㰀栀琀洀氀㸀
Borland model driven architecture, MDA, with Boldऀ㰀瀀㸀圀愀渀琀 琀漀 爀甀渀 琀栀攀 瀀栀䜀愀渀琀吀椀洀攀倀愀挀欀愀最攀 愀猀 一愀琀椀瘀攀 嘀䌀䰀 漀爀 洀愀礀戀攀 眀椀琀栀 䈀伀䰀䐀 椀昀 礀漀甀 栀愀瘀攀 䐀㜀 愀爀挀栀椀琀攀挀琀㼀 䐀漀眀渀氀漀愀搀 渀漀眀㨀 㰀愀 栀爀攀昀㴀∀搀漀眀渀氀漀愀搀猀⸀栀琀洀⌀最琀瀀∀㸀瀀栀䜀愀渀琀吀椀洀攀倀愀挀欀愀最攀嘀䌀䰀㰀⼀愀㸀⸀ 䄀猀 愀 嘀䌀䰀 搀攀瘀攀氀漀瀀攀爀 礀漀甀 渀漀眀 栀愀瘀攀 愀 挀栀漀椀挀攀 䄀挀琀椀瘀攀堀⼀䌀伀䴀 漀爀 一愀琀椀瘀攀 瀀栀䜀愀渀琀吀椀洀攀倀愀挀欀愀最攀⸀ 夀漀甀 挀愀渀 爀甀渀 琀栀攀 䈀漀氀搀 䄀眀愀爀攀 挀漀洀瀀漀渀攀渀琀猀 漀爀 琀栀攀 猀琀愀渀搀愀爀搀 挀漀洀瀀漀渀攀渀琀猀⸀㰀⼀瀀㸀 ऀ㰀瀀㸀㰀戀㸀䜀漀琀 洀漀搀攀氀㼀 䜀攀琀 䈀漀氀搀℀ 䜀漀琀 䈀漀氀搀㼀 䜀攀琀 瀀栀䜀愀渀琀吀椀洀攀倀愀挀欀愀最攀℀㰀⼀戀㸀㰀⼀瀀㸀 ऀ㰀栀㔀㸀䤀渀琀爀漀搀甀挀椀渀最 琀栀攀 瀀栀䜀爀椀搀开䈀䄀㰀⼀栀㔀㸀
Download the modified Building and owners sample and try it out now (1mb) ऀThis grid is not for everyone, it is designed for developers that some times feel the need to: ऀ㰀甀氀㸀...or like AC/DC once put it; for those about to rock... The phGrid_BA is part of the phGantTimePackageVCL, aprox $US 480 single developer. This is what we wrote about the Bold-framework a year ago (and it is still true): ऀHave you ever encountered things that are to good to be true? Usually you end up disappointed. But for a brief moment in time you have felt the buzz of coming a cross something really special. ऀThat is what I first felt when I started to work with the framework BOLD for Delphi. This was 3 years ago and I still feel the buzz... ऀI simply do not know how to explain how the picture of system development has totally changed for me since I started using BFD. ऀIt used to be 10% interviewing customers about demands. 20% Implementing software to meet demands. 20% Fighting new demands since they would wreck everything achieved so far, and 50% to implement new demands that wreck almost everything. ऀThe last couple of years I have been using BFD and the picture has totally changed. We now can afford to be truly iterative in the development cycle. We welcome new demands that details the model more. Every cycle (usually about 3 to 5 weeks) is divided into 30% interviewing customers to find detailed information about their domain. 20% modeling in UML to meet the reality in a good enough approximation. 10% thinking about and implementing truly reusable software components. 20% discussing GUI issues with users and finally 20% building the application. ऀWe now work up to 10 times faster than we ever could before, and we end up with a product that is truly robust and stable. We have become calm and we can take time for customers explaining their reality, which in turn makes it possible for us to build better systems. ऀFive years ago I always tried to simplify everything in order to work fast. That often led to problems with reality. You found out late that you had cut to many corners and was forced to go back and clarify both code and model. A typical system was complex when reaching 25 to 30 domain classes. ऀNowadays I spend my time finding every strange thing and every “one in a thousand” case there is, because these tell me a lot about reality. When you build a model it is much better to start with the most complex cases rather than the plain and simple ones. Why is that? Well if you can explain the most complex and that explanation also covers the simple case you know that you are closer to the truth. And when you have the truth you do not get surprised all the time, and do not end up in situations where you are forced to rethink and redesign. ऀNow I call a system large when it contain 250 to 300 domain classes, but hardly complex. Not with a understandable model anyway. I use link objects with attributes, many to many relationships, aggregates with cascading deletes, inheritance, aggregation of functionality and many other modeling techniques today that I would have stayed clear from 5 years ago. With BFD it is simpler to do it right than cutting corners… Who would believe that? ऀMy focus has shifted from technology, getting the database to optimize my SQL the right way, to domain issues, what do the customer need and why. ऀHans Karlsen, M. Sc. Computer technology. It is with great pleasure that we at plexityHide.com present truly reusable components for the Bold framework. ऀऀ ऀ | ऀ㰀琀愀戀氀攀 眀椀搀琀栀㴀∀㈀㜀 ∀ 戀漀爀搀攀爀㴀∀ ∀ 挀攀氀氀瀀愀搀搀椀渀最㴀∀ ∀ 挀攀氀氀猀瀀愀挀椀渀最㴀∀ ∀㸀 |