Actually, with a little experimentation I was able to get those cuts a lot smoother while retaining the same tortured shapes. So I'll have better piece masks to work with.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Looks nice!
But I wonder if it would look clearer with white outlines for the pieces instead of black/grey? I guess it depends on the picture. This picture is quite dark, so white would probably work better there. But it would not on a brighter picture.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
They're dark gray (#202020). That seems to be optimum. Before they were #666666. That color change, in combination with tweaking the shapes, increased smoothness so I can get away with no antialiasing. That color smooths against black and does very little damage against the picture.
Now I'm arranging those shapes into a single mask image , 5 across x 11 down, 64 pixels apart corner to corner with a 32-pixel border, floodfilled outside the pieces with masking color. Each in turn will be overlaid on the picture image to fill the piece for a snapshot that will be placed in that position of the mask as it resides in memory.
A second memory bitmap created by the program will use the same arrangement to keep track of what is behind each piece on the screen, ignoring the cursor, which gets its own special treatment.
All very simple yet tricky to implement.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
To me, the biggest issue with this was cursor speed, since this is one of those apps particularly well suited to mouse operation, like MS-Paint and the like. I speeded it up by making the cursor a sprite. Changing that to an RLE sprite, which is theoretically faster, improved on that a little, and then I pretty much doubled that by doubling the increment distance upon key repeat. So it has fine and coarse movement. With some fancy coding, I could crank that up even more, if necessary, once everything else is working.
Anyway, that part done, and it tells me how to approach dragging of pieces, which is the method used in the periodic table puzzle, also designed for speed.
For now, regular work calls.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Well, it's a real link. The button for inserting links isn't working at the moment. I took that as a sign the old system was back. Anyway, I think that sample looks pretty normal.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
There's still plenty to do on it. I've got the details on how written down as notes, to be programmed and tested when I have time. There's math in a lot of places, which all has to be valid and correct, for tricky handoffs among memory bitmaps, for example
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
There seems to be advertising for jigsaw puzzles on the project page lately. I wonder what brought that on. . . .
Anyway, it's working, and I added routines to keep track of layers, since pieces get strewn around of top of other pieces, so the priority is necessary when initializing an erase buffer for moving a piece,since that consist of everything below the piece but not the piece itself. Then, Updating the buffer as the piece moves is easy.
It still has a few bugs, but getting there.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
It just struck me that the code used in this app could perhaps eventually be used for the Starship Creator app which was discussed a while ago. Instead of puzzle pieces there would be Starship-pieces which "snaps" together in the same way when placed next to each other. Then the whole picture could be saved as a bmp so you can send the completed ships to your friends. :) Perhaps not a high priority program, but it would be cool.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
That's possible, with modifications There is an existing thing called Vance's Toolkit, which is just an image file of starship parts but all TOS-era, which anyone can download from Cyngus and use to paste together virtual kitbashes. But I could put together a bunch of parts of 24th-century ships like that. How hard could that be? The advantage of Vance's Toolkit is that they are all drawn with the same line thickness and in consistent sizes, etc. so they look right when pasted together.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm starting over from the beginning with this jigsaw app with a new design that will allow more screen space for the strewn pieces-lots of extra work but hopefully better results when it's done. So far I like it a lot better.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
So far I've used cropped TrekCore screencaps. I was thinking of having the program resize or crop user-supplied images, but you get the best pictures by cropping them beforehand with a human eye to decide on best picture composition. They're all jpegs.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
So what I'm doing for the supplied pictures is downloading them as .bmp, cropping them on MS-Paint, then converting to high-quality jpeg with other software, because MS-Paint doesn't produce good enough jpeg quality.
To make cropping easy, I made a .bmp file that's a red 361 x 481 rectangle on a black background, If I set the background color for the picture to black and transparent paste to on, I can paste in that frame and move it around to compose the picture nicely before cropping. That's quick and easy.
The selection menu has a little low-key but nice animation. When you press the ENTER key, the LCARS buttons disappear, and the picture slides into place and gets cut up into jigsaw pieces, which then get strewn left and right of the work area.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I might be able to create a few images of that type by rendering 3D-models of ships and then place a public domain space image in the background. NASA has lots of those that I believe are free to use and finding free 3D models shouldn't be too hard. I can have a look and see what I find.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
That could produce some great pictures, actually, as long as you use the best meshes available and get good composition within the 360 x 380 frame of the picture. Jigsaw puzzles have a special need for artistically composed pictures. A good reference for that is National Geographic's online jigsaw page, which uses pictures from Your Shot.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
That does raise some cool possibilites, ships in front of planets, of course, but also things like a close-up of a shuttlecraft in front of one tower of the Golden Gate or skimming over a moonscape.
Any many of the ship shots should show the ship tilted a bit, because of the vertical frame for the puzzles.
By the way, I'm not including anything from the reboot in LCARS 24.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Here's One Small Step:
: http://lcars24.com/schem-Jigsaw.html
Actually, with a little experimentation I was able to get those cuts a lot smoother while retaining the same tortured shapes. So I'll have better piece masks to work with.
I 've updated that link with a new screenshot, looking much better:
: http://lcars24.com/schem-Jigsaw.html
Looks nice!
But I wonder if it would look clearer with white outlines for the pieces instead of black/grey? I guess it depends on the picture. This picture is quite dark, so white would probably work better there. But it would not on a brighter picture.
