Menu

data types - bitching

Soren Bro
2019-01-12
2019-01-12
  • Soren Bro

    Soren Bro - 2019-01-12

    This is madness.

    There are seemingly 3-4 variables I can choose from that can represent all
    basic types:

    _varian_t
    VARIANT
    COleVariant
    CComVariant

    I've been told (on MSDN Forums no less) that at least CComVariant is
    compatible with MFC but, it has the same problem with booleans as the
    others so in which way it's "compatible" I don't know.

    Each one seems to encapsulate the one above it. To some degree at least. As
    I said, for historical respons, you have to use VARIANT_TRUE and
    VARIANT_FALSE to represent the built-in bool values true and false.
    _varian_t sees -1 as false and 0 as true (or the other way around. I forget
    right now. I'm not home.)

    The madness lies in the fact that I have to test _variant_t before sending
    it to CMFCPropertyGridProperty (which expects _variant_t) AND THEN CAST IT
    BACK TO _variant_t like so:

    [...]
    _varian_t v (VARIANT_TRUE)

    SomeFunc ( (_variant_t)(t.boolVal == VARIANT_TRUE ? true : false) ):
    [...]

    This works (??!!) and it makes absolutely no sense at all.

    Mind you, I'm not asking anything. I'm Just bitching.

    Regards,
    Søren

    --
    Søren Bro Thygesen

     
    • Soren Bro

      Soren Bro - 2019-01-12

      Of course, and needless to say, this took me no end of grief to work out. :(

      R

      On Saturday, January 12, 2019, sbrothy@gmail.com wrote:

      This is madness.

      There are seemingly 3-4 variables I can choose from that can represent all
      basic types:

      _varian_t
      VARIANT
      COleVariant
      CComVariant

      I've been told (on MSDN Forums no less) that at least CComVariant is
      compatible with MFC but, it has the same problem with booleans as the
      others so in which way it's "compatible" I don't know.

      Each one seems to encapsulate the one above it. To some degree at least.
      As I said, for historical respons, you have to use VARIANT_TRUE and
      VARIANT_FALSE to represent the built-in bool values true and false.
      _varian_t sees -1 as false and 0 as true (or the other way around. I forget
      right now. I'm not home.)

      The madness lies in the fact that I have to test _variant_t before sending
      it to CMFCPropertyGridProperty (which expects _variant_t) AND THEN CAST IT
      BACK TO _variant_t like so:

      [...]
      _varian_t v (VARIANT_TRUE)

      SomeFunc ( (_variant_t)(t.boolVal == VARIANT_TRUE ? true : false) ):
      [...]

      This works (??!!) and it makes absolutely no sense at all.

      Mind you, I'm not asking anything. I'm Just bitching.

      Regards,
      Søren

      --
      Søren Bro Thygesen

      --
      Søren Bro Thygesen

       
    • Ted Matavka

      Ted Matavka - 2019-01-12
      <meta http-equiv="Content-Security-Policy" content="script-src 'self'; img-src * cid: data:;"><style id="user-content-outgoing-font-settings">#response_container_BBPPID{font-family: initial; font-size:initial; color: initial;}</style>
      Please make sure to document all of the below in a format that will ve useful to newbies coming to work on Hermes Mail.  I predict a surge in interest as soon as we release a binary... and then a gradual tapering off.  So let's make use of that surge while we can.

      Sent from my BlackBerry — the most secure mobile device
      From: sbrothy@users.sourceforge.net
      Sent: 12 January 2019 03:05
      To: general@discussion.hermesmail.p.re.sourceforge.net
      Reply to: general@discussion.hermesmail.p.re.sourceforge.net
      Subject: [hermesmail:discussion] data types - bitching

      This is madness.

      There are seemingly 3-4 variables I can choose from that can represent all
      basic types:

      _varian_t
      VARIANT
      COleVariant
      CComVariant

      I've been told (on MSDN Forums no less) that at least CComVariant is
      compatible with MFC but, it has the same problem with booleans as the
      others so in which way it's "compatible" I don't know.

      Each one seems to encapsulate the one above it. To some degree at least. As
      I said, for historical respons, you have to use VARIANT_TRUE and
      VARIANT_FALSE to represent the built-in bool values true and false.
      _varian_t sees -1 as false and 0 as true (or the other way around. I forget
      right now. I'm not home.)

