Add 'read()' to the naming scheme, with the logic that it reads the whole stream until end of message or file. And add the methods for TFile to the stdlib.

How to reproduce:

type a = enum b,c,d
a.items()

My code -> https://gist.github.com/1110749 (although, the code was altered for issue #42)
When you try to pass a function to createThread, if it depends on generics for one of its arguments, the compiler will not accept it.

parallel.nim(24, 17) Error: type mismatch: got (TThread[TThreadFuncArgs[int]], proc(TThreadFuncArgs[int]), TThreadFuncArgs[int])
but expected one of:
createThread(t: var TThread[TMsg], tp: proc (TMsg){.thread.}, param: TMsg)
createThread(t: var TThread[TMsg], tp: proc (){.thread.})

type
  TEnum = enum
    has_value, empty
  TA = object
    case discriminator: TEnum
    of has_value:
      x : char
    of empty:
proc n(x : char): ref TA =
  new(result)
  result.discriminator = has_value
  result.x = x
var a = n('j')
echo repr(a)

Makes:
Hint: used config file '/Users/eric/Documents/Private/Code/nimrod/Compiler/Nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: test [Processing]
Error: internal error: computeRecSizeAux()
Traceback (most recent call last)
nimrod.nim(86)(?):           nimrod
nimrod.nim(69)(?):           HandleCmdLine
main.nim(201)(?):            MainCommand
main.nim(121)(?):            CommandCompileToC
main.nim(95)(?):             CompileProject
main.nim(90)(?):             CompileModule
passes.nim(139)(?):          processModule
passes.nim(104)(?):          processTopLevelStmt
sem.nim(202)(?):             myProcess
sem.nim(188)(?):             SemStmtAndGenerateGenerics
semstmts.nim(828)(?):        semStmt
semstmts.nim(563)(?):        SemTypeSection
semstmts.nim(558)(?):        typeSectionFinalPass
semtypinst.nim(15)(?):       checkConstructedType
types.nim(969)(?):           computeSize
types.nim(955)(?):           computeSizeAux
types.nim(850)(?):           computeRecSizeAux
types.nim(837)(?):           computeRecSizeAux
types.nim(861)(?):           computeRecSizeAux
msgs.nim(577)(?):            InternalError
msgs.nim(524)(?):            rawMessage
msgs.nim(521)(?):            rawMessage
msgs.nim(478)(?):            handleError
Error: unhandled exception: [Assertion failure] file: rod/msgs.nim line: 478
 [EAssertionFailed]

import macros
proc testProc: string {.compileTime.} =
  result = ""
  result = result & ""
macro test(n : stmt): stmt =
  result = newNimNode(nnkStmtList)
  echo testProc()
test:
  "hi"

Produces (NB: line numbers in eval are probably wrong, I played around with it):
Hint: used config file '/Users/eric/Documents/Private/Code/nimrod/Compiler/Nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: test [Processing]
Hint: macros [Processing]
Error: command line(0)(?): internal error: getStrValue
Traceback (most recent call last)
nimrod.nim(86)(?):           nimrod
nimrod.nim(69)(?):           HandleCmdLine
main.nim(201)(?):            MainCommand
main.nim(121)(?):            CommandCompileToC
main.nim(95)(?):             CompileProject
main.nim(90)(?):             CompileModule
passes.nim(139)(?):          processModule
passes.nim(104)(?):          processTopLevelStmt
sem.nim(202)(?):             myProcess
sem.nim(188)(?):             SemStmtAndGenerateGenerics
semstmts.nim(813)(?):        semStmt
semstmts.nim(32)(?):         semCommand
semstmts.nim(24)(?):         semExprNoType
semexprs.nim(1021)(?):       semExpr
semexprs.nim(934)(?):        semMacroStmt
sem.nim(109)(?):             semMacroExpr
evals.nim(1104)(?):          eval
evals.nim(1085)(?):          evalAux
evals.nim(1043)(?):          evalAux
evals.nim(776)(?):           evalMagicOrCall
evals.nim(450)(?):           evalEcho
evals.nim(443)(?):           getStrValue
msgs.nim(573)(?):            InternalError
msgs.nim(552)(?):            liMessage
msgs.nim(478)(?):            handleError
Error: unhandled exception: [Assertion failure] file: rod/msgs.nim line: 478
 [EAssertionFailed]
Works if you invert and use result = "" & result though.

