Login
Username:

Password:

Remember me



Lost Password?

Register now!

Sections

Who's Online
148 user(s) are online (98 user(s) are browsing Forums)

Members: 2
Guests: 146

Georg, Yodaphone, more...

Headlines

Forum Index


Board index » All Posts (joerg)




Re: SDK-Gurus: ScummVM_OS4 needs your help
Just can't stay away
Just can't stay away


@hnl_dk

Quote:

hnl_dk wrote:
@joerg

atan2( 0, 0) = 0.000000
atan2( 0, 1) = 0.000000
atan2( 1, 0) = 1.570796
atan2( 1, 1) = 0.785398
atan2( 0,-1) = 3.141593
atan2(-1, 0) = -1.570796
atan2(-1,-1) = -2.356194

that are the correct results (according to glibc)

I've fixed it in newlib.library 52.7, but atan2(0,0) still returns NaN, just like in libnix, which is correct IMHO and the 0 from glibc and clib2 are wrong!?

Go to top


Re: SDK-Gurus: ScummVM_OS4 needs your help
Just can't stay away
Just can't stay away


@Jack

Quote:

Jack wrote:
@hnl_dk and raziel

It could be easily verified by compiling the atan2 test program (I would test 7 calls, pairs of args such as:(1,1) (0,1) (-1,1) (-1,0) (-1,-1) (0, -1) and (1, -1).

Compile it vs clib2 and via newlib and check results.

And what are the correct results?

atan2() is broken in at least 2 of 3 AmigaOS C libraries
6.RAM Disk:> type test.c
#include <math.h>
#include <stdio.h>

int main(void)
{
printf("atan2( 0, 0) = %f\n", atan2( 0, 0));
printf("atan2( 0, 1) = %f\n", atan2( 0, 1));
printf("atan2( 1, 0) = %f\n", atan2( 1, 0));
printf("atan2( 1, 1) = %f\n", atan2( 1, 1));
printf("atan2( 0,-1) = %f\n", atan2( 0,-1));
printf("atan2(-1, 0) = %f\n", atan2(-1, 0));
printf("atan2(-1,-1) = %f\n", atan2(-1,-1));
return 0;
}
6.RAM Disk:> gcc -Os -Wall test.c -o test
6.RAM Disk:> test
atan2( 0, 0) = NaN
atan2( 0, 1) = 3.141593
atan2( 1, 0) = -1.570796
atan2( 1, 1) = 0.785398
atan2( 0,-1) = 3.141593
atan2(-1, 0) = -1.570796
atan2(-1,-1) = -2.356194
6.RAM Disk:> gcc -Os -Wall test.c -o test -mcrt=clib2 -lm
6.RAM Disk:> test
atan2( 0, 0) = 0.000000
atan2( 0, 1) = 0.000000
atan2( 1, 0) = 1.570796
atan2( 1, 1) = 0.785398
atan2( 0,-1) = 3.141593
atan2(-1, 0) = -1.570796
atan2(-1,-1) = -2.356194
6.RAM Disk:> m68k-amigaos-gcc -Os -Wall test.c -o test -noixemul -lm
6.RAM Disk:> test
atan2( 0, 0) = NaN
atan2( 0, 1) = 0.000000
atan2( 1, 0) = 1.570796
atan2( 1, 1) = 0.785398
atan2( 0,-1) = 3.141593
atan2(-1, 0) = -1.570796
atan2(-1,-1) = -2.356194

Go to top


Re: Optimizing OS final install
Just can't stay away
Just can't stay away


@lazi

Quote:
I am surprised how much memory is used after booting OS4 final on my micro A1.
There is certainly some commodities in the wbstartup, so it should be normal, but 182 megs free from the installed 256 is surprised me compared to my classic Amiga experiences.

Do you have SFS partitions? If yes you could remove it's global cache, diskcache.library.kmod, from your kicklayout to get about 20 MB more free memory (or less free memory if you have a lot of SFS partitions, without diskcache.library each one will use it's own, partition local caches instead), but much slower disk speed.

If you don't use SFS removing diskcache.library.kmod doesn't make much difference, you'd only get 15 KB more free RAM, or about 150 KB if you remove SmartFileSystem as well.

Go to top


Re: SFS2 with OS 4.0 Final
Just can't stay away
Just can't stay away


@PEB

Quote:
Actually, I was planning on experimenting with burning >4GB .iso files using FryingPan at the moment.

No idea if FryingPan supports files > 4 GB or not, I've never used it and don't need it since I've written my own DVD burning tool some years ago , but it's very unlikely that the AmigaOS4 version of FryingPan supports large files already. Check it's documentation, at least 99% of the software which doesn't mention anything about 64 bit file size support in it's documentation doesn't work with files > 4GB.

Go to top


Re: SFS2 with OS 4.0 Final
Just can't stay away
Just can't stay away


@PEB

Quote:
I guess I misunderstood this quote from Hyperion's website:
"64 bit support for file sizes."

