Microsoft's C++ bigotry - Page 7


DevX Home    Today's Headlines   Articles Archive   Tip Bank   Forums   

Page 7 of 43 FirstFirst ... 5678917 ... LastLast
Results 91 to 105 of 633

Thread: Microsoft's C++ bigotry

  1. #91
    Jonathan West Guest

    Re: Microsoft's C++ bigotry


    "Jason Sobell iGadget" <iGadget_@hotmail.com> wrote in message
    news:3e267387@tnews.web.devx.com...
    > Kent, you seem to be missing the point... Only joking
    >
    > The differences between VB6 and VB.NET are enormous, and most of the [VB
    > using] verbose people in this group seem to be split into three main
    > clusters:
    > 1. Those who swear VB.NET is brilliant and deny that it is difficult to

    pick
    > up, defending its every nuance
    > 2. Those who love VB6 and swear at VB.NET, claiming it should never have
    > existed
    > 3. Those who approve VB.NET for its improvements over VB6, and accept that
    > major changes were necessary


    Either I'm misunderstanding your definition of category 2, or I think you've
    missed one.

    4. Those who approve of moving VB to the .Net platform, but swear at all the
    non-platform changes that were made to the language.

    I'm going to explain this in a bit more detail, because it's a concept a
    number of people (apparently including quite a few within Microsoft) have
    some difficulty grasping.

    There are criticisms that you can make of the platform - the size of the
    runtime, the lack of native-mode compilation and other things. These are not
    relevant to the language issue, though they often get mixed up in it. I'm
    merely mentioning them here to make it clear that for the purpose of this
    discussion, I'm accepting the platform as it is.

    One of the main ideas behind the invention and development of high-level
    languages is that the syntax of the language can be made independent of the
    platform to a great extent. A 32-bit integer is just as much a 32-bit
    integer in DOS, any flavor of Windows, a Mac or a mainframe. A high-level
    language that gives a specific name to a 32-bit integer can give the *same*
    name to that datatype on all the different platforms for which compilers are
    developed.

    The consequence of this is that if you target a language at a new platform,
    it is possible to rewrite the compiler while leaving the language syntax the
    same. The only areas where you might come a bit unstuck with that principle
    is dealing with the interfaces with the platform. But if handled correctly,
    this is more a matter of different libraries being available rather than
    having to change the core syntax itself.

    There is nothing in the change to the .Net platform that requires a change
    in the default calling convention from ByRef to ByVal, or requires that the
    argument list for subroutines be enclosed in parentheses, or requires that
    16-bit integers be called Short instead of Integer.

    These changes (which are just 3 examples out of many other similar changes
    that exist) have no effect on compatibility with the platform - they are
    trivial for a compiler to cope with. They have no effect on performance,
    either version of the source code can compile to the same object code. The
    *only* effect they have is on the ability to make use of code written in the
    earlier version of the language. That effect is negative - importing
    projects is much more difficult, and importing code fragments unmodified is
    in most cases impossible.

    The problem with the mindset as it seems to exist within Microsoft is that
    VB has only ever really existed on one platform at a time, and so the
    distinction between platform and language has become quite inappropriately
    blurred. The fact that 32-bit is the "most performant" integer datatype on
    32-bit Windows platforms doesn't mean that it has to be called Integer. The
    fact that the platform codes a Boolean True as 0x0001 rather than as 0xFFFF
    doesn't mean than CInt(True) had also to take the value 1 instead of -1. You
    can decouple the syntax as presented to the programmer from what is going on
    under the hood. In Microsoft's own interest (i.e. getting lots of VB
    developers to port their applications to .Net) this decoupling ought to have
    been done, but wasn't.


    --
    Regards
    Jonathan West


  2. #92
    David Rothgery Guest

    Re: Microsoft's C++ bigotry


    "Jonathan West" <jwest@mvps.org> wrote in message
    news:3e2c5970$1@tnews.web.devx.com...
    > There is nothing in the change to the .Net platform that requires a change
    > in the default calling convention from ByRef to ByVal, or requires that

    the
    > argument list for subroutines be enclosed in parentheses, or requires that
    > 16-bit integers be called Short instead of Integer.


    Yeah, but the problem was that defaults to ByRef and 16-bit Integers were
    stupid (the change to 32-bit Integers really should have been made in VB4;
    the ByRef default never should have happened), and the thousand and one
    variations on how to call a method were insane.

