Thu, Sep 18, 2014, 8:44 PM EDT - U.S. Markets closed

Recent

% | $
Quotes you view appear here for quick access.

Versant Corp. Message Board

  • kennytilt kennytilt Nov 23, 1999 2:01 PM Flag

    Get a grip, netinvestor2000

    ...good luck to VSNT, I gotta few shares, but
    misinformation is not nice.

    ODIS has 40% of the ODB
    market and recently was awarded "Best ODB" by somebody
    (not ODI <g>)...don't make me dig it
    up.

    Agreed, ORCL sleeps with the fish. :)

    SortNewest  |  Oldest  |  Most Replied Expand all replies
    • crunch gives a good explanation of problems in
      using Versant JVI in IBM VAJava. Fundamentally, from
      what I understand, the IBM VAJava IDE is written in
      SmallTalk. Versant's concept is to create Java class files
      and apply "byte code enhancement". This causes
      problems because the IBM IDE cannot currently deal
      seamlessly with the enhanced class files, which must be
      external to the IDE. Debugging is essentially impossible
      once you cross into the enhanced class files. Good
      application architecture can ease this problem. I think it is
      an open question whether IBM will cooperate with
      Versant to allow better integration. My take is that IBM
      wants to push their DB2 Universal product and will be a
      bully.

    • Visual Age works great with RDBMS using JDBC. You
      can code, run, debug etc with abandon, all in the
      IDE. And Visual Age Enterprise edition includes an
      object/relational mapping tool.

      Now, I'm no fan of
      relational databases when it comes to coding in an oo
      language like java. One of the stated benifits of using an
      ODBMS with an oo language is that (theoretically) one
      can eliminate a ton of code that is necessary to map
      relational data to objects.

      I buy this argument,
      especially since I've already been down that road 2-3 times.
      And if all other things were equal (that is, if
      developing Versant applications in Visual Age was as smooth
      as developing a relational database application in
      Visual Age), I'd say the case is closed -- Versant would
      win.

      But things are NOT equal, because the
      VAJ/Versant integration is not there. Thus, a VAJ developer
      must make the following trade off: either

      a)
      use a relational database and gain a solid
      development environment at the expense of having to write
      object-relational mapping code, or

      b) go with Versant and
      accept a compromised, flakey development enviroment that
      slows you down by impeding the natural code/test/debug
      workflow cycle that all programmers use multiple times
      each day.

      I noticed I've never really directly
      answered your main question before, so I will now: the
      Versant/Visual Age integration problems are BOTH fundamental in
      nature (that is, the entire concept of forcing the
      programmer to manually post-process his class files to
      achieve persistence and to manually export his source
      files to debug, is ill-conceived at best) AND IN
      ADDITION there are a bunch of bugs in the integration that
      makes one unsure of what the hell is going
      on.

      Anyway, I can't state this any clearer. I'm done with
      this.

    • [I don't know what is wrong with Yahoo but I
      can't read my message back. Hmmm, may be they use a
      relational?]

      Thank you for the clarification. What you said is fine.
      Now the real question is how well is Visual Age
      integrated with the relationals? Do they use JDBC? And then
      what? Do they write a whole bunch of code to map their
      tabular data to java objects or does VA do something for
      them so that they can get the mapping for free (as it
      is with VSNT)! And of course there is no object
      cache I suppose. So what I am interested in knowing is
      whether you are comparing apples with apples, or are you
      just saying that integration with VSNT is
      uncomfortable. In the 2nd case the integration with the
      relationals may be more uncomfortable and still not perform.
      In that case I would still be long on VSNT because
      it is where the future is (they will eventually fix
      their problems). Your opinion?

    • Thank you for the clarification. What you said is
      fine. Now the real question is how well is Visual Age
      integrated with the relationals? Do they use JDBC? And then
      what? Do they write a whole bunch of code to map their
      tabular data to java objects or does VA do something for
      them so that they can get the mapping for free (as it
      is with VSNT)! And of course there is no object
      cache I suppose. So what I am interested in knowing is
      whether you are comparing apples with apples, or are you
      just saying that integration with VSNT is
      uncomfortable. In the 2nd case the integration with the
      relationals may be more uncomfortable and still not perform.
      In that case I would still be long on VSNT because
      it is where the future is (they will eventually fix
      their problems). Your opinion?

