Original bug ID: 1591
Reporter: administrator
Status: closed
Resolution: fixed
Priority: normal
Severity: minor
Category: ~DO NOT USE (was: OCaml general)

Bug description

At the end of this message is a uuencoded gzipped tar file containing
a small set of sources for reproducing a compiler problem. 61 lines
total. Here's the terminal output when it breaks:

lobus:/s/conscious> ./compile
rm: No match.
setenv OCAMLC /usr/bin/ocamlc
/usr/bin/ocamlc -c s_conscious_data.ml
/usr/bin/ocamlc -c conscious_data.mli
/usr/bin/ocamlc -c conscious_data.ml
/usr/bin/ocamlc -c s_conscious.ml
/usr/bin/ocamlc -c conscious.mli
/usr/bin/ocamlc -c conscious.ml
The implementation conscious.ml does not match the interface conscious.cmi:
Module type declarations do not match:
module type Unintern_strategy =
sig
val unintern_eventdesc :
([> ANY] as 'a) list list -> 'a Conscious_data.eventdesc end does not match module type Unintern_strategy = sig val unintern_eventdesc : ([> ANY] as 'a) list list -> 'a Conscious_data.eventdesc
end
Modules do not match:
sig
val unintern_eventdesc :
([> ANY] as 'a) list list -> 'a Conscious_data.eventdesc end is not included in sig val unintern_eventdesc : ([> ANY] as 'a) list list -> 'a Conscious_data.eventdesc
end
Values do not match:
val unintern_eventdesc :
([> ANY] as 'a) list list -> 'a Conscious_data.eventdesc is not included in val unintern_eventdesc : ([> ANY] as 'a) list list -> 'a Conscious_data.eventdesc
rm -f conscious.cmi conscious_data.cmi s_conscious.cmi s_conscious_data.cmi conscious_data.cmo s_conscious.cmo s_conscious_data.cmo
lobus:/s/conscious>

The problem is that it fails when saying that two values with
apparently identical types do not match.

I can cause this behavior with the Debian 3.06-15 build, or with a CVS
image I updated a few hours ago.

There are two occurrences of

type process_identifier = int

in the source. If I delete them and replace all occurrences of
"process_identifier" with "int", the problem goes away. There are a
number of other small changes I can make to the source to make the
problem go away, but this one is the strangest because it shouldn't
affect the meaning at all.

Tim Freeman tim@fungible.com
Which is worse: ignorance or apathy? Who knows? Who cares?
GPG public key fingerprint ECDF 46F8 3B80 BB9E 575D 7180 76DF FE00 34B1 5C78

begin 664 ouch.tar.gz
M'XL(!5E<3X``^U864_;0!#.LW_%5%3BJ.)C?4E104*T#Y5Z/*^5B%S7HA M"SXB[YJ">OSV[JZ=Q($D5A$.(/Q%L:W,9&;D];<SGTVKUSILVT-AZ,NS[82^ M73]/T;/#P/50@-PPD+^'GAWTP&^_M%ZOX+G#W!DK5^3?87"M,B63)A,6TQ MA^W(%?57K[_O^;7U]Y1_Z(<]L%NL:897OOY;;ZP12RU!^!CZ-T:>0/\<]DR2 M,'W,C"U.!4VOX=O1X9?/1V":VC\C.(G)7:,U+BZH)>^45:1C3*YHU->.?9QF M:9]<<XNE\G;',8WJ4>X$*7A>M[[]51K^0)\'Y(LY81E!1]&6&SB1<=[IE9 M@WUE_-6AUT15?[I_#Y]ZD=?M$RY+NPBS?+6M@#%#S5O+?L5W)?]^SY0<%
M<O]W'#=P.OYOF9$)S2-C+V_!M&$@_)DBFP"@G+AZ"."RR*],HBZUL01-\+8 MTT1XSD]WAR:H_C_?N]K)H?D?K.8_\KU9_[<=S?_0M3O^;P))%A4QA:,/L]'
M"[W1,"J;N)U0^)ZR5-\'7*18T$O;J6_;!L&P#6.H9A:Z35-140Y&4@+P,[) MP=GAUQ^G@#ELXUV(&1?ZH*W]GCG:SF+(*AMB40[LJ=!M-"UCD/VLE1Q/_ M;114_$>AKZX=Y-A^Q_]-@*4D+B(*Q[7!][ACVJM!C?]31?3H.9KX'WCS]S\( MN8K_(>KZ_T:@F_LDSPCE?,@BV7O9.:.Y[.ZRHQN&-M,;U=MQ/,SR8=GG9<O? MAY.SCY4!LO-E,7[#V:?*_51W\3+<SKOMT2[@]+:<%EA4)5,/HAPNY"5LCU3\ M]TLSJU&B3'GY61RTM&IW">Y0D60E5N9*-+2M2(48^(=<151,JQEE419G6 M-13X8@[:G"I5RO'%CGO"#V?+!2BZIB-+S+=K(XJ>E0DD'RNSDO?R%Z^^+*
MM!9?>;21HXG_KC]V>7$<^ZOB"=1G./93#]3!8/Z8%VB&,HN3SS:M'
MZQ5"DT9H5DFBGZ).),:Q1-$]]LY\AEO3_1Q<!C?.ZT[Y[[C*SW'ML./
M1K!D_J]HWK'E-6"A_[>D)KX'R*W-O\CQ7_DN1W_-X'E_;]!%50.CZ(+IL%6 MJ8+%*7ZM+E"^ZW0!S"?R_]$&#U,'Y5P"-=PI;Z5:>)A>*/.I[U,_4ATZ=.C0 ,X07@'_A.C=4*```
`
end