有兴趣可以了解一下老外的开发的心路历程，了解他们为什么要开发这个软件，了解他们在开发之前都做了些什么，很有意思。

这是从dabo的书(未公开)上摘录的，感兴趣的可以去访问：

http://paul.dabodev.com/doc/book/html/dabo-book.html

History of Dabo

Dabo is the result of a few years of research, starting in 2001 when I
started taking an active interest in the Linux operating system and
open source software in general. I had been using Microsoft Visual
FoxPro to develop data-aware business applications for my clients, and
with mixed messages coming from Microsoft and the FoxPro community as
to the long-term viability of FoxPro as a product, I started looking
for alternatives from the open source community, alternatives that
would permit the development of powerful database applications for
multi-platform deployment.

My quest led me first to Borland Delphi, which had just recently
announced a prerelease version of Kylix, the Linux version of Delphi.
This product would allow deployment of a semi-common codebase to one
flavor of Linux (RedHat 7.x) and Windows. It had good data-aware
controls, but only in the Enterprise version. The execution
performance was pitiful, and code wasn't truly portable between
platforms. The lack of a commitment from Borland to support Macintosh
was the straw that made me look elsewhere.

The next stop was to take a serious look at Java, which does have a
clean and elegant language and does run pretty much equally on all
platforms, thanks to the Java Virtual Machine. I developed some
prototype applications using the Swing components which performed
equally horribly on all three platforms. But, it was easy development
and deployment, although database integration was pretty obtuse
compared to what I was used to in Visual FoxPro. My overall feeling
was that Java may really take over the world, but it has a long way to
go performance-wise, and it really isn't fun to code.

Somewhere about this time, I found out that Kylix was using a toolkit
called Qt to provide the user interface components, and that Kylix was
using the last-generation (Qt 2.x) instead of the newer, much nicer Qt
3.x version. I downloaded the GPL'd version of Qt for Linux, and
followed some tutorials, and was able to build some very impressive
C++ applications that performed very well. I never got my head around
C++, however, so I felt like I'd be painting myself into a corner by
pursuing this angle. Also, database support required the Enterprise
version of Qt, which is something like $1200 per platform. Hardly free
and open source.

Eventually, I came to know of a programming language called Python,
which Ed Leafe had been using to power his website using a product
called Zope. I found Python to be very intuitive to learn, with an
easy readable syntax not unlike FoxPro. Python, on top of being free
and open source, also comes with "batteries included", meaning that
most everything you'd want to do comes in the standard library,
including building user interfaces. Python also comes with three
native data types that have to do with sequences - in other words,
Python can represent database tables and fields natively. Come to find
out, Python also provides an API for connecting to all kinds of
database servers. In other words, Python comes with a lot of the
pieces I'd been searching for over the past few years.

While it was great finding that Python had a lot of the pieces to my
puzzle, it was another thing entirely realizing that putting all the
pieces together into a workable whole would prove to be anything but
easy. Yes, Python can connect to any database and retrieve and update
data. Yes, Python has a graphical user interface. No, connecting to a
given database isn't the same as connecting to a different database.
And no, the user interface that comes with Python is not very modern.

In the Fall of 2003, I set out to create a framework for developing
data-aware applications in Python, using a GUI toolkit I'd just
learned about called wxWindows. I'd actually heard about wxWindows
before, but had discounted it thinking that it was only for the
Windows platform. They've since changed their name to wxWidgets at
Microsoft's request, which may keep future developers from being as
confused as I was. I named this framework 'Dabo', because it sounded
fun and reminds me of words like 'data', 'business', 'application',
and 'objects'. Also, we were watching Star Trek: Deep Space Nine at
the time and I liked the Dabo Girls.

I ended up learning the wxWidgets toolkit pretty well, but created a
pile of spaghetti code that had no separation between the database,
the business rules, and the user interface. It was really a mess and
completely unmaintainable. I set the project down for a few months,
until I was contacted in March of 2004 by Ed Leafe, a long-time FoxPro
guru that was looking for ways to move his skillset over to open
source, multiplatform development - he was looking for the same things
I'd been seeking over the previous couple years. I decided to share my
code with him, along with a sample application, and he very
diplomatically explained all the design problems with my approach. We
came up with an agreement to redesign Dabo from the ground up, with a
3-tier model.

By May of 2004 we announced our work to the public, got a website and
mailing lists, and encouragement from diverse areas of the open
source, Python, and FoxPro communities. As I write this in September
2004, Dabo is under active development, the user interface is 80%
abstracted, 3 databases are supported, and a user-interface graphical
designer is underway. It is already possible to create powerful
data-aware applications and to deploy them to Windows, Linux, and
Macintosh.
-- 
I like python! 
My Donews Blog: http://www.donews.net/limodou
New Google Maillist: http://groups-beta.google.com/group/python-cn

嗯嗯！！
整理进来吧……！
http://wiki.woodpecker.org.cn/moin/DaBo
老外真是非常爱护自个儿的作品哪！ limodou 一提，就立即开始修正自个儿的代码！

