Bug#1136716: A non-superficial test case for guile-fluidsynth

2026-05-20 Thread Tommi Höynälänmaa



David Pirotte kirjoitti 20.5.2026 klo 22.56:

Hello,


What is wrong with using AI? I did use ChatGPT to generate a draft
for the test case.

I don't want to discuss this subject, sorry.
Please never ever send me AI generated code and/or patch, tx.

Also, so you save some of your precious time, I suggest you ask me if I
need/want some code/patch before 'doing anything' - most of the time, I
do/will prefer to do it myself.


Ok.





What about the change I wrote to the .ffi file compilation?

I suspect this patch was also AI generated. But nonetheless, I'll write
a patch myself, thanks.



Ok. I didn't use AI with it, though.




David



    - Tommi

--
Kotisivu / Homepage: http://www.iki.fi/tohoyn/
Sähköposti / E-Mail: [email protected]
GPG-sormenjälki / GPG fingerprint:
55F4 2477 7155 3528 5CB2 2B7A BB86 1FDE 4046 0F83
FT, Debian-ylläpitäjä / PhD, Debian Maintainer



Bug#1136716: A non-superficial test case for guile-fluidsynth

2026-05-20 Thread David Pirotte
Hello,

> What is wrong with using AI? I did use ChatGPT to generate a draft
> for the test case.

I don't want to discuss this subject, sorry.
Please never ever send me AI generated code and/or patch, tx.

Also, so you save some of your precious time, I suggest you ask me if I
need/want some code/patch before 'doing anything' - most of the time, I
do/will prefer to do it myself.

> What about the change I wrote to the .ffi file compilation?

I suspect this patch was also AI generated. But nonetheless, I'll write
a patch myself, thanks.

David


pgpz5g9bYfKGy.pgp
Description: OpenPGP digital signature


Bug#1136716: A non-superficial test case for guile-fluidsynth

2026-05-19 Thread Tommi Höynälänmaa

Hello

David Pirotte kirjoitti 20.5.2026 klo 1.48:

Hi Tommi,


The test case and the soundfont file it uses are attached. I’ve also
included the |.wav| file used to generate the |.sf2| file.

Thanks, but this is unnecessarily (far too) complicated, as it somehow
reproduces tests that are part of libfluidsynth and nyacc themselves.

Also, I use (unit-test) from guile-lib, as you know, which imnsho is
quite superior to sfrfi-64, as it uses goops and therefore is entirely
customizable.

Last, I somehow (very much) doubt you wrote these tests by yourself:
for info, I wont' accept any AI generated code in any of my projects.

Just be a little patient here, I'll write a test suite, a simple test
suite - because there is no need for a complicated one - that covers
what needs to be covered in Guile-Fluisynth, but not 'more'.

All that needs to be tested is the higher level api that
guile-fluidsynth provides - in short, what the example in the manual
proposes - if that works, everything thing works fine, in particular
(but none exhaustive list):

1- can it import the init module
- then it found libfluisynth
- it found the default  soundfont file
2- can it make a synthesizer
- then all settings related lower level functions are
  running fine, as well as the guilie-fluisynth goops
  higher level api
3- can it play a note
dito, the guilie-fluisynth goops higher level api to
play a soiund work fine

I'll ping you as soon as I will have written the test suite.

Thanks,
David


Ok.

What is wrong with using AI? I did use ChatGPT to generate a draft for 
the test case.


What about the change I wrote to the .ffi file compilation?

- Tommi

--
Kotisivu / Homepage:http://www.iki.fi/tohoyn/
Sähköposti / E-Mail:[email protected]
GPG-sormenjälki / GPG fingerprint:
55F4 2477 7155 3528 5CB2 2B7A BB86 1FDE 4046 0F83
FT, Debian-ylläpitäjä / PhD, Debian Maintainer


Bug#1136716: A non-superficial test case for guile-fluidsynth

2026-05-19 Thread David Pirotte
Hi Tommi,

> The test case and the soundfont file it uses are attached. I’ve also 
> included the |.wav| file used to generate the |.sf2| file.

Thanks, but this is unnecessarily (far too) complicated, as it somehow
reproduces tests that are part of libfluidsynth and nyacc themselves.

Also, I use (unit-test) from guile-lib, as you know, which imnsho is
quite superior to sfrfi-64, as it uses goops and therefore is entirely
customizable.

Last, I somehow (very much) doubt you wrote these tests by yourself:
for info, I wont' accept any AI generated code in any of my projects.

Just be a little patient here, I'll write a test suite, a simple test
suite - because there is no need for a complicated one - that covers
what needs to be covered in Guile-Fluisynth, but not 'more'.

All that needs to be tested is the higher level api that
guile-fluidsynth provides - in short, what the example in the manual
proposes - if that works, everything thing works fine, in particular
(but none exhaustive list):

1- can it import the init module
- then it found libfluisynth
- it found the default  soundfont file
2- can it make a synthesizer
- then all settings related lower level functions are
  running fine, as well as the guilie-fluisynth goops
  higher level api
3- can it play a note
dito, the guilie-fluisynth goops higher level api to
play a soiund work fine

I'll ping you as soon as I will have written the test suite.

Thanks,
David


pgpKGcdpxWHhr.pgp
Description: OpenPGP digital signature