DIEGO's PYTHON PLUG-INS recently released

[dinasset]

Hi all !
I think that -having passed time and tests- my recently released filters in python concerning wireframe mapping may be "promoted" to go under my main menu entry : DIEGO and not DIEGO_TEST.
Hence I repost here all three with the following notes:
1- NO CHANGES in functionality
2- MENU PLACEMENT now follows my std classification, thus: under DIEGO and a specific subfolder WIREMAP in this case
3- file name follows my std naming conventions, i.e. the prefix in front of the filter name indicates clearly were it will appear once installed: DIEGO_WIREMAP_filtername, in this case
Because of point 1, there is no need of re-download/install them, unless you -like me- want to have one single (DIEGO) menu entry, where you have already (if downloaded) DIEGO_COLOURS_Mod_LCS, DIEGO_QUANTIZE_CrystallizeN ...)

Here the 3 attachments:

DIEGO_WIREMAP_WireframeMap_Garagecoder.7z

(2.31 KiB) Downloaded 316 times

DIEGO_WIREMAP_WireframeMap_Garagecoder_2.7z

(2.47 KiB) Downloaded 301 times

I didn't receive any further comment on post #26, thus i report directly here the amended version of Lyle method.

DIEGO_WIREMAP_WireframeMap_Lyle.7z

(3.14 KiB) Downloaded 283 times

[lylejk]

Definitely appreciated your efforts, Diego. Thanks for what you gave us here.

[dinasset]

thanks Lyle.
For all those who decide to re-download/install those 3 filters, please remove previous versions in order to avoid conflicts/duplications: the internal name of the filters are the same, only the menu entry has changes and the external name of the filter, for clarity.

[Kimba]

I am confused. You have listed two Garagecoder filters and 1 Lyle but I have 2 Lyle filters and one Garagecoder.

[Issabella]

Thank you Dinasset for a great job. Congrats!
Kimba if I were you, I'd remove your old files and put these new ones.

[dinasset]

those are the current versions
maybe that:
- you kept an old version of Lyle style
- you didn't download the 2nd GC style

[SysTemX]

Dinasset - Thankyou.
Kimbra - One of the garageCoders appear to be for Python2.7.

[Kimba]

those are the current versions
maybe that:
- you kept an old version of Lyle style
- you didn't download the 2nd GC style

Yes Dinasset that's probably what happened. They all work well except that Golden option in Garagecoder2. Its hardwired for a gradient or pattern I don't have.

[Pat625]

Kimba, I downloaded the gold pattern from the thread on wireframe 2.

gold-32246.pat

(768.03 KiB) Downloaded 289 times

[Kimba]

Thanks Pat.

[jazzon]

dinasset: Comment then Small Question, then a big one...

Firstly they work great! Even on 1.8GB images (albeit slowly, but at that size I do not expect lightning...)

Question 1:
What is the source and license on the gold pattern?

and 2:
I realize it is a bit late in the game for requests but....

Your wiremap_garagecoder scripts both fail for me on the merge down function. I believe I even know why. The images I work with often have extreme numbers of layers, and as such the layers all appear in groups. Might I request one of three things?
A..Add a PF_BOOL for source layer is part of group (then in script make a copy of source and move it to top as first step);
or,
B..Add a PF_BOOL for source layer is part of group (then in script use gimp_image_reorder_parent on your layer to merge down prior to the merge.)
or,
C..Simply assume this is the case and check before the merge down and act appropriately. This has the added bonuses of being completely user transparent (and interface transparent), and making Your script fully layer group compatible I believe.

Truth is I am capable of making these changes myself, but these are Your babies. GPL or no, I hate to intrude on another's work...
(Must admit some laziness here as well ...)

[jazzon]

pseudo code for C above

Code:

moving_layer_parent = pdb.gimp_item_get_parent(moving_layer)
target_layer_parent = pdb.gimp_item_get_parent(target_layer)

if not target_layer_parent == moving_layer_parent:
  target_position = pdb.gimp_image_get_item_position(image, target_layer)
  pdb.gimp_image_reorder_item(image, moving_layer, target_layer_parent, target_position - 1) # it is - and not + I THINK

#Now do your merge

[dinasset]

answers to Jazzon:
1 - it's so much time I have that pattern in my Gimp that I do not remember the origin
2 - thanks for your indications, but I do not manage "groups of layers" in any of my filters, thus sorry to say that I will not apply modifications for that specific purpose; you obviously are totally free of creating your versions of those two filters (open source is really open, I have many of existant filters modified by me for my purposes); you may even publish your revised versions here for other chatters interested in
Thanks for your nice words

[jazzon]

dinasset: No Problem, and will do. Was just trying to avoid consumer confusion on version. I'll add jaz_ to the names of my version, internally and to files, so others can separate them.

Biggest hope was just avoiding another code project

[jazzon]

dinasset,
As referenced in the PM I sent You,

bug_test.zip

Sample image for testing and the script

(1.43 MiB) Downloaded 102 times

here is the code fix, and test file, for the Lyle variant. Tweak and tune as you see fit. It is safe to install side-by-side with yours. This one appears in the same menu You coded it for.

To test, run original "Lyle" script with divisor 1. Then try this one with divisor 1. This one does not include Layer Group settings as You said You had no need.


Having trouble uploading ... logging out and back...brb

EDIT: OK, that was weird... 7 tries and it said not uploaded. Log out and back in and here it is! Oh well.

[dinasset]

Jazzon, in the attachment I don't see your input image file to get such a strange outcome.
I performed tests with divisor = 1 (even if not in the logic of Lyle!) and got "normal" results, with my input image.
Could you try with a simple image input yourself?

[dinasset]

BTW, I made two minor modifications to that filter:
1- set minimun divisor to 2 (Lyle tutorial indicates precisely that image has to be dowsized at that step, thus setting the minimum to 1 was a mistake - procedurally speaking)
2- modified the BlackLayer creation to use dwidth and dheight, as 3d Elevation filter input parms (it was originally width*2 and height*2 on both statements, then I had to take care of abends when 3d Elevation received a value greater than 4096, but I forgot to adjust the corresponding BlackLayer size)

[jazzon]

dinasset
Hmmmm..... I just downloaded the file to my desktop and it showed two files inside. Both extracted ok for me. Trying to add one here, but upload is giving me fits. Just check back here in a while. It will be in this post.

[EDIT: First I couldn't get it to upload, then it arrives twice. I downloaded both to seperate spots, and the images exist in each and open in gimp for me, and look right. Hopefully one will work for you.]

[dinasset]

there must be something wrong, because you put an .xcf file which presumably is a multilayer gimp file, but I get only a single pasted layer (no need for .xcf to send only one layer...); anyhow, is this strange "zigzag" image your input?

this is what I see

test_image.png (954.06 KiB) Viewed 1814 times

[jazzon]

Yes, thats the image. I converted to single layer and no grpoups to make it easier for you to test. Try running Your base script against it with a Divisor of one. (I know you changed the code already ... ) In fact I still have several versions installed. Give me a sec and I will try to get an image of a bad run up here for ya.