Sometimes recursive generic types still do not work, see the ignored tests for examples.

Filename: nimrod_0.8.10.exe
affected file: koch.exe

TR/ATRAPS.Gen

OS: Windows XP Pro SP3

Code:
method m(i: int): int =
return 5

Result:
Mike-Benfields-MacBook:Code mike$ nimrod compile --run sample.nim
Hint: used config file '/Users/mike/Applications/nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: sample [Processing]
No stack traceback available
SIGSEGV: Illegal storage access. (Attempt to read from nil?)

Nimrod 0.8.10.
Mac OS 10.6.3.

The code:

import
  macros, strutils

macro test_macro*(n: stmt): stmt =
  result = newNimNode(nnkStmtList)
  var ass : PNimrodNode = newNimNode(nnkAsgn)
  add(ass, newIdentNode("str"))
  add(ass, newStrLitNode("after"))
  add(result, ass)
when isMainModule:
  var str: string = "before"
  test_macro(str):
    var i : integer = 123
  echo str

Doesn't compile with:
Hint: system [Processing]
Hint: macro_gen [Processing]
Hint: macros [Processing]
Hint: strutils [Processing]
Hint: parseutils [Processing]
Traceback (most recent call last)
nimrod.nim(86)(?): nimrod
nimrod.nim(69)(?): HandleCmdLine
main.nim(201)(?): MainCommand
main.nim(121)(?): CommandCompileToC
main.nim(95)(?): CompileProject
main.nim(90)(?): CompileModule
passes.nim(139)(?): processModule
passes.nim(104)(?): processTopLevelStmt
sem.nim(182)(?): myProcess
sem.nim(168)(?): SemStmtAndGenerateGenerics
semstmts.nim(825)(?): semStmt
semstmts.nim(45)(?): semWhen
semstmts.nim(814)(?): semStmt
semstmts.nim(808)(?): semStmt
semstmts.nim(32)(?): semCommand
semstmts.nim(24)(?): semExprNoType
semexprs.nim(1022)(?): semExpr
semexprs.nim(935)(?): semMacroStmt
sem.nim(109)(?): semMacroExpr
evals.nim(1084)(?): eval
evals.nim(1065)(?): evalAux
evals.nim(1023)(?): evalAux
evals.nim(806)(?): evalMagicOrCall
evals.nim(1023)(?): evalAux
evals.nim(745)(?): evalMagicOrCall
evals.nim(284)(?): evalCall
evals.nim(1065)(?): evalAux
evals.nim(1045)(?): evalAux
evals.nim(364)(?): evalAsgn
assign.nim(171)(?): FieldDiscriminantCheck
Error: unhandled exception: assignment to discriminant changes object branch [EInvalidField]

When running the example at http://force7.de/nimrod/httpserver.html , and accessing localhost from Chromium (specifically), I get an

Error: unhandled exception: index out of bounds [EInvalidIndex]

at the line in httpserver.nim in next() that reads

if cmpIgnoreCase(data[0], "GET") == 0 or cmpIgnoreCase(data[0], "POST") == 0:

This is due to data having no elements, due to buf being the null string. It seems that, rather than sending a request with "GET / HTTP/1.1" as the first line, Chromium is sending a zero-length line. I don't know if this is proper behaviour, but in any case my program should not crash. So I propose changing this to something like

if data.len() == 0 or cmpIgnoreCase(data[0], "GET") == 0 or cmpIgnoreCase(data[0], "POST") == 0:

after cloning the git, I've tried to do bootstrapping as shown in the wiki but got the following error:

c:\Users\root\Nimrod\lib\pure\osproc.nim(323, 67) Error: type mismatch: got (seq
[string], openarray[string])
but expected one of:
&(x: string, y: char): string
&(x: char, y: char): string
&(x: string, y: string): string
&(x: char, y: string): string
&(x: T, y: seq[T]): seq[T]
&(x: seq[T], y: seq[T]): seq[T]
&(x: seq[T], y: T): seq[T]

