Yehouda Harpaz

There really isn't a "use case". We just run tests where we quickload "all" the systems in quicklisp (and when we can find them run their tests). As it happen...

It is the call to (SETF TRIVIAL-INDENT:INDENTATION) inthe short backtrace that I post. The last think that I "typed" (actually happened by the running of the tests) was `(ql:quickload "softdrink")`,...

Ok, thanks. I am actually testing what I get from quicklisp, so will actually test the changes after the next quicklisp release.

It also doesn't work on LispWorks (and never did), because LispWorks also does not allow constant symbols as slot names.

I didn't know what a pull request is until 10 minutes ago. I don't actually work with a branch of the source code, I downloaded it from quicklisp.

Changing to **:lispworks** (without a version) is even better. It is not going to work on earlier versions of LispWorks anyway.

Without SSL it changes the way it works in LispWorks, but still works ok. With SSL it breaks because the underlying code that it uses to support SSL (CL+SSL) is...

I don't actually make changes on github, but there are two PRs that fix it in the "Pull requests" tab. The one by tfiala looks better https://github.com/edicl/drakma/pull/129.

It is not obvious to me either what is the right fix. Using the keyword :id for initarg is more in line with the rest of common-lisp. Maybe just document...

No. I test code that I load by quicklisp, and don;t work on github. I come here (github) only to hassle people.