Loading ...
Sorry, an error occurred while loading the content.

Re: [bafuture] Wow how far have we come with Open Source software ...

Expand Messages
  • wayne radinsky
    ... Yeah I know -- Douglas Crockford (of Innovation and Conformity in a Microsoft World fame) said that Microsoft, by doing this, may end up doing the one
    Message 1 of 11 , Jun 1, 2003
    • 0 Attachment
      --- Troy Gardner <TheGreyman@...> wrote:
      >> It looks to me like the way things are going to shake out,
      >> in the next year or two, will look like this. There's going
      >> to be basically three platforms for developing internet
      >> apps, Microsoft's "dot-NET", Java, and PHP/postnuke from
      >> the open source camp. Either you'll develop in the C#
      >> language on Microsoft's "dot-NET" development environment,
      >> or you'll develop in the Java language
      >
      > Few comments. Supposedly .NET framework on linux is
      > supposed to be out sometime soon. There's already an open
      > source IDE for it, so semi-open source, like java control
      > of where it goes mainstream is in Microsoft's hands but the
      > cost of use is't necessarily tied to purchasing it.

      Yeah I know -- Douglas Crockford (of "Innovation and
      Conformity in a Microsoft World" fame) said that Microsoft,
      by doing this, may end up doing the one thing that nobody
      else can do: create a migration path from Windows to Linux.
      Wouldn't that be funny?

      > I'm not really up on PHP so I don't know what it's true
      > capabilities are but I do hear from PHP people (quite
      > passionate about it usually) about it's wonders...and then
      > they tell me about building framework components (like
      > servlets and application servers) that are already quite
      > robust in Java so I smirk a bit :) I wonder if the wide
      > proliferation of PHP is due to it's use on the User
      > Interface integration with data layer, which in general
      > seems to be the most fluid (so many ways it can be done).

      My impression is that PHP is rough around the edges. There
      are all kinds of quirks in the language. (For example, if
      you use a "for" loop on an array, you directly access and
      change the array elements, whereas if you use "foreach" it
      will access a copy of the array.) These kinds of quirks are
      just asking for hours banging your head against the wall,
      in my opinion, but PHP programmers learn to deal with them.
      Java on the other hand is very clean and consistent in its
      design. At least the language itself -- not necessarily
      true of the code libraries people wrote in the language.

      I'm not sure exactly what you mean by "User Interface
      integration with data layer". The main thrust of postnuke
      is to cleanly separate functionality, such as a discussion
      board or a classified ad system, with the graphic look of
      the site, which is called the "theme" or "template".

      > The distinction as whether to use PHP or N-tier solutions
      > (java) to me at this stage is a choice for right tool for
      > the right job, as you mentioned most people don't need to
      > full power and 'immature/unstable' open-source is good
      > enough upto a point when you start finding the limitations
      > and run out of workarounds for them.

      I agree completely with you on that. I think Java is way
      ahead when it comes to full n-tier apps. There's nothing
      comparable to Java RMI in the PHP realm. You could do
      n-tier distributed apps in PHP, of course, but it would
      take some hacking.

      > While I'm sure there a few I don't know of many people
      > using any fully open source project, in particular I wonder
      > how many PHP or .NET projects have scaled to the point
      > there are hierarchies in each tier of designer, developer,
      > dba. When will somebody like IBM who isn't in the tech
      > sector switch to all open source products?

      The dynamics seem to work like this: As computing becomes
      more 'ubiquitous', as ubiquitous as, say, electricity...
      there's a lot of functionality that is used by "everybody"
      -- a sufficiently large number of people. So you can think
      of that functionality as being "just part of the
      infrastructure". So for a company they want functionality
      that is "just part of the infrastructure" and nothing
      strategic to do with their company. This is the area where
      open source seems to thrive. So for example, everyone with
      a computer needs an operating system, and thus we get
      Linux, the open source operating system, for everyone to
      use. With regard to PHP/postnuke, the scenario would be
      something like, for example, a discussion board. Suppose
      that for some reason "every" company wants a discussion
      board on their website, for customer feedback or whatever,
      but that it's not their core business, their core business
      is selling tires or whatever, and they want their
      discussion board customized with their logo, their site's
      navigation and graphic design, etc. What they don't need is
      a full-blown n-tier development environment for doing huge
      complex enterprise applications, i.e. Java/WebLogic.

      The question is what this situation will look like 5, 10,
      15 years down the line? Will the high- and low-ends remain
      separate niches, or will they expand to compete directly
      against each other? My brave prediction is that eventually
      they will compete.

      > It occurs to me that the predominate problem in scaling
      > open source seems to be more database related. One of the
      > last projects I was one went sequentially from MySQL to
      > Postgres to oracle trying to keep opensource but in the end
      > couldn't do it for various usage/technical reasons and
      > switched to Oracle.

      Out of curiousity.. what were the reasons?

      > Another example is Filemaker which has been around quite a
      > while now, and for a new project makes many medium
      > difficulty things very easy, including generating a usable
      > web front end automatically. A small company could have a
      > web based whatever up in a day or two. But a few things
      > trivial in a 'real' database are difficult to impossible in
      > Filemaker, at this point it might make sense given
      > commercial databases costs to jump to MySQL or Postgres,
      > with significant development overhead. MySQL and Postgres
      > is sufficent for most and continuing to get better, but at
      > the same time Oracle is spending gobs of money on
      > development as well. So it long way to go before it's
      > competing with Oracle on the high end enterprise wide
      > features, neither party is standing still in the
      > developement path.

      Everything I've encountered so far is either MySQL or
      Oracle. (Or MS SQL Server, of course). I don't know anyone
      using Postgresql, would be happy to hear what it's like.

      > I'm particularly happy with Flash as a user interface, many
      > PDA's and Cellphones support it, with quicktime it supports
      > high quality video. Scripting has improved by leaps and
      > bounds every release. There are several nice examples of
      > Rich Internet Applications built with it, and coming,
      > though it certainly has a good deal of inertia (it's just
      > for intro's) to overcome.

      Flash may win as the client side UI. But Mozilla is still a
      contender. Have you used XUL? The user interface to Mozilla
      itself is done entirely in XUL. You can re-skin Mozilla and
      stuff. It's pretty fun to play with.

      >> language. One guy I know (Bill Grosso who runs the SDForum
      >> emerging technology SIG, to be precise) once commented to
      >> me on how amazing it is that for $300, you can get a PC
      >> from Wal-Mart, install JBOSS on it, and have the equivalent
      >> of an enterprise-class Java server from 5 years ago.
      >
      > I use Jboss/Apache/Tomcat, OpenSMF (an opensource port of
      > the Flash communications API which allows binary
      > communication (rather than XML) with minimal overhead to
      > flash clients), some XML, and Flash and a yet to be
      > decided database for my own upcoming projects, along with a
      > freeware but excellent Java Ide (Eclipse which used to be
      > IBM's VisualAge team), the only money I've spent is on the
      > Flash tool which I'm using for clientside UI. I would have
      > killed (not really) for tools like these 4 years ago, it
      > would have cost $30K plus for licenses for what was
      > available then and still not done the same thing, or
      > allowed me to do it in the same time. On that note, if you
      > have any projects you need done with these or similar
      > technologies give me a holler.

