I've mostly got rhlaunch working under UAE, the only issue I have is that on some games I get the following error:

IMG_20140622_075845_zpsa9d5ad6f.jpg

I'm tipping I need some form of backup KS1.3 rom loaded for certain games, however I have no idea how to do so, any advice?

Cool_amigaN, is it possible for you to post a step by step updated guide of what you have done just in case others want to do the same?

The more detailed information we have, the better.

[ Edited by madCrow 22.06.2014 - 06:07 ]

Quote:

http://www.whdload.de/docs/en/need.html

Sorry jPV,

I already know I need to use a different version of kickstart, my issue is how do I implement differing KS versions in UAE? Is it possible, for example, to have a backup KS for the software that needs it?

Your problem was that you need different kickstart images for WHDLoad, not for UAE. You need those kickstart images in your Amiga setup where WHDLoad can look them for the games which need those. Read the WHDLoad docs how to have several different versions available and how to name them.

Having different kickstarts for UAE won't help you here. But if you want those too, it's possible to make different config files and provide them to UAE when starting it. Can't check just now what was the option, but -f or something. But as said, this won't help the problem you described which is purely just WHDLoad issue.

Thanks jPV, copied all the nessecary ROMS and relocation files to the emulated A1200 and renamed the files accordingly and everything is working perfectly!

Once again, thank you very much for your time (and patience).

Now to install iGame.

One more question...

Is it possible to configure E-UAE so that the mouse pointer stays within the confines of the emulation screen? As when the mouse pointer hits either the left or right hand side edges another mouse pointer that goes outside the emulation screen and makes things very confusing.

Also, how can I make my desktop look prettier? The standard WB3.1 is a little bland and lacking in color.

[ Edited by madCrow 24.06.2014 - 14:05 ]

@madCrow

There are lots of Workbench (3.x) enhancing packs available: From AmigaSYS (my old time favorite) to ClassicWB and ofc the most notable AmiKit.

However, the more you skin/edit/add third party programs to the default environment, the more bloatware it gets (emulation gets "heavier", more cpu cycles are needed etc).

Frankly, the only reason I need 3.x is to boot straight to WHDload. If that's the case for you as well, then I suggest to have the more stripped version of classic AOS as possible.

Hi,
I'm Waitsnake from Germany and I'm new to MOS and hope to learn a bit more about this operating system family and also amiga systems. As kid I start with C64 and than unfortunately go direct to PC (Win & Linux) and later to Mac and I miss all the Amiga stuff. ;-(


As mentioned in this topic before I also tried to install iGame, UAELoad, WHDLoad, E-UAE with WB 3.1 and so forth according to an manual I found here: E-UAE_installation_on_MorphOS.

I use MOS 3.6 with Chrysalis pack 3.6.1 beta installed. It comes with E-UAE 0.8.29-WIP4 from 01.07.2014.

I had the same issues that WB 3.1 was not able to install when using as virtual hard disk a directory from MOS (filesystem2=rw,DH0:System:HardDisks/W orkbench,1) and it allways seems to hangup the complete emulation at around 12% of install progress when the installer start to create folders and icons on the file system.

In the logfile I saw some strange error messages that seems to correlate with the problem: „FILESYS: UNKNOWN PACKET 1f".

I try to look for the source code to find this error message. I only could found the source code of an older version (0.8.28) linked to this page here: www.amigaforever.com.

In the file filesys.c You see that 0x1f=31 is defined as ACTION_INHIBIT and the big switch inside the function handle_packet() does not handle this type and the switch will run into the default-case that only returns false and print this error message.

In the file history the author even mentioned: „Known limitations: * Does not support ACTION_INHIBIT (big deal).“. So that seems an important issue of E-UAE anyway.


Than I found some more information about that ACTION_INHIBIT here: Amiga DOS packets (see at "dos.library/ACTION_INHIBIT" ). It seems to be an kind of flush action that writes all file buffers to hard disc when requested, like the sync command in Linux systems.


With a google search I found the source code of fs-uae. Inside the winuaechangelog.txt you found that they implement this ACTION_INHIBIT with Beta 1 and with Beta 2 they fix some infinite loop that this function causes.

