly. works fine as standalone html page in Chrome, IE, and FF. will
> have to put this one on hold until i can find out whats happening ... i
> /really/ feel this is a wiki-ism ... stay tuned
>
> here's a screenshot of what i going for:
>
OK. Strange. I see the table OK in FF on the wiki talk page.

One thing I notice looking at the html again is I set the border not 
very thick at: border="2px solid black". Probably something like 5px 
would be better, but this might not be related at all to what you see.

Please let me know if I can help.

Bill P.

Post a reply to this message

On 9/24/2016 1:37 PM, William F Pokorny wrote:
> OK. Strange. I see the table OK in FF on the wiki talk page.
>
> One thing I notice looking at the html again is I set the border not
> very thick at: border="2px solid black". Probably something like 5px
> would be better, but this might not be related at all to what you see.

the table as it appears is a bit clunky and i was going for something a 
bit sleeker looking ... did you notice that border="2px solid black" is 
having difficulty? try changing black to red or solid to dashed and see 
what i mean. btw: going from 2px to 5px works tho. i even tried using 
inline style="yada-yada" and no joy either. that is what i meant by 
wiki-ism ... somethings work some don't. since inline style is not 
considered best practice these days i created the following style:


.lined {
   margin: auto;
   background-color: transparent;
   text-align: center;
   border-bottom: 1px solid #efefef;
   border-collapse: collapse;
}

the wiki doesn't render faint grey lines that border-bottom: 1px solid 
#efefef; specifies but works great as standalone html page ... again 
with the wiki-ism

Post a reply to this message

On 09/25/2016 08:21 AM, Jim Holsenback wrote:
> On 9/24/2016 1:37 PM, William F Pokorny wrote:
>> OK. Strange. I see the table OK in FF on the wiki talk page.
>>
>> One thing I notice looking at the html again is I set the border not
>> very thick at: border="2px solid black". Probably something like 5px
>> would be better, but this might not be related at all to what you see.
>
> the table as it appears is a bit clunky and i was going for something a
> bit sleeker looking ... did you notice that border="2px solid black" is
> having difficulty? try changing black to red or solid to dashed and see
> what i mean. btw: going from 2px to 5px works tho. i even tried using
> inline style="yada-yada" and no joy either. that is what i meant by
> wiki-ism ... somethings work some don't. since inline style is not
> considered best practice these days i created the following style:
>
>
> .lined {
>   margin: auto;
>   background-color: transparent;
>   text-align: center;
>   border-bottom: 1px solid #efefef;
>   border-collapse: collapse;
> }
>
> the wiki doesn't render faint grey lines that border-bottom: 1px solid
> #efefef; specifies but works great as standalone html page ... again
> with the wiki-ism

OK. I changed to a border of 5 and dropped the border color 
specification as it seemed to be ignored in the wiki as you suggested.

I'm not though seeing anything like the image you posted. The table now 
and prior looked alright to me. I've tested the current talk page in IE, 
Edge, Chrome, Firefox and Silk. All look OK to me.

I have noticed for other tables in our wiki we have gone without borders 
- the pavement pattern for example. Doing that here too is an option.

Bill P.

Post a reply to this message

On 9/25/2016 11:10 AM, William F Pokorny wrote:
> I have noticed for other tables in our wiki we have gone without borders
> - the pavement pattern for example. Doing that here too is an option.

well ... maybe toooo i'm hung up on having wiki presentation looking as 
close as possible to the distribution, so how's this sound as an 
alternate proposal?

how about modifying the file ~scenes/textures/pattern/tiling.pov that's 
already in the distribution! i see #declare TilingPattern = Value; so 
just test that variable in a switch/case block then group like scaling 
factors together by the case test (there are several) and assign scaling 
to a new variable.

