25-Oct-1999 (mnd) This file contains a list of all the Aldor bugs that appear in the bugs directory. They have been split into groups depending on how important they are and most are marked with an indication of how hard they might be to fix. ------------------------------------------------------------------------- Bugs in rough order of importance ([] indicates difficulty of fixing): [*] = hard [!] = moderately difficult [+] = not too difficult [-] = easy [?] = don't know Numbers of bugs outstanding (194): 45 important 34 not too important 38 good to fix 29 future work 48 ignored Numbers of bugs closed (154): 7 need closing 10 need closing but not fully dealt with 5 closed by trashing 69 closed and emailed 63 closed but can't email Important: 1260 [*] : hash code collision 1256 [*] : extending domain with domain export causes segfault 1255 [*] : two meanings for an operator 1253 [*] : missing export grabbed from outer scope 1251 [*] : core dump: domains defined in wrong order 1247 [*] : compiler fails to fill in proper const num (broken again) 1205 [*] : state local to domains gets lost 1197 [*] : passing Fortran functions to Fortran fails 1191 [+] : bit$Machine differs for SInt and BInt 1190 [!] : lazy function used before forced in domain init 1188 [*] : () -> T and T -> () generate the same hash code 1184 [*] : inlining in extensions doesn't work 1183 [?] : dependent types: related to bug1192 but not fixed ? 1181 [*] : confusions with extend and defaults 1178 [?] : problem with domain constants 1177 [+] : wrong extension (typo) can destroy important files 1168 [*] : segfaults when optimised above -Q2 (try -ginterp for clues) 1167 [!] : domain-valued function => compiler segfault 1153 [+] : More non-map category exports (see 1148) 1150 [+] : Specific length SInts (discussed at Edinburgh) 1148 [*] : Non-map category exports (see 1153) 1142 [-] : Pointer/long conversion; Linux okay - 64-bit specific? 1134 [!] : Overloading of domain names (bug or feature?) 1129 [!] : Bug in gen0Syme 1116 [+] : Segfault during domain initialisation (Summit/Linux) 1115 [!] : Domain initialisation problem? Related to inheritence bug? 1114 [!] : Defaults problem: segfault at runtime 1101 [*] : Conditional exports problem 1098 [!] : Side conditions not picked up 1097 [?] : Type inference problems (remove cond-apply fix and check) 1096 [?] : Either SMW is wrong or lots of hard work required ;) 1071 [?] : Missing export - inheritence bug? 1068 [?] : Philosophical question about sample for empty domain 1067 [?] : Category defaults fetched too late 991 [-] : Semantics of rem in Integer are wrong for Euclidean Domain 965 [*] : BinaryPowering bug (see bug 928) 928 [*] : BinaryPowering bug (see bug 965) 876 [!] : Association of space is wrong in lambdas 847 [*] : Attempt to overload constant C 827 [*] : Exporting domain constructors doesn't work 825 [!] : `Foo with with' segfaults the compiler 822 [?] : Compiler binds statically; user expects dynamic (see bug 1096) 818 [*] : Missing exports: Tuple(Type) problem? 513 [*] : Segfaults unless -Q2; demonstrates static binding Not too important: 1259 [!] : @ confuses tinfer in presence of enumeration 1258 [!] : same as bug 1257 1257 [!] : can't define with multi value 1252 [*] : missing export of complicated domain 1222 [*] : compile-time export not found when optimised 1249 [!] : gen0Syme bug again: complicated function return type 1248 [*] : curried functions returning category values have no exports 1245 [!] : inconsistency in error reports 1244 [!] : complicated categories confuse the compiler 1241 [!] : X$% doesn't work and we don't import enough meanings 1240 [!] : domain valued category exports and defaults problem 1198 [!] : repeated if's not cancelled (was fixed: see jflowFlipNegates()). 1196 [*] : embeddings fail in presence of `pretend' 1173 [!] : default/constant init ordering problem (see monom, needs Salli) 1133 [*] : Nasty tricks with domains 1086 [!] : Strange effects of local on type inference 1057 [?] : Irrelevent & out-of-date libraries ought not to be fatal 1056 [ ] : Empty bug report (see bug 1057) 1054 [?] : Constant export has value of free variable 922 [?] : Updating free variables used in dom consts (see bug 1054) 1131 [?] : Diamond extensions: how to drop extensions 1040 [?] : Optimiser bug; needs minisumit to test 1023 [?] : Stack growth excessive: bug message 982 [?] : Extend twice in the same file. 978 [?] : todd-coxeter.as and todd-coxeter-test.as hash the same 976 [?] : Hash codes identical for foobar, foobarbar, foobarbarbar ... 