      People I've talk with say Eclipse is the best IDE and
      WebLogic is the best platform. I've only played with
      JBuilder, and, quite frankly, it wasn't the best experience.

      > Raid, storage, gigabit ethernet (or firewire networking)
      > has also plummetted to the point where it's quite possible
      > to get what used to be high end only computing on a
      > commidity pc, and serve far more data than most consumer
      > DSL's connections can support outgoing. I just read that a
      > mainstream water cooled PC is going to be available ...this
      > was at one time reserved for cars and crays!

      Whoa... are PC's getting that hot already?

      I've heard one guy joke about putting the CPU on top of the
      PC case, so we could warm our coffee on it.

      But he was talking about the Itanium, which is supposed to
      run too hot because it's a poorly designed chip.

      The energy consumption of these systems is an important
      piece of data -- it relates to Chaisson's "phi"
      (ergs/sec/gram) measurement.

      Remember the Google talk I went to? Google wanted to pack
      PC's in twice as densely as the facilities could power
      them... the reason they have so much "empty space" in their
      facilities is because the facility has a limit to the power
      per square foot. It's A Sign O The Times.

      > Tangentially, I also am amazed that one can get TV/DVD
      > combos (and toolboxs and other items) in the checkout lines
      > of some larger supermarket chains. Maybe it's just me but I
      > normally don't think of getting eggs and vegtables and
      > electronics items at the same time, I've certainly never
      > gotten a shopping list with it on it yet "Honey pick up a
      > galleon of milk and a new DVD player"..are these the new
      > impulse buys? I know years ago wealthy used to be more than
      > one TV's per family, then TV's with cable/VCR and now it's
      > starting to happen with computers, I wonder how many years
      > it will be before I see cellphones, pda's and computers
      > right up there between the hagen daaz and national enquirer.

      Yeah I saw DVD players at the supermarket and had more or
      less the same reaction! I think you're right, disposable
      cellphones next to the national enquirer aren't too far off.

      You know what else is amazing? One friend of mine is
      interested in home theater and he's found that the cost of
      a home theater systems is falling like a rock -- if you
      skip the high-end stores and build the system from PC
      hardware. DLP projectors that cost over $2000 6 months ago
      are $999 now. So you get a PC, put in a HDTV tuner card, a
      Faroudja board, a Radion video card, THX 5.1 surround
      speakers for PC... I'm not sure offhand what that adds up
      to but it's not that much... It won't be long before people
      will never buy a TV again, just use DLP projectors.



      --- Troy Gardner <TheGreyman@...> wrote:
      > > It looks to me like the way things are going to shake out,
      > > in the next year or two, will look like this. There's going
      > > to be basically three platforms for developing internet
      > > apps, Microsoft's "dot-NET", Java, and PHP/postnuke from
      > > the open source camp. Either you'll develop in the C#
      > > language on Microsoft's "dot-NET" development environment,
      > > or you'll develop in the Java language
      >
      > Few comments. Supposedly .NET framework on linux is supposed
      > to be out sometime
      > soon. There's already an open source IDE for it, so semi-open
      > source, like java
      > control of where it goes mainstream is in Microsoft's hands
      > but the cost of use
      > is't necessarily tied to purchasing it.
      >
      > I'm not really up on PHP so I don't know what it's true
      > capabilities are but I
      > do hear from PHP people (quite passionate about it usually)
      > about it's
      > wonders...and then they tell me about building framework
      > components (like
      > servlets and application servers) that are already quite
      > robust in Java so I
      > smirk a bit :) I wonder if the wide proliferation of PHP is
      > due to it's use on
      > the User Interface integration with data layer, which in
      > general seems to be
      > the most fluid (so many ways it can be done).
      >
      > The distinction as whether to use PHP or N-tier solutions
      > (java) to me at this
      > stage is a choice for right tool for the right job, as you
      > mentioned most
      > people don't need to full power and 'immature/unstable'
      > open-source is good
      > enough upto a point when you start finding the limitations and
      > run out of
      > workarounds for them.
      >
      > While I'm sure there a few I don't know of many people using
      > any fully open
      > source project, in particular I wonder how many PHP or .NET
      > projects have
      > scaled to the point there are hierarchies in each tier of
      > designer, developer,
      > dba. When will somebody like IBM who isn't in the tech sector
      > switch to all
      > open source products?
      >
      > It occurs to me that the predominate problem in scaling open
      > source seems to
      > be more database related. One of the last projects I was one
      > went sequentially
      > from MySQL to Postgres to oracle trying to keep opensource but
      > in the end
      > couldn't do it for various usage/technical reasons and
      > switched to Oracle.
      >
      > Another example is Filemaker which has been around quite a
      > while now, and for a
      > new project makes many medium difficulty things very easy,
      > including generating
      > a usable web front end automatically. A small company could
      > have a web based
      > whatever up in a day or two. But a few things trivial in a
      > 'real' database are
      > difficult to impossible in Filemaker, at this point it might
      > make sense given
      > commercial databases costs to jump to MySQL or Postgres, with
      > significant
      > development overhead. MySQL and Postgres is sufficent for most
      > and continuing
      > to get better, but at the same time Oracle is spending gobs of
      > money on
      > development as well. So it long way to go before it's
      > competing with Oracle on
      > the high end enterprise wide features, neither party is
      > standing still in the
      > developement path.
      >
      > I'm particularly happy with Flash as a user interface, many
      > PDA's and
      > Cellphones support it, with quicktime it supports high quality
      > video. Scripting
      > has improved by leaps and bounds every release. There are
      > several nice examples
      > of Rich Internet Applications built with it, and coming,
      > though it certainly
      > has a good deal of inertia (it's just for intro's) to overcome
      >
      > > language. One guy I know (Bill Grosso who runs the SDForum
      > > emerging technology SIG, to be precise) once commented to
      > > me on how amazing it is that for $300, you can get a PC
      > > from Wal-Mart, install JBOSS on it, and have the equivalent
      > > of an enterprise-class Java server from 5 years ago.
      >
      > I use Jboss/Apache/Tomcat, OpenSMF (an opensource port of the
      > Flash
      > communications API which allows binary communication (rather
      > than XML) with
      > minimal overhead to flash clients), some XML, and Flash and a
      > yet to be
      > decided database for my own upcoming projects, along with a
      > freeware but
      > excellent Java Ide (Eclipse which used to be IBM's VisualAge
      > team), the only
      > money I've spent is on the Flash tool which I'm using for
      > clientside UI. I
      > would have killed (not really) for tools like these 4 years
      > ago, it would have
      > cost $30K plus for licenses for what was available then and
      > still not done the
      > same thing, or allowed me to do it in the same time. On that
      > note, if you have
      > any projects you need done with these or similar technologies
      > give me a holler.
      >
      > Raid, storage, gigabit ethernet (or firewire networking) has
      > also plummetted to
      > the point where it's quite possible to get what used to be
      > high end only
      > computing on a commidity pc, and serve far more data than most
      > consumer DSL's
      > connections can support outgoing. I just read that a
      > mainstream water cooled PC
      > is going to be available ...this was at one time reserved for
      > cars and crays!
      >
      > Tangentially, I also am amazed that one can get TV/DVD combos
      > (and toolboxs and
      > other items) in the checkout lines of some larger supermarket
      > chains. Maybe
      > it's just me but I normally don't think of getting eggs and
      > vegtables and
      > electronics items at the same time, I've certainly never
      > gotten a shopping list
      > with it on it yet "Honey pick up a galleon of milk and a new
      > DVD player"..are
      > these the new impulse buys? I know years ago wealthy used to
      > be more than one
      > TV's per family, then TV's with cable/VCR and now it's
      > starting to happen with
      > computers, I wonder how many years it will be before I see
      > cellphones, pda's
      > and computers right up there between the hagen daaz and
      > national enquirer.
      >
      > Troy.
      >
      > =====
      > Troy Gardner http://www.troyworks.com
      >
      > "How you live your seconds, is how you live your days, is how
      > you live your life..."
      >