Post a reply to this message

On 9/25/2016 1:48 PM, Jim Holsenback wrote:
> how about modifying the file ~scenes/textures/pattern/tiling.pov

attached is modified file ... all you have to do is plugin TilingPattern 
and TileTypes respectively from the Tiling Pattern image in the docs. i 
plugged in a few values and i agree with what i see others ... wellll 
maybe not

give it a whirl and see if the results are what you intended

Post a reply to this message

Attachments:
Download 'tiling371.pov.txt' (4 KB)

On 9/25/2016 4:23 PM, Jim Holsenback wrote:
> attached is modified file ...

that i'll teach me to do it from memory ... attached is corrected 
version .... #case (1) | #case (2) NOT #case (1|2)

Post a reply to this message

Attachments:
Download 'tiling371.pov.txt' (4 KB)

Am 25.09.2016 um 22:37 schrieb Jim Holsenback:
> On 9/25/2016 4:23 PM, Jim Holsenback wrote:
>> attached is modified file ...
> 
> that i'll teach me to do it from memory ... attached is corrected
> version .... #case (1) | #case (2) NOT #case (1|2)

Actually it should be `#case (1) #case (2)`. Nothing in between. If you use

    #case (1)
      ANYTHING
    #case (2)
      WHATEVER
    #break

you'll effectively get `ANYTHING WHATEVER` for a value of 1, and
`WHATEVER` for a value of 2.

Post a reply to this message

On 9/25/2016 6:03 PM, clipka wrote:
> Am 25.09.2016 um 22:37 schrieb Jim Holsenback:
>> On 9/25/2016 4:23 PM, Jim Holsenback wrote:
>>> attached is modified file ...
>>
>> that i'll teach me to do it from memory ... attached is corrected
>> version .... #case (1) | #case (2) NOT #case (1|2)
>
> Actually it should be `#case (1) #case (2)`. Nothing in between. If you use
>
>     #case (1)
>       ANYTHING
>     #case (2)
>       WHATEVER
>     #break
>
> you'll effectively get `ANYTHING WHATEVER` for a value of 1, and
> `WHATEVER` for a value of 2.

doh ... tested with 2nd case and it passed. when i changed to 1st case 
it belched on |

attached is repaired version

Post a reply to this message

Attachments:
Download 'tiling371.pov.txt' (4 KB)

On 09/25/2016 07:45 PM, Jim Holsenback wrote:
> On 9/25/2016 6:03 PM, clipka wrote:
>> Am 25.09.2016 um 22:37 schrieb Jim Holsenback:
>>> On 9/25/2016 4:23 PM, Jim Holsenback wrote:
>>>> attached is modified file ...
>>>
>>> that i'll teach me to do it from memory ... attached is corrected
>>> version .... #case (1) | #case (2) NOT #case (1|2)
>>
>> Actually it should be `#case (1) #case (2)`. Nothing in between. If
>> you use
>>
>>     #case (1)
>>       ANYTHING
>>     #case (2)
>>       WHATEVER
>>     #break
>>
>> you'll effectively get `ANYTHING WHATEVER` for a value of 1, and
>> `WHATEVER` for a value of 2.
>
> doh ... tested with 2nd case and it passed. when i changed to 1st case
> it belched on |
>
> attached is repaired version

Not opposed to adding the information in the tiling.pov example.

I am opposed to this additional line in the example:

scale TileScale

which will distort what users see when exploring the tiling patterns.

Perhaps make it a comment :

// scale TileScale   // Unit square scale for this pattern if needed.

thought I recommend deleting the line.

Folks need to know the unit tile scale if wrapping the tiles around a 
cylinder or doing something like Jaime's wallpaper technique. Just 
setting up the switch is enough to provide the information assuming 
there is a mention in the documentation that is where the information 
is.

I'm also a little unsure of the new additional scale up of <5,1,5> 
you've added just below. Feels large. Does it look OK with all 27 tiling 
patterns?

Aside: There will be a similar set of 3 scalings for the pavement 
pattern, but I've just put in a pull request to add a pavement.pov 
example - so I'll go modify that pull request now instead of proposing 
wiki updates.

Will you be putting in the tiling.pov pull request yourself, or would 
you like me to do it?

Bill P.

