Duy Nguyen

ASAN_OPTIONS is mentioned in gc .travis.yaml and there's code in the GC to recognize address_sanitizer feature. I'd say just try it out. Probably will work out of the box. Oh.....

Just tried with CFLAGS=-fsanitize=address and ran make test. We may have a problem in `count_size_and_length`.

@shirok CiSE generated code probalby confuses the address sanitizer (or worse, we're relying on undefined behavior). sys-strftime generates this code ``` { char tmpbuf[256]; strftime(tmpbuf,sizeof(tmpbuf),format,&(SCM_SYS_TM_TM(tm))); { SCM_RESULT=(tmpbuf); goto SCM_STUB_RETURN; }...

Correction, `make -C src test` passes. Testing external modules will still fail with the same problem (CiSE generates lots of stuff). But since the core code passes asan, I'm quite...

Thanks. sys-ctermid needs fxing too because to pass the test suite with address sanitizer (it's used in termios tests). And I think the potential problem is also in sys-gethostname and...

Is this to override the default value in`*history-size*`, or do we want separate history sizes for file and in memory? Bash separates the two with `HISTSIZE` and `HISTFILESIZE`

I could try to take a stab at this, but first, should we provide REPL control in ~/.gaucherc (less poluting ~/.*shrc). That kind of control could be used from the...

As long as we are sure no race condition (even without synchronization), I think changing an active REPL session is ok. For resizing the history vector in line-edit object, one...

> It is advantageous to make srfi/.scm implementation and srfi-.scm alias Out of curiosity, what are the advantages?

The good news is python isn't doing any better here. And in gosh if I Ctrl-C, then it gets back to normal. Ther user just has to know to type...