      __________________________________
      Do you Yahoo!?
      Yahoo! Calendar - Free online calendar with sync to Outlook(TM).
      http://calendar.yahoo.com
    • Michael Korns
      On Sunday, June 01, 2003 1:05 AM wayne radinsky wrote Re: [bafuture] Wow how far have we come with Open Source software ... ... Here are some prices: a..
      Message 2 of 11 , Jun 1, 2003
      • 0 Attachment
        On Sunday, June 01, 2003 1:05 AM wayne radinsky wrote
        Re: [bafuture] Wow how far have we come with Open Source software ...

        >You know what else is amazing? One friend of mine is
        >interested in home theater and he's found that the cost of
        >a home theater systems is falling like a rock -- if you
        >skip the high-end stores and build the system from PC
        >hardware. DLP projectors that cost over $2000 6 months ago
        >are $999 now. So you get a PC, put in a HDTV tuner card, a
        >Faroudja board, a Radion video card, THX 5.1 surround
        >speakers for PC... I'm not sure offhand what that adds up
        >to but it's not that much... It won't be long before people
        >will never buy a TV again, just use DLP projectors.

        Here are some prices:
        a.. DealTime has a 600x800 pixel DLP projector for $899; but, a reasonably good 1280x720 projector still costs $1899.
        b.. HDTV Authority has an HiPix DTV-200 PC card for $399.
        c.. A Faroudja board can be had at www.buymedicalvideo.com for $1999.
        d.. PriceGrabber.com has a Radeon video card for $376.
        e.. A Dell Dimension 2350 home multimedia PC with surround sound speakers, a 19 inch flast panel display, plus home movie editing software, etc. $1598.

        So why not just get the HiPix HDTV PC card and the Dell Dimension? How does the low end DLP projector output compare with a 19 inch flat panel display?


        *******************************
        Michael F. Korns
        1 Plum Hollow Drive
        Henderson, Nevada 89052
        (702) 837-3498
        mkorns@...
        www.korns.com
        www.InvestByAgent.com
        *******************************





        [Non-text portions of this message have been removed]
      • Troy Gardner
        ... Indeed. The latest version of http://www.mikeslist.com predicted that Microsoft comes out with their own distribution of Linux (of course with some
        Message 3 of 11 , Jun 1, 2003
        • 0 Attachment
          > Yeah I know -- Douglas Crockford (of "Innovation and
          > Conformity in a Microsoft World" fame) said that Microsoft,
          > by doing this, may end up doing the one thing that nobody
          > else can do: create a migration path from Windows to Linux.
          > Wouldn't that be funny?

          Indeed. The latest version of http://www.mikeslist.com predicted that Microsoft
          comes out with their own distribution of Linux (of course with some proprietary
          enhancements). He makes a reasonable argument for it, and I can see how it
          might come to pass though it would I think not be a house divided against
          itself for the people involved in the OS development. Would be pretty trippy
          though.

          > My impression is that PHP is rough around the edges. There
          > are all kinds of quirks in the language.

          This I have heard as well, and these seem to be strongly polarizing some people
          REALLY love these quirks, and others (especially coming from other camps)
          REALLY hate the non-compatibility with other code mindsets, as not everyone has
          the luxury of only choosing one language or the other to develop in -this
          dissonance I think is harder in computer language than human language as the
          systax and keywords can be identical but do subtly different things.

          > Java on the other hand is very clean and consistent in its
          > design. At least the language itself -- not necessarily
          > true of the code libraries people wrote in the language.

          This is partly due because of the way Java came about I suppose. The language
          itself was given a bit of non-commercial breathing room to mature before the
          whole schlew of core libraries came to pass, which IMO were probably written by
          a less 'hard core' and experienced team and less invested in the philosophy of
          why and what problems Java was trying to solve. I think more than a few
          libraries were due to people demanding the feature and standardization, and sun
          responding as best they could...java's success was far from garuanteed in the
          beginning.

          > I'm not sure exactly what you mean by "User Interface
          > integration with data layer". The main thrust of postnuke
          > is to cleanly separate functionality, such as a discussion
          > board or a classified ad system, with the graphic look of
          > the site, which is called the "theme" or "template".

          Hmm maybe I'll have to read up on it more. What I was intending was the degree
          of application logic encoded into the visual pages/templates themselves. It's
          the difference between having an minimal script/SQL statement embedded on the
          page, rather than being only passed static data from some application control
          logic which itself may have to go through a data abstraction layer..even if
          they are on the same physical system.

          > infrastructure". So for a company they want functionality
          > that is "just part of the infrastructure" and nothing
          > strategic to do with their company. This is the area where
          > open source seems to thrive.
          <snip> Suppose
          > that for some reason "every" company wants a discussion
          > board on their website, for customer feedback or whatever,
          > but that it's not their core business, their core business
          > is selling tires or whatever, and they want their
          > discussion board customized with their logo, their site's
          > navigation and graphic design, etc. What they don't need is
          > a full-blown n-tier development environment for doing huge
          > complex enterprise applications, i.e. Java/WebLogic.

          True, and that may work well for the upper tiers but it's my experience that
          generally companies pay for data acquisition and use it for historical
          analysis/future trending (better than no information) and are loathe to throw
          it away. Thus it tends to grow rapidly (at least for me)in size and reaches
          past the limits for many smaller db's as well as lacks the tools to help
          efficently mine the information. They also search for integration of dissimlar
          datasources to do things like the true cost of a resources across the
          enterprise, workflow (thus the success of ERP packages like SAP). Even for
          things as 'lite' as discussion boards if it has ties into commercial interests
          (which it does or it wouldn't likely exist) there's some manager there wanting
          reports on it to help prioritize future development ie.. number of customers
          complaining about whatever since a certain date, who bought whatever version of
          whatever.