974 [?] : Integer doesn't have coerce: Integer -> % 932 [!] : Conditions in categories not passed to tinfer 915 [*] : Anonymous categories segfault the compiler 885 [*] : Type inference problem - types as exports, cascading exports 859 [-] : scobind doesn't check that goto target is in same scope 667 [*] : Original bug not found but foamExprType bug hit instead 505 [*] : Now compiles but segfaults at runtime 504 [*] : Not tested but same idea as bug505 Good to fix: 1225 [!] : tinfer problem reporting error 1224 [*] : confusion with embeddings 1223 [*] : curried functions or tinfer problem 1218 [!] : types based on enumerations don't modularise 1216 [*] : named enumerations don't work across files (see 1218) 1211 [*] : type dependencies are stuffed / too eager (see bug 1022) 1209 [*] : similar to 1207 and 1208 but with differences. 1208 [*] : identical to 1207 with slightly different formatting 1207 [*] : wierd, not fixed 1201 [?] : #include non-piled code into a piled file 1199 [*] : multi-pass inlining required 1195 [-] : fputss() in output.c is broken (fputs doesn't return # chars put) 1194 [?] : conditional categories confuse tinfer 1180 [?] : not fixed but might not be a bug (using function not domain) 1169 [!] : keyword argument/free variable confusion? 1163 [!] : 32/64 bit differences with assemble/dissemble (don't email) 1176 [+] : dispose! on constant string 1161 [!] : inherited exports not inlined 1151 [+] : Defaults problem (not a bug?) 1117 [!] : Named return values can be used in expressions! 1084 [!] : Non-const in type expression -> segfault 1082 [?] : Non-const in type expression (see 1084, 1081) 1081 [?] : Variations on non-const type expression (see 1082, 1084) 1022 [?] : Dependent types, convoluted domain definitions (see 1211) 1020 [?] : More dependent types, complicated Rep 1014 [?] : Multi/const definitions don't work well 945 [?] : Multi/const definitions don't work well (see bug 1014) 944 [?] : Multi/const definitions don't work well 970 [!] : Using (x, y) as for "variable" 955 [*] : Tagged tuple/tuple type aren't equal 951 [-] : Redefining functions in interpreter complains wrongly 911 [!] : Types evaluated too early 895 [!] : Interpreter ought not to die on call to C 828 [?] : If fixed then confusing error message else need fixing 821 [-] : Missing dipose!$Rec from Machine (axllib and basicmath) 664 [!] : Playing games with apply, space and . and floats 660 [!] : More games with apply and floats 516 [!] : Can build category bad using `add' ; check `add add' Future work: 1231 [*] : pushing Record/Tuple Type too far 1230 [*] : xaxiomxl doesn't work 1202 [*] : parsing of "with" could do with a warning 1158 [-] : Hacked - needs proper fix (don't send fix message) 1126 [*] : Pushing Record/Tuple(Type) a bit too far? 1122 [-] : Similar to the `FooDom: FooCat == add' bug 1089 [?] : Extending and using domain in same module doesn't work 1073 [*] : Confusion over % when types export types 957 [-] : Interpreter chokes on bad AO files 939 [*] : Partially fixed? Type inference just ain't good enough 927 [*] : Can't call C from interpreter 908 [?] : Unhelpful error message? 905 [!] : Can't use default-types for parameters in lambdas 901 [?] : Compiler could do better type inference on % 879 [-] : Rule in Sec 5.1 of user guide ignored by compiler 877 [-] : :* and : introduce distinct locals (wrong) 871 [?] : Problem with Ratio() in axllib; fixed for Fraction() in BM 870 [*] : Better pruning of possible types listed on type error 832 [!] : Type inference of generators/collections could be improved 814 [?] : Multi's in with-body 722 [!] : Can't use :* for type decls in function signatures 721 [*] : Restriction on redef of special ops (eg from Record) 714 [*] : Cascaded imports don't work 710 [?] : I don't understand the description of the bug 707 [?] : Possibly fixed or perhaps more work required 668 [*] : Bugs in test program - original problem needs fixing 629 [-] : Error for sqrt(negNum) could be more informative 600 [-] : Indication of error position on line could be improved 198 [*] : Doesn't inline function defined in funny way Ignored (need libaxiom, other libraries or missing header files). Some of these need to be checked and classified (eg Summit bugs): 1254 : segfault compiling a domain (need libadk) 1187 : optimiser bug (needs Salli) (fixed?) 