    --
    Dave Rothgery
    drothgery@alum.wpi.edu



  3. #93
    Mike Mitchell Guest

    Re: Microsoft's C++ bigotry

    On Sun, 19 Jan 2003 23:01:55 -0000, "Jonathan West" <jwest@mvps.org>
    wrote:

    >They have got themselves stuck in a nasty hole. They no doubt want and need
    >to get the VB6 codebase on board, and yet now they have released VB.NET,
    >there is a growing codebase of VB.NET code out there as well. They can
    >hardly alienate the early adopters that they have recruited so far, so
    >VB.NET with its existing syntax is hardly likely to die in the short term.


    Depends on how many actual, genuine (as opposed to dabblers) VB.Net
    development "shops" there are out there, actively engaged in selling
    VB.Net apps or providing apps in-house for corporate IT/MIS. What
    would you rather have: 3 million VB programmers happily coding with a
    "real" VB7, or just a few thousand doggedly, against all odds, trying
    to sell the 20 MB runtime to their prospects?

    >Therefore, the only solution available to Microsoft is to bring out a
    >VBClassic.Net based on the .Net runtime, with a syntax that is as consistent
    >with VB6 as can reasonably be achieved, and with extensions that give full
    >access to the framework.


    Nah! I'm quite sure that would just be seen as yet another fudge. "as
    can reasonably be achieved" doesn't sound too promising to my mind.
    And there's still the .Net stuff underlying it all, something I don't
    believe the vast majority of corporates are really willing to buy into
    (yet) (because they can't see why they should). (Because they couldn't
    care a bean about web services.) (Because not even Microsoft can help
    'em out here.)

    No, the only thing they could do is keep VB.Net, naturally (I would
    never wish it to disappear, given that there are a number of
    programmers who *have* adopted it), but re-introduce a new classic
    version, based on VB6, 100% compatible with VB6, but maybe with a
    better menu designer and the bugs fixed. Everything else in VB6++
    would work absolutely fine already, so there's no need to bust a gut
    trying to "improve" it. And then, over the course of, say, the next
    five years, corporates, software houses, and individuals could
    <i>gradually</i> make the shift. The main problem with .Net/VB.Net has
    been the forced pace at which such a major paradigm change was thrust
    upon us.

    MM

  4. #94
    Mike Mitchell Guest

    Re: Microsoft's C++ bigotry

    On Mon, 20 Jan 2003 12:02:24 -0600, Dan Barclay <Dan@MVPs.org> wrote:

    >Do you think, having looked at the path MS provided for transition,
    >that these people really want a move that will (possibly/likely) set
    >them up for this situation again?


    And this is where Microsoft is really between a rock and a hard place.
    The less VB.Net is able to capture the enthusiasm of a considerable,
    significant band of brothers, happily coding away, the more likely it
    is that Microsoft will order the grim reaper to pay another visit.
    Well, the potential is there that even Microsoft would eventually have
    to throw in the towel, otherwise they'd have a lame dog product
    limping around after them into the future wherever they go. But in
    order for the above not to happen, they need to generate that
    enthusiasm! In the millions of programmers who have just had the rug
    pulled out from under. Not exactly the most trusting bunch of would-be
    converts, I'd guess.

    I still say that the entire .Net thing is not needed. Whatever $ signs
    Bill might have on his eyeballs when he dreams at night with respect
    to web services and pay-by-the-minute, I just can't see how web
    services are taking off to any significant extent and, moreover, I
    reckon there are thousands of corporates for which web services will
    forever remain on someone's esoteric wish-list, i.e. could be nice,
    but we just don't need it.

    MM

  5. #95
    Jonathan West Guest

    Re: Microsoft's C++ bigotry


    "David Rothgery" <drothgery@alum.wpi.edu> wrote in message
    news:3e2c607a$1@tnews.web.devx.com...
    >
    > "Jonathan West" <jwest@mvps.org> wrote in message
    > news:3e2c5970$1@tnews.web.devx.com...
    > > There is nothing in the change to the .Net platform that requires a

    change
    > > in the default calling convention from ByRef to ByVal, or requires that

    > the
    > > argument list for subroutines be enclosed in parentheses, or requires

    that
    > > 16-bit integers be called Short instead of Integer.

    >
    > Yeah, but the problem was that defaults to ByRef and 16-bit Integers were
    > stupid (the change to 32-bit Integers really should have been made in VB4;
    > the ByRef default never should have happened), and the thousand and one
    > variations on how to call a method were insane.


    You've missed the point in exactly the way I described. The datatypes
    already existed with well-defined names. A change in platform does *not*
    require a change in language syntax or name of datatypes in a high-level
    language. If using 32-bit integers as counters is faster than using 16-bit,
    then you tell people about it in the docs. You don't break code by changing
    datatypes. If people don't read the docs, then you shrug it off with the
    thought that a counter or two is usually a trivial performance issue when
    you are shovelling large numbers of objects around.