感谢，四处发现好东西的探险行者哪！

在 05-8-10，limodou<limodou at gmail.com> 写道：
> 有兴趣可以了解一下老外的开发的心路历程，了解他们为什么要开发这个软件，了解他们在开发之前都做了些什么，很有意思。
> 
> 这是从dabo的书(未公开)上摘录的，感兴趣的可以去访问：
> 
> http://paul.dabodev.com/doc/book/html/dabo-book.html
> 
> History of Dabo
> 
> Dabo is the result of a few years of research, starting in 2001 when I
> started taking an active interest in the Linux operating system and
> open source software in general. I had been using Microsoft Visual
> FoxPro to develop data-aware business applications for my clients, and
> with mixed messages coming from Microsoft and the FoxPro community as
> to the long-term viability of FoxPro as a product, I started looking
> for alternatives from the open source community, alternatives that
> would permit the development of powerful database applications for
> multi-platform deployment.
> 
> My quest led me first to Borland Delphi, which had just recently
> announced a prerelease version of Kylix, the Linux version of Delphi.
> This product would allow deployment of a semi-common codebase to one
> flavor of Linux (RedHat 7.x) and Windows. It had good data-aware
> controls, but only in the Enterprise version. The execution
> performance was pitiful, and code wasn't truly portable between
> platforms. The lack of a commitment from Borland to support Macintosh
> was the straw that made me look elsewhere.
> 
> The next stop was to take a serious look at Java, which does have a
> clean and elegant language and does run pretty much equally on all
> platforms, thanks to the Java Virtual Machine. I developed some
> prototype applications using the Swing components which performed
> equally horribly on all three platforms. But, it was easy development
> and deployment, although database integration was pretty obtuse
> compared to what I was used to in Visual FoxPro. My overall feeling
> was that Java may really take over the world, but it has a long way to
> go performance-wise, and it really isn't fun to code.
> 
> Somewhere about this time, I found out that Kylix was using a toolkit
> called Qt to provide the user interface components, and that Kylix was
> using the last-generation (Qt 2.x) instead of the newer, much nicer Qt
> 3.x version. I downloaded the GPL'd version of Qt for Linux, and
> followed some tutorials, and was able to build some very impressive
> C++ applications that performed very well. I never got my head around
> C++, however, so I felt like I'd be painting myself into a corner by
> pursuing this angle. Also, database support required the Enterprise
> version of Qt, which is something like $1200 per platform. Hardly free
> and open source.
> 
> Eventually, I came to know of a programming language called Python,
> which Ed Leafe had been using to power his website using a product
> called Zope. I found Python to be very intuitive to learn, with an
> easy readable syntax not unlike FoxPro. Python, on top of being free
> and open source, also comes with "batteries included", meaning that
> most everything you'd want to do comes in the standard library,
> including building user interfaces. Python also comes with three
> native data types that have to do with sequences - in other words,
> Python can represent database tables and fields natively. Come to find
> out, Python also provides an API for connecting to all kinds of
> database servers. In other words, Python comes with a lot of the
> pieces I'd been searching for over the past few years.
> 
> While it was great finding that Python had a lot of the pieces to my
> puzzle, it was another thing entirely realizing that putting all the
> pieces together into a workable whole would prove to be anything but
> easy. Yes, Python can connect to any database and retrieve and update
> data. Yes, Python has a graphical user interface. No, connecting to a
> given database isn't the same as connecting to a different database.
> And no, the user interface that comes with Python is not very modern.
> 
> In the Fall of 2003, I set out to create a framework for developing
> data-aware applications in Python, using a GUI toolkit I'd just
> learned about called wxWindows. I'd actually heard about wxWindows
> before, but had discounted it thinking that it was only for the
> Windows platform. They've since changed their name to wxWidgets at
> Microsoft's request, which may keep future developers from being as
> confused as I was. I named this framework 'Dabo', because it sounded
> fun and reminds me of words like 'data', 'business', 'application',
> and 'objects'. Also, we were watching Star Trek: Deep Space Nine at
> the time and I liked the Dabo Girls.
> 
> I ended up learning the wxWidgets toolkit pretty well, but created a
> pile of spaghetti code that had no separation between the database,
> the business rules, and the user interface. It was really a mess and
> completely unmaintainable. I set the project down for a few months,
> until I was contacted in March of 2004 by Ed Leafe, a long-time FoxPro
> guru that was looking for ways to move his skillset over to open
> source, multiplatform development - he was looking for the same things
> I'd been seeking over the previous couple years. I decided to share my
> code with him, along with a sample application, and he very
> diplomatically explained all the design problems with my approach. We
> came up with an agreement to redesign Dabo from the ground up, with a
> 3-tier model.
> 
> By May of 2004 we announced our work to the public, got a website and
> mailing lists, and encouragement from diverse areas of the open
> source, Python, and FoxPro communities. As I write this in September
> 2004, Dabo is under active development, the user interface is 80%
> abstracted, 3 databases are supported, and a user-interface graphical
> designer is underway. It is already possible to create powerful
> data-aware applications and to deploy them to Windows, Linux, and
> Macintosh.
> --
> I like python!
> My Donews Blog: http://www.donews.net/limodou
> New Google Maillist: http://groups-beta.google.com/group/python-cn
> 
> _______________________________________________
> python-chinese list
> python-chinese at lists.python.cn
> http://python.cn/mailman/listinfo/python-chinese
> 
> 
> 


-- 
[Time is unimportant, only life important!]