Well I really don't know much about that filesystem stuff and I'm not sure if the missing implementation of ACTION_INHIBIT in the E-UAE of MOS really causes the hangup of the emulation. I just hope that someone with more knowledge will pay a bit attention on it and take a look if it is a issue or not.

The source code of ACTION_INHIBIT in FS-UAE can be found here: filesys.cpp

A good workaround of this problem is to avoid file system access and use a hardfile (hdf) instead of that. Unfortunely file system is needed to connect iGame in MOS to the games installed in E-UAE.

[ Edited by Waitsnake 21.07.2014 - 21:56 ]

Hi Waitsnake and welcome!
Your error occours when you try install Workbench 3.1 directly from Ambient. If you perform the same action inside E-UAE installation goes flawless. You can create Workbench and Games directories directly from Ambient but you have to install Workbench using E-UAE emulation.
Inside the guide read page 22, from line 1 to 7: if you are using iGame launcher you can use a hardfile (hdf) with virtual hard disk (filesystem) provided that both are "2". So you can have: hardfile2= and filesystem2=. In fact usingiGame launcher you must use virtual hard disk only to store WHDLoad games while Workbench can be installed inside a hardfile.

Edit:
sorry Waitsnake, I was reading your post again: you wrote that emulation hangs at 12%. So , does this problem happen when you are installing Workbench 3.1 after you have started E-UAE?

[ Edited by Jambalah 22.07.2014 - 19:48 ]

Hi Jambalah,
thanks for your answer and sorry. I guess I miss the point in my last post.

1.) I tried to install WB 3.1 in E-UAE on MOS and I was using filesystem2 to mount directories from MOS into E-UAE (for System and Games).
I tried the installation a couple of times and always at 12% the complete emulation hangs, well even the MOS hangs more or less (windows moving around still reacts, but any other action hangs as well). I take a closer look at installation and see that first 11% of installation do nothing else than interact with user. The copy of files and directories just starts at 12% and this is the moment when E-UAE hangs.

2.) I just boot from a WB 3.1 start diskette and try only try to copy 1 or 2 files (e.g. the calculator app or something else) from the start diskette to the folders mounted by filesystem2. Sometimes the copy works. But more often this simple file copy also hangs up the E-UAE. So problem is not the WB 3.1 installer itself but it seems to be an issue of E-UAE when using the feature of filesystem2 and copy any file to it.

In both cases I see these „FILESYS: UNKNOWN PACKET 1f" error messages in E-UAE console output at the time of first file access to an filesystem2 mounted device. So I was thinking something could be wrong with E-UAE and I was hoping one of the developers of this software is active in this forum and he could take an eye on it. As I mentioned in my last post I was guessing the issue of E-UAE seems to be related to that missing implementation of the ACTION_INHIBIT feature from Amiga DOS.


3.) I was creating an empty hardfile on an other operating system and than use this file with the hardfile2 feature of E-UAE and than I could install WB 3.1 without any problem. So the issue is only related to filesystem2 and hardfile2 works a treat.


Problem I see in future with this workaround of 3.) is that even if I have now a working WB 3.1 installation using hardfile2 I still need the filesystem2 for the game folder that bridges iGame application on MOS with WB 3.1 on E-UAE. And I saw in 2.) that any simple write access from WB 3.1 to an filesystem2 mounted device could lead to an hangup and this could really ruin your day.

@Waitsnake

"always at 12% the complete emulation hangs, well even the MOS hangs more or less"
"FILESYS: UNKNOWN PACKET 1f"

Same for me in my E-UAE setup with AmigaOS 3.0. Copying files hangs the system.
But, strangly, it works with E-UAE/AmigaOS 3.9, no hangs, no "unknown packet" messages, etc.

I remember a user had the same problem on AmigaOS 4.1 (I read this on AmigaWorld or Amigans some weeks ago).

Quote:

Anyone can comment on the above, please?