FAILURE

after looking into the file I've do the following, changed:
if false: # poUseShell in options:
cmdl = buildCommandLine(getEnv("COMSPEC"), @["/c", command] & args)
else:
cmdl = buildCommandLine(command, args)
into:
#if false: # poUseShell in options:
# cmdl = buildCommandLine(getEnv("COMSPEC"), @["/c", command] & args)
#else:
cmdl = buildCommandLine(command, args)

because it produce the same effect, but now compiles.

now when I try again I get the following error (after rod/nimrod.exe is done - iteration 1):
Hint: operation successful (41491 lines compiled; 72 sec total) [SuccessX]
iteration: 2
Traceback (most recent call last)
koch.nim(195) koch
koch.nim(138) boot
koch.nim(123) bootIteration
os.nim(653) moveFile
os.nim(167) OSError
Error: unhandled exception: Access is denied.
[EOS]

P.S:
I've tested the newly compiled rod/nimrod.exe and it seems to work.

In the tutorial there's the example "procedural type":

type
  TCallback = proc(x: int)

proc echoItem(x: int) =
  echo(x)

proc forEach(callback: TCallback) =
  const
    data = [2, 3, 5, 7, 11]
  for d in items(data):
    callback(d)

forEach(echoItem)

Why is it necessary to give parameters names when only declaring a type?
I feel like I should be able to do:

type
  TCallback = proc(int)

Is this in the works?

Mac can not bootstrap since 974a143 :
$ sh build.sh
gcc -w -O3 -fno-strict-aliasing -Ibuild -c build/3_2/nim__dat.c -o build/3_2/nim__dat.o
gcc -w -O3 -fno-strict-aliasing -Ibuild -c build/3_2/system.c -o build/3_2/system.o
build/3_2/system.c:379: error: thread-local storage not supported for this target
build/3_2/system.c:386: error: thread-local storage not supported for this target
build/3_2/system.c:387: error: thread-local storage not supported for this target

See http://lists.apple.com/archives/xcode-users/2006/Jun/msg00550.html

(edit: fixed sha-1)

EDIT: The issue is still present with a bit more complicated scenario

type
  TTestObj = object of TObject
    x: string

proc MakeObj(): TTestObj =
  result.x = "Hello"

var global = MakeObj()

var count = 0
proc raiseRandomly =
  inc count
  if count mod 2 == 0:
    raise newException(ESystem, "")

proc foo =
  while true:
    raiseRandomly()
    global = MakeObj()

while true:
  try:
    foo()
  except:
    nil

The quicksort benchmark: https://bitbucket.org/jmb/shootout/src/741575f6602f/programs/quicksort/quicksort.nimrod-2.nimrod crashes with the following backtrace:

Traceback (most recent call last)
quicksort.nim(95)        worker
inboxes.nim(193)         send
inboxes.nim(166)         rawSend
SIGSEGV: Illegal storage access. (Attempt to read from nil?)

From the manual:
import macros
macro case_token(n: stmt): stmt =
# creates a lexical analyzer from regular expressions
# ... (implementation is an exercise for the reader :-)
nil

case_token: # this colon tells the parser it is a macro statement
of r"[A-Za-z_]+[A-Za-z_0-9]*":
  return tkIdentifier
of r"0-9+":
  return tkInteger
of r"[\+\-\*\?]+":
  return tkOperator
else:
  return tkUnknown

Doesn't parse with:
Hint: used config file '/Users/eric/Documents/Private/Code/nimrod/Compiler/Nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: test [Processing]
Hint: macros [Processing]
Error: test.nim(8): expression expected, but found '[same indentation]'

Example:

type
  TEnum = enum
    a=2, b=4
var foo: TEnum = TEnum(0)
echo($foo)

Output:

4 (invalid data!)

Expected output:

b

I get the following error when attempting to compile.

log.c: In function 'Initlogger_45041':
log.c:318: error: request for member 'Sup' in something not a structure or union
Error: execution of an external program failed

It probably has something to do with my InitLogger function:
https://gist.github.com/1106939

According to git bisect, the bad commit is 1986469582280ae0eda91f1d337e9dc0759f8b77. I booted with the revision before and rebuilt koch there too, but that doesn't help.

