I tried some reading and writing of files with [textfile].

Reading file:

""" tried /home/claude/workinprogress/copyme/gravity/patches/../data/cm_g_kicker-0000.reservoir and succeeded """

This works as I think it should, some feedback on file IO.

An even better way of getting feedback would be an extra [textfile] outlet that outputs [read 0( or [read 1( depending on whether the file loaded successfully or not, perhaps with the absolute path name - that way the patch knows if it worked, rather than just the console-watching user knowing, and help catch bugs with relative paths and patches in different directories etc. Feedback at the point of error.

I think there should still be console output, but perhaps saying "tried to read" instead of just "tried" - tried what?

Reading non-existent file:

""" tried /home/claude/workinprogress/copyme/gravity/patches/../data/cm_g_kicker-0067.eveloop and failed tried /usr/lib/pd/extra/../data/cm_g_kicker-0067.eveloop and failed error: ../data/cm_g_kicker-0067.eveloop: can't open error: ../data/cm_g_kicker-0067.eveloop: read failed """

Two "error:" messages seems a little redundant.

Writing file successfully:

""" """

No output - no way of knowing if the file was written successfully, except maybe using [shell] to get a directory listing.

I think console output like "tried to write [absolute path] and succeeded/failed" would be useful, and the outlet for success/failure outputting [write 0( or [write 1(, for the reasons above.

Writing empty file successfully:

""" error: ../data/cm_g_kicker-0000.reservoir: write failed """

This is a bug. The file is written to disk successfully (all 0 bytes without errors). Reading the 0 byte file works like any other file reading, if successful it clears the [textfile]. It is only the console message that is wrong.