1186 : bad C generation (needs Salli) (fixed?) 1185 : compiled produced code that GCC 2.8.0 chokes on (ignored) 1174 : export not found when optimised (fixed?) 1141 : Missing a header file 1135 : Needs Sumit to demonstrate 1118 : User-defined library not supplied (libfoo) 1109 : Needs minisummit to test 1052 : Obsolete? Needs to link with libaxiom 1042 : Needs minisumit 1037 : Aldor twice as slow as AXIOM, 7 times slower than Maple (minisumit) 1036 : Needs minisumit 1035 : Same as bug 1036, different signal received 1032 : Obsolete? Needs to link with libaxiom 1026 : Obsolete? Needs to link with libaxiom 985 : Suggest using BinaryPowering as default for ^ in Monoid 977 : Possibly obsolete comments about mod_* 963 : Needs libaxiom 959 : Needs libaxiom 953 : Needs libaxiom: bug error when out-of-date .asl found 952 : Needs libaxiom 946 : Library not supplied; obsolete? 921 : Semantics of symbol % and value of % when extending (fixed?) 920 : See bug 921 (fixed?) 862 : Obsolete? Uses libasinit 840 : Needs libaxiom (obsolete/fixed?) 838 : Identical to bug 836 836 : Needs libaxiom (see bug 838) 835 : DoubleFloat and literals not inlined in libaxiom (obsolete?) 809 : Possibly fixed; possibly needs tweak to interpreter 808 : Needs libaxiom, possibly fixed 807 : Nice feature not a bug! Improves the error messages 806 : Not a bug? 778 : XAxiomXL bug 760 : Segfaults with -Q3 702 : Not a bug? (space associates to the right) 687 : Fields are modules over Z hence n*x for F:Field, n:F, x:I 649 : Needs libaxiom 645 : Compiles okay but generates strange code 639 : XAxiomXL bug 630 : Games with apply and list: neat if it can be fixed 628 : Needs libaxiom 547 : Nasty recursive Rep definition: `Rep ==> foo(%)' in foo(T) 545 : Probably related to for.as bug (can't replicate this one) 296 : Missing important header file 294 : Missing important header file 293 : Missing important header file Need closing: 1236 : fiDFloDissemble uses Bool not FiBool (fixed) 1229 : comment on bug 1228 (fixed) 1228 : user probably has wrong glibc version of compiler (fixed) 1226 : problems with float conversion and ANSI C (fixed) 1220 : user doesn't understand Aldor semantics (fixed) 1217 : missing export segfaults compiler (fixed) 1072 : Missing export - types exporting type constants? (fixed) Need closing but don't email: 892 : Expecting too much of Tuple(Type)? (fixed) Need closing but not fully dealt with: 1234 : using glibc 2.0 compiler on 2.1 system (fixed) 1233 : user writing to constant string (fixed) 1232 : inliner not as good as it ought to be (fixed) [wait for MB verify] 1203 : inliner not as good as it ought to be (fixed) [wait for MB verify] 1136 : BasicMath linkage problem under Solaris (fixed?) 1121 : Interpreter segfault (fixed (check)) 1108 : Expecting too much of Tuple(Type)? 1083 : Const defn in a with-body (fixed) 1074 : Seems to have been fixed (no code to test) 1062 : Fixed (or AIX-specific bug?) Closed by trashing (see trash directory): 1155 : Rob Dyson test of axlbug 1100 : Empty bug report - where bug with no explanation of what it is 1050 : Reply to axlbug by SMW's vacation program ;) 1049 : Empty file 1048 : Empty file Closed and emailed (if appropriate): 1250 : hash$Integer never terminates (fixed) 1247 : compiler fails to fill in proper const num (fixed and broken again) 1246 : constants defined in with-bodies cause problems (fixed) 1243 : domain valued category exports problem (see 1242) (fixed) 1242 : segfault (fixed) and errors fixed by adding explicit "with {}" 1239 : tibupComma segfaults when dependent type tinfer fails (fixed) 1239 : segfault when reporting multiple meanings error (fixed) 1238 : identical to 1239. (fixed) 1237 : XByte is signed not unsigned (fixed) 1235 : semantics of shift() wrong under interpreter (fixed) 1227 : simple version of 1221 (fixed) 1221 : domain-valued functions/defaults problem (fixed) 1219 : category definition prevents inlining (fixed) 1215 : void return in non-void context (fixed) 1214 : select statements not working (fixed) 1213 : long file names and `ar' archive handling broken (fixed) 1212 : eager domain constructor/defaults problem (fixed) 1210 : curried function type checking segfaults compiler (fixed) 1206 : where statements broken (fixed) 1204 : tinfer of for-loops broken (fixed) 1200 : failure to inline from parents (fixed) 1193 : inliner doesn't create temporary for