???
I've created two new empty directories to use as virtual hard disk partitions to perform a Workbench 3.1 installation (I did the same when I was writing the guide, see snapshots) and... it doesn't work anymore... I have the same problem above mentioned during installation which stops at 12% while installer is copyng WBStartup.info inside hard disk. How is it possible? I'm using the same adf images I used before and I tried Wb installation on the 3.5 too (I haven't upgraded MorphOS on my MDD yet). I've tried with filesystem and filesystem2 and always failed. I'm really disappointed...

A complete installation of Workbench 3.1 via E-UAE work for me on a HDF but i must rename "Locale3.1" volume to "Locale", bizarre...

After some days of tests I discovered that I can't install both Workbench 3.0 and 3.1 if I use E-UAE-0.8.29-WIP3 and WIP4. This morning I tried with E-UAE-0.8.28-2 and all went flawless trying to install WB3.0 and WB 3.1 (I've performed installation of both system three times with no problems). The FILESYS: UNKNOWN PACKET 1f error (if this is an error) was always prompted but installations always went ok. So I ask to Waitsnake and Papiosaur and all people that want to give a try to repeat what I tested so to confirm that all works with this E-UAE version.
I will try with E-UAE-0.8.29-PPCJIT_BETA03_MORPHOS next so to make a complete test.

<offtopic>
I had some big trouble with my MacMini G4 last few days. The hard disk stopped working and gives sounds like from a head crash. I removed the drive and just boot from CD without problem. Than I ordered a new SSD drive and it arrives yesterday.
I try it and find out that as long as the new SSD or the old HD is inside (both use IDE bus) not even the CD drive is working. I try all the tricks (press c, press alt or use Openfirmware commands to boot from CD) but all fail. I also reset the NVRAM and the PRAM without making it better.

CD drive is only working if I remove the HD or SDD from IDE bus. Maybe the IDE controller has an error or the power rail due to an defect cap I don't know. I ordered this 9 year old mac mini just a month ago at Ebay.

Because I want to cross check SSD and HD with other computer I ordered an USB to IDE adaptor yesterday, but it needs some days till the adapter will arrive.

Somehow in the evening I managed it to install MOS on an external USB drive and configure the Openfirmware to boot always from USB. My feeling is that when using USB the system is not as stable as with IDE before. But this impression could be wrong since I start to play around with different screens. And maybe just the feature of multiple screens is unstable anyway and it is not related to the USB drive.

After I have checked the SSD in the next days I will decide if I order a new PPC hardware or not. I still don't know what I will do next.
</offtopic>


Anyway with the USB drive workaround solution I try to install WB 3.1 with E-UAE-0.8.29-PPCJIT_BETA03_MORPHOS and E-UAE-0.8.28-2.

E-UAE-0.8.28-2 was working without problems (good find Jambalah thanks), but E-UAE-0.8.29-PPCJIT_BETA03_MORPHOS has the same issues as E-UAE-0.8.29-WIP3 and WIP4 and sucks at 12%.

[ Edited by Waitsnake 27.07.2014 - 12:21 ]

> maybe just the feature of multiple screens is unstable anyway

I'm not aware of any unstability induced by using multiple screens.

Oh, there are many ways to get problems. E.g. this one here:
1.) MOS prefs -> Screens -> New sceen
2.) Open Odyssey
3.) Right click to go to "MUI..." prefs of Odyssey -> Screen -> assign to new screen from 1.)
4.) select "Bearbeiten ..." (guess in english it is called "Change ...") to change settings of this screen
5.) change "background picture" of this screen -> where Texture.png is preselect change it to maybe Winter.jpg than OK -> OK -> Save
6.) very often now I get a new window called "log server" with "Public Screen Manager" and "Unknown Exception"
After this exception most times my main Ambient screen is gone and I only have the Odyssey screen left, but with that screen I can only surf the web till I close Odyssey an than I can not do anything else than move the mouse arround or make a cold restart with the power button.

Or afer the reboot from last screen crash (see 1.) to 6.)) I do:
7.) open Odyssey in its own screen now with it's own Winter background -> nice no problem till now
8.) go back to Ambient screen while Odyssey screen is still open
9.) open Morphos prefs -> Screen -> make a new third screen maybe for MPlayer
10.) close Screens preferences with "save"
11.) Again I get "Log Server" window with "<>" and "Unknown Exception"


I don't know if it could be related that my hole MOS system dunning on a slow sdcard inside USB stick and system using all the ti,e Poseidon USB stack while normaly system would run over IDE hard drive?

