Re: [Inkscape-devel] New tiling interface proposal
![](https://secure.gravatar.com/avatar/6818bab4da85a3a138cdfa35cfc7a64f.jpg?s=120&d=mm&r=g)
Yes, the Rosetta group has two basic patterns: simple rotations (restricted
to angles that evenly divide into 360) and rotation plus
reflection - I
didn't see the reflection option in your design, perhaps
I missed it.
It's in the Tiling dialogue mock-up for Radial tiling, and in the description of the work sequence for the "Add tiling type" section. I guess I should make a table with the transformation options, and add the Frieze options while I'm at it.
On the previews: I'm actually trying to keep the guides themselves as uncluttered as possible where extra indicators are concerned. That's because one look at the detailed wikipedia base tiles convinced me that they're more confusing than helpful.
As I said, there already are previews, before you create the guides. I think that they're enough to give users an idea of what's going on. Besides, the shapes themselves usually give a clue as to what the transformation will be (if it's a triangle, you know that it isn't simple translation).
The guides can be transformed like any other object, so you can resize, rotate and flip them like normal objects.
I believe the guide points feature is more useful than giving more indicators on the guides themselves: at the end of the day, when the user is designing a seamless tile, what he needs to know most is how a line going out of the tile will connect to the next tile. What happens to the corner of the base tile isn't as useful.
Here's the description again:
When a guide point is created on one side of a guide, counterparts are automatically created on all other sides of the basic guide. All changes to the original will be reflected onto the new tile. What's more, you can click and drag this guide to create a guide
segment. By snapping a node on a guide point and snapping its handle to
the end of the guide segment, and by repeating the
operation with the
node's counterpart, the user can create smooth
transitions from one tile to the next.
Guide points can be moved along the corners of the guides. When you do so, all its counterpoints are updated dynamically. I guess the current mock-ups aren't clear enough, I'll have to make a few new ones. Look at this program:
http://www.cgl.uwaterloo.ca/~csk/software/penrose/
Unlike this program, the guide points can be moved along the sides of the guide. Like this program, by dragging on some of them, a segment is created that also appears on all the guide points' counterparts. Do you understand the concept now? You can know exactly how a line going out of the tile will go back in.
As for the tiling shadow, the problem is how many tiles will be displayed. Controlling the number of tiles must be done from the tiling dialogue rather than from the tiling guide tool. I guess I could add a "show tiles outline" option to the tiling dialogue.
![](https://secure.gravatar.com/avatar/1a85cf363775fb0a88737e81fc2ac17a.jpg?s=120&d=mm&r=g)
/"On the previews: I'm actually trying to keep the guides themselves as uncluttered as possible where extra indicators are concerned. That's because one look at the detailed wikipedia base tiles convinced me that they're more confusing than helpful." /
I agree that the wallpaper descriptions in Wikipedia are cluttered and hard to follow. Makes me a little crosseyed actually.
I also agree that displaying the symmetry information should be optional.
After playing around with the descriptions in wikipedia, I think the main reason they are so cluttered (and therefore hard to follow) is because they indicate every type of symmetry in the pattern. There are several symmetries which are related. For example, a vertical reflection followed by a horizontal reflection will result in a 180 degree rotation. In the Wikipedia entry, all of these symmetry properties are represented by a symbol or a line.
Just displaying the minimum symmetry operations necessary to create the pattern, greatly reduces the information you need to display. Of course, the minimum number of operations is not necessarily unique. There may be several ways to create the pattern from the base tile. Sticking to the operations that are easy to perform mentally is maybe a good guideline.
I have made a first stab at doing this for the 17 wallpaper patterns. These may not be the best representations and there are a few patterns where I think the design could be made simpler (pgg and p2 for example) but the picture gives the basic idea.
http://inkscape.13.n6.nabble.com/file/n4642287/symmetry3.png
-- View this message in context: http://inkscape.13.n6.nabble.com/New-tiling-interface-proposal-tp4570734p464... Sent from the Inkscape - Dev mailing list archive at Nabble.com.
![](https://secure.gravatar.com/avatar/1a85cf363775fb0a88737e81fc2ac17a.jpg?s=120&d=mm&r=g)
Here is a better copy of the image in my previous post:
-- View this message in context: http://inkscape.13.n6.nabble.com/New-tiling-interface-proposal-tp4570734p464... Sent from the Inkscape - Dev mailing list archive at Nabble.com.
![](https://secure.gravatar.com/avatar/bbafda23c6de22ec5c35a50971921f21.jpg?s=120&d=mm&r=g)
On 2012-22-03 9:37 AM, veronika wrote:
Here is a better copy of the image in my previous post:
-- View this message in context: http://inkscape.13.n6.nabble.com/New-tiling-interface-proposal-tp4570734p464... Sent from the Inkscape - Dev mailing list archive at Nabble.com.
This SF email is sponsosred by: Try Windows Azure free for 90 days Click Here http://p.sf.net/sfu/sfd2d-msazure _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel
Hi Veronika,
This is just a tiny idea to improve the pattern recognition in your illustration: Spirals might be just a little bit too intensive visually. The key part is the orientation and 'spin' right? How about reducing the spins or perhaps using a shape that basically resembles a serif font comma sign [ , ] instead?
Cheers,
alex
![](https://secure.gravatar.com/avatar/1a85cf363775fb0a88737e81fc2ac17a.jpg?s=120&d=mm&r=g)
Great suggestion. Here is a revised version of the diagram using a comma instead of a spiral:
-- View this message in context: http://inkscape.13.n6.nabble.com/New-tiling-interface-proposal-tp4570734p464... Sent from the Inkscape - Dev mailing list archive at Nabble.com.
![](https://secure.gravatar.com/avatar/1a85cf363775fb0a88737e81fc2ac17a.jpg?s=120&d=mm&r=g)
Whoops, there was an error in p4g which is fixed in the following image:
-- View this message in context: http://inkscape.13.n6.nabble.com/New-tiling-interface-proposal-tp4570734p464... Sent from the Inkscape - Dev mailing list archive at Nabble.com.
![](https://secure.gravatar.com/avatar/014b988254e77ad61312267eb1d2d696.jpg?s=120&d=mm&r=g)
That was a great suggestion, definitely communicates things more clearly imho.
I do have to throw out there at this point, the proposed UI for tiling at http://wiki.inkscape.org/wiki/index.php/Tiling_tool is incredibly complex. One of the benefits to a dedicated tool is actually being able to manipulate parameters on-canvas. I'm not saying that having UI available for precision is bad, but I think being able to quickly manipulate various parameters using on-canvas controls would be more user-friendly.
Now, I'm not suggesting this as an appropriate route to go for this tool, but for some parameters (where "handles" on-canvas wouldn't work) you could possibly go the route of Gimp 2.7 in the text tool (UI shows up on-canvas) http://i.imgur.com/wyuov.png
Basically, in the end, the less dialogs that are required, the better.
Cheers, Josh
On Thu, Mar 22, 2012 at 3:12 PM, veronika <vmi@...2827...> wrote:
Great suggestion. Here is a revised version of the diagram using a comma instead of a spiral:
-- View this message in context: http://inkscape.13.n6.nabble.com/New-tiling-interface-proposal-tp4570734p464... Sent from the Inkscape - Dev mailing list archive at Nabble.com.
This SF email is sponsosred by: Try Windows Azure free for 90 days Click Here http://p.sf.net/sfu/sfd2d-msazure _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel
participants (4)
-
Aleksandar Kovač
-
Josh Andler
-
Valerie
-
veronika