          > The question is what this situation will look like 5, 10,
          > 15 years down the line? Will the high- and low-ends remain
          > separate niches, or will they expand to compete directly
          > against each other? My brave prediction is that eventually
          > they will compete.

          I agree. Any application requires a degree of power/sophistication/complexity
          in what actually implements it, with an upper limit I think of most mainstream
          business (e.g. how much code (say named Business2003) does it take to process a
          customer, order, inventory, etc. across two or more locations). Given continued
          resources to develop I have no doubt that if a feature set of oracle and mysql
          are captured today in 10 years MySQL2013 might have comparible features of
          Oracle2003, which provided it's just a weblog (a relatively simple application)
          either already do and will continue to do, and before only Oracle2003 may have
          worked now either will work. Perhaps 95% of the features required by
          Business2013 will run on MySQL2013, but maybe in order to be competitive it
          absolutely still requires that 5% of features missing. Who knows? Along those
          lines, given the profliferation of smaller independent projects (just for fun)
          it wouldn't surprise me if someone in 2013 in a team the fraction the size
          builds something comparible to MySQL2003 in their spare time using more
          powerful tools, better hindsight, a different view of what's possible and less
          restrictions as far as an installed userbase that has to be supported.

          > > It occurs to me that the predominate problem in scaling
          > > open source seems to be more database related. One of the
          > > last projects I was one went sequentially from MySQL to
          > > Postgres to oracle trying to keep opensource but in the end
          > > couldn't do it for various usage/technical reasons and
          > > switched to Oracle.
          >
          > Out of curiousity.. what were the reasons?

          I don't remember the full details. MySQL didn't (and as far as I know) still
          doesn't support transactions which were necessary for the type of application
          we were writing. Also it could only handle tens of gigabytes and we were
          dealing with multimedia. This article goes into some of the MySQL
          shortcomings/advantages.

          http://www.oreillynet.com/pub/a/network/2000/06/16/magazine/mysql.html?page=3

          Postgres has a more fully flushed out SQL implementation, while also free and
          open source like MySQL. Conversely there is more overhead in learning it (but
          not as much as Oracle) I think they now support subselect queries but didn't a
          year or two ago and the workarounds for that were just to tedious. I don't
          think they support stored procedures yet, which come in handy for performance
          tuning as it' obliviates sending data back and forth across the wire between
          database and application/webserver tiers. Also the need for
          failover/replication, backup and recovery wasn't as full featured as we needed
          it to be...this is one of the areas that oracle does really well. Most of these
          features most sites don't need.

          http://www.postgresql.org/

          Comparing it to cars: MySQL is like the aircooled volkswagon bug. Easy to work
          on, accessible to anyone, just not the most high performance thing in the
          world.

          Postgres is the economy car, water cooled fuel injected for better performance,
          significantly more complex than the air cooled VW bug, but not having the all
          the features of a more expensive model.

          SQL Server would be a mid-high end car with a built in computer automotive
          tech. Lots of built in intelligence as far as how to optimally tune it, so it
          takes care of itself, it can automatically add indexes and the like.

          Oracle is a the sports car engine, stock out of the factory works well with
          lots of options. It has lots of knobs dials and adjustments one can make so the
          true performance is revealled and but that requires a special (and usually
          expensive) pit team to do so.

          Versant/Objectstore/Jasmine are like Delorians or wankle rotary engines-
          beautiful and realtively rare/odd creatures (Object Oriented databases) that
          can do incredible things (e.g. the objectstore database grows by terabytes at
          Stanfords Particle Accelerator everytime they use it), as not many understand
          how to use them or what makes them special (e.g. eliminating the tedious Object
          to relational mapping).

          > Flash may win as the client side UI. But Mozilla is still a
          > contender. Have you used XUL?

          No, hadn't heard of it before. From reading limited it doesn't sound like you
          can do the same things, and first glance it's not immediately apparent how it's
          superior to CSS/XSLT.

          Flash's weakness is rich text handling (scrolling large pages of text isn't
          it's forte') and it does't look like XUL supports graphics or rather has to use
          SVG for that aspect, so I'm not sure they are in the same market. SVG and Flash
          have some overlap, except as far as I know developing content for SVG
          interactive animation is still in the early stages compared to flash. Flash
          supports rich communication with servers for many different datatypes as well
          as components for UI elements (list boxes and the like).

          > People I've talk with say Eclipse is the best IDE and
          > WebLogic is the best platform. I've only played with
          > JBuilder, and, quite frankly, it wasn't the best experience.

          I was using (and loving) IBM's VisualAge (quirky though) as it supported
          incremental compiling which speeded up application development immensely
          (allowing working on the Tomcat server code while it was actually running
          without having to restart). Unfortunately they phased it out and many of the
          libraries I use need a newer version of the JVM (in order to support
          incremental compiling it was tied to the VM, which I believe was actually
          written in Smalltalk).

          RE: Water cooled PC's
          > Whoa... are PC's getting that hot already?

          depends on how you look at it. While air cooling works Fan noise, or just plain
          geekiness are the reasons why jump to water cooling. I have a few PC's in my
          bedroom/recordign studio and those little fans on the CPU are the noisiest of
          the bunch. Besides water is much more efficent at removing heat than air is,
          which means reduced heat, and longer life of operation.
          http://neasia.nikkeibp.com/wcs/leaf?CID=onair/asabt/news/247756

          > I've heard one guy joke about putting the CPU on top of the
          > PC case, so we could warm our coffee on it.

          this is an old joke. Whatever the current hottest chip is gets inserted in
          that. Coffee cups despite getting colder faster than most would like are
          actually poor heat sinks, and the cost of chips versus coffee (even starbucks)
          hasnt' been worth it. Haven't heard anybody useing chips as cookware yet
          though I'm sure someone has done it: "hey run that graphics app...the egg yolk
          is right over the floating point processor and it's not getting done enough.."

          > But he was talking about the Itanium, which is supposed to
          > run too hot because it's a poorly designed chip.

          I always get a kick out of hearing people talk about 325 million transistor
          done on billion dollar chips like it was a Yugo off of an Mexico assembly line.

          > The energy consumption of these systems is an important
          > piece of data -- it relates to Chaisson's "phi"
          > (ergs/sec/gram) measurement.

          While I haven't heard anything lately, the article was suggesting that power
          dissapation and internconnetions were becoming more of the design limitations
          rather than the ultimate limit of transistor size.

          > You know what else is amazing? One friend of mine is
          > interested in home theater and he's found that the cost of
          > a home theater systems is falling like a rock --

          Being a movie/virtual reality loving geek, I was early on the home theater/pc
          convergence curve (hooking up a 'real' stereo system, replacing the offboard
          surround sound processor with a sound card based one, using a projector, tivo
          like tv capture cards), and continue to love that falling rock though it
          squashes my investments values, and I'm no longer quite ahead as I was a few
          years ago.

