IBrowse 2.5.2 / AmiSSL 4.4 and Images…..

[futaura]

Euro72
Multiscan
PAL
VGAOnly

graphics.library 45.27 (2018-09-20)
There is no cybergraphics.library

Thanks for your patience duga - having re-read the entire thread, I now realize that I have been repeating myself a lot!

Going back to the https://safir.amigaos.se/main.php site, I can see a problem too on my AGA boot. I presume that some of the PNG files look fine (e.g. the header/logo at top/left and the news item images), but it is all the table/page backgrounds that are not shown correctly? That is what it looks like here.

[futaura]

Euro72
Multiscan
PAL
VGAOnly

graphics.library 45.27 (2018-09-20)
There is no cybergraphics.library

Thanks for your patience duga - having re-read the entire thread, I now realize that I have been repeating myself a lot!

Going back to the https://safir.amigaos.se/main.php site, I can see a problem too on my AGA boot. I presume that some of the PNG files look fine (e.g. the header/logo at top/left and the news item images), but it is all the table/page backgrounds that are not shown correctly? That is what it looks like here.

No problem. It's easier for me to be confident that the fault is in the program now when I have two functional A1200 with different CPU's and OS versions.

Yes, it sounds like you see exactly the same problem as me.


Some stuff that might be of interest, regarding what I call the "garbled graphics" problem:
http://forum.amiga.org/index.php?topic=74037.msg843629#msg843629
ECS and RTG are fine, but not AGA: http://forum.amiga.org/index.php?topic=74037.msg844604#msg844604
https://eab.abime.net/showpost.php?p=1391719&postcount=12

Pleased to say I've fixed this particular issue. It was a bug in the page/table background render code - I had optimised this some years ago, and it was totally unrelated to more recent changes in the image decoders themselves. So, it affects JPEG, GIF and XBM, not only PNG. Images were loaded correctly - it was just that a check was missed in the render code that meant that the OS BltBitMap() was mistakenly being used on one of IBrowse's custom fastmem bitmaps, instead of IBrowse's replacement routine. A nice easy fix in the end. That's why it worked with the fastmem setting off, and also using external decoders (as datatype images go into chipram too).

This fix will be present in IBrowse 2.5.3 (http://www.ibrowse-dev.net/2.5/history.php).