Layer Group Variants of Diegos Wireframe Plugins (Support)

jazzon · **Posted:** Mon Feb 01, 2016 8:14 am **(#1)**

LINK --> These scripts are based on Diego's WireframeMap scripts from this thread. <-- LINK

I needed to make dinasset's plugins layer group safe. In the process I made a couple extra features and made teh Lyle variant more "aspect ratio" aware. All three have been heavily tested and work. Visually they are almost identical in output, the only difference being the aspect ratio stuff on the "Lyle" variant.

My version of "Lyle" will still accept a divisor of 1, and compensates for some strangeness that it caused when used on images with non vertical borders. (Like this one)

All three have had DIEGOs mod_LCS installed as a function, so the plugin is not REQUIRED, although it is still PREFERRED. The plugin tries to use it, if it goes un-found it uses the internal code version.This appears to not work as I had planned. Will be updating this as soon as I figure out how to handle the testing of the presence of the mod_LCS plugin. For now the plugin is still a requirement, and is available in this thread (below) which is this link. Fixed in 1.1.4 (1.2.4 Lyle)

All three have a second form now (second plug in of each) which auto flattens just the working layers.

All six plugins now appear under Filters -> Render -> Wireframe.

These plugins can be safely installed along side those by dinasset.

The most current versions will be available here. (Feb 06 2016, 03:20)
Release notes ( in this thread)
Garagecoder_I Variant version 1.1.4
Garagecoder_II Variant version 1.1.4 For 5 minutes a faulty version was here. If Yours does not work, grab this one.
Lyle Variant version 1.2.4

This thread is also to be the "Official support thread" for the plug ins. So If these cause any problems, let me know here, or email me at the address given in the first few lines of the source code files. (They can be opened in any text editor.)

Kimba · **Posted:** Mon Feb 01, 2016 3:00 pm **(#2)**

Thanks Jazzon, They all work. The Lyle threw an error when I tried to test them on the same image one after the other, but after testing Lyle on a new image it worked. And thanks for putting them in a sensible spot in the filters list and for having permissions already set.

jazzon · **Posted:** Mon Feb 01, 2016 7:31 pm **(#3)**

The error was about "not pickling layer groups" right? (Please say yes ...) That's the default error for an entire group being selected instead of a layer being selected. The only thing I could do to fix that is detect the group in advance and either :
1 .. Assume You mean the top layer in the group and auto select it; or
2 .. Copy the group to another group, flatten it, and work on that layer; or,
3 .. Throw my own error.

1 and 2 make dangerous assumptions, and 3 just repeats standard behaviour. So I do not normally mess with that error.

(Honestly the permissions thing I never thought about. Must be the default behaviour of my compression software.)

Kimba · **Posted:** Mon Feb 01, 2016 10:14 pm **(#4)**

No, this is the error. As I said processing it on its own with no other layers it was fine.

jazzon · **Posted:** Tue Feb 02, 2016 12:01 am **(#5)**

Hmmmm... Funny thing is that is the same error I got which first caused me to delve into this filter.

Looking into it.

Dont suppose You could tell me the steps You took just before it occured? What was selected in the image, channel dialogue, and layer dialogue?

jazzon · **Posted:** Tue Feb 02, 2016 12:33 am **(#6)**

I believe I have the issue isolated. Updating the scripts to use distinct layer names, and ensure the merge operation starts with a certain layer selected. Fix should be up in half an hour or so.

jazzon · **Posted:** Tue Feb 02, 2016 1:10 am **(#7)**

Proposed fixes. If You could test them out Kimba I would appreciate. I did as You said You did, and ran them (all 6) back-to-back-to-back-etc. and got a similar error. It was caused by a merge down operation being performed from the wrong spot, and trying to go a layer to far.

These versions correct the merge down start point, and give all the layers unique names. I hadn't thought of various scripts being run on the same image, so the names were never altered.
(Will move them to the first post after I here from You, or a couple days without hearing anything.)

A repeat of what You said caused it, now runs with no error in my tests. Let me know how yours go.

jazzon · **Posted:** Tue Feb 02, 2016 1:30 am **(#8)**

Posted the new versions to first post anyway, as they work, even if I have missed Kimba's bug (He/She hasn't had time to let me know anything yet.)

jazzon · **Posted:** Tue Feb 02, 2016 9:39 pm **(#9)**

Finally finished the updates!

It's been 19 hours since I last reported, and I have been on these 95% of that time. Part of one bug was very elusive, and I had to track down some information which wasn't easy to find.

All scripts are fixed. This is the release notes from the Lyle varient, and all of the same information (except the bug issue) applies to all of the scripts. (different version numbers though.

Release Notes wrote:

Bug Fix in response to issue reported by Kimba on gimpchat.com forums.
When executed from within an image with many layers, the resultant image was wildly off.

While this was not a true bug, as the original code was never intended to be used on an image where layers were expected to be retained, I foresaw others reporting this as an issue as well, so modified the code in advance.