      The madness lies in the fact that I have to test _variant_t before sending
      it to CMFCPropertyGridProperty (which expects _variant_t) AND THEN CAST IT
      BACK TO _variant_t like so:

      [...]
      _varian_t v (VARIANT_TRUE)

      SomeFunc ( (_variant_t)(<wbr></wbr>t.boolVal<wbr></wbr> == VARIANT_TRUE ? true : false) ):
      [...]

      This works (??!!) and it makes absolutely no sense at all.

      Mind you, I'm not asking anything. I'm Just bitching.

      Regards,
      Søren

      --
      Søren Bro Thygesen


      data types - bitching


      Sent from <wbr></wbr>sourceforge.net<wbr></wbr> because you indicated interest inhttps://sourceforge.net/p/hermesmail/discussion/general/

      To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/

       
      • Soren Bro

        Soren Bro - 2019-01-12

        Yeah. That's one of the reasons I dump so much crud on the list. Because to
        be honest I have no time (what with all the crap I've run into. Én passant:
        Also, I happen to know that Mr. MacLean is not a big fan of COM.) to write
        proper documentation.

        Much will be found in the code (including a ton of "//TODO: "s) and a lot
        here.

        Regards.

        On Saturday, January 12, 2019, Ted Matavka nmatavka@users.sourceforge.net
        wrote:

        <meta content="script-src 'self'; img-src * cid: data:;" http-equiv="Content-Security-Policy"><style id="user-content-outgoing-font-settings">#r<br> esponse_container_BBPPID{font-family: initial; font-size:initial; color:<br> initial;}</style>
        Please make sure to document all of the below in a format that will ve
        useful to newbies coming to work on Hermes Mail. I predict a surge in
        interest as soon as we release a binary... and then a gradual tapering
        off. So let's make use of that surge while we can.

        Sent from my BlackBerry — the most secure mobile device
        From: sbrothy@users.sourceforge.net
        Sent: 12 January 2019 03:05
        To: general@discussion.hermesmail.p.re.sourceforge.net
        Reply to: general@discussion.hermesmail.p.re.sourceforge.net
        Subject: [hermesmail:discussion] data types - bitching

        This is madness.

        There are seemingly 3-4 variables I can choose from that can represent all

        basic types:

        _varian_t

        VARIANT

        COleVariant

        CComVariant

        I've been told (on MSDN Forums no less) that at least CComVariant is

        compatible with MFC but, it has the same problem with booleans as the

        others so in which way it's "compatible" I don't know.

        Each one seems to encapsulate the one above it. To some degree at least. As

        I said, for historical respons, you have to use VARIANT_TRUE and

        VARIANT_FALSE to represent the built-in bool values true and false.

        _varian_t sees -1 as false and 0 as true (or the other way around. I forget

        right now. I'm not home.)

        The madness lies in the fact that I have to test _variant_t before sending

        it to CMFCPropertyGridProperty (which expects _variant_t) AND THEN CAST IT

        BACK TO _variant_t like so:

        [...]

        _varian_t v (VARIANT_TRUE)

        SomeFunc ( (_variant_t)(<wbr></wbr>t.boolVal<wbr></wbr> == VARIANT_TRUE ?
        true : false) ):

        [...]

        This works (??!!) and it makes absolutely no sense at all.

        Mind you, I'm not asking anything. I'm Just bitching.

        Regards,

        Søren

        --

        Søren Bro Thygesen


        data types - bitching
        https://sourceforge.net/p/hermesmail/discussion/general/thread/be13465ec7/?limit=25#03b0


        Sent from <wbr></wbr>sourceforge.net<wbr></wbr> because you indicated
        interest inhttps://sourceforge.net/p/hermesmail/discussion/general/

        To unsubscribe from further messages, please visit
        https://sourceforge.net/auth/subscriptions/



        data types - bitching
        https://sourceforge.net/p/hermesmail/discussion/general/thread/be13465ec7/?limit=25#03b0/f35a


        Sent from sourceforge.net because you indicated interest in
        https://sourceforge.net/p/hermesmail/discussion/general/

        To unsubscribe from further messages, please visit
        https://sourceforge.net/auth/subscriptions/

        --
        Søren Bro Thygesen

         
        • Soren Bro

          Soren Bro - 2019-01-12

          Putting it all together it should be possible ro extrapolate proper docs.