    As for ByRef, it was a sensible enough choice at the time, when thinking in
    terms of Subs being an extension of the old Gosub statement. As the state of
    the art advanced, offering ByVal as an alternative was a sensible
    *extension* of the syntax that didn't change the behavior of existing code.
    If you are designing a whole new language (like C#) then you make those
    decisions according to the current state of thinking, and hope that they
    wear well. If you are wanting to extend a language that already exists, then
    if you want to keep existing developers (and their existing applications) on
    board, you really have to live by those earlier decisions, whether in
    hindsight you think they were good or not.

    With VB.NET, Microsoft never quite did work out whether they were extending
    an existing language or creating a new language. It has suffered from this
    lack of clarity in its design philosophy.

    --
    Regards
    Jonathan West


  6. #96
    Mike Mitchell Guest

    Re: Microsoft's C++ bigotry

    On 20 Jan 2003 03:50:15 -0800, "T. Hoskins" <thoskins@nospam.com>
    wrote:

    >As for the "ease of migration" question you posed, "has anyone in this newsgroup
    >communicated with any of the developers working on VB.NET on this issue?".
    >Whatever happened to the Microsoft employee (he had a name that started with
    >the letter Y?) who use to visit this newsgroup? Didn't he write in a post
    >that he was going to be checking in on this newsgroup from time to time?


    The Y fellah (doubtless a perfectly nice chap!) was asked by Phil (I
    believe, but shoot me if I'm wrong about that) to use an
    "announcements" ng instead of this one. Also, I couldn't stand much
    more of the self-aggrandizement for MS, so I might have complained
    just a tad...

    Agree with most of your other points! What some commentators often
    overlook is that Microsoft showed themselves willing and able to
    continually improve VB from version 1 through version 6. Thus, they
    would have been equally capable of continuing this gestation through
    several more versions, always providing as near as dammit 100%
    compatibility. Any of the shortcomings in VB6 could have been improved
    upon, maybe after genuine consultation with representative users (and
    not an incestuous in-house crowd with a particular aim in mind). Just
    look how much improved VB6 was over VB1!

    MM

  7. #97
    Mike Mitchell Guest

    Re: Microsoft's C++ bigotry

    On Mon, 20 Jan 2003 12:47:09 -0000, "Jonathan West" <jwest@mvps.org>
    wrote:

    >He's moved to another job within Microsoft. However, there are others at
    >Microsoft who are still lurking.


    Hello, there!

    MM

  8. #98
    Jonathan West Guest

    Re: Microsoft's C++ bigotry


    "Mike Mitchell" <kylix_is@yahoo.co.uk> wrote in message
    news:cimo2vodp32mpk2vre1osnith5r6gfsu5o@4ax.com...
    > On Sun, 19 Jan 2003 23:01:55 -0000, "Jonathan West" <jwest@mvps.org>
    > wrote:
    >
    > re-introduce a new classic
    > version, based on VB6, 100% compatible with VB6, but maybe with a
    > better menu designer and the bugs fixed. Everything else in VB6++
    > would work absolutely fine already, so there's no need to bust a gut
    > trying to "improve" it. And then, over the course of, say, the next
    > five years, corporates, software houses, and individuals could
    > <i>gradually</i> make the shift. The main problem with .Net/VB.Net has
    > been the forced pace at which such a major paradigm change was thrust
    > upon us.


    The problem with this is that the Ruby platform really is worn out as a base
    for further progress. I'm not surprised at the fact, it's a well-known
    phenomenon of software engineering. The principles of software entropy were
    described 30 years ago by Fred Brooks in The Mythical Man-Month and haven't
    changed much since. Therefore, the platform had to be replaced, and as it
    was going to be replaced, the replacement might as well be as capable and
    robust as modern software design could make it. You may argue as to whether
    Microsoft succeeded in that aim with the .Net platform, but I have no doubt
    they were right to try. Simply rewriting Ruby from scratch would not have
    been a sensible move.

    Having a VBClassic.Net that is syntax-compatible, and whose behavior changes
    wrt VB6 only to the extent absolutely necessary because of the nature of the
    platform would also allow people to choose when and how to make the
    transition, especially as the transition would be much simpler had the
    language been as I described. This is what Microsoft started out doing (at
    least that's what the beta documentation indicated), but they got
    sidetracked because of the fun involved in making "improvements" to the
    language.

    Notice that I'm making a distinction between the language syntax (which may
    be a bit old-fashioned, but has a coherent and implemental set of rules) and
    the underlying platform, which is now heavily patched and not usable as a
    base for further major development.