➜  Nimrod  ./koch boot                                          
iteration: 1
compiler/nimrod0 cc   compiler/nimrod.nim
Hint: used config file '/home/tass/dev/nimrod/Nimrod/config/nimrod.cfg' [Conf]
Hint: used config file 'compiler/nimrod.cfg' [Conf]
Hint: system [Processing]
Error: system module needs 'pushFrame'
FAILURE

Nimrod/build dir contents are now

csources.zip
cycle.h
empty.txt
nimbase.h

there are some files and empty.txt seems no longer needed.

and other projects often use to commit empty dir by addnig `.gitignore'.

thank you. :-)

When compiling the mandelbrot benchmark: https://bitbucket.org/jmb/shootout/src/741575f6602f/programs/mandelbrot/mandelbrot.nimrod the compiler crashes with the following backtrace:

Hint: used config file '/home/jerome/prog/nimrod/hg/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: mandelbrot [Processing]
Hint: math [Processing]
Hint: os [Processing]
Hint: strutils [Processing]
Hint: parseutils [Processing]
Hint: times [Processing]
Hint: posix [Processing]
Traceback (most recent call last)
nimrod.nim(88)           nimrod
nimrod.nim(69)           HandleCmdLine
main.nim(201)            MainCommand
main.nim(121)            CommandCompileToC
main.nim(95)             CompileProject
main.nim(90)             CompileModule
passes.nim(139)          processModule
passes.nim(104)          processTopLevelStmt
cgen.nim(935)            myProcess
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(753)        genStmts
ccgstmts.nim(183)        genWhileStmt
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(753)        genStmts
ccgstmts.nim(183)        genWhileStmt
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(751)        genStmts
ccgstmts.nim(202)        genBlock
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(753)        genStmts
ccgstmts.nim(183)        genWhileStmt
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(750)        genStmts
ccgstmts.nim(764)        genStmts
ccgstmts.nim(738)        genAsgn
ccgexprs.nim(1680)       expr
ccgexprs.nim(783)        genCall
ccgexprs.nim(269)        initLocExpr
ccgexprs.nim(1637)       expr
cgen.nim(691)            genProc
cgen.nim(685)            genProcNoForward
cgen.nim(616)            genProcAux
ccgstmts.nim(750)        genStmts
ccgstmts.nim(758)        genStmts
ccgstmts.nim(159)        genReturnStmt
ccgstmts.nim(764)        genStmts
ccgstmts.nim(738)        genAsgn
ccgexprs.nim(1687)       expr
ccgexprs.nim(1538)       genTupleConstr
SIGSEGV: Illegal storage access. (Attempt to read from nil?)

import  macros

macro test*(a: stmt): stmt =
  proc testproc(recurse: int) =
    echo "Thats weird"
    var o : PNimrodNode = nil
    echo "  no its not!"
    o = newNimNode(nnkNone)
    if recurse > 0:
      testproc(recurse - 1)
  testproc(5)

test:
  "hi"

If you remove the = nil, or replace it with newNimNode(nnkNone) it works...

Fails with
Error: test_case_of.nim(6): internal error: evalAux: nkNone

gcc.exe -c -w -IC:\Users\alex\Downloads\Araq-Nimrod-ffa6519\Araq-Nimrod-ffa6519\
lib -o C:\Users\Alex\Desktop\vlc\nimcache\irc.o C:\Users\Alex\Desktop\vlc\nimcac
he\irc.c
C:\Users\Alex\Desktop\vlc\nimcache\nim__dat.c:182: error: field `Data' has incom
plete type
C:\Users\Alex\Desktop\vlc\nimcache\irc.c:88: error: field `Data' has incomplete
type
Error: execution of an external program failed

GIST -> https://gist.github.com/1140601

If one of the actions already running tries to send a new action to the pool after the main thread has called join, then the new actions may not get called.

import actors

var
  a: TActorPool[int, void]
createActorPool(a)

proc task (i: int) {.thread.} =
  echo i
  if i != 0: a.spawn (i-1, task)

a.spawn (10, task)
a.join()

This code should print all numbers between 0 and 10 in no particular order, instead it only prints "10".

echo parseExpr("foo(bar)").toLisp() 

produces

nnkCall(nnkIdent(foo), nnkIdent(bar))

but should

nnkCall(nnkIdent("foo"), nnkIdent("bar"))

The current build system for the nimrod compiler looks rather improvised. How about using a build system like CMake? This would enable easy cross-compiling and user familiarity.

