[Kexi-devel] About GSoC

[Jarosław Staniek]

Hugo Parente Lima said the following, On 2006-05-09 13:30:

> Hi,
> 
> I read your comment on my google summer of code project submission, I will 
> try to answer the questions.
> 
> 
>>>Expected features:
>>>- Database modeling capabilities (create tables, relations 1:1, 1:n,
>>>n:n, regions, labels).
>>
>>Note: I see duplication here. In KDE there is at least Kexi that tries to
>>implement this. Please do not consider starting this from scratch.
> 
> 
> I think that Kexi is a data oriented database program, I plan to do a 
> structure oriented database program, I have only the version 0.8.1 of Kexi 
> installed on my system and I cant find (maybe because I'm lazy) how to 
> visualize the project database diagram (the main objective of my future app).
> 
> 
>>>- Reverse engineering (Generate the database model based on tables
>>>already created in a database).
>>
>>I'd be interested in this if developed with cooperation with the Kexi
>>Project. Eventually take a look at Umbrello, IIRC they played with reverse
>>engineering?
> 
> 
> Yeah, they have reverse engineering, but for generate UML diagram based on 
> C++/Java/PHP/etc source codes.
> 
> 
>>>Kexi is not in kdesdk, because it is a MS-Access like tool, not really
>>>suitable for database modeling.
>>
>>To date. Note that by making your tool very specialized you'll end up
>>supporting only one db backend, i.e. a tool like MySQL Query Browser. Kexi
>>is plugin-based, and alrady have db abstraction and introspection into the
>>db apatiblities (KEXIDB), all this is what you may need to finish the
>>project.
>>
>>I have nothing against mentoring the project if you're not going to start
>>from scratch.
> 
> 
> I agree with you, start from scratch is a lot of work and I perfectly 
> understand the preference for help applications against creating new 
> applications from scratch.
> 
> My plan is to start from scratch (or semi-scratch, since I already have small 
> pieces of code done) because I dont see it as duplication of code... I see it 
> as a complement to Kexi, Kexi is data oriented with focus on any desktop as 
> end-user, my proposal is database-structure oriented with focus on the 
> developer as end-user.

Hello
Could you please explain what's the difference betwen data-orientation and 
structure-orientation? Kexi is not a spreadsheet too, it makes heavy use of 
relationships and that's structure, isn't it?

If you mean the logical view versus the physical one, it's also something we 
want to have largely used in Kexi. For example, support for many-to-many 
relatonships generated automatically is in progress.

Summing up, at Kexi we could benefit from any features not deeply connected to 
a given app, so I am looking forward to see your work. Good luck!

-- 
regards / pozdrawiam,
  Jaroslaw Staniek / OpenOffice Polska

Sponsored by OpenOffice Polska to work on
* Kexi & KOffice: http://www.kexi-project.org | http://koffice.org/kexi
* KDE3 & KDE4 Libraries For Developing MS Windows Applications:
                   http://www.kdelibs.com/wiki
See also:
* Kexi For MS Windows: http://kexi.pl/wiki/index.php/Kexi_for_MS_Windows
* Kexi Support:        http://www.kexi-project.org/support.html