Or maybe my hole system has some serious hardware problems? This is the thing that MOS relates to Hardware that is not longer manufactured and I had to buy an 9 year old computer at Ebay with no clue about it's conditions. It's a bit sad because I really like the look and feel of MOS since I play around with it for a month or so... I even would prefer it to OSX that I use since 10.3. if only all things would be a bit more stable.

If it is only my hardware I would look for a nother "new" old hardware at ebay... but there is allways the risk that this mashine has also problems since it will be at least 9 years old too.

[ Edited by Waitsnake 27.07.2014 - 15:16 ]

> I don't know if it could be related that my hole MOS system dunning on a slow sdcard inside
> USB stick and system using all the ti,e Poseidon USB stack while normaly system would
> run over IDE hard drive? Or maybe my hole system has some serious hardware problems?

I followed your steps while running MorphOS from PATA and couldn't reproduce your result. I think you should file a bug report. This way, the MorphOS Team could determine with your help whether it's a hardware problem at your end or a MorphOS bug.

Quote:

This is nothing unusual and not a hardware failure. In the early days ide devices had many issues, when being used on the same IDE bus. It got better and better over the years, but never was perfect.

Especially the drives used in Apple devices are often problematic in this case. It also happens, when using SSDs in PowerBooks.

These devices simply do not like each other. Another issue is that in many cases you cannot change the master/slave mode, as it is hard coded/soldered or stored in firmware. Check if there is a jumper on the CD drive and remove/plug it, if possible.

Replacing the harddrive along with the CDROM drive however is the best solution in most cases.

Geit

Finally I buy a second MacMini at Ebay. Exact same model as before (1,5GHz 1GB RAM with 80 GB HDD). I found out that RAM of this 2nd MM is defect and CD-ROM reads much much worser than the CD from my 1st MM. But this are issues that I can fix with replacement parts in future if I will.

Anyway I used this 2nd MM to cross check/replace all the components step by step with the 1st MM (the power supply, the HDD, the SSD, the CD-Drive, the logic board and this mini bridge board that connects CD/HDD to logic board).

Finally I come to the conclusion that the problem that causes the trouble was the SSD. So the suggestion of geit was correct:
Quote:

The original 80GB HDD works in both machines and CD drive is than not blocked, while with the SSD the CD drive was missing. For those how wants to avoid the same problems I use an "KINGSTON SSDNow KC380 120GB 1,8 microSATA3" together with an "Adaptare 46234 Adapter". Maybe it is this IDE 44 pin to Micro-SATA adapter that causes this problems with both MM.


Than just for luck I ordered an other SSD this time an "Transcend TS32GPSD330 interne SSD 32GB" that has direct an IDE 44 pin interface and not needs an adapter as my 1st SSD.
And what should I say? This SSD works perfect in both MM and not have any side effects to the CD drive.

Quote:

Do you use MOS 3.6 with Chrysalis pack 3.6.1 beta ?
Because during my hardware experiments and fresh installations I realize that with an blank/clean MOS 3.6 I had also NO issues with the screens. Just after installing Chrysalis pack that was of cause an still an beta I get this problems.

Anyway now it is MOS 3.7 out and also Chrysalis pack 3.7.1 beta and I will repeat my tests with this version in the next days. In Chrysalis pack 3.7.1 Papiosaur removed something in the Ambient prefs that "caused problems". Maybe this causes problems with the screens as well? We will see.

[ Edited by Waitsnake 07.08.2014 - 13:46 ]

> Do you use MOS 3.6 with Chrysalis pack 3.6.1 beta ?

No Chrysalis here.

> with an blank/clean MOS 3.6 I had also NO issues with the screens. Just after
> installing Chrysalis pack that was of cause an still an beta I get this problems.

This explains why I don't have these problems.

What is the problem with pack 3.7.1 please ?