How to reproduce:

import macros
const x =  parseStmt("""
echo "42"
""").lispRepr()
echo x

output.atEnd(output)

sucks. Probably other procs than close and atEnd in PStream require the argument to be passed again.

The documentation of 'const' is not up to date: It is possible that it is evaluated at runtime; 'const' means 'readonly' then. Unfortunately, what is and what will be compile time evaluable heavily depends on the implementation.

Hi.

Following code:

var q: seq[int]
q = @[1,2,3,4,5]
q[high(q)+1] = 10
q[high(q)+2] = 666

echo($high(q))
echo($len(q))
echo($q[5])

C:\nimrod>nimrod c -d:release --opt:size hello
4
5
10

It means that high and len functions are based on static information about sequences?
Maybe echo($high(q)) should print 6 and echo($len(q)) should 7 ?

And if I compile without -d:release and --opt:size (just pure nimrod c hello.nim), I get the following error:

Traceback (most recent call last)
hello.nim(10) hello
Error: unhandled exception: index out of bounds [EInvalidIndex]

Where I'm wrong?

example:

>>> proc echo_int(a,b:int,x,y:var int) =
...     x = a 
stdin(14, 0) Error: tabulators are not allowed
>>> proc echo_int(a,b:int,x,y:var int) =
...     x = a
stdin(15, 0) Error: tabulators are not allowed
stdin(14, 4) Error: redefinition of 'echo_int'

In example, "echo_int" proc is Error.then, "echo-int" should be un-defined,and could redefine.

1. Having an xml comment on the same line as an xml node causes the parser to fail to parse the next node correctly.
2. PXmlNode.text seems to be broken.

When trying to call a func like:

for func in items(emitter.events[event]):
        func(args) #call function with args.

I get func is not callable.

Also, when I try to pass a List to a table, it doesn't behave correctly.

append(emitter.events[event], func)

events is a Table and what its returning is a DoublyLinkedList which I try to append the proc to.

Gist -> https://gist.github.com/1101726

The compile pragma seems to truncate .cpp files to just .c during compile time.

The following code does not compile:

import actors

var
  p: TActorPool[int, void]
createActorPool(p)
for i in 0 .. < 300:
  p.spawn(i, proc (x: int) {.thread.} = echo x)
p.join()

Using current HEAD (3d36182)

iota github/Araq/Nimrod % ./bin/nimrod c rod/c2nim/c2nim.nim
Hint: used config file '/home/manveru/github/Araq/Nimrod/config/nimrod.cfg' [Conf]
Hint: used config file 'rod/c2nim/c2nim.cfg' [Conf]
Hint: system [Processing]
Hint: c2nim [Processing]
Hint: strutils [Processing]
Hint: parseutils [Processing]
Hint: os [Processing]
Hint: times [Processing]
Hint: posix [Processing]
Hint: parseopt [Processing]
Hint: llstream [Processing]
Hint: ast [Processing]
Hint: msgs [Processing]
Hint: options [Processing]
Hint: lists [Processing]
Hint: nstrtabs [Processing]
Hint: nhashes [Processing]
Hint: nversion [Processing]
Hint: crc [Processing]
Hint: ropes [Processing]
Hint: platform [Processing]
/home/manveru/github/Araq/Nimrod/rod/ropes.nim(215, 18) Hint: 'RopeSeqInsert(rs: var TRopeSeq, r: PRope, at: Natural)' is declared but not used [XDeclaredButNotUsed]
Hint: idents [Processing]
Hint: rnimsyn [Processing]
Hint: scanner [Processing]
Hint: lexbase [Processing]
Hint: wordrecg [Processing]
/home/manveru/github/Araq/Nimrod/rod/scanner.nim(174, 14) Hint: 'findIdent(L: TLexer, indent: int): bool' is declared but not used [XDeclaredButNotUsed]
/home/manveru/github/Araq/Nimrod/rod/rnimsyn.nim(414, 26) Hint: 'rfNoIndent' is declared but not used [XDeclaredButNotUsed]
/home/manveru/github/Araq/Nimrod/rod/rnimsyn.nim(238, 11) Hint: 'popCom(g: var TSrcGen)' is declared but not used [XDeclaredButNotUsed]
/home/manveru/github/Araq/Nimrod/rod/rnimsyn.nim(111, 12) Hint: 'putLong(g: var TSrcGen, kind: TTokType, s: string, lineLen: int)' is declared but not used [XDeclaredButNotUsed]
Hint: clex [Processing]
Hint: cparse [Processing]
Hint: pegs [Processing]
Hint: unicode [Processing]
Hint: astalgo [Processing]
Hint: rodutils [Processing]
/home/manveru/github/Araq/Nimrod/rod/astalgo.nim(234, 19) Hint: 'strTableToYaml(n: TStrTable, marker: var TIntSet, indent: int, maxRecDepth: int): PRope' is declared but not used [XDeclaredButNotUsed]
Hint: strtabs [Processing]
Hint: hashes [Processing]
rod/c2nim/cpp.nim(281, 22) Error: type mismatch: got (string, string, int, int)
but expected one of: 
parsePeg(pattern: string, filename: string, line: int, col: int): TPeg

