Visual C++ .NET 2003 compatibility

Collapse
This topic is closed.
X
X
 
  • Time
  • Show
Clear All
new posts
  • Filip Konvicka

    #1

    Visual C++ .NET 2003 compatibility

    Hello,

    I'm using VC++ 6.0. Sadly, it lacks some of STL classes,
    like hash_set. So I've decided to upgrade to VC++ .NET
    2003.

    My questions are:
    - can I produce PEs for x86, like I did with VC++ 6.0,
    with .NET version of VC?
    - will my apps run on Windows 95 (I mean 4.0.0.95 or
    whatever version is the oldest)?

    Thanks for any information.

    Best Regards,
    Filip

  • Ken Alverson

    #2
    Re: Visual C++ .NET 2003 compatibility

    "Filip Konvicka" <anonymous@disc ussions.microso ft.com> wrote in message
    news:2ba3701c39 32a$a2a1d450$a6 01280a@phx.gbl. ..[color=blue]
    >
    > I'm using VC++ 6.0. Sadly, it lacks some of STL classes,
    > like hash_set. So I've decided to upgrade to VC++ .NET
    > 2003.[/color]

    Please note that hash_set *IS NOT* an STL class. It may be one in the next
    iteration of the standard, but it is currently an extension if and when it is
    implemented. Also, as an extension, the different implementations by
    different vendors are mutually incompatible, so don't expect any code you
    write using hash_set to be portable, even if both platforms have a hash_set.

    That said, VC 7.1 does provide a hash_set, but it's not in the std::
    namespace...it is, instead, in stdext::.
    [color=blue]
    > My questions are:
    > - can I produce PEs for x86, like I did with VC++ 6.0,
    > with .NET version of VC?[/color]

    Yes.
    [color=blue]
    > - will my apps run on Windows 95 (I mean 4.0.0.95 or
    > whatever version is the oldest)?[/color]

    I don't know of any specific reasons why it shouldn't, as long as you limit
    yourself to features and APIs that are supported in Win95.

    Ken


    Comment

    • Carl Daniel [VC++ MVP]

      #3
      Re: Visual C++ .NET 2003 compatibility

      Ken Alverson wrote:[color=blue]
      > "Filip Konvicka" <anonymous@disc ussions.microso ft.com> wrote in[color=green]
      >> - will my apps run on Windows 95 (I mean 4.0.0.95 or
      >> whatever version is the oldest)?[/color]
      >
      > I don't know of any specific reasons why it shouldn't, as long as you
      > limit yourself to features and APIs that are supported in Win95.[/color]

      Win95 is not an officially suported platform for VC7{.1} applications. That
      said, most applications will work on Win95 just fine, as long as you limit
      yourself to Win95 APIs.

      There are issues with MFC requiring OLEACC.DLL which may not be present on
      Win95. The workaround for this is to staically link to MFC and delay load
      oleacc.dll, assuming you don't actually need any of the functionality that
      oleacc.dll provides (which you won't if you're targeting windows 95).

      -cd


      Comment

      • David Lowndes

        #4
        Re: Visual C++ .NET 2003 compatibility

        >There are issues with MFC requiring OLEACC.DLL which may not be present on[color=blue]
        >Win95. The workaround for this is to staically link to MFC and delay load
        >oleacc.dll, assuming you don't actually need any of the functionality that
        >oleacc.dll provides (which you won't if you're targeting windows 95).[/color]

        I think the 7.1 MFC already resolved that issue by delayloading
        OLEACC.DLL itself.

        Dave
        --
        MVP VC++ FAQ: http://www.mvps.org/vcfaq

        Comment

        • Carl Daniel [VC++ MVP]

          #5
          Re: Visual C++ .NET 2003 compatibility

          David Lowndes wrote:[color=blue][color=green]
          >> There are issues with MFC requiring OLEACC.DLL which may not be
          >> present on Win95. The workaround for this is to staically link to
          >> MFC and delay load oleacc.dll, assuming you don't actually need any
          >> of the functionality that oleacc.dll provides (which you won't if
          >> you're targeting windows 95).[/color]
          >
          > I think the 7.1 MFC already resolved that issue by delayloading
          > OLEACC.DLL itself.[/color]

          I think you're right :)

          -cd


          Comment

          • Filip Konvicka

            #6
            Re: Visual C++ .NET 2003 compatibility

            Thank you all, that's exactly what I needed to know.

            To Ken: well, it fits within the STL just well and most compilers support
            it; from what I know, I don't expect significant differences.

            To Carl and David: thank you for pointing me to possible problems; I'll test
            that to be sure that it works.

            Best Regards,
            Filip

            "Carl Daniel [VC++ MVP]" <cpdaniel_remov e_this_and_nosp am@mvps.org.nos pam>
            pí¹e v diskusním pøíspìvku news:uqNXl6zkDH A.2080@TK2MSFTN GP10.phx.gbl...[color=blue]
            > David Lowndes wrote:[color=green][color=darkred]
            > >> There are issues with MFC requiring OLEACC.DLL which may not be
            > >> present on Win95. The workaround for this is to staically link to
            > >> MFC and delay load oleacc.dll, assuming you don't actually need any
            > >> of the functionality that oleacc.dll provides (which you won't if
            > >> you're targeting windows 95).[/color]
            > >
            > > I think the 7.1 MFC already resolved that issue by delayloading
            > > OLEACC.DLL itself.[/color]
            >
            > I think you're right :)
            >
            > -cd
            >
            >[/color]


            Comment

            • Ken Alverson

              #7
              Re: Visual C++ .NET 2003 compatibility

              "Filip Konvicka" <filip.konvic ka at {I.HATE.SPAM} seznam dot cz> wrote in
              message news:OebdI26kDH A.2536@tk2msftn gp13.phx.gbl...[color=blue]
              > Thank you all, that's exactly what I needed to know.
              >
              > To Ken: well, it fits within the STL just well and most compilers support
              > it; from what I know, I don't expect significant differences.[/color]

              If you're just using it with built in types for keys, you probably won't see
              significant differences. If you start trying to use custom types as keys and
              need to write hashing functions and comparators, you'll start to see major
              differences.

              I agree hash_map, in some form, is a good fit for STL. It just isn't
              standard...yet.

              Ken


              Comment

              Working...