| |||||||||||||
From a user...From a user...Posted Mar 22, 2007 1:33 UTC (Thu) by dskoll (subscriber, #1630)Parent article: SQL-Ledger and LedgerSMB: a study in security reporting
We purchased the SQL-Ledger manual about a year ago. It wasn't bad and I wanted to help a developer of software we use extensively. That being said: SQL-Ledger's code is absolutely horrible. It's truly putrid. Nevertheless, from an end-user point of view, it works quite well, and it would be a waste of our time to try to fix it up. We recently switched to Ledger-SMB, and I'm hopeful the code quality will improve. I gather that the Ledger-SMB developers' immediate goal is simply to clean up the worst of SQL-Ledger, but I really hope they give it a thorough rewrite. It should be written using a proper framework like Catalyst; all existing code should be chucked out and only the database schema retained. This would never happen with only one project, but now that there's a competing project... maybe my dream will come true.
(Log in to post comments)
From a user... Posted Mar 24, 2007 18:04 UTC (Sat) by ccyoung (guest, #16340) [Link] not only the code - the UI is imho unusable.
those who use my sites, highly interactive and esthetically pleasing, were horrified when I showed them SQL-Ledger as an alternative.
$900 for a programming manual. $200 for users manual. I wish someone would give me $200K so I could get a good F/OSS accounting system out there.
From a LedgerSMB Developer Posted Mar 31, 2007 18:01 UTC (Sat) by einhverfr (guest, #44407) [Link] I would suggest joining the LedgerSMB community if you are looking for an alternative. At the moment the code and UI are fairly similar but this is starting to change. LedgerSMB 1.2.0 will be released on Monday and will be the last of our releases where we merge security changes and work done elsewhere into the LSMB codebase. From here on out, we will be re-engineering the entire application in place.
I would suggest that one part of the problem with the UI is that the code is horrible, as others have suggested. Hence modernizing the UI is not really possible with the current codebase. While our user interface may not change drastically until 2.0, we are putting in place now the framework to make this sort of change feasible. In fact, after 2.0, it will be quite possible to re-theme the application however one would like using a combination of CSS and templates.
Our new architecture is more object oriented, more modular, MVC-like, etc. and hence will allow for better separation of interface from mechanism. It will also allow for full user-level security to be enforced (which the SQL-Ledger codebase does not).
One final comment. SQL-Ledger 2.8.0 was released yesterday. One of the surprises was an apparent license change (no longer GPL).
LedgerSMB rewrite Posted Apr 8, 2007 20:59 UTC (Sun) by einhverfr (guest, #44407) [Link] We actually have begun the process of a complete rewrite of LedgerSMB. Although we have opted not to use Catalyst or other ready-made frameworks, we have spent a great deal of time evaluating them. We are, however, making increasing use of CPAN modules.
By 2.0, LedgerSMB will be entirely MVC-oriented, but we will be doing a few things differently than other MVC frameworks:
1) The entire data model (or nearly all of it) will be fully encapsulated in the database system. We have opted to do ORM-stuff by mapping object methods to stored procedures, and not doing direct table mapping. This allows us to ensure that we have an ideal database structure. Our current approach allows for dynamically associating such methods, which makes custom reporting something that is really easy.
2) Security can be managed easily via database access controls
3) We make use of existing standards wherever possible (so the templates make use of PI tags extensively).
|
|||||||||||||