End of Copyright year should be 2011.

diff --git a/lib/copying.txt b/lib/copying.txt
index 6fcdca1..ab2de06 100755
--- a/lib/copying.txt
+++ b/lib/copying.txt
@@ -1,6 +1,6 @@
 =======================================================
                   The Nimrod Runtime Library
-             Copyright (C) 2004-2010  Andreas Rumpf
+             Copyright (C) 2004-2011  Andreas Rumpf
 =======================================================

 This is the file copying.txt, it applies to the Nimrod Run-Time Library

Env: Ubuntu 11.01.

I use Arrow Key in Nimrod REPL,I don't expected the behavior.

>>> ^[[A^[[D^[[D^[[D

I think UP or Down Key is "History",And Right or Left Key is "Move Cursor".

This code:
import macros

type 
    TA = tuple[a : int]
    PA = ref TA

macro test*(a: stmt): stmt =
  var val : PA
  new(val)
  val.a = 4

test:
  "hi"

Produces upon compilation:
Hint: used config file '/yadayada/Nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: test [Processing]
Hint: macros [Processing]
stack trace: (most recent call last)
test.nim(12): file: test.nim, line: 12
Error: test.nim(10): attempt to access a nil address

type
  TEnum = enum
    a, b
var val = {a, b}
echo repr(val)

results in:
Traceback (most recent call last)
test.nim(5):              test
repr.nim(102)(?):            reprSet
repr.nim(81)(?):             reprSetAux
SIGSEGV: Illegal storage access. (Attempt to read from nil?)
Hint: used config file '/Users/eric/Documents/Private/Code/nimrod/Compiler/Nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: test [Processing]
Hint: test.nim(2): 'TEnum' is declared but not used [XDeclaredButNotUsed]
gcc -c -w -fasm-blocks -O1 -I/Users/eric/Documents/Private/Code/nimrod/Compiler/Nimrod/lib -o nimcache/nim__dat.o nimcache/nim__dat.c
gcc -c -w -fasm-blocks -O1 -I/Users/eric/Documents/Private/Code/nimrod/Compiler/Nimrod/lib -o nimcache/test.o nimcache/test.c
gcc  -ldl   -o test  nimcache/test.o nimcache/nim__dat.o nimcache/lib/system.o
Hint: operation successful (5559 lines compiled; 1 sec total) [SuccessX]
./test
Error: execution of an external program failed

Experimenting with Nimrod today, I was able to make the compiler segfault by attempting to compile:

type Node = tuple[left: ref Node]

proc traverse(root: ref Node) =
  if root.left != nil: traverse(root.left)

This code is nonsense, but on the principle that the compiler should never segfault, I'm suggesting this is a bug. I'm not certain what in the code is causing it, but it seems to at least require: the tuple (the equivalent object does not cause a segfault), the recursive reference in the tuple, and the recursive proc.