          >if you
          > skip the high-end stores and build the system from PC
          > hardware. DLP projectors that cost over $2000 6 months ago
          > are $999 now. So you get a PC, put in a HDTV tuner card, a
          > Faroudja board, a Radion video card, THX 5.1 surround
          > speakers for PC... I'm not sure offhand what that adds up
          > to but it's not that much... It won't be long before people
          > will never buy a TV again, just use DLP projectors.

          That's even if you need all of that. Many cases the Faroudja board isn't
          necessary, and DVD's on a laptop (say the Apple Titanium) with a decently large
          display, with a selfpowered speaker system are good enough for most people's
          use (to replace a conventional TV when it's just them watching). Even having
          access to a projector I find the image on my LCD monitor superior for most
          movies.

          Projectors are cool but when OLED finally makes it to roll to roll
          manufacturing you will get HDTV wallpaper with intgrated surround sound (using
          NXT type front speakers and/or ultrasonic surrounds e.g.
          http://www.atcsd.com/tl_hss.html possible in rollup poster tubes for under
          $300-500, possibly with associated electronics. These won't have bulbs to burn
          out either.

          My beast of a projector was actually something like $250,000 (!) when it was
          new now on ebay you can find them for less than a grand.
          http://www.projectorcentral.com/projector_details.cfm?part_id=842

          What's amazing to me is power density of projectors... a microportable that
          wieghts less than the projector lens pounds of mine has more light output than
          mine did at the top of it's day, and is far quieter (and might actually fit in
          my room) makes me wish at times I didn't invested in one of those instead :/ A
          recent speaker at the La futurists had a mini vaio and a microprojector, all
          said with cords couldn't have wieghed more than my current laptop (7#'s), which
          aside from sound could have been home theater in a knapsack. I know of a few
          people working on laser projectors, which in theory could be the same size and
          more powerful with no bulbs to burn out, and higher color purity than lamp
          based technologies. Future projectors will resemble flashlights more than
          briefacases, and probably be able to be powered for limited durations by
          batteries....it will be an interesting world when people have the ability to
          truly show what they mean anywhere anytime, it will result in new laws and
          regulations against 'video graffiti' and 'video distraction' (e.g. tv's ads
          while driving)

          Troy.

          =====
          Troy Gardner http://www.troyworks.com

          "How you live your seconds, is how you live your days, is how you live your life..."
        • wayne radinsky
          ... It all depends on whether you want an up-close TV experience or a sit-back home theater experience. I m not a home theater expert, I can tell you that
          Message 4 of 11 , Jun 1, 2003
          • 0 Attachment
            > Here are some prices:
            > a.. DealTime has a 600x800 pixel DLP projector for $899;
            > but, a reasonably good 1280x720 projector still costs $1899.
            > b.. HDTV Authority has an HiPix DTV-200 PC card for $399.
            > c.. A Faroudja board can be had at www.buymedicalvideo.com
            > for $1999.
            > d.. PriceGrabber.com has a Radeon video card for $376.
            > e.. A Dell Dimension 2350 home multimedia PC with surround
            > sound speakers, a 19 inch flast panel display, plus home movie
            > editing software, etc. $1598.
            >
            > So why not just get the HiPix HDTV PC card and the Dell
            > Dimension? How does the low end DLP projector output
            > compare with a 19 inch flat panel display?

            It all depends on whether you want an up-close "TV"
            experience or a sit-back "home theater" experience. I'm
            not a home theater expert, I can tell you that what my
            friend is looking at is the MyHD card, the Radeon 9000, the
            Holo3dGraph Faroudja card, and the LogiTech Z680 THX
            compliant 500 watt 5.1 surround sound system. As for the
            projector, he's undecided because the prices are changing
            so fast, and he's looking for people that test them and
            make sure they live up to their advertised claims (e.g.
            measure the actual lumens and contrast). Yes you are going
            to get a better experience with the high res projectors. I
            don't know how much a difference it makes. Is it like the
            difference between sitting in the first row vs the 20th, or
            like the difference between a regular theater and an IMAX?
            In any case, I don't think it will be very long before the
            high res projectors are cheap enough that most people won't
            care. The sudden price changes of projectors really make
            me wonder what Texas Instruments is thinking.





            --- Michael Korns <mkorns@...> wrote:
            > On Sunday, June 01, 2003 1:05 AM wayne radinsky wrote
            > Re: [bafuture] Wow how far have we come with Open Source
            > software ...
            >
            > >You know what else is amazing? One friend of mine is
            > >interested in home theater and he's found that the cost of
            > >a home theater systems is falling like a rock -- if you
            > >skip the high-end stores and build the system from PC
            > >hardware. DLP projectors that cost over $2000 6 months ago
            > >are $999 now. So you get a PC, put in a HDTV tuner card, a
            > >Faroudja board, a Radion video card, THX 5.1 surround
            > >speakers for PC... I'm not sure offhand what that adds up
            > >to but it's not that much... It won't be long before people
            > >will never buy a TV again, just use DLP projectors.
            >
            > Here are some prices:
            > a.. DealTime has a 600x800 pixel DLP projector for $899;
            > but, a reasonably good 1280x720 projector still costs $1899.
            > b.. HDTV Authority has an HiPix DTV-200 PC card for $399.
            > c.. A Faroudja board can be had at www.buymedicalvideo.com
            > for $1999.
            > d.. PriceGrabber.com has a Radeon video card for $376.
            > e.. A Dell Dimension 2350 home multimedia PC with surround
            > sound speakers, a 19 inch flast panel display, plus home movie
            > editing software, etc. $1598.
            >
            > So why not just get the HiPix HDTV PC card and the Dell
            > Dimension? How does the low end DLP projector output compare
            > with a 19 inch flat panel display?
            >
            >
            > *******************************
            > Michael F. Korns
            > 1 Plum Hollow Drive
            > Henderson, Nevada 89052
            > (702) 837-3498
            > mkorns@...
            > www.korns.com
            > www.InvestByAgent.com
            > *******************************
            >
            >
            >
            >
            >
            > [Non-text portions of this message have been removed]
            >
            >


            __________________________________
            Do you Yahoo!?
            Yahoo! Calendar - Free online calendar with sync to Outlook(TM).
            http://calendar.yahoo.com
          • wayne radinsky
            ... Could you ask the people that love the quirks why they love them? Honestly, I think the only reason a person would love the quirks is that they have piles
            Message 5 of 11 , Jun 1, 2003
            • 0 Attachment
              >> My impression is that PHP is rough around the edges. There
              >> are all kinds of quirks in the language.
              >
              > This I have heard as well, and these seem to be strongly
              > polarizing some people REALLY love these quirks, and others
              > (especially coming from other camps) REALLY hate the
              > non-compatibility with other code mindsets, as not everyone
              > has the luxury of only choosing one language or the other
              > to develop in -this dissonance I think is harder in
              > computer language than human language as the systax and
              > keywords can be identical but do subtly different things.

