I'm not alone! <div>FWIW, I still have that problem in 5.2.0.6.</div><div>And I confirm the ".1" minimalist test case.</div><div><br></div><div>When running 'racket' from the command line, this does not occur however:</div> <div><div>laurent:~$ racket</div><div>Welcome to Racket v5.2.0.6.</div><div>Edit ~/.racketrc to modify inits.</div><div>> .1</div><div>0.1</div><div>> </div></div><div><br></div><div>But this still happens with gracket:</div> <div><div>Welcome to Racket v5.2.0.6.</div><div>This is a simple window for evaluating Racket expressions.</div><div>Quit now and run DrRacket to get a better window.</div><div>The current input port always returns eof.</div> <div>> .1</div><div>0.0</div><div>> </div></div><div><br></div><div>Probably irrelevant: Could the GPU be in the loop somehow?</div><div><br></div><div>Good luck,</div><div>Laurent</div><div><br></div><div><br><div class="gmail_quote"> On Tue, Dec 27, 2011 at 14:26, Marijn <span dir="ltr"><<a href="mailto:hkBst@gentoo.org">hkBst@gentoo.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> -----BEGIN PGP SIGNED MESSAGE-----<br> Hash: SHA1<br> <br> On 24-12-11 16:23, Matthew Flatt wrote:<br> > At Wed, 14 Dec 2011 16:49:23 +0100, Marijn wrote:<br> >> #lang racket (* .1 pi)<br> >><br> >> produces 0.0 reliably in a freshly started drracket even when<br> >> rerunning it, but:<br> >><br> >> $ racket -e '(* .1 pi)' 0.3141592653589793<br> >><br> >> and the program:<br> >><br> >> #lang racket pi (* .1 pi)<br> >><br> >> on the first run (drracket again) produces:<br> >><br> >> 3.141592653589793 0.0<br> >><br> >> while on subsequent runs it produces:<br> >><br> >> 3.141592653589793 0.3141592653589793<br> ><br> > Thanks --- this is very helpful!<br> ><br> > I'm still stumped, unfortunately. It looks the same as PR 12070,<br> > which we never figured out:<br> ><br> > <a href="http://bugs.racket-lang.org/query/?debug=&database=default&cmd=view+audit-trail&cmd=view&pr=12070" target="_blank">http://bugs.racket-lang.org/query/?debug=&database=default&cmd=view+audit-trail&cmd=view&pr=12070</a><br> <br> Interesting!,<br> ><br> though it is a bit hard to read the Audit Trail because<br> everything is jumbled together.<br> <br> I reproduce that bug and step 7) and 8) can be replaced with a single<br> re-evaluate for me.<br> <br> It seems to indicate very clearly that the problem is in the floating<br> point start-up somehow and has nothing to do with `pi' or variable<br> references.<br> <br> Another possibility would be that multiplication (*) was at fault, but<br> my new smallest test-case rules that out:<br> <br> #lang racket<br> .1<br> <br> => 0.0<br> <br> And actually it works also directly in the REPL without running any<br> program:<br> <br> Welcome to DrRacket, version 5.2.0.7--2011-12-15(-/f) [3m].<br> Language: racket; memory limit: 128 MB.<br> > .1<br> 0.0<br> > 1<br> 1<br> > .1<br> 0.0<br> > .2<br> 0.0<br> > .3<br> 0.0<br> > .4<br> 0.0<br> > .5<br> 0.0<br> > .6<br> 0.0<br> > .7<br> 0.0<br> > .8<br> 0.0<br> > .9<br> 0.0<br> > 1.0<br> 1.0<br> > .1<br> 0.1<br> <br> So should we look at the byte-code produced for this (how does one do<br> that and what is the expected code)?<br> Is there some way to invoke racket from the command line in such a way<br> that it behaves like the drracket initial repl?<br> How does one run the byte-code?<br> <br> What external suspects are there? I suppose gcc, but are there any<br> other? My CFLAGS are set to a very conservative AFAIK<br> CFLAGS="-march=native -ggdb -O2 -pipe", currently on version 4.5.3.<br> <br> Marijn<br> <br> <br> -----BEGIN PGP SIGNATURE-----<br> Version: GnuPG v2.0.18 (GNU/Linux)<br> Comment: Using GnuPG with Mozilla - <a href="http://enigmail.mozdev.org/" target="_blank">http://enigmail.mozdev.org/</a><br> <br> iEYEARECAAYFAk75x5IACgkQp/VmCx0OL2xZRACcC8dubYCSNXScyOG8wdJQxxN1<br> xAgAnRVKv8YpTWODEKn2qEMMhcXHZLSk<br> =n+xT<br> -----END PGP SIGNATURE-----<br> </blockquote></div><br></div>