Post a reply to this message

Jim Holsenback <spa### [at] nothanksnet> wrote:
> On 9/25/2016 6:03 PM, clipka wrote:
> > Am 25.09.2016 um 22:37 schrieb Jim Holsenback:
> >> On 9/25/2016 4:23 PM, Jim Holsenback wrote:
> >>> attached is modified file ...
> >>
> >> that i'll teach me to do it from memory ... attached is corrected
> >> version .... #case (1) | #case (2) NOT #case (1|2)
> >
> > Actually it should be `#case (1) #case (2)`. Nothing in between. If you use
> >
> >     #case (1)
> >       ANYTHING
> >     #case (2)
> >       WHATEVER
> >     #break
> >
> > you'll effectively get `ANYTHING WHATEVER` for a value of 1, and
> > `WHATEVER` for a value of 2.
>
> doh ... tested with 2nd case and it passed. when i changed to 1st case
> it belched on |
>
> attached is repaired version

First of all, thanks for supplying this test of the tiling pattern. I wasn't
aware of it, or had long since forgotten.

Something I was looking at by using that scene file was to find out how those
patterns would repeat as, well, tiles. Noticing you were rotating the original
tile to make 4 instances I wanted to see them side by side without rotation.

I replaced the scaling of box texture with the following, commenting out the
rotated boxes to put in moved boxes. I used animation to create the series of
images, which is why frame_number is here. That was put into the TilingPattern
declare also. Here's the other change at end of the box object:

#if (frame_number=19|frame_number=20)
scale TileScale /* no differences of tiling pattern 19 and 20 at this scale */
#else // some other scales do not tile repeatedly
scale TileScale/2 /* other patterns tile repeatedly except Penrose 25, 26 and 27
*/
#end
  }
  scale <5,1,5>
} // box

object {TestCase} // about to move tiles around, not rotate
object {TestCase translate <-5,0,0>}
object {TestCase translate <-5,0,-5>}
object {TestCase translate <0,0,-5>}

/*
#local YDir = 0;
#while (YDir < 360)
  object {TestCase rotate y*YDir}
  #declare YDir = YDir + 90;
#end
*/

Now a quandary about those 19 and 20 patterns. Apparently they each track
somewhat opposite (not exact) directions but are seen as identical with this
particular scale of just TileScale, the original. However it can be seen if
scaled smaller, except not able to repeat from what I've seen. I already tried
mirroring, too, but maybe not enough things to discover the repeating pattern
for individual tiles.

Had me thinking it might be one of those Penrose patterns before I realized
those were said to be the last three. Of course I don't know what any of these
are really, besides interesting!

And this was a good opportunity to say thanks to everyone still working on the
POV-Ray stuff.

Bob

Post a reply to this message

On 9/26/2016 5:38 AM, William F Pokorny wrote:
> Not opposed to adding the information in the tiling.pov example.
>
> I am opposed to this additional line in the example:
>
> scale TileScale
>
> which will distort what users see when exploring the tiling patterns.
>
> Perhaps make it a comment :
>
> // scale TileScale   // Unit square scale for this pattern if needed.
>
> thought I recommend deleting the line.

hmmm ... /thought/ this was the whole point of this addition

>
> Folks need to know the unit tile scale if wrapping the tiles around a
> cylinder or doing something like Jaime's wallpaper technique. Just
> setting up the switch is enough to provide the information assuming
> there is a mention in the documentation that is where the information is.
>
> I'm also a little unsure of the new additional scale up of <5,1,5>
> you've added just below. Feels large. Does it look OK with all 27 tiling
> patterns?

your addition ... didn't you test /all/ those out for yourself. that's 
why i threw this back into your court. some of the results looked 
questionable to me. as for scale <5,1,5> ... either that or move camera 
closer is all ... your code so do what you think is best

>
> Aside: There will be a similar set of 3 scalings for the pavement
> pattern, but I've just put in a pull request to add a pavement.pov
> example - so I'll go modify that pull request now instead of proposing
> wiki updates.
>
> Will you be putting in the tiling.pov pull request yourself, or would
> you like me to do it?

be my guest ...

Post a reply to this message

On 9/26/2016 5:52 AM, omniverse wrote:
> Something I was looking at by using that scene file was to find out how those
> patterns would repeat as, well, tiles. Noticing you were rotating the original
> tile to make 4 instances I wanted to see them side by side without rotation.

the rotation was a basic attempt to see if the single /scaled/ tiles 
indeed can be tiled and to me some do and some don't that is why i 
tossed back and hoped OP would respond. i'm missing the utility of this 
change

the original version so this scene file placed the pattern on a plane 
and the tiling of the pattern /was/ obvious.

