Description of appl_control maybe incomplete

[th-otto]

I'm currently investigating a problem with taskbar, and from the disassembly i can see that the ap_cout parameter that is passed to appl_control always points to an array of 2 shorts. So maybe the description at

tos.hyp/gem/aes/appl_/appl_control.ui

Line 76 in 040c7ba

For (!I)ap_cout(!i) a pointer to an integer is to be passed

is incomplete. Can someone check whether N.AES actually writes something to ap_cout[1]? If so, that should atleast be mentioned there, even if it is always zero.

[xdelatour]

Check with a GEM app with menubar:

• APC_INFO (14): the first short ap_cout[0] is filled with infos ($2), the second short ap_cout[1] contains garbage ($48)

short info;
r = appl_control(apid, 14, (void*)&info);
printf("%x\n", info); /* 2 (APCI_HASMBAR) */

• APC_MENU (15): expected return value=OBJECT *

OBJECT *addr;
r = appl_control(apid, 15, (void*)&addr);
printf("address=%.08lx\n", (long)addr); /* same as (long)addr_menu of the installed menubar */

• APC_WIDGETS (16): expected return value: fill a buffer where ap_cout points to

short widgets[]={-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,0};
r = appl_control(apid, 16, widgets);
for (i=0;i<12;i++) printf("%d\n", widgets[i]);
Outputs: 2, 3, 30, 19, 4, 10, 13, 11, 15, 18, 16, 0

• APC_APP_CONFIG (17): expected return value: none, ap_cout is a pointer to a string

char *enable_dbg="app_debug";
r = appl_control(apid, 17, enable_dbg);

Returns 0 (failure) but APC_APP_CONFIG is missing from doc

[EDIT] In the MyAES programmer guide, APC_APP_CONFIG is 50, not 17

[th-otto]

Ah, thanks. Do you maybe have a similar setup as geneva for N.Aes?

[xdelatour]

Unfortunately no, not yet. Also on my ToDdo list. I have to change my mint.cnf each time to switch from XaAES to N.AES

[xdelatour]

Do you maybe have a similar setup as geneva for N.Aes?

Available on my website through a link on the wiki. It's surely possible to do better, maybe a single setup with all the aes...

[th-otto]

Nice, thank you. BTW is there maybe a problem with http caching on your website? I first didn't see the archive, until i hit F5 to reload the page.

BTW2, disabled menu items look at bit strange in N.Aes:

Is that normal, or maybe a bug in fVDI?

[xdelatour]

BTW is there maybe a problem with http caching on your website?

It's possible, I'm updating several pages at the moment. I did not look into the management of the cache of Grav

Is that normal, or maybe a bug in fVDI?

I don't know. My N.AES install comes from multiple sources (the .SYS file was found on a ftp, the skins somewhere else and the configuration file from yet another place). Its also why the tools (nclosure...) are missing.

By the way, fVDI have a lot of problem.

• oAESis doesn't work with the current snapshot. I thought it was because of the mouse driver but the mouse works with a different fvdi.prg (I can't remember if it's from an old snapshot or if I recompiled it)
• AES 4.1 with default fVDI settings (32b) displays all white desktop icons. No problem with 16b colors, or without fVDI.

I'm not sure I'll be done by tonight (I'm busy this afternoon), just missing MagiC and MultiGem for now, MagiC can't access hostfs and MultiGEM has to boot from a floppy.

[lpgb]

BTW2, disabled menu items look at bit strange in N.Aes:

Is that normal, or maybe a bug in fVDI?

ARAnyM with only NVDI 5.03. fVDI totally disabled. N.AES in 256 color mode.

Maybe it is related to N.AES setting? This can be set in n_aes.cnf as well.

[xdelatour]

Maybe it is related to N.AES setting? This can be set in n_aes.cnf as well.

Thanks, you're right. My n_aes.cnf has disablecolor = lcyan
With 9, it looks like:

As my config file was found on Internet, I have no idea about the default settings

[mikrosk]

I have to change my mint.cnf each time to switch from XaAES to N.AES

Not like it solves all the issues but I have learned to use different config files and include them. That way you can prepare N.AES-specific settings, XaAES-specific settings etc and just change one line (include xaaes.cnf instead of include naes.cnf).

Another trick I'm using is starting AESes manually, also via separate scripts. That works even for different xaaes versions (one can supply path to xaaes to xaloader.prg).

[xdelatour]

Hi @th-otto, I updated n_aes.cnf to disable custom colors. All default values are available in n_aes.hyp

@mikrosk I tried (unsuccessully) to build vlogin many months (or years) ago. Black screen only and no mouse.

[mikrosk]

@xdelatour hmm, I'm pretty sure that I did see vlogin working at some point. Surely around the time of 1.16 release, maybe you can try that version?

[skaftetryne]

Can someone check whether N.AES actually writes something to ap_cout[1]?

Taskbar has called appl_control/APC_INFO with a pointer to an int since the beginning, as this is how it's documented in the N.AES docs. For all the modes where ap_cout isn't documented to be used for anything I pass a pointer to a long, I can't remember why I did this 25+ years ago, but possibly because it didn't work if I passed a NULL-pointer.