For 64 bit (files >= 4GB) support you need 3 things:
- dos.library functions for large files (included in AmigaOS4 Final, and partially in Update4 already).
- A file system which supports large files (for example SFS2, SFS 1.x and FFS2 don't support files >= 4 GB).
- Applications supporting large files by using the new dos.library functions (I only know 4 which do).

Go to top


Re: SFS2 with OS 4.0 Final
Just can't stay away
Just can't stay away


@PEB

Quote:

PEB wrote:
Is SFS2 completely useless (unnecessary) under OS 4.0 Final?

No, but it has some bugs, at least ACTION_GET_FILE_SIZE64 (IDOS->GetFileSize()) is broken, and it doesn't support the new directory scanning API of AmigaOS4 final yet, only the 64 bit functions which were in Update4 already.

But even if it wouldn't have the bugs and missing features, for normal users it's of course still useless since there is next to no software which supports files > 4GB, but it was never for normal users anyway: It's for developers to be able to test the 64 bit support they added in their software, but that didn't happen (yet?) AFAIK. Except for lha and DvPlayer which support large files now, and of course some of my own software, for example a DVD burning tool which doesn't support image files > 4GB would be quite useless

As long as there is no AmigaOS software which supports large files the large file support in dos.library and file systems like SFS2 will stay useless

Go to top


Re: Privoxy
Just can't stay away
Just can't stay away


@Raziel

Quote:

Raziel wrote:
btw: Is there a way to get rid og the annyoing shell window
telling me that the config has been loaded?

Run <>NIL: doesn't do it? It's the only thing that prevents
it from being perfect to me...off to configure

Redirect stderr as well: Run <>NIL: *>< Privoxy.
(or add *>NIL: instead, *>< is the same in this case and just shorter: it redirects stderr to stdin, which is NIL: as well)

Go to top


Re: What happened to IExec->StartDMA()...?
Just can't stay away
Just can't stay away


@ssolie

Quote:
Are you sure you can do DMA drivers without the updated SDK on the new kernel?

For hardware which supports scatter/gather DMA: Yes.

For hardware which doesn't but needs contiguous physical memory for DMA you need AllocVecTags() now and need the new SDK, up to Update4 you could use a MEMF_PUBLIC buffer instead.

Go to top


Re: SCSI issues, help please
Just can't stay away
Just can't stay away


@tonyw

Quote:

tonyw wrote:

The HD drivers all announce themselves on startup. If you had a serial terminal, you could look for the "lsi53..." greeting message when the system boots (they're all very friendly). You would have to set the kernel command line to include the word "serial".

Or simply use "C:DumpDebugBuffer >RAM:bootlog" instead after booting from an IDE HD or CD drive and check it, a serial terminal is only required if you can't boot at all.

Go to top


Privoxy
Just can't stay away
Just can't stay away


I've build Privoxy 3.0.6 for AmigaOS4 and uploaded it to my homepage.
I didn't do much tests yet and didn't check the changes in the Privoxy code since my last AmigaOS version (3.0.3), but it seems to work without problems.

Go to top


Re: Collector 3.5 and Index Files loading
Just can't stay away
Just can't stay away


@Pierre55

Quote:
The index file is on my SYS (SmartFilesystem 1.267), I
try to load the file from RAM and still loading very slow.

I've downloaded gfx/misc/Collector.lha from aminet to check what the problem is, but when creating the thumbnails the included MPImage.library crashes sooner or later. Is there another MPImage.library somewhere without these bugs?

Go to top


Re: Collector 3.5 and Index Files loading
Just can't stay away
Just can't stay away


@Pierre55

Quote:
Since the OS4 Final my index file from Collector 3.5
load 10 times slower....

Which file system are you using on the partition on which this index file is?

Go to top


Re: Hyperion Games and OS4 Final?
Just can't stay away
Just can't stay away


@ikir

Quote:
Shogo, after installing powerpc.library from upd4 starts and seems to work in software mode. MiniGL crash.

I'm surprised you got something working at all, the old powerpc.library versions which were included in Update3 and Update4 do not work on AmigaOS4 final. Even if you don't get immediate crashes for some reason you'll get a very unstable system if you install them and run WarpOS software.

Go to top


Re: OS4 and cdrom trouble...
Just can't stay away
Just can't stay away


Quote:

Skov wrote:
Well guys, thanks for all your help so far! Now I got the right version back in the system.
Idetools shows my drive nicely.

Flags : $0000011D - present, supports DMA, removable media, ATAPI, interrupts used,
Xfer mode : best pio 12 (PIO 4, 16MB/s) / best dma 66 (UDMA 2, 33 MB/s) / current 12...
Total blocks : 0
Blocksize : 0
....
Power mode : 2 /idle (ready for operation)
IO1 /IO2 / BMCR @ : $1F0 / $3F6 / $CC00

But it still doesn't recognize any cd's... Drive's broken or what?

If IDETool shows 0 for "Total blocks" and "Blocksize" even if a CD is in the drive it's very likely broken, but to make sure it's really the drive and not a problem in the AmigaOS4 IDE drivers you could abort booting in U-Boot and enter "ide info" in the U-Boot shell (if the CD wasn't in the drive yet use "ide reset" after inserting it first), that will display which drives U-Boot detects and their sizes, if U-Boot can't get the size of an inserted CD either the drive is broken.

