I apologize if this is the inappropriate forum for this discussion
Reading through the archives I noticed there were attempts at regression testing Pd but I couldn't find any concrete results of these tests. I suspect the scope of the problem might have been beyond the scope of Pd itself thus it wasn't worth the effort. I do not as I am an outsider to this project.
What I'd like to know is what functionality would one want to test and provide regression tests for in Pd? What has been done in the past? What difficulties have arisen in unit testing or regression testing. If one wanted to unit test what type of assertions would one need, and if one wanted to regression test what kind of functionality would be required from the test framework?
When you're developing your applications and instruments do you ever wish you didn't have to listen to it over and over again to verify nothing serious has changed? Do you ever wish you could just test a filter or generator to see if it was producing the right sound rather than listening to it. Do you ever wish you could apply test first methodologies to your filters, signals generators and instruments?
Such a tool would be great but how would you want to use it? Would you want to integrate it into your testing framework or would you prefer it provides it's own test framework. What would you want the test to cover. Would you want it to cover IPC and latency issues? What common functionality would you want from it. What kind of assertions would you like to see.
There are difficulties in testing audio, you need complex functions to figure out: how loud it is; whether or not certain frequencies are too loud; is it producing a proper distribution of elements,.
What paradigms would like it to use. Would you prefer a procedural interface or an OO interface. What languages would want the test suite to directly support?
It's an interesting problem and if it sufficiently explored and implemented it could definitely help the audio programming community.
If you are interesting in providing input regarding the requirements of such a system please read the attached Invitation Letter as I am executing a study on what the open source audio community would like to see in the way of audio testing.
abram
========= Invitation Letter =============
Dear Potential Participant:
Hello, my name is Abram Hindle; I am currently a masters student in the faculty of Computer Science at the University Of Victoria. I am investigating the requirements of a Audio Testing Suite. as part of a directed studies course supervised by Dr. George Tzanetakis .
The purpose of this research is to extract the testing requirements of audio programmers. By gathering the requirements from potential users of an audio test suite we can build an effective audio test suite. This will help enable audio programmers regression test their code as well as employ test first methodology to audio development.
Participants should be experienced in computer instrument design or audio programming. Participants will be asked to fill out a questionnaire. Data from the questionnaire will be recorded for future use but the results anonymized to protect the participants identity.
Participation in this study is completely voluntary and the participant may withdraw at any time without any consequences. Data will be collected only with your consent. Participants will participate semi-anonymously.
Your participation is very important as it can improve the state of software engineering in the audio domain. You may contact the PI, Abram Hindle by email abez@uvic.ca or by telephone 250 721 9066 for further information about the research.
Thank you for your assistance.
You may participapate by visiting http://abez.ca/~abez/audiotest/
Sincerely,
Abram Hindle Graduate Student of the Faculty of Engineering Department of Computer Science University of Victoria Ph: (250)-721-9066