Re: Do consumers even want web services? - Page 3


DevX Home    Today's Headlines   Articles Archive   Tip Bank   Forums   

Page 3 of 19 FirstFirst 1234513 ... LastLast
Results 31 to 45 of 279

Thread: Re: Do consumers even want web services?

  1. #31
    Mark Jerde Guest

    Re: Do consumers even want web services?

    Dave,

    > > when Bill Gates saw Alan Cooper's product he
    > > said, "This advances the state of the art."

    <snip>
    > from The Inmates Are Running The Asylum (page 57)
    > Bill said: "How did you do that?"


    Perhaps the conversation included both sentences. ;-)
    I checked the index in "About Face" but Ruby isn't listed.

    Mark Jerde
    Biometrics - www.idtechpartners.com



  2. #32
    Jonathan Allen Guest

    Re: Do consumers even want web services?

    > Do you realize that "an order of magnitude" generally means a factor of
    > 10? So "several" is a factor of over 100? And this means that you're
    > claiming that you developed something in 1/100th the time it would have
    > been possible by other methods. Such claims are sheer hyperbole.


    Those numbers are reasonable. Adding a Web Service interface and the
    matching client code takes a couple minutes. Doing the same using a TCP port
    could take me a couple hours. So for me, it literally is two orders of
    magnitude faster.

    --
    Jonathan Allen


    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote in message
    news:3C8AA58E.EFEE1C0@hal-pc.org...
    > Ed Courtenay wrote:
    > > Yes, it definately helps to solve a business need. A recent project
    > > undertaken by my company involved communicating with several disparate
    > > systems, none of which were under our direct control.

    >
    > > Web services provided
    > > a quick and easy way of exposing methods on these systems

    >
    > If the methods were "exposed" by using Web Services, then they must have
    > previously existed (in a non-Web Services form). So didn't they already
    > have checks to their input to ensure that it's valid? If not, then you
    > just had crappy code that didn't check it's inputs.
    >
    > You undoubtedly could have made these same methods available by any
    > number of different means; Web Services is just the one you used.
    >
    > > and reduced the overall development time by several orders of
    > > magnitude.

    >
    > Do you realize that "an order of magnitude" generally means a factor of
    > 10? So "several" is a factor of over 100? And this means that you're
    > claiming that you developed something in 1/100th the time it would have
    > been possible by other methods. Such claims are sheer hyperbole. And
    > further, how did you compute this reduction in effort? Did you try to do
    > the development several ways and then compare costs?
    >
    > Your words *do* support the contention that you are a numerical
    > illiterate, however.




  3. #33
    Dave Keighan Guest

    Re: Do consumers even want web services?

    Hi Mark

    > Perhaps the conversation included both sentences. ;-)

    Alan: "Hey Bill look at this."
    Bill: "How'd you do that?"
    Yea, there was probably a lot more to it ... probably a conversation more
    suitable for the OR that it was for publication by Alan.
    Perhaps;
    Bill: "This advances the state of the art ... can I have it?"
    Alan: "FOAD Bill, make your own ... lazy bugger."
    <grin>

    OK, enough, sorry.
    --
    Dave Keighan

  4. #34
    Jay Glynn Guest

    Re: Do consumers even want web services?

    Since you obviously know the answer to some of your "questions" I'm not
    going to bother. I can sum up by saying the productivity gains I have seen
    are very real. Reduced lines of code for similiar functionality are very
    real as well. Some becuause of web services some without. Web services is
    the marketing slogan MSFT has chosen for .NET. I think that's a shame
    because it takes away from the real benefit of .NET. You don't want to take
    advantage of that then that's OK with me. I know what the gains are, I don't
    need a study.

    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote in message
    news:3C8AA183.EA01AEE6@hal-pc.org...
    > Jay Glynn wrote:
    > > It does very littel that can't be done now. The point is it does it much
    > > easier and with much less effort on the part pf the developer. <snipped>
    > > What .NET brings to the table is developer productivity. It has been
    > > reported by some that they see a productivity gain by a factor of 2 or

    3.
    > > That is inline with me experience.

