Loading ...
Sorry, an error occurred while loading the content.

[eiffel-nice-library] Re: ISE's position

Expand Messages
  • marcel.satchell@eiffel.co.uk
    ... raised. If ... could have ... The semantics of features such as `has and `capacity may not have major implementation technical problems, but they do have
    Message 1 of 3 , Sep 30, 1999
    • 0 Attachment
      > because from what has been discussed no major implementation technical
      > problems for the ISE Eiffel runtime or for EiffelBase has been
      raised. If
      > there had one, ISE would have posted a request explaining why it
      could have
      > been an issue.

      The semantics of features such as `has' and `capacity' may not have
      major implementation technical problems, but they do have major problems
      for the abstractions and legacy software. Is ISE really prepared to see
      the semantics of `has' or `capacity' changed? It has been pointed out to
      me by others that ISE's current kernel is not ELKS95 compliant. Will we
      see a situation where ISE says OK to ELKS2000 and then ignores it? I
      would very much like to know whether ISE is prepared to change
      EiffelBase in this sort of way since it has ramifications for any work
      I might undertake in order to help out with the library. If the
      goalposts change regularly it is difficult to contribute anything
      worthwhile.

      Marcel Satchell
    • Roger Browne
      ... I guess Emmanuel was assuming that the 2000 vintage would not modify has , but would add a pair of new features (e.g. has_value / has_object ) so as not
      Message 2 of 3 , Oct 1, 1999
      • 0 Attachment
        marcel.satchell@... wrote:

        > The semantics of features such as `has' and `capacity' may not have
        > major implementation technical problems, but they do have major problems
        > for the abstractions and legacy software...

        I guess Emmanuel was assuming that the 2000 vintage would not modify
        'has', but would add a pair of new features (e.g.
        'has_value'/'has_object') so as not to break ISE's existing use of
        'has'.

        Of course these new features could break existing software which uses
        those names for its own purposes - but most users would gladly use a
        different name in their own code in return for a rigorous, interoperable
        kernel class.

        Now that Emmanuel has indicated ISE's goodwill, I suggest we seize the
        opportunity and proceed rapidly towards vintage 2000.

        I suggest we look in more detail at STRING, as most of the focus so far
        has been on ARRAY. I also suggest that other classes, and the questions
        of Eiffel/C interfacing, be left for vintage 2001. Much better to get a
        small but useful improvement out the door, than to risk spending another
        five years without any tangible result.

        Regards,
        Roger
        --
        Roger Browne - roger@... - Everything Eiffel
        6 Bambers Walk Wesham PR4 3DG UK - +44 1772 687525
      Your message has been successfully submitted and would be delivered to recipients shortly.