parameter (fixed) 1192 : tforms escape to stabFile() (fixed) 1189 : assert as last statement in function causes segfault (fixed) 1182 : Salli code not optimised like Axllib is (fixed) 1179 : same as bug1192 (fixed) 1175 : syntax errors too terse (fixed) 1172 : can't conditionally define function with if-else (fixed) 1171 : undefined symbol in Solaris GMP build (fixed) 1170 : conditional categories problem (related to 1212) (fixed) 1166 : bad macro for double-word multiply under Linux/GCC (fixed) 1165 : such-that clauses not working (fixed) 1164 : DFlo non-map exports broken (fixed) 1162 : unicl cannot create executable file called `for' (fixed) 1160 : Defaults/inheritence problem 1159 : Easy to hack (forbid local doms), ... (fixed) 1157 : EOF bug 1156 : Problem with Solaris build of 1.1.11e? (fixed: MB email) 1154 : ? bug 1144 : Function body tinfered with no types causes segfault (fixed) 1143 : Strange compiler message; code not supplied; obsolete? 1140 : Segfault under AIX with 1.1.10c (fixed) 1139 : Redefining << compiles but breaks if ~(T has BasicType) (fixed) 1137 : Documentation for =$Union needs explaining 1132 : Problem with category defaults/conditions (fixed) 1130 : Compiler needs to be reassured about side-conditions (fixed) 1127 : Problems with conditional exports (fixed) 1125 : Compiler never terminates with -Q2 (fixed) 1123 : Defaults problem (fixed) 1120 : The "1 possible type" error (fixed) 1119 : Question about warning message (fixed) 1113 : Compiler stores string literals in read-only memory map (fixed) 1111 : UNICL mangling its arguments (fixed) 1092 : Segfault if -V wasn't specified on command line (fixed) 1090 : Definition order of exports messed up (fixed) 1087 : Using constant before it is defined (fixed) 1080 : Implicit set within multi-assign (not implemented error) 1077 : Strange interactions between =>/const/macro (fixed) 1075 : Different behaviour when compiling files separate/together (fixed) 1070 : Problems with default arguments (fixed) 1069 : Types and #library problem (fixed) 1066 : Fixed 1061 : Fixed (see bug 1080) and fixed 1058 : Does really nasty things with libraries/extends (fixed?) 1055 : Macro gens nested seq in add-body (fixed) 1053 : Inliner problem (fixed) 969 : Conditional const export not found error when not needed (fixed) 880 : Assignment/define using void function segfaults compiler (fixed) 872 : Radix integers were completely broken; radix floats still are Need closing but don't email (email addresses are in old JANET format): ???? : Dependent types: not fixed but added compiler warning 1138 : Pushing categories-as-values too far (commented on bug1130) 1047 : Side conditions not checked (fixed) 1046 : Free-variable substitution not working (fixed) 1044 : Export not found (fixed) 1043 : Documented feature: can't use # at start of line in -gloop 1038 : Definitions move at top-level (fixed) 1039 : Inlining oddities for strange domain constructors (fixed) 1031 : Obsolete? v1.1.13 under OSF - bad library distribution? 1027 : Can't extend Boolean (fixed) 1015 : See bug 1038.as (fixed) 1009 : List Cross() doesn't work (fixed) 990 : Apology 989 : Accidental email - mail alias problems 988 : Checking of constness in types wee bit too eager (fixed) 987 : Similar to bug 1038 (fixed) 980 : Obsolete 971 : Fixed? 962 : Fixed 958 : Fixed 954 : Confusion with ==>/:=/== (fixed) 942 : Fixed 941 : Redefining macros doesn't raise warning (fixed) 940 : Fixed 934 : Fixed? Description of bug is unclear 926 : Fixed 910 : Fixed 898 : Fixed 891 : Fixed 887 : Fixed 867 : Fixed 852 : Segfault compiling collection used as value (fixed) 850 : Fixed 849 : Fixed 848 : Fixed 826 : Not a bug: parents checked before defaults 823 : Fixed? 810 : Fixed? 798 : Fixed 715 : Fixed 696 : Fixed 689 : Fixed? 688 : Function bodies with no braces were a problem 671 : Fixed? Unsure what the problem is 646 : Fixed 634 : Fixed 626 : Fixed 625 : Fixed? 593 : Fixed? 558 : Fixed? Try under IRIX? 538 : Fixed? 528 : Fixed 517 : Obsolete: no longer legal Aldor syntax? 494 : Obsolete (Themos fix) 482 : Messages about AIX port 481 : Messages about OSF port 480 : Messages about AIX port 428 : Assumed to be fixed (default7.as in test suite) 421 : Fixed; can't iterate over types 412 : Obsolete? 209 : Fixed? 107 : Test message? 