Nimrod was compiled on Ubuntu 10.04, on an AMD-64 architecture. I'm compiling the above code like so:

james@bast:~/nim_hw$ nimrod c hw.nim
Hint: used config file '/home/james/nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: hw [Processing]
Segmentation fault

Trying to build a generic hash table using generics. The compiler crashes with an unhandled exception (line 879 in rod/ast.nim).

The code to be compiled is here: http://pastebin.com/aaXdUDjX

I get the mentioned error with the following code:

proc `||->`* [T] (func: proc(): T, callback: proc(val: T)): TThread[TThreadFuncArgs[T]] =

The error is related to generics because when I change T to int in the return value (T is suppose to be int anyway), it'll work.

The full code can be found here: https://gist.github.com/1110749

This compiles:
import macros
proc makeMacro: PObject =
result = nil
var p = makeMacro
This doesn't:
import macros
proc makeMacro: PNimrodNode =
result = nil
var p = makeMacro

Dump:
Hint: used config file '/Users/eric/Documents/Private/Code/nimrod/Compiler/Nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: test [Processing]
Hint: macros [Processing]
Hint: strutils [Processing]
Hint: parseutils [Processing]
Error: internal error: mapType
Traceback (most recent call last)
nimrod.nim(85)(?): nimrod
nimrod.nim(69)(?): HandleCmdLine
main.nim(194)(?): MainCommand
main.nim(120)(?): CommandCompileToC
main.nim(95)(?): CompileProject
main.nim(90)(?): CompileModule
passes.nim(134)(?): processModule
passes.nim(98)(?): processTopLevelStmt
cgen.nim(917)(?): myProcess
ccgstmts.nim(719)(?): genStmts
cgen.nim(677)(?): genProc
cgen.nim(659)(?): genProcNoForward
cgen.nim(654)(?): genProcPrototype
ccgtypes.nim(523)(?): genProcHeader
ccgtypes.nim(181)(?): genProcParams
ccgtypes.nim(122)(?): isInvalidReturnType
ccgtypes.nim(105)(?): mapType
msgs.nim(487)(?): InternalError
msgs.nim(457)(?): rawMessage
msgs.nim(454)(?): rawMessage
msgs.nim(410)(?): handleError
Error: unhandled exception: [Assertion failure] file: rod/msgs.nim line: 410
[EAssertionFailed]

Many script programming languages are commonly able to comment out w/ '#'.
Nimrod is also able to do it.

And many script programming languages are able to use w/ #!shebang.

!shebang is i.e.

#!/usr/bin/env gosh

(define (fib n)
  (if (< n 2) 1
    (+ (fib (- n 2)) (fib (- n 1)))))

(print (fib 30))

But Nimrod cant kick w/ #!shebang.

To reproduce, compile (eg with nimrod c test.nim) following source:
method somethin(obj: TObject) =
echo "do nothing"
Generates
Hint: used config file '/Users/eric/Documents/Private/Code/nimrod/Compiler/Nimrod/config/nimrod.cfg' [Conf]
Hint: system [Processing]
Hint: test [Processing]
Traceback (most recent call last)
nimrod.nim(85) nimrod
nimrod.nim(69) HandleCmdLine
main.nim(194) MainCommand
main.nim(120) CommandCompileToC
main.nim(95) CompileProject
main.nim(90) CompileModule
passes.nim(137) processModule
passes.nim(91) closePasses
cgen.nim(965) myClose
cgen.nim(604) genProcAux
ccgstmts.nim(679) genStmts
ccgstmts.nim(120) genIfStmt
ccgexprs.nim(257) initLocExpr
SIGSEGV: Illegal storage access. (Attempt to read from nil?)

Using install.sh, I get the following error:

cp: doc/altsyn.txt: No such file or directory

Error when linking nimrtl on mac os x:

Undefined symbols:
  "_environ", referenced from:
      _Getenvvarsc_40240 in os.o
      _Getenvvarsc_40240 in os.o
ld: symbol(s) not found

It seems from the man:

Shared libraries and bundles don't have direct access to environ, which is only available to the loader ld(1) when a complete program is being linked.
The environment routines can still be used, but if direct access to environ is needed, the _NSGetEnviron() routine, defined in <crt_externs.h>, can be used to retrieve the address of environ at runtime.

I'm willing to fix, but would need some pointers