    --
    Regards
    Jonathan West


  9. #99
    Dan Barclay Guest

    Re: Microsoft's C++ bigotry

    On Mon, 20 Jan 2003 21:29:18 +0000, Mike Mitchell
    <kylix_is@yahoo.co.uk> wrote:

    > But in
    >order for the above not to happen, they need to generate that
    >enthusiasm! In the millions of programmers who have just had the rug
    >pulled out from under. Not exactly the most trusting bunch of would-be
    >converts, I'd guess.


    To win over the (they think few) developers with the billions of lines
    of code on the table, they only have to provide a clear and trusty
    path forward. That is easily said, but it will take some doing on
    their part to make it so. That "trusty" part is the kicker now.

    >I still say that the entire .Net thing is not needed. Whatever $ signs
    >Bill might have on his eyeballs when he dreams at night with respect
    >to web services and pay-by-the-minute, I just can't see how web
    >services are taking off to any significant extent and, moreover, I
    >reckon there are thousands of corporates for which web services will
    >forever remain on someone's esoteric wish-list, i.e. could be nice,
    >but we just don't need it.


    I can see some uses for the technology, even in rich-client
    applications. We'll be doing more distributed apps in the future,
    whether with .net or on our own.

    The key issue here is whether they plan to invite us along on their
    new "Strategic Direction". The question of the soundness of this
    direction is a different (perhaps larger, but different) question. If
    I could trust them to cover the platforms I needed, protect my code
    from intrusion, and to continue to support rich clients with the
    technology, it would have a lot going for it in spite of its size and
    complexity.

    Regardless of all that, VB is the tool MS has to move a very large
    contingent of their developers toward their next strategy, and the
    next, and the next (whatever it may be). They're throwing that
    leverage away.

    Dan
    Language Stability is a *feature* I wish VB had!
    (#6)
    Error 51
    Error 3
    Error 9
    ....

  10. #100
    David Rothgery Guest

    Re: Microsoft's C++ bigotry


    "Jonathan West" <jwest@mvps.org> wrote in message
    news:3e2c6861@tnews.web.devx.com...
    > With VB.NET, Microsoft never quite did work out whether they were

    extending
    > an existing language or creating a new language. It has suffered from this
    > lack of clarity in its design philosophy.


    I guess I was always in the camp that wanted a Basic.NET -- something with a
    VB-like syntax, but acted like a modern OO language (and therefore wouldn't
    retain bad VB Classic design decisions). VB.NET mostly accomplishes this
    (enough that I'd much rather deal with a few minor annoyances than the more
    significant ways C# annoys me). So the Beta 2 rollbacks were a bad thing
    IMO, and I think Option Strict On should be the default.

    --
    Dave Rothgery
    drothgery@alum.wpi.edu



  11. #101
    Jonathan West Guest

    Re: Microsoft's C++ bigotry


    "David Rothgery" <drothgery@alum.wpi.edu> wrote in message
    news:3e2c7585$1@tnews.web.devx.com...
    >
    > "Jonathan West" <jwest@mvps.org> wrote in message
    > news:3e2c6861@tnews.web.devx.com...
    > > With VB.NET, Microsoft never quite did work out whether they were

    > extending
    > > an existing language or creating a new language. It has suffered from

    this
    > > lack of clarity in its design philosophy.

    >
    > I guess I was always in the camp that wanted a Basic.NET -- something with

    a
    > VB-like syntax, but acted like a modern OO language (and therefore

    wouldn't
    > retain bad VB Classic design decisions). VB.NET mostly accomplishes this
    > (enough that I'd much rather deal with a few minor annoyances than the

    more
    > significant ways C# annoys me). So the Beta 2 rollbacks were a bad thing
    > IMO, and I think Option Strict On should be the default.


    Well, that's fine. I don't grudge you your new language. I suspect that if
    VB.NET had been as I described it, you would have turned Option Strict on by
    default and coded away as happily as you are doing now, if not in VB, then
    in C#.

    I just wish the new languages could have been provided without depriving
    existing application owners of a useful VB6 upgrade path. What has happened
    is that Microsoft has for all practical purposes written two new Java-killer
    languages and neglected their own customer base in the process. You could
    say that they have shot themselves in the foot by so doing, but I think it
    might be a little higher up, and more towards the centreline...

    --
    Regards
    Jonathan West


  12. #102
    Kent Guest

    Re: Microsoft's C++ bigotry


    David,

    I think a lot of us feel the same way you do. If presented as a new product
    VB.Net MAY have had a better chance in the market. Who can say for sure
    though?

    The beauty of VB over C in the early windows days was that VB made a very
    difficult task (creating windows applications using the modern methods of
    the time) very simple.