101 : Test message Bugs in reverse date order ([] indicates urgency): [*] = important [!] = good to fix [-] = unimportant [?] = something strange [ ] = nothing to do -1260 [*] : hash code collision -1259 [!] : @ confuses tinfer in presence of enumeration -1258 [!] : same as bug 1257 -1257 [!] : can't define with multi value -1256 [*] : extending domain with domain export causes segfault -1255 [*] : two meanings for an operator -1254 [*] : segfault compiling a domain -1253 [*] : missing export grabbed from outer scope -1252 [!] : missing export of complicated domain -1251 [*] : core dump: domains defined in wrong order -1250 [*] : hash$Integer never terminates (fixed) -1249 [!] : gen0Syme bug again: complicated function return type -1248 [!] : curried functions returning category values have no exports -1247 [*] : compiler fails to fill in proper const num (broken again) -1246 [*] : constants defined in with-bodies cause problems (fixed) -1245 [!] : inconsistency in error reports -1244 [!] : complicated categories confuse the compiler -1243 [!] : domain valued category exports problem (fixed) -1242 [*] : segfault (fixed) and errors (fixed) -1241 [!] : X$% doesn't work and we don't import enough meanings -1240 [!] : domain valued category exports and defaults problem -1239 [!] : tibupComma segfaults when dependent type tinfer fails (fixed) -1238 [!] : identical to bug1239 (fixed) -1237 [*] : XByte is signed not unsigned (fixed) -1236 [*] : fiDFloDissemble uses Bool not FiBool (fixed) -1235 [*] : semantics of shift() wrong under interpreter (fixed) -1234 [ ] : using glibc 2.0 compiler on 2.1 system (fixed) -1233 [ ] : user writing to constant string (fixed) -1232 [!] : inliner not as good as it ought to be (fixed) -1231 [-] : pushing Record/Tuple Type too far -1230 [-] : xaxiomxl doesn't work -1229 [ ] : comment on bug 1228 (fixed) -1228 [ ] : user probably has wrong glibc version of compiler (fixed) -1227 [!] : eager domains and initialisation problems (fixed) -1226 [*] : problems with float conversion and ANSI C (fixed) -1225 [!] : tinfer problem reporting error -1224 [!] : confusion with embeddings -1223 [!] : curried functions or tinfer problem -1222 [*] : compile-time export not found when optimised -1221 [*] : domain-valued functions/defaults problem (fixed) -1220 [?] : user doesn't understand Aldor semantics (fixed) -1219 [!] : category definition prevents inlining (fixed) -1218 [!] : types based on enumerations don't modularise -1217 [!] : missing export segfaults compiler (fixed) -1216 [!] : named enumerations don't work across files (see 1218) -1215 [!] : void return in non-void context (fixed) -1214 [*] : select statements not working (fixed) -1213 [!] : long file names and `ar' archive handling broken (fixed) -1212 [*] : Eager domain constructor/defaults problem (fixed) -1211 [!] : type dependencies are stuffed / too eager (see bug 1022) -1210 [*] : curried function type checking segfaults compiler (fixed) -1209 [!] : similar to 1207 and 1208 but with differences. -1208 [!] : identical to 1207 with slightly different formatting -1207 [!] : wierd, not fixed -1206 [*] : where statements broken (fixed) -1205 [*] : state local to domains gets lost -1204 [*] : tinfer of for-loops broken (fixed) -1203 [!] : inliner not as good as it ought to be (fixed) -1202 [!] : parsing of "with" could do with a warning -1201 [!] : #include non-piled code into a piled file -1200 [*] : failure to inline from parents (fixed) -1199 [!] : multi-pass inlining required -1198 [!] : repeated if's not cancelled (was fixed: see jflowFlipNegates()). -1197 [*] : passing Fortran functions to Fortran fails -1196 [!] : embeddings fail in presence of `pretend' -1195 [!] : fputss() in output.c is broken (fputs doesn't return # chars put) -1194 [!] : conditional categories confuse tinfer -1193 [*] : inliner doesn't create temporary for parameter (fixed) -1192 [*] : tforms escape to stabFile() (fixed) -1191 [*] : bit$Machine differs for SInt and BInt -1190 [*] : lazy function used before forced in domain init -1189 [*] : assert as last statement in function causes segfault (fixed) -1188 [*] : () -> T and T -> () generate the same hash code -------------------------------^ 2000 ^------------------------------- -1187 [!] : optimiser bug (needs Salli) (fixed?) -1186 [!] : bad C generation (needs Salli) (fixed?) -1185 [ ] : compiled produced code that GCC 2.8.0 chokes on (ignored) -1184 [!] : inlining in extensions doesn't work -1183 [*] : dependent types: related to bug1192 but not fixed ? -1182 [*] : Salli code