In the provided example, yes I'm seeing the stale zo issue. In my "real" code, each submodule also requires a common module as well and in this situation I get a cyclic require error as well when check-syntax is performed as well as the zo error at the bottom of DR. I don't know if there are two distinct issues or one domino-ing the other.<div> <div> <div class="gmail_quote">On Tue, Aug 28, 2012 at 12:27 PM, Sam Tobin-Hochstadt <<a href="mailto:samth@ccs.neu.edu" target="_blank">samth@ccs.neu.edu</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div class="HOEnZb"><div class="h5">On Tue, Aug 28, 2012 at 12:23 PM, Eric Dobson <<a href="mailto:eric.n.dobson@gmail.com">eric.n.dobson@gmail.com</a>> wrote: > Possibly, how is it not working for you? > > With: > #lang typed/racket/base > > (module mytypes racket/base > > (provide S) > > (struct S ())) > > (module tyuser racket/base > > (require > (submod ".." mytypes)) > > (define (mkS value) > (S))) > > (note submodules are now in plain racket) > > > I get: > tmp.rkt:17:5: link: module mismatch; > possibly, bytecode file needs re-compile because dependencies changed > importing module: 'tyuser > exporting module: 'tmp > exporting phase level: 0 > internal explanation: variable not provided (directly or indirectly) > at: make-S > in: make-S.1 > context...: > /Users/endobson/proj/racket/plt/collects/compiler/cm.rkt:308:0: compile-zo* > /Users/endobson/proj/racket/plt/collects/compiler/cm.rkt:508:26 > /Users/endobson/proj/racket/plt/collects/compiler/cm.rkt:501:42 > /Users/endobson/proj/racket/plt/collects/compiler/cm.rkt:466:0: > maybe-compile-zo > /Users/endobson/proj/racket/plt/collects/compiler/cm.rkt:579:2: do-check > /Users/endobson/proj/racket/plt/collects/compiler/cm.rkt:653:4 > /Users/endobson/proj/racket/plt/collects/compiler/commands/make.rkt:77:8: > for-loop > /Users/endobson/proj/racket/plt/collects/compiler/commands/make.rkt: > [running body] > /Users/endobson/proj/racket/plt/collects/raco/raco.rkt: [running body] > /Users/endobson/proj/racket/plt/collects/raco/main.rkt: [running body] > > Which looks like a bug in TR handling of submodules. </div></div>I believe that this is a bug in Racket, although I'm not sure. You can see a simpler example here: <a href="http://drdr.racket-lang.org/25284/collects/tests/typed-racket/succeed/internal-typed-submod.rkt" target="_blank">http://drdr.racket-lang.org/25284/collects/tests/typed-racket/succeed/internal-typed-submod.rkt</a> <div class="HOEnZb"><div class="h5"> > > On Tue, Aug 28, 2012 at 7:32 AM, Ray Racine <<a href="mailto:ray.racine@gmail.com">ray.racine@gmail.com</a>> wrote: >> The following does not work. Bug? >> >> #lang typed/racket/base >> >> (module mytypes typed/racket/base >> >> (provide >> (struct-out S)) >> >> (struct: S ())) >> >> (module tyuser typed/racket/base >> >> (require >> (submod ".." mytypes)) >> >> (: mkS (Any -> S)) >> (define (mkS value) >> (S))) >> >> ____________________ >> Racket Users list: >> <a href="http://lists.racket-lang.org/users" target="_blank">http://lists.racket-lang.org/users</a> >> > ____________________ > Racket Users list: > <a href="http://lists.racket-lang.org/users" target="_blank">http://lists.racket-lang.org/users</a> </div></div>-- sam th <a href="mailto:samth@ccs.neu.edu">samth@ccs.neu.edu</a> </blockquote></div> </div></div>