    VB.Net does make it significantly easier to create .Net apps. It's just
    about as easy to use C#. Really they should not have made a VB at all and
    just ran with C# and left well enough alone. But then again, would another
    C like language running over a VM really have been enough to dethrone Java?
    probably not.

    Kent

    >I guess I was always in the camp that wanted a Basic.NET -- something with

    a
    >VB-like syntax, but acted like a modern OO language (and therefore wouldn't
    >retain bad VB Classic design decisions). VB.NET mostly accomplishes this
    >(enough that I'd much rather deal with a few minor annoyances than the more
    >significant ways C# annoys me). So the Beta 2 rollbacks were a bad thing
    >IMO, and I think Option Strict On should be the default.



  13. #103
    John Butler Guest

    Re: Microsoft's C++ bigotry


    "Mike Mitchell" <kylix_is@yahoo.co.uk> wrote in message
    news:bvpo2vg1lc5b28qlhlmnhasfqb7t6tm8s6@4ax.com...
    > I still say that the entire .Net thing is not needed.


    You see Mike...that is the kind of thing only someone who hasn't used .net
    could say. You're still all hung up on the web services marketing hype. I've
    used .Net for a while, have even built a few web services which are deployed
    as part of a commercial app....but the Framework itself is what is so great.

    I could list some of the things that are good...but I know I'm wasting my
    time with you...so I won't bother. You don't really hate .NET...you just
    hate Microsoft, particularly it's marketing department...that is what it
    boils down to in the end. Even Microsoft has distilled it's whole web
    services hype by now...it's dropping of the whole .NET moniker on everything
    is clear proof that sanity is prevailing...and yet you still ramble on about
    web services....How many "Netizens" on this forum have rambled on about Web
    services?? Not many.....lots of people have been positive about the
    platform...but that is because they've used it and experienced some of the
    benefits that a well thought out class library can bring about.

    The .NET platform is a great thing. Think what you like about VB.NET...but
    criticising the literally thousands of improvements and enhancements offered
    by the platform...is just ignorant.

    rgds
    John Butler



  14. #104
    John Butler Guest

    Re: Microsoft's C++ bigotry


    "Kent" <kp@kp.org> wrote in message news:3e2b34e3$1@tnews.web.devx.com...
    >
    > "John Butler" <nospamjrbutler@btinternet.com> wrote:


    > With Microsoft at the wheel you can almost bet on it. Someone else with
    > comeup with something better for them to steal and there will be no reason
    > for them to continue on with poor old .Net or whatever the M$ marketing

    people
    > are calling it then.


    Are you trying to out-hate Microsoft against Mike? I don't have a problem
    with criticisms agains Microsoft...they certainly leave themselves open to
    it sometimes, especially their marketing dept....but when you are
    consistenly anti...it taints one's ability to analyse your posts in a
    neutral way. It's obvious you have issues with MS...so why bother even
    arguing with you..since you're just going to rant.

    No offence intended...just trying to point out that if you take up an
    extreme stance...it makes it harder to really discuss much with you. I've
    always tried to take a middle road approach..yet I'm consistenly branded a
    "Netizen" because I refuse to endulge in the now fashionable Microsoft
    bashing and *like* using .NET. Two years ago, I started out arguing with
    Zane and others against .NET. They kept telling me to try it...which I did.
    I'm glad I followed their advice...because they were right...

    rgds
    John Butler



  15. #105
    John Butler Guest

    Re: Microsoft's C++ bigotry


    "Kent" <kp@kp.org> wrote in message news:3e2b2eb8$1@tnews.web.devx.com...
    >
    > I know what you're saying, but I cannot see how 90% can be achieved when

    there
    > is so much reliance on libraries in .Net.


    What is bad about a reliance on libraries? At least there is only one
    framework to deal with. When last did you look at Java? You didn't use COM
    libraries in VB6? Our projects were riddled with dozens of COM add-ins, to
    do all the things VB6 didn't do. Unless you were building some mickey mouse
    app that didn't do much at all, you would have had 5 or more references to
    third party controls and libraries in any given VB6 commercial app....and
    the corresponding DLL **** and huge installation projects
    (Wise/Installshield) to try and manage it all on 95/98/NT/2000 platforms.

    rgds
    John Butler





Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
HTML5 Development Center
 
 
FAQ
Latest Articles
Java
.NET
XML
Database
Enterprise
Questions? Contact us.
C++
Web Development
Wireless
Latest Tips
Open Source


   Development Centers

   -- Android Development Center
   -- Cloud Development Project Center
   -- HTML5 Development Center
   -- Windows Mobile Development Center