          R

          On Saturday, January 12, 2019, sbrothy@gmail.com wrote:

          Yeah. That's one of the reasons I dump so much crud on the list. Because
          to be honest I have no time (what with all the crap I've run into. Én
          passant: Also, I happen to know that Mr. MacLean is not a big fan of COM.)
          to write proper documentation.

          Much will be found in the code (including a ton of "//TODO: "s) and a lot
          here.

          Regards.

          On Saturday, January 12, 2019, Ted Matavka nmatavka@users.sourceforge.net
          wrote:

          <meta content="script-src 'self'; img-src * cid: data:;" http-equiv="Content-Security-Policy"><style id="user-content-outgoing-font-settings">#response_container_BBPPID{font-family:<br> initial; font-size:initial; color: initial;}</style>
          Please make sure to document all of the below in a format that will ve
          useful to newbies coming to work on Hermes Mail. I predict a surge in
          interest as soon as we release a binary... and then a gradual tapering
          off. So let's make use of that surge while we can.

          Sent from my BlackBerry — the most secure mobile device
          From: sbrothy@users.sourceforge.net
          Sent: 12 January 2019 03:05
          To: general@discussion.hermesmail.p.re.sourceforge.net
          Reply to: general@discussion.hermesmail.p.re.sourceforge.net
          Subject: [hermesmail:discussion] data types - bitching

          This is madness.

          There are seemingly 3-4 variables I can choose from that can represent all

          basic types:

          _varian_t

          VARIANT

          COleVariant

          CComVariant

          I've been told (on MSDN Forums no less) that at least CComVariant is

          compatible with MFC but, it has the same problem with booleans as the

          others so in which way it's "compatible" I don't know.

          Each one seems to encapsulate the one above it. To some degree at least.
          As

          I said, for historical respons, you have to use VARIANT_TRUE and

          VARIANT_FALSE to represent the built-in bool values true and false.

          _varian_t sees -1 as false and 0 as true (or the other way around. I
          forget

          right now. I'm not home.)

          The madness lies in the fact that I have to test _variant_t before sending

          it to CMFCPropertyGridProperty (which expects _variant_t) AND THEN CAST IT

          BACK TO _variant_t like so:

          [...]

          _varian_t v (VARIANT_TRUE)

          SomeFunc ( (_variant_t)(<wbr></wbr>t.boolVal<wbr></wbr> == VARIANT_TRUE
          ? true : false) ):

          [...]

          This works (??!!) and it makes absolutely no sense at all.

          Mind you, I'm not asking anything. I'm Just bitching.

          Regards,

          Søren

          --

          Søren Bro Thygesen


          data types - bitching
          https://sourceforge.net/p/hermesmail/discussion/general/thread/be13465ec7/?limit=25#03b0


          Sent from <wbr></wbr>sourceforge.net<wbr></wbr> because you indicated
          interest inhttps://sourceforge.net/p/hermesmail/discussion/general/

          To unsubscribe from further messages, please visit
          https://sourceforge.net/auth/subscriptions/



          data types - bitching
          https://sourceforge.net/p/hermesmail/discussion/general/thread/be13465ec7/?limit=25#03b0/f35a


          Sent from sourceforge.net because you indicated interest in
          https://sourceforge.net/p/hermesmail/discussion/general/

          To unsubscribe from further messages, please visit
          https://sourceforge.net/auth/subscriptions/

          --
          Søren Bro Thygesen

          --
          Søren Bro Thygesen

           
          • Ted Matavka

            Ted Matavka - 2019-01-12
            <meta http-equiv="Content-Security-Policy" content="script-src 'self'; img-src * cid: data:;"><style id="user-content-outgoing-font-settings">#response_container_BBPPID{font-family: initial; font-size:initial; color: initial;}</style>
            Perhaps now would be a bad time to mention that I plan a migration to BitBucket in the not so near future.  Code will migrate easily but the discussions will be a pain.

            Not to worry; this is in the far off but only just foreseeable future.  Only because I have been receiving unimaginable amounts of flak on Reddit from people who really should know better.