Post a reply to this message

Am 26.09.2016 um 11:52 schrieb omniverse:

> Had me thinking it might be one of those Penrose patterns before I realized
> those were said to be the last three. Of course I don't know what any of these
> are really, besides interesting!

The Penrose tilings are a family of curious tilings discovered by Sir
Roger Penrose in the 1970's.

The Penrose tilings have special properties that make them interesting
from both an aesthetic and mathematical point of view:

To the artist, the Penrose tilings have the rare property of exhibiting
pentagonal symmetry while at first glance appearing to be repetitive
(spoiler alert: they aren't; that would be impossible).

To the mathematician, the Penrose tilings have the rare property of
being /regular/ (meaning they follow a fixed set of construction rules
rather than randomly filling the plane with the tiles available) but at
the same time /aperiodic/ (meaning that they don't repeat -- no matter
what distance and direction you translate them, there are always tiles
that don't match up).

The three tilings are actually closely related; you could transform any
one of them into the other merely by applying a few simple substitution
rules.

There are other regular aperiodic tilings, but the Penrose tilings are
arguably the most aesthetic ones.

Post a reply to this message

I've just noticed that at the bottom of the table on the Wiki discussion
page the tilings 25 to 27 are referred to as "penrose tiling"; since
they are named after Sir Roger Penrose, they should be written with an
uppercase P.

Post a reply to this message

On 09/26/2016 07:15 AM, Jim Holsenback wrote:
> On 9/26/2016 5:38 AM, William F Pokorny wrote:
>
> hmmm ... /thought/ this was the whole point of this addition
>
...
> your addition ... didn't you test /all/ those out for yourself. that's
> why i threw this back into your court. some of the results looked
> questionable to me. as for scale <5,1,5> ... either that or move camera
> closer is all ... your code so do what you think is best
>
> be my guest ...

OK. I'll put in a pull request to update tiling.pov.

What we have today is OK & looks good if all people want to do is tile a 
floor with one of the patterns.

Where we want to do more with tiling and pavement patterns, it is often 
necessary to know the basic tile's scaling. It is this latter 
'information' we are aiming to make easily accessible somewhere.

These scalings do not necessarily mean repeatability of the unit square 
after scaling - though that is often the case.

I'll change penrose to Penrose in tiling.pov.

FYI. Jérôme created the tiling and pavement patterns and is our expert.

Bill P.

Post a reply to this message

On 9/27/2016 9:12 AM, William F Pokorny wrote:
> FYI. Jérôme created the tiling and pavement patterns and is our expert.

yes of course i'm aware of that ... when i said "your code" i was 
referring to the SDL (scaling) code not the application code and ....

FYI: I was the one who added the documentation for those patterns :-)

Post a reply to this message

Jim Holsenback <spa### [at] nothanksnet> wrote:
> On 9/27/2016 9:12 AM, William F Pokorny wrote:
> > FYI. Jérôme created the tiling and pavement patterns and is our expert.
>
> FYI: I was the one who added the documentation for those patterns :-)

Obviously you and others have been very busy!

I hadn't seen the pavement patterns yet (that I could recall) so when looking
for a sample scene, not finding one, I thought the templates subfolders were
empty since no pov file showed up trying File|Open.

That had me looking at the actual subfolders to finally see everything via web
browser!

Almost speechless. Wow. Simply wow. I had no idea that much work had been done
on the documentation and examples. I was probably about to start a long hiatus
from POV-Ray during the time these things were beginning to develop for online
help.

Using POV less, mainly to keep a couple graphical charts going, is boring.
There's so much more to it!
Thanks to C. Lipka for getting me into UberPOV recently when I noticed it being
talked about.
Oh, and thanks for that Penrose explanation. I had seen something on TV about it
years ago, so knew of the person/name, but only remembered it as a geometric
curiosity.

A tremendous thanks to all for the efforts.

Bob

Post a reply to this message