As of June 12, 2010 the following issues are outstanding:

• should SXF and VAL primitives be pervasives or in module COMPILER ?

• should type constructor SET OF only be followed by namedType ?

• should SXF/VAL primitives be pervasives or in module COMPILER ?

• should SET OF only be followed by namedType ?

• should SET OF only be followed by namedType ?

• should SXF/VAL primitives be pervasives or in module COMPILER ?

• should SET OF only be followed by nameType ?

As of June 11, 2010 the following issues are outstanding:

• confirm removal of usage cases for anonymous types

As of June 5, 2010 the following issues are outstanding:

• should digitCapacity have a binding or should it be a magic name or a pragma?

Bindings

• should digitCapacity have a binding or should it be a magic name?

• use [&] for digitCapacity constant?

Operator Bindings

• use [&] for digitCapacity constant?

• native subrange type constructor: TYPE S = [n .. m] OF INTEGER ?

• should open array autocast be marked and imported?

• review and confirm convertibility rules for SYSTEM types

As of June 4, 2010 the following issues are outstanding:

• confirm that single line comment prefix is //

• should single line comment be prefixed // or !!

• no outstanding issues

• should open array autocast be marked and imported

Review and Confirm

• review and confirm convertibility rules for SYSTEM types
• review and confirm removal of usage cases for anonymous types

As of June 3, 2010 the following issues are outstanding:

• should single line comment be prefixed // or !!

• no outstanding issues

• use [&] for digitCapacity constant?

• is NIL a pervasive or a reserved word?

As of June 2, 2010 the following issues are outstanding:

Lexis

• should ~ and & synonyms be removed?

• review and confirm removal of usage cases for anonymous types

• review and confirm definition of NIL

• can signatures for [?], [!] and [FOR] bindings be shared between static and dynamic ADTs

• review and confirm convertibility rules for SYSTEM types

• confirm move of type bound procedure facility from core language to an OO extension layer

• Rick to review and confirm ComplexMath

• should the type bound procedure facility be in the core language or in an OO extension layer

• review and possibly reduce the usage cases for anonymous types

• review convertibility rules for SYSTEM types

• Rick to review ComplexMath

• is ComplexMath OK now?

• should type bound procedure facility be in the core language or in an OO extension layer

• define constant for the meta-data (non-digit) overhead in an SXF string, probably in COMPILER

Documentation

• detailed description for initialisation order (wiki)
• FOR..IN loop narrative (wiki and language document)

• complete the Pathnames module

• complete the Pathnames module

• binding concatenation to + for string ADTs would be nice to have
• binding dot product to + for vector ADTs would be nice to have

• maybe the VAL macro should be removed

• should the VAL macro be removed

• review convertibility of SYSTEM types

• review definition of NIL

• TO DO: pseudo module COROUTINES

• finalise semantics for Exceptions
• TO DO: COROUTINES

Procedures

• should open array autocast be marked and imported: FROM SYSTEM IMPORT AUTOCAST

As of May 24, 2010 the following issues are outstanding:

• can signatures for [.], [!] and [FOR] bindings be shared between static and dynamic ADTs

• binding concatenation to + for string ADTs
• binding dot product to + for vector ADTs

• should there be native syntax for subrange types: [n .. m] OF INTEGER

• should the VAL macro be removed

• can signatures for [.], [!] and [FOR] bindings be shared between static and dynamic AD Ts?

• review ComplexMath
• finalise IO library, in particular TextIO

Keep-in-mind items

• binding concatenation to + for string AD Ts?
• binding dot product to + for vector AD Ts?

• finalise IO library, in particular TextIO

• finalise IO library, in particular Text IO?

• syntax for constructing bitsets, tentatively: SET [n]
• introduction of builtin associative array, tentatively: yes

• can matrices be handled using V-Type?
• where should semantic type specifiers go? RECORD and OPAQUE or module header?

• do we need a constant for the meta-data (non-digit) overhead in an SXF string? where should it live?

• do we need a constant for the meta-data (non-digit) overhead in an SXF string? where?

As of April 22, 2010 the following issues are outstanding:

• revisit VAL function macro

• do we need a constant for the meta-data (non-digit) overhead in an SXF string?

As of April 14, 2010 the following issues are outstanding:

• revise ComplexMath

• revise Complex Math?

• should the [{}] binding be removed

As of April 13, 2010 the following issues are outstanding:

• finalise COMPLEX and LONGCOMPLEX
• finalise IO library

• can matrices be handled using V-Type

• finalise procedures to convert to and from intermediate format for scalar values

Library

• finalise BCD and LONGBCD
• finalise COMPLEX and LONGCOMPLEX

• finalise procedures to convert to and from intermediate format for scalar values

• add fromCARD, fromINT, fromLONGCARD, fromLONGINT to [::] bindings

• add fromCARD, fromINT, fromLONGCARD, fromLONGINT conversion bindings

Grammar

• replacing IMMUTABLE with CONST, using Oberon-2 semantics

• should collection types be the default, if not, how to mark a type as a collection

Type Conversion

• introduction of a digitCapacity property for scalar types
• terminology for the intermediate format used to convert between scalars

• introduction of bindings for collection types
• should collection types be the default, if not, how to mark a type as a collection

• introduction of COUNT function for collections

Types

• syntax for constructing bitsets
• introduction of builtin associative array

• introduction of COUNT for enumerations and collections

• should V-Type allow bindings to * and / or not
• introduction of Collection type
• Collections are always opaque

• introduction of V-Type
• should V-Type allow bindings to * and / or not

As of April 12, 2010 the following issues are outstanding:

Pervasives

• introduction of pervasive COUNT for enumerations and collections

Operator Bindings

• introduction of V-Type, bindings for * and / or not