Re: New features for Delphi
- From: "Mirco Malaguti" <mirco.cut.away.this.malaguti@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>
- Date: 27 Aug 2006 03:00:17 -0700
Hi Graham,
I disagree. There are plenty of open-source databases around (Firebird, MySQL, PostgreSQL for client/server; SQLite and FlashFiler for the desktop). Every Windows computer also comes with the JET engine.
The reason I see Paradox so valuable in tandem with Delphi is NOT for its database particular qualities.
You're right: open source db-engines are powerful and reliable. In fact I happily use Firebird as a back-end
In my view, one of Delphi's strengths is the way it can interact with a variety of databases; tying it too closely to one, like Paradox, would make Delphi less attractive, not more.
I agree, but the strength of the Delphi/Paradox pair is NOT about Paradox as a database, the BDE.
Rather, it's on the ability to leverage each other best qualities. I try to explain.
Given a client/server database application as a development case, Paradox is extremely productive for
the client front-end part:
- the visual form designer is simple and straightforward
- the object model, the RTL, is well architected, feature complete and simple to understand (in essence it's the VCL's ancestor)
- the language (ObjectPAL) is a nice and useful compromise between a true compilable one and a scripting one, in terms of simplicity and power
- the development model is flexible and powerful, because it offers true event management for the UI and true modularity for the funcionalities
- it's based on a fast runtime, so the application is fast enough and inherently portable (though Paradox was never ported anywhere out of Windows)
For the server part it has the advantages and drawbacks of the BDE, which you surely know.
The downside of Paradox is pretty evident: its development was stopped when it was sold to Corel in 1997. They have done nothing
to add value and keep it in line with modern technologies. They've simply done a bug-fixing work, most notably on the reporting functionalities.
At present time, our Paradox-based applications are happily used by our customers (well, as happy as WE can make them...<g>).
So, where can the DevCo team can add value for Paradox users/developers and for themselves?
IMHO, simply(!) addressing the downsides:
- put modern feature in the form designer keeping it simple
- rework the object model toward the current Delphi one
- consider enhancing the ObjectPAL language, for them is a few more than a child's play :)
- same consideration for the development model
- make the runtime portable and port it anywhere, Mac, Linux, BSD-likes and portable OSes included!!! :))
- rework the BDE part or replace it with current Borland/DevCo technology, enhancing back-end connection capabilities
- and something new: EFFECTIVELY open it to integrate with software object developed outside, the current OLE-based thing is too constrained and cumbersome
This last point is where the integration with Delphi can really exploits each other value in a consistent way
Is there a market? Not huge, maybe some tens of thousands, but significant and with MUCH room to grow, because
they are developers in any skill grade.
Is it worth the effort? AFAIK Paradox is a ten years old product but don't really the state of its sources and resources.
Maybe I'm thinking about the whole thing too much from my own position and perspective, but in Paradox's newsgroups (news://pnews.thedbcommunity.com)
many people thinks in not so different way.
Would be glad to hear some opinions.
Mirco
.
- Prev by Date: Re: To pay or not to pay?
- Next by Date: Re: Any document previews?
- Previous by thread: Re: New features for Delphi
- Next by thread: Re: New features for Delphi
- Index(es):
Relevant Pages
|
|
