gsoftwaredevelopment / midicar-player Goto Github PK

currently, for lengths greater than 30 characters, it spoils the appearance of the screen :P

I enabled 1MB in U1MB.

1. When SDX, PBI and HDD are disabled in U1MB and SDX is enabled in the SIDE2 cart and I start MPCar Player, it shows 1033 KB free memory
2. When SDX, PBI and HDD are enabled in U1MB and SDX is disabled in the SIDE2 cart and I start MPCar Player, it shows only 89KB free memory

Implement a watchdog for the drivers that would be responsible for keeping an eye on the stability of the program, e.g.
For the MIDIMate driver, it is unconnected, the program is unable to send information over the serial port because there is no clocking from the device (clock on Clock-In SIO) This causes information to go into the FIFO buffer, and this buffer is not emptied regularly. An overflow of the FIFO occurs, and as a result, the driver proceeds to FLUSH the buffer. This operation is primitive and exposes a byte to potentially initiate a data send and waits for the buffer to empty. It does this without any TimeOut, resulting in a suspended player.

Such a Watchdog, could keep an eye on such events and more.

• moving procedures to a more suitable place in the code
• full handling of keyboard events from the key table. The ESC and RETURN keys have events assigned to them like the rest of the program functions.
• Changed the way the volume indicators are displayed. From now on they are oriented horizontally.
• removal of replacement code for the bug in the MP repo
• other, minor fixes

• too small a buffer for the list
• no list limitation

Opcja szybkiego wyszukiwania w liście wyświetlonych plików.

when the next in the list is a catalogue, it moves to it and recursion occurs. An invalid path is created

there are possible situations in which the bar is "broken", and its area is not cleared when it is launched

I undid the buffer increase. I forgot that the used area is for the song data :P

There were also a few oversights on my part, such as using variables of type Byte to denote the list offset (shFile), the number of list items (totalFiles) or variables pointing to the currently selected and played list item (curFile, curPlay) As a result, this caused the number of loaded list items to be artificially limited or displayed incorrectly.

Finally, this fix also brings an enhancement that increases the number of elements that can be stored in the list buffer.

There are no keys to quickly navigate large lists of files, e.g.

• jump about a page back and forth
• jump to the beginning of the list and to the end
• maybe a Search option?

version designations in the program, so that there are no problems with distinguishing between them

there was a bug, which, when the list is filled to the penultimate position, causes that, the list scrolls further :p

1 There has been a change in the way the selection list cursor is generated. It is now based on the DLI interrupt, which modifies the colors on the line where the cursor is. As a result, bugs have come to light, including one related to a bug in the 6502 processor, where when NMI and IRQ interrupts occur simultaneously, the NMI interrupt is not executed.

A patch has been made, but it does not always work - inquiry into the causes is underway.

2. the tempo determination for the song has been improved. now, the data from the Time Signature meta event is also taken into account. this allowed to properly calculate the MIDI clock interval, which reduced the likelihood of bad song playback.

3. The number of clock cycles (CPU) that POKEY Timer1 handles has been reduced.

4. Abandoned support for primary memory as an area for MID file data. From now on, it became a requirement to have a minimum of 64KB of secondary memory. Sorry.

currently it is set to every 3rd frame, which translates into smoother decoding of the MID stream