    >
    > Every "silver bullet" software technology is accompanied by such claims,
    > e.g.
    > "a productivity gain by a factor of 2 or 3"
    > "50% less code"
    > "deploys 100% faster", etc.
    > Such claims are almost always false. And those that are true are
    > irrelevant when other factors are considered.
    >
    > But let's go throught the exercise anyway, since you claim to know
    > someone who really knows something about "productivity gains" using Web
    > Services and since you claim to have seen such gains yourself:
    > Who did the studies?
    > How did they do them?
    > Did they develop their product with separate teams in two ways:
    > a) using Web Services,
    > b) without Web services?
    > Did they then deploy both products and compare their performance?
    > Did they then total the costs of development and support using the two
    > methodologies (with and w/o Web Services)?
    > Did they maintain the apps over a period of years and keep records of
    > the total costs of ownership?
    > Finally, where are their published numbers, data and statistics to back
    > up the "increased productivity" claims?
    >
    > I believe that no one has done any of the above whatsoever and that any
    > productivity claims are BS.
    >
    > > In your thinking we would all still be coding assembler.

    >
    > Straw man argument: not my words.
    >
    > > .NET and web services provides a solution that is
    > > quick and easy to implement.

    >
    > Same old restatement of the Microsoft party line in the Microsoft party
    > style: "If you repeat it often enough, they will come."




  5. #35
    Mark Jerde Guest

    Re: Do consumers even want web services?

    Dave --

    > OK, enough, sorry.


    <g> No, no "sorry" needed. I bought the "Inmates" book about 3 weeks ago,
    also clearance, but haven't read as far as the dedication.

    -- Mark






  6. #36
    Jay Glynn Guest

    Re: Do consumers even want web services?

    I was speaking in terms of ASP.NET in general.

    "Jonathan Allen" <greywolfghost@cox.net> wrote in message
    news:3c8ab079@10.1.10.29...
    > > And performs much much better.

    >
    > Not necessarily. With web services, all requests have to be converted to

    and
    > from XML, which is slow to parse. Passing binary data would give you much
    > better performance in the general case.
    >





  7. #37
    Robert Lantry Guest

    Re: Do consumers even want web services?


    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote:
    >Jonathan Allen wrote:
    >> > Do you realize that "an order of magnitude" generally means a factor

    of
    >> > 10? So "several" is a factor of over 100? And this means that you're
    >> > claiming that you developed something in 1/100th the time it would have
    >> > been possible by other methods. Such claims are sheer hyperbole.

    >>
    >> Those numbers are reasonable. Adding a Web Service interface and the
    >> matching client code takes a couple minutes. Doing the same using a TCP

    port
    >> could take me a couple hours. So for me, it literally is two orders of
    >> magnitude faster.
    >> Jonathan Allen

    >
    >I will now mathematically *prove* that what you say is not only
    >impossible, but that it is always *faster* to write a standard function
    >than to write a Web Service, regardless of language:
    >
    >1. Writing a web service requires
    >a) development of a SOAP XML interface,
    >b) development of code for function.
    >
    >2. Writing a function requires (only)
    >b) development of code for function.
    >
    >3. Let us designate T(x) = the time to do x.
    >
    >Then T(a) + T(b) > T(b) always.
    >Q.E.D.
    >
    >Via a similar proof, we can prove that a Web Service will alway execute
    >in time and memory exceeding a function call.


    Call your same function over a network.

    You've just falling into the "willfully incompotent" bucket.

    You have no clue...worse, you are willfully being clueless. Worse, you honestly
    equate a web-service to a function call, which is either willfull or just
    stupid. I think we can assume the later with no argument.

  8. #38
    Jay Glynn Guest

    Re: Do consumers even want web services?

    I find it truly amazing that someone can blow off a comment, certain that I
    havn't done my *homework*. Explain to me please how it is that you *know*
    what I have been doing for the past year and a half. There is absolutely
    nothing misleading about what I have stated in this thread or any other
    thread in this group. Some may be opinion, but none have been misleading.
    The gains in productivity, the reduced lines of code can be proven over and
    over again. I've given examples in this group. There seems to be only one
    person with empty talk in this thread and it is not me.

    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote in message
    news:3C8ACD17.51E4B59B@hal-pc.org...
    > Jay Glynn wrote:
    > > Since you obviously know the answer to some of your "questions" I'm not
    > > going to bother.

    >
    > Yes, I do, because I am certain that you haven't done your homework and
    > that your earlier claims are simply garbage. They are worse than noise,
    > because they are misleading, and are intentionally so.
    >
    > > I can sum up by saying the productivity gains I have seen
    > > are very real. Reduced lines of code for similiar functionality are very
    > > real as well. Some becuause of web services some without. <snip>
    > > I know what the gains are, I don't need a study.