not optimised like Axllib is (fixed) -1181 [!] : confusions with extend and defaults -1180 [?] : not fixed but might not be a bug (using function not domain) -1179 [*] : same as bug1192 (fixed) -1178 [*] : problem with domain constants -1177 [*] : wrong extension (typo) can destroy important files -1176 [!] : dispose! on constant string -1175 [!] : syntax errors too terse (fixed) -1174 [?] : export not found when optimised (fixed?) -1173 [*] : default/constant init ordering problem (see monom, needs Salli) -1172 [!] : can't conditionally define function with if-else (fixed) -1171 [*] : undefined symbol in Solaris GMP build (fixed) -1170 [*] : Conditional categories problem (related to 1212) (fixed) -1169 [!] : keyword argument/free variable confusion? -1168 [*] : segfaults when optimised above -Q2 (try -ginterp for clues) -1167 [*] : domain-valued function => compiler segfault -1165 [*] : such-that clauses not working (fixed) -1164 [*] : DFlo non-map exports broken (fixed) -1163 [!] : 32/64 bit differences with assemble/dissemble (don't email) -1162 [!] : unicl cannot create executable file called `for' (fixed) -1161 [!] : inherited exports not inlined -1160 [ ] : Fixed by mnd but not closed -1159 [-] : Easy to hack (forbid local doms), hard to fix (fixed) -1158 [-] : Hacked - prints "oops" message (don't send fix message) -1157 [ ] : Fixed (themos?) but not closed -1156 [*] : Problem with Solaris build of 1.1.11e? (fixed) -1155 [?] : Rob Dyson testing axlbug but not mentioning this ... -------------------------------^ 1999 ^------------------------------- -1154 [ ] : "Bob's" experimental code: fixed (mnd) -1153 [?] : I don't quite understand this one -1151 [!] : Might be trivial to fix (not a bug?) -1150 [*!] : Specific length SInts (discussed at Edinburgh) -1148 [*] : Possibly hard to fix -1144 [*] : Function body tinfered with no types causes segfault (fixed) -1143 [?] : Strange compiler message; code not supplied; obsolete? -1142 [*] : Seems fine under Linux - 64-bit specific? 1141 [!] : Missing a header file -1140 [?] : Seems to have been fixed -1139 [!] : Redefining << compiles but breaks if ~(T has BasicType) (fixed) -1138 [-] : Pushing categories-as-values too far -1137 [-] : Documentation for =$Union needs explaining -------------------------------^ 1998 ^------------------------------- -1136 [?] : Seems to have been fixed 1135 [!] : Needs Sumit to demonstrate -1134 [*] : Overloading of domain names -1133 [*] : Nasty tricks with domains -1132 [?] : Fixed: problem with category defaults/conditions (fixed) 1131 [?] : Diamond extensions: how to drop extensions -1130 [-] : Compiler needs to be reassured about side-conditions (fixed) -1129 [*] : Bug in gen0Syme -1127 [!*] : Problems with conditional exports (fixed) -1126 [-] : Pushing Record/Tuple(Type) a bit too far? -1125 [?] : Compiler never terminates with -Q2 (fixed) -1123 [-] : Defaults problem (fixed) -1122 [-] : Similar to the `FooDom: FooCat == add' bug (easy to fix?) -1121 [?] : interpreter segfault (fixed) -1120 [?] : the "1 possible type" error (fixed) -1119 [?] : Question about warning message (fixed) 1118 [?] : User-defined library not supplied (libfoo) -1117 [!] : Named return values can be used in expressions! -1116 [*] : Needs Summit and Linux -1115 [*] : Domain initialisation problem? Related to inheritence problem? -1114 [*] : Defaults problem -1113 [*] : Attempt to write to read-only data segment -------------------------------^ 1997 ^------------------------------- -1111 [?] : UNICL mangling its arguments (fixed) 1109 [?] : Needs minisummit to test -1108 [!] : Expecting too much of Tuple(Type)? -1101 [!] : Conditional exports problem -1100 [?] : Empty bug report - where bug with no explanation of what it is -1098 [*] : Side conditions not picked up -1097 [?] : Type inference problems (remove cond-apply fix and check) -1096 [?] : Confusion over Aldor binding mechanism (see bug 822) -1092 [?] : Fixed: segfault if -V wasn't specified on command line -1090 [*] : Definition order of exports messed up (fixed) -1089 [?] : Extending and using domain in same module doesn't work -1087 [*] : Using constant before it is defined (fixed) -1086 [!