vqt_name incorrectly indexes the fonts about emutos

[mfro0]

Checked with fVDI and TOS output is the exact same than EmuTOS output for me (checked in Hatari with Falcon and STE emulation, but I would assume that shouldn't make a difference).

TOS output:

30 new screen fonts loaded
font 0, 6x6 system font, id = 1
font 1, 6x6 system font, id = 1
font 2, Comic Sans MS Regular, id = 7059
font 3, Courier New Bold, id = 7203
font 4, Trebuchet MS Bold, id = 7612

EmuTOS output:

30 new screen fonts loaded
font 0, 6x6 system font, id = 1
font 1, 6x6 system font, id = 1
font 2, Comic Sans MS Regular, id = 7059
font 3, Courier New Bold, id = 7203
font 4, Trebuchet MS Bold, id = 7612

...

I admit that I fixed your code first as there were small imperfections (probably not the cause of the differences you are experiencing, but anyway):

• vqt_fontinfo() array indexing is zero-based
• fonts loaded with vst_load_fonts() are appended to the font list. The number of fonts already present before vst_load_fonts() is available from work_out[10] after opening the virtual workstation.

To create a list of all available fonts, your loop should look like this, then:
for( font_idx=0; font_idx<new_fonts+work_out[10]; font_idx++ ) {

To further track down the differences you are observing, your GDOS setup and fonts are most likely required. Would you be able to make them available somehow?

[tevorbl]

Hi,
Thanks for the response.

My GDOS setup is the standard GDOS 1.1, fonts are standard atari fonts distributed with GDOS.
Here's a link -> http://www.atarimania.com/utility-atari-st-gdos-distribution-disk_25889.html
My assign.sys file is already posted above.

The test should be repeatable with any GDOS and any gem fonts.

{Edit:
I wonder, in your check above, is there only one instance of each font (face) in the fontchain?
The issue I am seeing would occur only if the fontchain held gem fonts of different sizes and with the same id.
The vdi_vqt_name() function is not skipping the second and following fonts with the same id, hence the difference in results.
It should be counting only the unique font id values.
Further, if you change the end condition in the loop in my test to say
for( font_idx=0; font_idx<new_fonts+20; font_idx++ ) {
you can see that the vdi_vqt_name() function is counting all gem fonts at all sizes.
]

[anodynesoftware]

Hi!
I'm one of the EmuTOS maintainers. I'll take a look at this and get back to you. For future reference, you'll probably get a faster response if you post to the EmuTOS list: [email protected]. That's where you'll also get to hear about new features, upcoming releases etc.

If you're interested, you can sign up at:
https://sourceforge.net/projects/emutos/lists/emutos-devel

Roger Burrows

[anodynesoftware]

Thanks again for the report. Your diagnosis was correct and this is now fixed in the repository. As far as using vwk->loaded_fonts is concerned, that only points to the loaded fonts, whereas vqt_name() must scan all the fonts including the system font.

[tevorbl]

Many thanks for the quick fix.
I'll start using this right away.