            Sent from my BlackBerry — the most secure mobile device
            From: sbrothy@users.sourceforge.net
            Sent: 12 January 2019 05:37
            To: general@discussion.hermesmail.p.re.sourceforge.net
            Reply to: general@discussion.hermesmail.p.re.sourceforge.net
            Subject: [hermesmail:discussion] Re: data types - bitching

            Putting it all together it should be possible ro extrapolate proper docs.

            R

            On Saturday, January 12, 2019, sbrothy@gmail.com wrote:

            Yeah. That's one of the reasons I dump so much crud on the list. Because
            to be honest I have no time (what with all the crap I've run into. Én
            passant: Also, I happen to know that Mr. MacLean is not a big fan of COM.)
            to write proper documentation.

            Much will be found in the code (including a ton of "//TODO: "s) and a lot
            here.

            Regards.

            On Saturday, January 12, 2019, Ted Matavka nmatavka@users.sourceforge.net
            wrote:

            <meta content="script-src 'self'; img-src * cid: data:;" http-equiv="Content-Security-Policy"><style id="outgoing-font-settings">#response_container_BBPPID{font-family:<br> initial; font-size:initial; color: initial;}</style>
            Please make sure to document all of the below in a format that will ve
            useful to newbies coming to work on Hermes Mail. I predict a surge in
            interest as soon as we release a binary... and then a gradual tapering
            off. So let's make use of that surge while we can.

            Sent from my BlackBerry — the most secure mobile device
            From: sbrothy@users.sourceforge.net
            Sent: 12 January 2019 03:05
            To: general@discussion.hermesmail.p.re.sourceforge.net
            Reply to: general@discussion.hermesmail.p.re.sourceforge.net
            Subject: [hermesmail:discussion] data types - bitching

            This is madness.

            There are seemingly 3-4 variables I can choose from that can represent all

            basic types:

            _varian_t

            VARIANT

            COleVariant

            CComVariant

            I've been told (on MSDN Forums no less) that at least CComVariant is

            compatible with MFC but, it has the same problem with booleans as the

            others so in which way it's "compatible" I don't know.

            Each one seems to encapsulate the one above it. To some degree at least.
            As

            I said, for historical respons, you have to use VARIANT_TRUE and

            VARIANT_FALSE to represent the built-in bool values true and false.

            _varian_t sees -1 as false and 0 as true (or the other way around. I
            forget

            right now. I'm not home.)

            The madness lies in the fact that I have to test _variant_t before sending

            it to CMFCPropertyGridProperty (which expects _variant_t) AND THEN CAST IT

            BACK TO _variant_t like so:

            [...]

            _varian_t v (VARIANT_TRUE)

            SomeFunc ( (_variant_t)(<wbr></wbr><wbr></wbr>t.boolVal<wbr></wbr><wbr></wbr> == VARIANT_TRUE
            ? true : false) ):

            [...]

            This works (??!!) and it makes absolutely no sense at all.

            Mind you, I'm not asking anything. I'm Just bitching.

            Regards,

            Søren

            --

            Søren Bro Thygesen


            data types - bitching
            https://sourceforge.net/p/hermesmail/discussion/general/thread/be13465ec7/?limit=25#03b0


            Sent from <wbr></wbr><wbr></wbr>sourceforge.net<wbr></wbr><wbr></wbr> because you indicated
            interest in<wbr></wbr>https://sourceforge.net/p/hermesmail/discussion/general/<wbr></wbr>

            To unsubscribe from further messages, please visit
            https://sourceforge.net/auth/subscriptions/



            data types - bitching
            https://sourceforge.net/p/hermesmail/discussion/general/thread/be13465ec7/?limit=25#03b0/f35a


            Sent from <wbr></wbr>sourceforge.net<wbr></wbr> because you indicated interest in
            https://sourceforge.net/p/hermesmail/discussion/general/

            To unsubscribe from further messages, please visit
            https://sourceforge.net/auth/subscriptions/

            --
            Søren Bro Thygesen

            --
            Søren Bro Thygesen


            data types - bitching


            Sent from <wbr></wbr>sourceforge.net<wbr></wbr> because you indicated interest inhttps://sourceforge.net/p/hermesmail/discussion/general/

            To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/

             

Log in to post a comment.