*] : Strange effects of local on type inference -1084 [!] : Non-const in type expression (see 1082, 1081) -1083 [!] : Const defn in a with-body (fixed) -1082 [!] : Non-const in type expression (see 1084, 1081) -1081 [!] : Variations on non-const type expression (see 1082, 1084) 1080 [-] : Implicit set within multi-assign (not implemented error) -1077 [*] : Strange interactions between =>/const/macro (fixed) -1075 [?] : Fixed: strange behaviour when compiling files separate/together -1074 [?] : Seems to have been fixed (no code to test) -1073 [-] : Confusion over % when types export types -1072 [*] : Missing export - types exporting type constants? (fixed) -1071 [*] : Missing export - inheritence bug? -1070 [?] : Problems with default arguments (fixed) -1069 [?] : Types and #library problem (fixed) 1068 [?] : Philosophical question about sample for empty domain -1067 [*] : Category defaults fetched too late -1066 [*] : Fixed -1062 [?] : Fixed (or AIX-specific bug?) -1061 [?] : Fixed (see bug 1080) and fixed -1058 [*] : Does really nasty things with libraries/extends (fixed?) -1057 [*] : Irrelevent & out-of-date libraries ought not to be fatal -1056 [?] : Empty bug report (see bug 1057) -1055 [*] : Macro gens nested seq in add-body (fixed) -1054 [*] : Constant export has value of free variable -1053 [?] : Inliner problem (fixed) 1052 [?] : Obsolete? Needs to link with libaxiom -1050 [?] : Reply to axlbug by SMW's vacation program ;) -1049 [?] : Empty file -1048 [?] : Empty file -1047 [?] : Side conditions not checked (fixed) -1046 [?] : Free-variable substitution not working (fixed) -------------------------------^ 1996 ^------------------------------- -1044 [?] : Export not found (fixed) -1043 [?] : Documented feature: can't use # at start of line in -gloop -1042 [*] : Needs minisumit -1040 [*] : Optimiser bug; needs minisumit to test -1039 [-] : Inlining oddities for strange domain constructors (fixed) -1038 [*] : Definitions move at top-level (fixed) -1037 [*] : Aldor twice as slow as AXIOM, 7 times slower than Maple 1036 [?] : Needs minisumit 1035 [?] : Same as bug 1036, different signal received 1032 [?] : Obsolete? Needs to link with libaxiom -1031 [?] : Obsolete? v1.1.13 under OSF - bad library distribution? -1027 [?] : Can't extend Boolean (fixed) 1026 [?] : Obsolete? Needs to link with libaxiom -1023 [*] : Stack growth excessive: bug message -1022 [!] : Dependent types, convoluted domain definitions (see 1211) -1020 [!] : More dependent types, complicated Rep -1015 [*] : See bug 1038.as (fixed) -1014 [!] : Multi/const definitions don't work well -1009 [?] : List Cross() doesn't work (fixed) 991 [?] : Semantics of rem in Integer are wrong for Euclidean Domain - 990 [?] : Apology - 989 [?] : Accidental email - mail alias problems - 988 [-] : Checking of constness in types wee bit too eager (fixed) - 987 [!] : Similar to bug 1038 (fixed) - 985 [-] : Suggest using BinaryPowering as default for ^ in Monoid - 982 [*] : Extend twice in the same file. - 980 [?] : Obsolete - 978 [*] : todd-coxeter.as and todd-coxeter-test.as hash the same 977 [?] : Possibly obsolete comments about mod_* - 976 [*] : Hash codes identical for foobar, foobarbar, foobarbarbar ... - 974 [*] : Integer doesn't have coerce: Integer -> % - 971 [?] : Fixed? - 970 [!] : Using (x, y) as for "variable" - 969 [!] : Conditional const export not found error when not needed (fixed) - 965 [*] : BinaryPowering bug (see bug 928) 963 [?] : Needs libaxiom - 962 [?] : Fixed 959 [?] : Needs libaxiom - 958 [?] : Fixed - 957 [-] : Interpreter chokes on bad AO files - 955 [!] : Tagged tuple/tuple type aren't equal - 954 [*] : Confusion with ==>/:=/== (fixed) 953 [?] : Needs libaxiom: bug error when out-of-date .asl found 952 [?] : Needs libaxiom - 951 [-] : Redefining functions in interpreter complains wrongly 946 [?] : Library not supplied; obsolete? - 945 [!] : Multi/const definitions don't work well (see bug 1014) - 944 [!] : Multi/const definitions don't work well - 942 [?] : Fixed - 941 [-] : Redefining macros doesn't raise warning (fixed) - 940 [?] : Fixed - 939 [-] : Partially fixed? Type inference just ain't good enough -------------------------------^ 1995 ^------------------------------- - 934 [?] : Fixed? Description of bug is unclear - 932 [*] : Conditions in categories not passed to tinfer - 928 [*] : BinaryPowering bug (see bug 965) - 927 [-] : Can't call C from interpreter - 926 [?] : Fixed - 922 [*] : Updating free variables used in dom consts (see bug 1054) 921 [?] : Semantics of symbol % and value of % when extending (fixed?) 