SpareMiNT - ISSUES

0000169: Cursor not position correctly in editfield

The cursor is not correctly positioned in edit fields with a mask that contains non-editable characters. E.g. if the editfield has a mask of "xx:xx:xx" (where ':' is not editable), the editable string is usually six characters ("xxxxxx"). In this case the cursor is not displayed past character position six. It looks like the length of the editable string is used to determine the end-position of the cursor, and not the length of the mask.

0000176: Fileselector doesn't list contents of /ram

The fileselector doesn't list the contents of u:\ram. I've tried with several programs, the bug occurs every time.

0000171: Bug with the file timestamps.

When I ask the date of a file( with Fxattr for example), the creation and modification dates appears to be switched.

This bug is easily reproductible with the Jinnee info box.

Also, when I look at files not touched for ages, the timestamps looks
sensible although swapped. If I touch one of these files, both
timestamps are modified and are identical.

0000174: non-XaAES fileselectors

Loading the NAES fileselector n_fsel.acc causes the fileselector not to appear in applications.
No error is reported, neither fileselector appears and the application shows the busy-bee mouse
pointer.

0000175: Media keys on modern keyboards

Almost all modern keyboards have a set of media keys for volume control etc.

It would be nice to have these enabled to control the sound output.

0000021: no cursor with NVDI installed

When booting mint with nvdi installed I have no cursor in the bash.

0000153: XaAES problem with GEM focus

I've noticed a problem with my Eureka software, that EB_Dump
from Emmanuel Baranger also suffers from ...

The problem is that the latest GEM window that is closed
with a GEM application, makes the focus of the GEM app to
be lost.

With EB_Dump, when one do a fullscreen screenshot, that is not
a problem, because after the dump of the screen is done, EB_Dump
automatically redraws its GEM interface window. But if one chooses
to partially dump the screen, it requires the user to interact
before the interface GEM window reappears. Because it has lost
the AES focus, one can't interact, and EB_Dump is dead locked.

The problem is also there with my proper Eureka software,
because there's two modes : graphics windowed drawings with
a GEM window, and GEM desktop interaction mode, with no GEM
window except from dialog boxes that are in fact non preemptive
GEM window boxes, and the Eureka menu bar.

With Eureka when the user switches from the graphic mode (drawings)
to the GEM desktop mode (menu commands interface mode) the last
GEM window is closed by the application, that subsequently loses
its focus.

It means that with Eureka and XaAES, I must install Taskbar
or zTask to manually recover the focus (and the Eureka GEM menu bar)
each time I come back from graphic screen mode.

But that won't work with EB_Dump.acc being an accessory, with
no menu bar etc. And the loss of the focus makes EB_Dump to be
unusable in the particular partial screen dump mode.

There's always been, so far as I remember, this problem with
the AES focus of GEM applications. I always had the same problem
with my Eureka software. I think that the XaAES way to handle the
app's focus should be modified in case its latest GEM window is closed.

The app shouldn't lose the focus and/or its menu bar, if its
latest GEM window is closed. That doesn't happen with many apps,
but we're both having the same problem with our apps, Eureka and
EB_Dump.

OZK proposed to me :

There is an 'app_option' argument called 'clwtna' (Close Last Window
Tops Next App). It is explained in the example xaaes.cnf file found in
the cvs source tree, or in the alpha packages.

But it has no effect whatsoever.

This problem doesn't happen with MyAeS, because OL corrected it
when I reported. So it shoudn't be problematic, at last.

0000073: Bus error when starting xaaes or naes

with the new mint version (current cvs), I quite often get a bus error, when trying to start xaaes/naes from the login shell.

0000149: SPACE + ALT should not be valid dialog key shortcut, when SPACE is first char

When SPACE is the first character in eg. an ALERTs button, XaAES can at times address ALT+SPACE as the keyboard shortcut for this button. While this ain\'t illogical, it gives odd results when the button text is padded with SPACES. Hence, the indicator for the keyboard shortcut will show up to the far right of the actual text, which looks completely wrong.

0000173: Texel 1.6 checboxes

The check boxes do not respond to a mouse click.

In Texel 2 demo. the check boxes are different and work ok.