    • Thank you for the clarification. What you said is
      fine. Now the real question is how well is Visual Age
      integrated with the relationals? Do they use JDBC? And then
      what? Do they write a whole bunch of code to map their
      tabular data to java objects or does VA do something for
      them so that they can get the mapping for free (as it
      is with VSNT)! And of course there is no object
      cache I suppose. So what I am interested in knowing is
      whether you are comparing apples with apples, or are you
      just saying that integration with VSNT is
      uncomfortable. In the 2nd case the integration with the
      relationals may be more uncomfortable and still not perform.
      In that case I would still be long on VSNT because
      it is where the future is (they will eventually fix
      their problems). Your opinion?

    • 100,000 shares today @ $ 7.2 or so it
      seems.

      People who throw around this kind of money with
      something like VSNT usually know what they are doing. Now
      that our little shakeout is over, we are back on the
      2-digit track.

    • Ok I'll repeat myself for your sake.

      In
      message 1704 I said "Versant can be forced to work within
      the Visual Age IDE, but only if you accept some
      fairly unnatural workarounds (e.g. - must export source
      code to debug; environment gets confused about what
      code it's working with; no support for Java2
      Collections, etc, etc). This is hardly "transparent
      persistence". "

      To that let me add the following:


      To make a class storable in the database, you must
      use Versant's "Enhancer" utility that exports the
      class out of Visual Age and tweaks its java class
      files. This is an entirely manual operation that must be
      done every time you modify the class. If you want to
      use the IDE's debugger on your enhanced code you must
      manually export the source code (this is a seperate
      operation from enhancing your classes) for your enhanced
      classes each time you modify your source; if you don't do
      this you'll crash your IDE when you trace into the
      your persistent classes. You have to put all classes
      that will be stored in the database into a separate
      project, and enhance them all at once. There are also a
      number of bugs (RMI-related?) that cause explained
      behavior and crashes in when you try to run your
      application from the IDE.

      While there are
      work-arounds for some of these issues, the typical programmer
      uses an IDE for rapid edit/test/debug cycles, and all
      this adds up to a serious hit to productivity. I'm
      certain you'd be happy if your workflow is to code for a
      long time in Visual Age, export your entire
      application and then test outside the IDE without the benefit
      of a debugger. That's certainly an option for
      trivial applications.

      But then, is there any such
      thing as a trivial WebSphere / Versant application?

    • Go VSNT!

    • Your point when you started out was not the
      general state of VSNT. You started by saying that the
      "Versant/IBM integration overrated". When pressed to provide
      specifics you have pretended to know what you are talking
      about without giving any details. Jasmine was a joint
      venture between CA and Fujitsu, not Hitachi. Those in the
      know always knew that it would never fly (too heavy).
      O2 had a lot of theoreticians working on it - never
      in the same ballpark as VSNT or ODIS. And the niche
      market you are talking about now includes the WEB. ODBMS
      will play the embedded db role because it is fast and
      fast to integrate with. Relationals will be legacy
      access dbs. There will be one ODB sold for every 2 web
      servers, in the near future.

      For the 3rd time I
      would like you to stick to your original point and tell
      us about the specifics of your assertion or loose
      your credibility on this board (at least with me).

    • I'd agree that "there is more to Versant then
      just the IBM relationship", but no, I don't believe
      Versant or any other OODBMS vendor can make it purely on
      their own as simply an OODBMS vendor. Versant is
      arguably the best OODBMS out there, but that's simply not
      enough anymore.

      OODBMSs have been out since the
      80s, and promised take over the database market. But
      to date they've made only minor inroads in niche
      markets against the relational juggernaut, and those
      niche markets are getting saturated (there are only so
      many telecom companies). Serious future growth depends
      entirely on getting their ODBMS technology to work well
      with tools that exploit the exploding demand for
      distributed internet/eBusiness based applications. That means
      EJB containers, that means application servers, that
      means XML. Only those ODBMS vendors that have good
      partnerships and/or products along these lines can
      realistically expect increasing revenues. The rest, well, you
      only need so many OODBMS vendors to address a niche
      market...

    • View More Messages
 

Trending Tickers

i
Trending Tickers features significant U.S. stocks showing the most dramatic increase in user interest in Yahoo Finance in the previous hour over historic norms. The list is limited to those equities which trade at least 100,000 shares on an average day and have a market cap of more than $300 million.