    >
    > This is empty talk: words with neither arguments nor facts to back them
    > up.




  9. #39
    Kunle Odutola Guest

    Re: Do consumers even want web services?


    "Mike Mitchell" <kylix_is@yahoo.co.uk> wrote in message
    news:3c8a623e.17830752@news.devx.com...
    > On Sat, 9 Mar 2002 14:47:10 -0000, "Kunle Odutola"
    > <kunle.odutola@<REMOVETHIS>okocha.freeserve.co.uk> wrote:
    >
    > >What did VB1-6 do that couldn't be done before.

    >
    > It allowed you to write Windows programs.....<SNIP>


    So, nothing then. Nothing that couldn't be done before.

    Kunle


  10. #40
    Kunle Odutola Guest

    Re: Do consumers even want web services?


    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote in message
    news:3C8AD36D.89D45F@hal-pc.org...

    > I will now mathematically *prove* that what you say is not only
    > impossible, but that it is always *faster* to write a standard function
    > than to write a Web Service, regardless of language:


    Moron.

    <SNIPPED further bullshit>

    Kunle



  11. #41
    Robert Lantry Guest

    Re: Do consumers even want web services?


    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote:
    >Jay Glynn wrote:

    <snip>
    >When someone foolishly *persists* in making claims that cannot be
    >demonstrated by logical argument and cannot bring forth factual
    >information (software, data and statistics) to support them, then those
    >claims are not only utterly useless garbage, they are also *misleading*
    >utterly useless garbage. A rational person who was not malevolent, would
    >realize his error and cease to make such claims; only a madman or
    >someone who wishes to mislead others would continue to make such claims.


    When someone foolishly runs off at the mouth like a rabid wombat with it's
    head stuck in a toaster...blah blah blah...you're a dummy and no real programmer
    and you wasted your time and I'm better than you are...and who cares what
    you think because only I know what I'm talking about...blah blah blah.

    In sum, you're another village idiot with an axe to grind...yet, oddly enough...no
    axe and no grindstone.

  12. #42
    Jonathan Allen Guest

    Re: Do consumers even want web services?

    > Actually, it is quite a bit more complicated, since a Web Service call
    > involves(and I'm simplifying here):


    Why are you boring me with the internal implementation? The whole reason I'm
    using a web service is so that I don't have to think about those things.


    > And we haven't begun to consider such situations as:
    > - one of the remote systems hosting a Web Service goes down,
    > - the network suddenly becomes busy because Ed in Accounting is
    > downloading porn,


    These problems would exist no matter what method I used to talk to remote
    servers.


    > - the UDDI server shuts down because script kiddies found a hole in it,
    > - a crook modified the UDDI entry for this Web Service to point to
    > his/her own Web Service, so that he could change the price of the car he
    > wanted to buy, forecast hailstorms in Nevada, or whatever.


    That is just plain stupid. Why would I use UDDI when I already know the URL
    for all the web services I need to communicate with?


    --
    Jonathan Allen


    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote in message
    news:3C8ACBCC.38CB351B@hal-pc.org...
    > Jonathan Allen wrote:
    > > Well, look at this way. In VB.Net, making a Web Service call is no more
    > > complicated than a regular function call.

    >
    > Actually, it is quite a bit more complicated, since a Web Service call
    > involves(and I'm simplifying here):
    > a) Requestor makes UDDI request for location of Web Service,
    > b) Return of location of a Web Service,
    > c) construction of SOAP message containing Web Service arguments,
    > d) passing SOAP message to Web Service,
    > e) Web Service parses SOAP request arguments,
    > f) Web Service computes result,
    > g) Web Service constructs and returns SOAP response,
    > h) Requestor parses SOAP response to get result,
    > wherein steps a, b, d, and g are done *over the network*. You could
    > possibly omit a and b, but it would make no difference.
    >
    > In contrast, a regular function call consists of:
    > a) preparation of arguments,
    > b) function call,
    > c) examined returned results
    > all done *in memory* on the same machine.
    >
    > Now here's a place where the usage of the phrase "two orders of
    > magnitude" is correct: we can safely say that execution of an in-memory
    > function call will easily execute in "two orders of magnitude" less time
    > than a Web Service function call, which is done over the network. Even
    > if everything was on the same machine, the additional overhead of
    > parsing and passing messages through the communications stack would be
    > pure overhead.
    > So a Web Service call requires (far) more memory, more CPU, and more
    > network overhead than an in-memory function call. Web Services are
    > *guaranteed* to eat up CPU, use up memory, and generate increased
    > network traffic. Web Services will be slooooooooooooooowwwwwwwwww! This
    > may be music to Intel and other hardware and telecommunications vendors
    > ears, but it is terrifying to good software developers.
    >
    > And we haven't begun to consider such situations as:
    > - one of the remote systems hosting a Web Service goes down,
    > - the network suddenly becomes busy because Ed in Accounting is
    > downloading porn,
    > - the UDDI server shuts down because script kiddies found a hole in it,
    > - a crook modified the UDDI entry for this Web Service to point to
    > his/her own Web Service, so that he could change the price of the car he
    > wanted to buy, forecast hailstorms in Nevada, or whatever.
    >
    > The concepts of Web Services quite clearly contains the elements of
    > their own demise.