They're dark gray (#202020). That seems to be optimum. Before they were #666666. That color change, in combination with tweaking the shapes, increased smoothness so I can get away with no antialiasing. That color smooths against black and does very little damage against the picture.
Now I'm arranging those shapes into a single mask image , 5 across x 11 down, 64 pixels apart corner to corner with a 32-pixel border, floodfilled outside the pieces with masking color. Each in turn will be overlaid on the picture image to fill the piece for a snapshot that will be placed in that position of the mask as it resides in memory.
A second memory bitmap created by the program will use the same arrangement to keep track of what is behind each piece on the screen, ignoring the cursor, which gets its own special treatment.
All very simple yet tricky to implement.
To me, the biggest issue with this was cursor speed, since this is one of those apps particularly well suited to mouse operation, like MS-Paint and the like. I speeded it up by making the cursor a sprite. Changing that to an RLE sprite, which is theoretically faster, improved on that a little, and then I pretty much doubled that by doubling the increment distance upon key repeat. So it has fine and coarse movement. With some fancy coding, I could crank that up even more, if necessary, once everything else is working.
Anyway, that part done, and it tells me how to approach dragging of pieces, which is the method used in the periodic table puzzle, also designed for speed.
For now, regular work calls.
About the piece outline colors, here's a sample with a space scene:
http://i299.photobucket.com/albums/mm309/LCARS24/Sample2.jpg
Well, it's a real link. The button for inserting links isn't working at the moment. I took that as a sign the old system was back. Anyway, I think that sample looks pretty normal.
An update:
: http://lcars24.com/schem-Jigsaw2.html
Oh, LCARS can do that?!! Very convincing screengrab.
There's still plenty to do on it. I've got the details on how written down as notes, to be programmed and tested when I have time. There's math in a lot of places, which all has to be valid and correct, for tricky handoffs among memory bitmaps, for example
There seems to be advertising for jigsaw puzzles on the project page lately. I wonder what brought that on. . . .
Anyway, it's working, and I added routines to keep track of layers, since pieces get strewn around of top of other pieces, so the priority is necessary when initializing an erase buffer for moving a piece,since that consist of everything below the piece but not the piece itself. Then, Updating the buffer as the piece moves is easy.
It still has a few bugs, but getting there.
It's getting much better. It's playable, and the pieces are slanted a little when first distributed, with each straightening out as it is picked up.
It just struck me that the code used in this app could perhaps eventually be used for the Starship Creator app which was discussed a while ago. Instead of puzzle pieces there would be Starship-pieces which "snaps" together in the same way when placed next to each other. Then the whole picture could be saved as a bmp so you can send the completed ships to your friends. :) Perhaps not a high priority program, but it would be cool.
That's possible, with modifications There is an existing thing called Vance's Toolkit, which is just an image file of starship parts but all TOS-era, which anyone can download from Cyngus and use to paste together virtual kitbashes. But I could put together a bunch of parts of 24th-century ships like that. How hard could that be? The advantage of Vance's Toolkit is that they are all drawn with the same line thickness and in consistent sizes, etc. so they look right when pasted together.
I'm starting over from the beginning with this jigsaw app with a new design that will allow more screen space for the strewn pieces-lots of extra work but hopefully better results when it's done. So far I like it a lot better.
It's debugged and working beautifully.
Here it is in play with the new design, which leaves more room to spread the pieces on both sides the target area:
: http://lcars24.com/schem-Jigsaw.html
Nice! Indeed more space and greater visibility for the pieces.
Btw, how does it handle the images? Does it scale the loaded image automatically to fit or is the image cropped?
So far I've used cropped TrekCore screencaps. I was thinking of having the program resize or crop user-supplied images, but you get the best pictures by cropping them beforehand with a human eye to decide on best picture composition. They're all jpegs.
So what I'm doing for the supplied pictures is downloading them as .bmp, cropping them on MS-Paint, then converting to high-quality jpeg with other software, because MS-Paint doesn't produce good enough jpeg quality.
To make cropping easy, I made a .bmp file that's a red 361 x 481 rectangle on a black background, If I set the background color for the picture to black and transparent paste to on, I can paste in that frame and move it around to compose the picture nicely before cropping. That's quick and easy.
The selection menu has a little low-key but nice animation. When you press the ENTER key, the LCARS buttons disappear, and the picture slides into place and gets cut up into jigsaw pieces, which then get strewn left and right of the work area.
It think at good theme for the puzzle pictures would be ship shots, like this:
But I would have to find 12 or 15 really good ones and have permission to use them.
: http://i299.photobucket.com/albums/mm309/LCARS24/JS-Defiant.png
Yes, that looks good.
I might be able to create a few images of that type by rendering 3D-models of ships and then place a public domain space image in the background. NASA has lots of those that I believe are free to use and finding free 3D models shouldn't be too hard. I can have a look and see what I find.
That could produce some great pictures, actually, as long as you use the best meshes available and get good composition within the 360 x 380 frame of the picture. Jigsaw puzzles have a special need for artistically composed pictures. A good reference for that is National Geographic's online jigsaw page, which uses pictures from Your Shot.
That does raise some cool possibilites, ships in front of planets, of course, but also things like a close-up of a shuttlecraft in front of one tower of the Golden Gate or skimming over a moonscape.
Any many of the ship shots should show the ship tilted a bit, because of the vertical frame for the puzzles.
By the way, I'm not including anything from the reboot in LCARS 24.