Modified format of the layer merging operation in the ancillary plugin. Plugin now uses a parasite to carry in a prefix code to apply to the layer names. Parasite is removed on successful completion, and through gimp undo operation.

Modified the main plugin to use this parasitic data for layer name prefixes. This makes null and void the changes in 1.1.1 to the layer names.

Cleaned up the code somewhat.

Modified the primary plugin to work on a new image, instead of in place. (Speed / memory optimization)

Transfer layers back to source image upon completion ensuring mode transfers as well; then adjust their parent and location as needed. Flattening occurs before transfer.

Requirement of having the mod_LCS plugin by DIEGO has been modified. The source code for that script has been merged into these files. The plugin will look for an installed version of the other plugin, and if it fails to find it, will then call on the source code version in here. Preference is always given to separate installation of mod_LCS so that if DIEGO updates it, it will trump the version coded in.

The file for the Garagecoder_II variant includes the Gold_322.pat file, which belongs in Your patterns directory.

Tested against: (All Tests Passed)

Image with one layer, no groups.
Image with many layers, no groups, top layer selected.
Image with many layers, no groups, bottom layer selected.
Image with many layers, no groups, intermediate layer selected.
All above tests inside a group, in an image with many groups.

Using:
A 66MB image, and a 2.4 GB image.

The scripts may now be safely used in any image environment.
With or without layer groups;
With or without other layers present;
In any position within a stack of layers (inside or outside a layer group).

The original source layer is unmodified.
New layers appear above the source layer.
If using one of the alternate forms of the script (Auto Collapsing) the WORK LAYERS only will be auto collapsed. All other layers in the image will remain untouched. The scripts will silently ignore orders to Flatten when used in this mode.
Scaling of the layers is done so that the end result returns to image size.

Kimba · **Posted:** Tue Feb 02, 2016 11:09 pm **(#10)**

Sorry I took so long to get back to you Jazzon . The filters now seem to be working well. The Lyle seems to throw a new colour into the pic but thats alright with me.

Garagecoders are fine.

Graechan · **Posted:** Wed Feb 03, 2016 1:18 am **(#11)**

just hobbled out of bed and logged on, these wire-frame plugins look interesting so i decided to try, but I get this message 'Procedure 'python-fu-mod-LCS' not found', where do I find LCS and where do I put it

oldmangrumpy · **Posted:** Wed Feb 03, 2016 1:23 am **(#12)**

Graechan - firstly welcome back and I hope you are getting better.

The python script is one of Dinassets's and to save us both searching through the forum posts I have attached it below -

Graechan · **Posted:** Wed Feb 03, 2016 1:45 am **(#13)**

OMG your a champ it was hard for me to think as I am not allowed to take painkillers for the leg

mackenzieh · **Posted:** Wed Feb 03, 2016 2:41 am **(#14)**

When I run Lyle's wire map (group safe) and say flatten image, it doesn't flatten it and the final result is a blurred image. Why is that?

doronf2 · **Joined:** Mar 06, 2015 **Posts:** 295

Hello friends.
Finally the effect works.
Thank you all for your help.
I am attaching two pictures.

Doron.

dinasset · **Posted:** Wed Feb 03, 2016 2:55 am **(#16)**

welcome back Graechan !
Happy to read again from You.
If you (as well as Mackenzieh) do not have layers groups, you may use directly my original versions.
Jazzon wrote his versions just for supporting layers groups.

Rod · **Joined:** May 16, 2010 **Posts:** 14709 **Location:** USA

Hey Graechan! Great to see you up and about.

SysTemX · **Joined:** Dec 08, 2013 **Posts:** 1088

Hey! Welcome back Graechan. I hope you are mending well.

jazzon · **Posted:** Wed Feb 03, 2016 7:08 am **(#19)**

Kimba I noticed the color change thing to. Maybe I am merging a layer that is meant to be delete?

GraechanHmmm. Now that I think about it I never tested without the mod installed, I ausumed the error trap would correct for it. The mod_LCS is a requirement of dinasset's original code and can be found in this post. I'll add that to the top post here as well.

OMG LOL thanks for posting that!

mackenzieh Just got back on, checking into it ... Are you sure You didn't run the Auto Collapse variant? That one ignores the flatten param, In fact I simply haven't removed it yet.

doronf2 Credits go to dinasset! All I did was make it group safe.

jazzon · **Posted:** Wed Feb 03, 2016 7:12 am **(#20)**

mackenzieh: DUUUH.... I have a block of code in the wrong spot. Was so concerned with testing the other stuff, I never tested this part. Fix will be in shortly. And I left in a piece of code from finding the other bug, which caused this one.

	Similar Topics	Replies
	Shortcuts to 2.10 documentation re: Layer & Layer-group modes	0
	Copy Layer Group (SOLVED)	0
	Does Intel Celeron B820 support Opengl3.3?	11
	*solved* GMIC has Flatpak support! But a few things missing/broken	9
	How do I script a new white layer to the bottom of the layer stack	7