              Could you ask the people that love the quirks why they love
              them? Honestly, I think the only reason a person would love
              the quirks is that they have piles of code that depend on
              it and also have invested time/energy learning them.

              Human language may be a lot more flexible -- in computer
              quirks like this get stuck because of the need for backward
              compatibility. Then again, human languages are full of
              quirks. Ask anyone who's had to learn English as a second
              language whether English has any quirks. And show me one
              human language that doesn't have irregular verbs. :) Maybe
              the real difference is that human language doesn't have to
              be as "mathematically" precise as computer code.

              >> Java on the other hand is very clean and consistent in its
              >> design. At least the language itself -- not necessarily
              >> true of the code libraries people wrote in the language.
              >
              > This is partly due because of the way Java came about I
              > suppose. The language itself was given a bit of
              > non-commercial breathing room to mature before the whole
              > schlew of core libraries came to pass, which IMO were
              > probably written by a less 'hard core' and experienced team
              > and less invested in the philosophy of why and what
              > problems Java was trying to solve.

              Hmm my perception was that the original Java designers were
              intentionally trying to fix the problems they had with C++.
              Java was carefully thought out that's why we see the clean
              design and logical consistency.

              PHP on the other hand, started as a simple script for some
              guys home page, and grew in bits and pieces from there,
              and that's why it's got so many quirks and inconsistencies.
              In fact PHP stands for "Personal Home Page". Oh, I know,
              people now say it stands for "PHP Hypertext Processor" but
              that's just a BS attempt to sanitize things for marketing
              purposes.

              > java's success was far from garuanteed in the beginning.

              I guarantee you Java would have failed if Netscape hadn't
              put it into the web browser and suddenly there were JVM's
              everywhere.

              >> I'm not sure exactly what you mean by "User Interface
              >> integration with data layer". The main thrust of postnuke
              >> is to cleanly separate functionality, such as a discussion
              >> board or a classified ad system, with the graphic look of
              >> the site, which is called the "theme" or "template".
              >
              > Hmm maybe I'll have to read up on it more. What I was
              > intending was the degree of application logic encoded into
              > the visual pages/templates themselves. It's the difference
              > between having an minimal script/SQL statement embedded on
              > the page, rather than being only passed static data from
              > some application control logic which itself may have to go
              > through a data abstraction layer..even if they are on the
              > same physical system.

              Well again you've lost me but I'll try to straighten out
              what *I* said and hopefully things will make sense. I
              should have said "from" instead of "with" -- that is
              "separate from", since "separate with" is gramatically
              incorrect. Argh, my posts are always fully of
              grammar/spelling errors. You all probably think I'm a 5th
              grader :) But yeah, the idea of postnuke is to completely
              *separate* the theme/template (visual look) *from* the code
              that implements the functionality. The "application logic"
              is contained in installable modules that are completely
              independent of the theme/template.





              --- Troy Gardner <TheGreyman@...> wrote:
              > > Yeah I know -- Douglas Crockford (of "Innovation and
              > > Conformity in a Microsoft World" fame) said that Microsoft,
              > > by doing this, may end up doing the one thing that nobody
              > > else can do: create a migration path from Windows to Linux.
              > > Wouldn't that be funny?
              >
              > Indeed. The latest version of http://www.mikeslist.com
              > predicted that Microsoft
              > comes out with their own distribution of Linux (of course with
              > some proprietary
              > enhancements). He makes a reasonable argument for it, and I
              > can see how it
              > might come to pass though it would I think not be a house
              > divided against
              > itself for the people involved in the OS development. Would be
              > pretty trippy
              > though.
              >
              > > My impression is that PHP is rough around the edges. There
              > > are all kinds of quirks in the language.
              >
              > This I have heard as well, and these seem to be strongly
              > polarizing some people
              > REALLY love these quirks, and others (especially coming from
              > other camps)
              > REALLY hate the non-compatibility with other code mindsets, as
              > not everyone has
              > the luxury of only choosing one language or the other to
              > develop in -this
              > dissonance I think is harder in computer language than human
              > language as the
              > systax and keywords can be identical but do subtly different
              > things.
              >
              > > Java on the other hand is very clean and consistent in its
              > > design. At least the language itself -- not necessarily
              > > true of the code libraries people wrote in the language.
              >
              > This is partly due because of the way Java came about I
              > suppose. The language
              > itself was given a bit of non-commercial breathing room to
              > mature before the
              > whole schlew of core libraries came to pass, which IMO were
              > probably written by
              > a less 'hard core' and experienced team and less invested in
              > the philosophy of
              > why and what problems Java was trying to solve. I think more
              > than a few
              > libraries were due to people demanding the feature and
              > standardization, and sun
              > responding as best they could...java's success was far from
              > garuanteed in the
              > beginning.
              >
              > > I'm not sure exactly what you mean by "User Interface
              > > integration with data layer". The main thrust of postnuke
              > > is to cleanly separate functionality, such as a discussion
              > > board or a classified ad system, with the graphic look of
              > > the site, which is called the "theme" or "template".
              >
              > Hmm maybe I'll have to read up on it more. What I was
              > intending was the degree
              > of application logic encoded into the visual pages/templates
              > themselves. It's
              > the difference between having an minimal script/SQL statement
              > embedded on the
              > page, rather than being only passed static data from some
              > application control
              > logic which itself may have to go through a data abstraction
              > layer..even if
              > they are on the same physical system.
              >
              > > infrastructure". So for a company they want functionality
              > > that is "just part of the infrastructure" and nothing
              > > strategic to do with their company. This is the area where
              > > open source seems to thrive.
              > <snip> Suppose
              > > that for some reason "every" company wants a discussion
              > > board on their website, for customer feedback or whatever,
              > > but that it's not their core business, their core business
              > > is selling tires or whatever, and they want their
              > > discussion board customized with their logo, their site's
              > > navigation and graphic design, etc. What they don't need is
              > > a full-blown n-tier development environment for doing huge
              > > complex enterprise applications, i.e. Java/WebLogic.
              >
              > True, and that may work well for the upper tiers but it's my
              > experience that
              > generally companies pay for data acquisition and use it for
              > historical
              > analysis/future trending (better than no information) and are
              > loathe to throw
              > it away. Thus it tends to grow rapidly (at least for me)in
              > size and reaches
              > past the limits for many smaller db's as well as lacks the
              > tools to help
              > efficently mine the information. They also search for
              > integration of dissimlar
              > datasources to do things like the true cost of a resources
              > across the
              > enterprise, workflow (thus the success of ERP packages like
              > SAP). Even for
              > things as 'lite' as discussion boards if it has ties into
              > commercial interests
              > (which it does or it wouldn't likely exist) there's some
              > manager there wanting
              > reports on it to help prioritize future development ie..
              > number of customers
              > complaining about whatever since a certain date, who bought
              > whatever version of
              > whatever.
              >
              >
              > > The question is what this situation will look like 5, 10,
              > > 15 years down the line? Will the high- and low-ends remain
              > > separate niches, or will they expand to compete directly
              > > against each other? My brave prediction is that eventually
              > > they will compete.
              >
              > I agree. Any application requires a degree of
              > power/sophistication/complexity
              > in what actually implements it, with an upper limit I think of
              > most mainstream
              > business (e.g. how much code (say named Business2003) does it
              > take to process a
              > customer, order, inventory, etc. across two or more
              > locations). Given continued
              > resources to develop I have no doubt that if a feature set of
              > oracle and mysql
              > are captured today in 10 years MySQL2013 might have comparible
              > features of
              > Oracle2003, which provided it's just a weblog (a relatively
              > simple application)
              > either already do and will continue to do, and before only
              > Oracle2003 may have
              > worked now either will work. Perhaps 95% of the features
              > required by
              > Business2013 will run on MySQL2013, but maybe in order to be
              > competitive it
              > absolutely still requires that 5% of features missing. Who
              > knows? Along those
              > lines, given the profliferation of smaller independent
              > projects (just for fun)
              > it wouldn't surprise me if someone in 2013 in a team the
              > fraction the size
              > builds something comparible to MySQL2003 in their spare time
              > using more
              > powerful tools, better hindsight, a different view of what's
              > possible and less
              > restrictions as far as an installed userbase that has to be
              > supported.
              >
              > > > It occurs to me that the predominate problem in scaling
              > > > open source seems to be more database related. One of the
              > > > last projects I was one went sequentially from MySQL to
              > > > Postgres to oracle trying to keep opensource but in the
              > end
              > > > couldn't do it for various usage/technical reasons and
              > > > switched to Oracle.
              > >
              > > Out of curiousity.. what were the reasons?
              >
              > I don't remember the full details. MySQL didn't (and as far as
              > I know) still
              > doesn't support transactions which were necessary for the type
              > of application
              > we were writing. Also it could only handle tens of gigabytes
              > and we were
              > dealing with multimedia. This article goes into some of the
              > MySQL
              > shortcomings/advantages.
              >
              >
              http://www.oreillynet.com/pub/a/network/2000/06/16/magazine/mysql.html?page=3
              >
              > Postgres has a more fully flushed out SQL implementation,
              > while also free and
              > open source like MySQL. Conversely there is more overhead in
              > learning it (but
              > not as much as Oracle) I think they now support subselect
              > queries but didn't a
              > year or two ago and the workarounds for that were just to
              > tedious. I don't
              > think they support stored procedures yet, which come in handy
              > for performance
              > tuning as it' obliviates sending data back and forth across
              > the wire between
              > database and application/webserver tiers. Also the need for
              > failover/replication, backup and recovery wasn't as full
              > featured as we needed
              > it to be...this is one of the areas that oracle does really
              > well. Most of these
              > features most sites don't need.
              >
              > http://www.postgresql.org/
              >
              > Comparing it to cars: MySQL is like the aircooled volkswagon
              > bug. Easy to work
              > on, accessible to anyone, just not the most high performance
              > thing in the
              > world.
              >
              >
              === message truncated ===


