A little bit stuff is happened in the last week, I want to report a bit about it.
The first very nice development is a bugfix in the 68k Free Pascal compiler, there was a long standing bug inside the PNG-loader. I noticed it with MUIMapparium, but it only showed on some computers, namely my Draco. Charlie tried to look into the issue long time ago, but not really found anything to fix.
But now he looked again into the issue and finally was able to track down the problem. The problem is not really fixed but there is a workaround included so now MUIMapparium works on my Draco, very nice.
MorphOS added an OpenSSL3 Amiga style library a while before, as soon that was available I started to try it to get it to work with Free Pascal on MorphOS. Sadly it did not work, always was crashing already when I tried to use even the simplest function, so I was lost. Charlie mentioned that there must be a change in our linker script of some sort to make it work and he will later care about it. But as often too much other stuff happened and this got delayed. Finally I started to implement the stuff via AmiSSL also on MorphOS. I implemented that for Amiga68k already so the switch to MorphOS is just not a big thing. But then Charlie said I should better use the openssl3.library instead and he will fix the problem with the linker script. And he did and it seemed there was even an easier solution for that and finally openssl3.library does work together with Free Pascal for MorphOS.
Charlie tried MCAmiga on his „new“ build Amiga2000. It only has a 68030 with 25 MHz and he noticed that the load is rather high (around 30%) when the program is doing nothing just waiting for the key. That is the case because it is based on old DOS behavior, therefore the key events and so on must be polled and not fired via events as it is common today. But he showed me a special function in the keyboard unit called WaitForSystemEvent() which is only available for Amiga system to be able to really wait for events and not poll them. That reduces the load when the program waits for input (to under 10%) which is nice of course.