  13. #43
    Jonathan Allen Guest

    Re: Do consumers even want web services?

    > You have no clue...worse, you are willfully being clueless. Worse, you
    honestly
    > equate a web-service to a function call, which is either willfull or just
    > stupid. I think we can assume the later with no argument.


    I give up. That loser has just won a permanent place in my kill file.

    --
    Jonathan Allen


    "Robert Lantry" <mirth@mirthy.com> wrote in message
    news:3c8ad0a6$1@10.1.10.29...
    >
    > "Michael D. Kersey" <mdkersey@hal-pc.org> wrote:
    > >Jonathan Allen wrote:
    > >> > Do you realize that "an order of magnitude" generally means a factor

    > of
    > >> > 10? So "several" is a factor of over 100? And this means that you're
    > >> > claiming that you developed something in 1/100th the time it would

    have
    > >> > been possible by other methods. Such claims are sheer hyperbole.
    > >>
    > >> Those numbers are reasonable. Adding a Web Service interface and the
    > >> matching client code takes a couple minutes. Doing the same using a TCP

    > port
    > >> could take me a couple hours. So for me, it literally is two orders of
    > >> magnitude faster.
    > >> Jonathan Allen

    > >
    > >I will now mathematically *prove* that what you say is not only
    > >impossible, but that it is always *faster* to write a standard function
    > >than to write a Web Service, regardless of language:
    > >
    > >1. Writing a web service requires
    > >a) development of a SOAP XML interface,
    > >b) development of code for function.
    > >
    > >2. Writing a function requires (only)
    > >b) development of code for function.
    > >
    > >3. Let us designate T(x) = the time to do x.
    > >
    > >Then T(a) + T(b) > T(b) always.
    > >Q.E.D.
    > >
    > >Via a similar proof, we can prove that a Web Service will alway execute
    > >in time and memory exceeding a function call.

    >
    > Call your same function over a network.
    >
    > You've just falling into the "willfully incompotent" bucket.
    >
    > You have no clue...worse, you are willfully being clueless. Worse, you

    honestly
    > equate a web-service to a function call, which is either willfull or just
    > stupid. I think we can assume the later with no argument.




  14. #44
    Ed Courtenay Guest

    Re: Do consumers even want web services?


    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote in message
    news:3C8AA58E.EFEE1C0@hal-pc.org...
    > Ed Courtenay wrote:


    > Your words *do* support the contention that you are a numerical
    > illiterate, however.


    Whose contention? When? Where?



  15. #45
    Ed Courtenay Guest

    Re: Do consumers even want web services?


    "Michael D. Kersey" <mdkersey@hal-pc.org> wrote in message
    news:3C8AA58E.EFEE1C0@hal-pc.org...
    > Ed Courtenay wrote:
    >
    > Do you realize that "an order of magnitude" generally means a factor of
    > 10? So "several" is a factor of over 100? And this means that you're
    > claiming that you developed something in 1/100th the time it would have
    > been possible by other methods. Such claims are sheer hyperbole. And
    > further, how did you compute this reduction in effort? Did you try to do
    > the development several ways and then compare costs?


    No, I'm a village idiot and make claims by sticking my finger in the air and
    triangulating the position of the nearest cow... Of course we've tried other
    methods; we could've knocked up some convaluted CORBA or DCOM
    implementation, or any number of other methodologies.

    The fact that I can produce a fully functional, fully described Web Service,
    that any number of clients can talk to (whether they be written in VB.NET,
    C# or Java) in five minutes flat is, IMHO, several orders of magnitude.




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