              __________________________________
              Do you Yahoo!?
              Yahoo! Calendar - Free online calendar with sync to Outlook(TM).
              http://calendar.yahoo.com
            • Brad
              ... I ve used Postgresql extensively. It is my open-source database of choice. Recently many people that I know have been switching over to it from MySQL. It
              Message 6 of 11 , Jun 1, 2003
              • 0 Attachment
                >
                >
                >Everything I've encountered so far is either MySQL or
                >Oracle. (Or MS SQL Server, of course). I don't know anyone
                >using Postgresql, would be happy to hear what it's like.
                >
                >
                I've used Postgresql extensively. It is my open-source database of
                choice. Recently many people that I know have been switching over to it
                from MySQL. It seems to me that it is gaining traction as the leading
                free database.

                As far as I'm concerned MySQL does not have adequate functionality for
                use in anything but the simplest applications. Postgresql has
                sufficiently powerful functionality to be used in almost any enterprise
                application. The only applications for which it would not suffice is
                those requiring the scalability and powerful functionality of Oracle.
                (In reality, there are very few databases that actually need all of the
                power that Oracle offers.) Though it is significantly less user-friendly
                than MySQL and its documentation is somewhat terse in some cases, for
                anyone with a good knowledge of SQL and some DBA skills, it isn't
                difficult at all.

                Postgresql is live in two enterprise applications that I've worked on,
                and I'm using it again in my current project. On this current project,
                if our company is successful we will be dealing with many, many
                terabytes of data and we will probably ultimately need to migrate to
                Oracle, but as a pre-funding company, Postgresql will meet our current
                requirements supremely.

                I would highly recommend Postgresql for any application requiring a
                database, save the very simplest and very largest.

                Brad
              • wayne radinsky
                Just remembered there s one thing I forgot to answer... ... XUL is in a similar spirit. It separates UI from graphics. If you want menus, for example, you ll
                Message 7 of 11 , Jun 2, 2003
                • 0 Attachment
                  Just remembered there's one thing I forgot to answer...

                  >> Flash may win as the client side UI. But Mozilla is still a
                  >> contender. Have you used XUL?
                  >
                  > No, hadn't heard of it before. From reading limited it
                  > doesn't sound like you can do the same things, and first
                  > glance it's not immediately apparent how it's superior to
                  > CSS/XSLT.

                  XUL is in a similar spirit. It separates UI from graphics.
                  If you want menus, for example, you'll use XUL "menu"
                  tags. If you later decide you want to change the way your
                  menus look, you can do that, because XUL knows which
                  elements in your UI are menus and which are something
                  else, like tables or listboxes.