0000172: Typo in window menu

The 'Advanced' window sub-menu has an item written 'keep *uner* others'.

0000148: Built-in support for taking screenshots

It would be cool if XaAES had native support for taking
screenshots, triggered by a certain key combination. File format preferebly being something that can be easily converted to other formats without loss. Lp\'s GFASNAP is dumping TGA, which could perhaps be a good solution here too.

0000161: mouse-klick does not move cursor in a standard form_do-dialog

under TOS, the keyboard-cursor is moved to the object on which a mouse-click occured, so keyboard-input goes to that object. since MiNT 1.16 the cursor can only be moved by cursor-keys, not by mouse-click.

Reproduce: Start pc.rpg (pure-c), select options->compiler and click on one of the paths. keyborad-imput always goes to the first path (Include directory), though the right form is slelected.

All form_do-dialogs show this behaviour.

0000086: placing iconified windows

It would be nice, if there were rules where iconified windows are placed (eg. have some distance from screen border, go from right to left, or top to bottom etc.)

0000145: bad ARGV support

there\'s bug in argv[] passing to application. I used teradesk to test it with > 128 characters long path, result is argv[0] = filename without path and argv[n], n>1 is empty. In n.aes it works OK.

0000147: Keyboard navigation does not wrap to beginning/end

When using TAB / TAB+SHIT in the new keyboard navigation scheme (which is absolutely awesome!), the moving of focused object is halted when the first/last object in the dialog is reached. On other systems, this would instead result in the focus being wrapped over to the top/bottom of dialog. IMO this would be to prefer for XaAES as well.

0000168: NULL is not filtered out when entering text in editfields

Some keys, like Insert and Help, returns a scancode but no ASCII code. These must be filtered out in form_keybd/form_do to prevent a NULL to be entered in the editfield. E.g. if you try to use Insert when the cursor is not at the end of the string, the string is truncated at the cursor position.

0000170: INSERT-key is a modified. Why? Should it be removed?

Why is INSERT a modifier? What is the purpose of this? It creates problems for Texel (Control+Insert does not work). I suggest to remove this feature. I have tested this by removing INSERT from the modifiers-list in keyboard.c, and have noticed no side-effects (except that Control+Insert works again in Texel).

0000166: EOL options in settings

This would be a useful option and allow creation of specific text type files, for example BASH shell scripts require \n (&OAh) (r the valid alternative \r\n - &0D0Ah), while msdos and windows batch files require \r\n EOL characters

something like:
use EOL
O - \r (default)
O - \n (linux)
X - \r\n (win/dos)

0000159: gcc 4.3: missing " in gemma/src/gemproto.h

gcc 4.3 can not compile stuff due to the asm macro using a multi-line string.

Patch attached.

0000089: Problems with keys

With the latest mint versions I have two problems concerning keys:
1. ALT don\'t work with any app. ALT-U for instance doesn\'t work for CAB, there are problems with Teradesk also.
Other apps like highwire work without problems
2. Sometimes a key is repeated once or twice. As this also happens with the login shell under mint, it might be no problem of any AES

Milan 060, mint 1.16.1 (CVS 6.9.2004), no key tbl used, NAES 2.0 / XAAES (CVS 6.9.2004)

0000158: gcc 4.3: can not build cflib

Just noticed cflib does not build for gcc 4.3.2:
cc1: error: unrecognized command line option "-fwritable-strings"

I have to clear the DEFINITIONS= in the Makefile

Is this option really necessary?

0000151: Unable to format floppy discs

Since version 1.16.1 floppy disc formatting fails.

0000167: Unicode format

the ability to save (and load) unicode text files as a setting (or autodetect)

to convert a plain ASCII text file to unicode &00h is placed before each character at the binary level and the file starts &0000h. ALL non-ASCII characters are mappable to unicode. Display of extended characters may require non-system font (TTF via fVDI or NVDI).

a setting like:

save as unicode []
or
use unicode []

0000136: Ctrl-Alt-Tab improvements

- Control-Alt-Tab should NOT include the accessories
- Graphical feedback for Control-Alt-Tab
- Holding down Shift should reverse the order of switching