Go to top


Re: Report your 'AmigaOS 4.0 Final' bugs here!
Just can't stay away
Just can't stay away


Quote:

orgin wrote:
@joerg

This is the problem that needs to be solved:

Detect when an icon has been dragged and dropped on a window and the position where is was dropped. If it was dropped in a specific zone, retrieve filename of said icon. Read icon image data from the icon file. Display icon images.

Are you saying that that is a problem that is _totally_ impossible to do in AmigaOS 4.0 without breaking backward compability?

Yes, because you do not drag and drop icons with the current AppWindow system, you drag and drop objects. As long as there actually is an object there is no problem, it only fails if you try to drag and drop an icon which does not have an object (and even in that case only for directories), something like that wasn't supposed to work at all ...

A new, additional method for draging and dropping icons, which don't belong to anything, would have to be added, but of course that could only work in new software which will have to support this new method once it's implemented.

Go to top


Re: Report your 'AmigaOS 4.0 Final' bugs here!
Just can't stay away
Just can't stay away


Quote:

orgin wrote:

As I said, there's no technical reason that makes it impossible to solve this problem.

There is no problem at all changing it if you don't care about backward compatibility, but if we change it no current software using an AppWindow will work any longer ... Try to exmplain that to AmigaOS users

Reread this thread again and check how many people complain that broken software doesn't work any longer on AmigaOS4 final, now imagine what would have happened if a lot of correct software would have stop working

Go to top


Re: Report your 'AmigaOS 4.0 Final' bugs here!
Just can't stay away
Just can't stay away


Quote:

orgin wrote:
@saimo

Well whatever the problem is (In my icon set all icons have a drawer associated with them and it still doesn't work all the time) having a file/directory shouldn't be an issue and should be corrected in a future update/release, as it would only confuse new users. There's no technical reason as to why it has to be that way, the icon graphics is associated with the info file, not the directory/file.

Wrong. When you drop an icon to an appwindow it get's an AppMessage with struct WBArg * am_ArgList which includes a lock (wa_Lock) to the current directory the object is in and the name (wa_Name) of the object (not it's icon, i.e. without ".info") in that directory. For tool and project icons without files that's possible, for example if you have a foo:bar/file.info without a foo:bar/file the workbench can sent the appmessage with the wa_Lock of "foo:bar" and "file" as wa_Name. For directory icons without matching directory it's not possible, for example if you have a foo:bar/dir.info directory icon but no foo:bar/dir directory the appmessage to be sent would have to have the wa_Lock of "foo:bar/dir" and wa_Name "", but there is no foo:bar/dir -> no lock possible -> no appmessage possible.

Go to top


Re: OS4 and cdrom trouble...
Just can't stay away
Just can't stay away


Quote:

Skov wrote:
idetool -l a1ide.device
*** Error : device version 51 / revision 51 does not match tool
version 51 / revision 49
*** Error : device version 51 / revision 51 does not match tool
version 51 / revision 49
*** Error opening device a1ide.device / unit 2

And so on...

Now what do I do?

Where did you get version 51.51 of a1ide.device from?

Install the a1ide.device.kmod from Update4 (version 51.49), or IDETool and the matching device drivers from http://os4depot.net/share/driver/storage/ide_driverpack.zip (version 51.54).

Go to top


Re: AOS4 graphics update is slower now
Just can't stay away
Just can't stay away


Quote:

saimo wrote:

I mean the (raw?) data stored on the disk that can be written only with MediaToolbox - i.e. not the file in Kickstart.

Did you use the Update4 version of MediaToolBox to install the PPC native SmartFileSystem into the RDB? The current version finally shouldn't allow that anymore ...

Anyway, the only reason that doesn't fail with AmigaOS4 final is that the same version of SFS was loaded as kickstart module already, file systems are only loaded from the RDB if there is none with the same dos type already, or the one which is there already has a smaller version.revisions - which is the case with Update4 => you can't boot.

PPC native file systems must not be stored into the RDB with MediaToolBox, that's not supported and never will be, only m68k file system can be stored in the RDB!

Go to top


Re: AOS4 graphics update is slower now
Just can't stay away
Just can't stay away


Quote:

saimo wrote:
@joerg

I've been too concise, sorry
I did put UF aside and re-install the previous, full U4 to the same partition (because I had no spare partitions). The only change with respect to my pre-UF setup was thus the new SFS installed on the disk: since I did not deem it safe to re-install the previous one, I gave up and restored the UF setup.

saimo

Sorry, I don't understand what you mean with new SFS installed on the disk. Did you use the Update4 kickstart modules except for SmartFileSystem and used the one included in the final version instead? That should have worked.
Unless you replaced the diskcache.library.kmod as well, the one included in the final version doesn't work on Update4.

Or did you install a m68k version of SFS which is newer than the PPC native one included in Update4 in the RDB of the HD?

Go to top



TopTop
« 1 ... 82 83 84 (85) 86 »




Powered by XOOPS 2.0 © 2001-2023 The XOOPS Project