920 [?] : See bug 921 (fixed?) - 915 [*] : Anonymous categories segfault the compiler - 911 [!] : Types evaluated too early - 910 [?] : Fixed - 908 [-] : Unhelpful error message? - 905 [-] : Can't use default-types for parameters in lambdas - 901 [-] : Compiler could do better type inference on % - 898 [?] : Fixed - 895 [!] : Interpreter ought not to die on call to C - 892 [!] : Expecting too much of Tuple(Type)? (fixed) - 891 [?] : Fixed - 887 [?] : Fixed - 885 [*] : Type inference problem - types as exports, cascading exports - 880 [*] : Assignment using void function segfaults compiler (fixed) - 879 [-] : Rule in Sec 5.1 of user guide ignored by compiler - 877 [-] : :* and : introduce distinct locals (wrong) - 876 [*] : Association of space is wrong in lambdas - 872 [**] : Radix integers completely broken - 871 [-] : Problem with Ratio() in axllib; fixed for Fraction() in BM - 870 [-] : Better pruning of possible types listed on type error - 867 [?] : Fixed 862 [?] : Obsolete? Uses libasinit - 859 [*] : scobind doesn't check that goto target is in same scope - 852 [*] : Segfault compiling collection used as value (fixed) - 850 [?] : Fixed - 849 [?] : Fixed - 848 [?] : Fixed - 847 [*] : Attempt to overload constant C 840 [?] : Needs libaxiom (obsolete/fixed?) 838 [?] : Identical to bug 836 836 [?] : Needs libaxiom (see bug 838) 835 [?] : DoubleFloat and literals not inlined in libaxiom (obsolete?) - 832 [-] : Type inference of generators/collections could be improved - 828 [!] : If fixed then confusing error message else need fixing - 827 [*] : Exporting domain constructors doesn't work - 826 [?] : Not a bug: parents checked before defaults - 825 [*] : `Foo with with' segfaults the compiler - 823 [?] : Fixed? - 822 [*] : Compiler binds statically; user expects dynamic (see bug 1096) - 821 [!] : Missing dipose!$Rec from Machine (axllib and basicmath) [easy] - 818 [*] : Missing exports: Tuple(Type) problem? - 814 [-] : Multi's in with-body - 810 [?] : Fixed? 809 [?] : Possibly fixed; possibly needs tweak to interpreter 808 [?] : Needs libaxiom, possibly fixed 807 [?] : Nice feature not a bug! Improves the error messages 806 [?] : Not a bug? - 798 [?] : Fixed 778 [?] : XAxiomXL bug 760 [?] : Segfaults with -Q3 - 722 [-] : Can't use :* for type decls in function signatures - 721 [-] : Restriction on redef of special ops (eg from Record) - 715 [?] : Fixed - 714 [-] : Cascaded imports don't work - 710 [-] : I don't understand the description of the bug - 707 [-] : Possibly fixed or perhaps more work required 702 [?] : Not a bug? (space associates to the right) - 696 [?] : Fixed - 689 [?] : Fixed? - 688 [!] : Outer meanings not visible from where body 687 [?] : Fields are modules over Z hence n*x for F:Field, n:F, x:I - 671 [?] : Fixed? Unsure what the problem is - 668 [-] : Bugs in test program - original problem needs fixing - 667 [*] : Original bug not found but foamExprType bug hit instead - 664 [!] : Playing games with apply, space and . and floats - 660 [!] : More games with apply and floats 649 [?] : Needs libaxiom - 646 [?] : Fixed 645 [?] : Compiles okay but generates strange code 639 [?] : XAxiomXL bug - 634 [?] : Fixed 630 [?] : Games with apply and list: neat if it can be fixed - 629 [-] : Error for sqrt(negNum) could be more informative 628 [?] : Needs libaxiom - 626 [?] : Fixed - 625 [?] : Fixed? - 600 [-] : Indication of error position on line could be improved - 593 [?] : Fixed? - 558 [?] : Fixed? Try under IRIX? -------------------------------^ 1994 ^------------------------------- 547 [?] : Nasty recursive Rep definition: `Rep ==> foo(%)' in foo(T) 545 [?] : Probably related to for.as bug (can't replicate this one) - 538 [?] : Fixed? - 528 [?] : Fixed - 517 [?] : Obsolete: no longer legal Aldor syntax? - 516 [!] : Can build category bad using `add' ; check `add add' - 513 [*] : Segfaults unless -Q2; demonstrates static binding - 505 [*] : Now compiles but segfaults at runtime - 504 [*] : Not tested but same idea as bug505 - 494 [?] : Obsolete (Themos fix) - 482 [?] : Messages about AIX port - 481 [?] : Messages about OSF port - 480 [?] : Messages about AIX port - 428 [?] : Assumed to be fixed (default7.as in test suite) - 421 [?] : Fixed; can't iterate over types - 412 [?] : Obsolete? 296 [?] : Missing important header file 294 [?] : Missing important header file 293 [?] : Missing important header file - 209 [?] : Fixed? 198 [?] : Doesn't inline function defined in funny way - 107 [?] : Test message? - 101 [?] : Test message