rle extension
#65
Posted 28 December 2005 - 02:34 AM
Image locations can be determined by looking at the style for the specific area.
For example, the picture on top of the contact list is mentioned in 4004 -> 947
So we then go to 4005 -> 947 and find the msnlogo element.
element[id=atom(msnlogo)]
{
content:rcimg(255);
}so since that is rcimg(255), and we know the images are stored in 4000, replacing 4000 -> 255 will replace the msn logo on the window frame.
EDIT: Don't worry, I'll keep trying to decode the format.
This post has been edited by TheSteve: 28 December 2005 - 02:35 AM
#66
Posted 28 December 2005 - 04:52 AM
ipab, on Dec 22 2005, 04:05 PM, said:
Hope this is good :)
I hope people will like as much as those who have already tried it.
http://forums.msnfan...showtopic=13988
#67
Posted 28 December 2005 - 10:28 AM
as for your Skinning project, this was an idea i'd spoken to ipab about some time ago... its great to see someone actually putting it together :) .
-The Unknown
This post has been edited by theunknown: 28 December 2005 - 10:33 AM
#69
Posted 28 December 2005 - 11:11 AM
this new program is interesting.. will be interesting to see how it works with full blown skins.
#70
Posted 15 January 2006 - 11:03 PM
It is about the structure of the file. Don't know if you have found this so I put here.
First of all, the header:
52 4C 45 = RLE (begins always by that) 01 (don't know, perhaps for next feature, simple separator or start of the hell brrrrrrr!!!) 08 (don't know, can be 00, 08, 20) XX or XXXX (it is the lenght of the data. The lenght of the data begins here) I think the max of the data that can be 64 ko. When it is XXXX, inverse to find the lenght.;) XX ( it is the width of the file) XX (Don't know. Can be 00, 80, C0, other?? That seems to be a round number like 00, 10, 20,...) XX ( it is the height of the file)
Now, the data:
In the first time, I have thought that id a matrix color but not at all, it is where each line is cut.
We have 4 cut style.
That always begin by 00, 10, 20 or 30.
When it is 00, we have always a serie of number which is followed like 00 1A 28 3C 46 5B, etc...
The last number is always lower than the overall length of the data.
When it is 10, the lenght of the data is always XXXX. It is also a serie if number but now, there is 2 elements like that 10 1A 00 56 00 8A 00 AA 00 FB 00 16 01 .... Inversed like the lenght of the data.
I don't find for 20 and 30, I think it is when it is a small picture like 1*1,2*2, 1*7....
Now, how to cut line by line:
Now we know where we must cut each line but where we must start and how count?
It is simple, the start is where the data starts ;) so by 00 or 10.
The simplest to show you how to make is to do it on a true file.
So I have selected the RLE file n°264, it is the separatorh.
Here is the data: 52 4C 45 01 00 13 01 30 05 00 06 07 08 09 F8 02 88 03 88 03 50 03 B0 01 Now the cut: 52 4C 45 = RLE 01 00 13 = lenght of the data 01 = width 30 05 = height 00 06 07 08 09 (09 is lower than 13) 0----------> 4 (not the cut, continue Always start by 0) F8 02 5-->6 (Yeah, the cut is here) 88 03 6-->7 (next cut, start with the number of the last data) 88 03 7-->8 50 03 8-->9 B0 01 9-->end of file Result: 52 4C 45 01 00 13 01 30 05 00 06 07 08 09 F8 02 88 03 88 03 50 03 B0 01
I have found another thing too, when the lenght of the data isn't near the lenght of the file, that wants to say that the file contents more of one picture in.
That can be another picture (see the RLE file n°472, it is ExpandCollapseIcon. We have inside the plus and minus icon) et part of the same picture.
Each part of the next picture or part of picture starts with the hex FE. We have 4 numbers after who is the lenght of data, the width, number don't know and the height.
Like that:
FE 3B 06 80 05.
Now we know how is the structure of the file, we can decompress the data but this, I can't do it. I hope you understand what I said and I have helped you.;)
In a few days, I will give a zip file where I'm sure the data is correctly cut.
#72
Posted 25 January 2006 - 07:02 PM
See this post for information:
http://www.messenger-blog.com/?p=176
Mind you, this does not decode the RLE format, this merely reconstructs a (quite accurate) PNG image with alpha channel from what the internal messenger library draws.
#76
Posted 29 January 2006 - 12:36 PM
You have a part in 201.rle where it says 24 24 24 24 24 24 24 24 24 24 , the 24 thingy is a transparent horizontal line here. 84 45 FD 09 is right after the 24 stuff and if you replace that with 24 24 24 also you only have a single lined _ button left. :)
84 45 FD 09
this stuff starts with 84 and ends with 09
if you change the 5 in 0 you will only have 1 pixel on that row, when you change it to 1 it will have 2 pixels in the row, etcetra, changing the rest will sometimes make weird black-blue gradients. *-)
Where in 201.rle there seems to be a palet indicated by the B0 (C0 creates pinkyelloworange dots on the normally blue places, other values might make the program fail to create a png of it), in 198.rle there are normal hex colors used: C8D0E8 F8F8F8 8090C8
You can't replace these colors to pure colors like FF or 00, so replace FF with like F8 and 00 with like 08 in your colors.
This post has been edited by Monkey: 29 January 2006 - 02:33 PM
#77
Posted 28 April 2006 - 03:05 AM

Sign In
Register
Help
This topic is locked

MultiQuote