                  --- Troy Gardner <TheGreyman@...> wrote:
                  > > Yeah I know -- Douglas Crockford (of "Innovation and
                  > > Conformity in a Microsoft World" fame) said that Microsoft,
                  > > by doing this, may end up doing the one thing that nobody
                  > > else can do: create a migration path from Windows to Linux.
                  > > Wouldn't that be funny?
                  >
                  > Indeed. The latest version of http://www.mikeslist.com
                  > predicted that Microsoft
                  > comes out with their own distribution of Linux (of course with
                  > some proprietary
                  > enhancements). He makes a reasonable argument for it, and I
                  > can see how it
                  > might come to pass though it would I think not be a house
                  > divided against
                  > itself for the people involved in the OS development. Would be
                  > pretty trippy
                  > though.
                  >
                  > > My impression is that PHP is rough around the edges. There
                  > > are all kinds of quirks in the language.
                  >
                  > This I have heard as well, and these seem to be strongly
                  > polarizing some people
                  > REALLY love these quirks, and others (especially coming from
                  > other camps)
                  > REALLY hate the non-compatibility with other code mindsets, as
                  > not everyone has
                  > the luxury of only choosing one language or the other to
                  > develop in -this
                  > dissonance I think is harder in computer language than human
                  > language as the
                  > systax and keywords can be identical but do subtly different
                  > things.
                  >
                  > > Java on the other hand is very clean and consistent in its
                  > > design. At least the language itself -- not necessarily
                  > > true of the code libraries people wrote in the language.
                  >
                  > This is partly due because of the way Java came about I
                  > suppose. The language
                  > itself was given a bit of non-commercial breathing room to
                  > mature before the
                  > whole schlew of core libraries came to pass, which IMO were
                  > probably written by
                  > a less 'hard core' and experienced team and less invested in
                  > the philosophy of
                  > why and what problems Java was trying to solve. I think more
                  > than a few
                  > libraries were due to people demanding the feature and
                  > standardization, and sun
                  > responding as best they could...java's success was far from
                  > garuanteed in the
                  > beginning.
                  >
                  > > I'm not sure exactly what you mean by "User Interface
                  > > integration with data layer". The main thrust of postnuke
                  > > is to cleanly separate functionality, such as a discussion
                  > > board or a classified ad system, with the graphic look of
                  > > the site, which is called the "theme" or "template".
                  >
                  > Hmm maybe I'll have to read up on it more. What I was
                  > intending was the degree
                  > of application logic encoded into the visual pages/templates
                  > themselves. It's
                  > the difference between having an minimal script/SQL statement
                  > embedded on the
                  > page, rather than being only passed static data from some
                  > application control
                  > logic which itself may have to go through a data abstraction
                  > layer..even if
                  > they are on the same physical system.
                  >
                  > > infrastructure". So for a company they want functionality
                  > > that is "just part of the infrastructure" and nothing
                  > > strategic to do with their company. This is the area where
                  > > open source seems to thrive.
                  > <snip> Suppose
                  > > that for some reason "every" company wants a discussion
                  > > board on their website, for customer feedback or whatever,
                  > > but that it's not their core business, their core business
                  > > is selling tires or whatever, and they want their
                  > > discussion board customized with their logo, their site's
                  > > navigation and graphic design, etc. What they don't need is
                  > > a full-blown n-tier development environment for doing huge
                  > > complex enterprise applications, i.e. Java/WebLogic.
                  >
                  > True, and that may work well for the upper tiers but it's my
                  > experience that
                  > generally companies pay for data acquisition and use it for
                  > historical
                  > analysis/future trending (better than no information) and are
                  > loathe to throw
                  > it away. Thus it tends to grow rapidly (at least for me)in
                  > size and reaches
                  > past the limits for many smaller db's as well as lacks the
                  > tools to help
                  > efficently mine the information. They also search for
                  > integration of dissimlar
                  > datasources to do things like the true cost of a resources
                  > across the
                  > enterprise, workflow (thus the success of ERP packages like
                  > SAP). Even for
                  > things as 'lite' as discussion boards if it has ties into
                  > commercial interests
                  > (which it does or it wouldn't likely exist) there's some
                  > manager there wanting
                  > reports on it to help prioritize future development ie..
                  > number of customers
                  > complaining about whatever since a certain date, who bought
                  > whatever version of
                  > whatever.
                  >
                  >
                  > > The question is what this situation will look like 5, 10,
                  > > 15 years down the line? Will the high- and low-ends remain
                  > > separate niches, or will they expand to compete directly
                  > > against each other? My brave prediction is that eventually
                  > > they will compete.
                  >
                  > I agree. Any application requires a degree of
                  > power/sophistication/complexity
                  > in what actually implements it, with an upper limit I think of
                  > most mainstream
                  > business (e.g. how much code (say named Business2003) does it
                  > take to process a
                  > customer, order, inventory, etc. across two or more
                  > locations). Given continued
                  > resources to develop I have no doubt that if a feature set of
                  > oracle and mysql
                  > are captured today in 10 years MySQL2013 might have comparible
                  > features of
                  > Oracle2003, which provided it's just a weblog (a relatively
                  > simple application)
                  > either already do and will continue to do, and before only
                  > Oracle2003 may have
                  > worked now either will work. Perhaps 95% of the features
                  > required by
                  > Business2013 will run on MySQL2013, but maybe in order to be
                  > competitive it
                  > absolutely still requires that 5% of features missing. Who
                  > knows? Along those
                  > lines, given the profliferation of smaller independent
                  > projects (just for fun)
                  > it wouldn't surprise me if someone in 2013 in a team the
                  > fraction the size
                  > builds something comparible to MySQL2003 in their spare time
                  > using more
                  > powerful tools, better hindsight, a different view of what's
                  > possible and less
                  > restrictions as far as an installed userbase that has to be
                  > supported.
                  >
                  > > > It occurs to me that the predominate problem in scaling
                  > > > open source seems to be more database related. One of the
                  > > > last projects I was one went sequentially from MySQL to
                  > > > Postgres to oracle trying to keep opensource but in the
                  > end
                  > > > couldn't do it for various usage/technical reasons and
                  > > > switched to Oracle.
                  > >
                  > > Out of curiousity.. what were the reasons?
                  >
                  > I don't remember the full details. MySQL didn't (and as far as
                  > I know) still
                  > doesn't support transactions which were necessary for the type
                  > of application
                  > we were writing. Also it could only handle tens of gigabytes
                  > and we were
                  > dealing with multimedia. This article goes into some of the
                  > MySQL
                  > shortcomings/advantages.
                  >
                  >
                  http://www.oreillynet.com/pub/a/network/2000/06/16/magazine/mysql.html?page=3
                  >
                  > Postgres has a more fully flushed out SQL implementation,
                  > while also free and
                  > open source like MySQL. Conversely there is more overhead in
                  > learning it (but
                  > not as much as Oracle) I think they now support subselect
                  > queries but didn't a
                  > year or two ago and the workarounds for that were just to
                  > tedious. I don't
                  > think they support stored procedures yet, which come in handy
                  > for performance
                  > tuning as it' obliviates sending data back and forth across
                  > the wire between
                  > database and application/webserver tiers. Also the need for
                  > failover/replication, backup and recovery wasn't as full
                  > featured as we needed
                  > it to be...this is one of the areas that oracle does really
                  > well. Most of these
                  > features most sites don't need.
                  >
                  > http://www.postgresql.org/
                  >
                  > Comparing it to cars: MySQL is like the aircooled volkswagon
                  > bug. Easy to work
                  > on, accessible to anyone, just not the most high performance
                  > thing in the
                  > world.
                  >
                  >
                  === message truncated ===


                  __________________________________
                  Do you Yahoo!?
                  Yahoo! Calendar - Free online calendar with sync to Outlook(TM).
                  http://calendar.yahoo.com
                Your message has been successfully submitted and would be delivered to recipients shortly.