Friday, February 8, 2008

Scarab should survive

Scarab is designed by Collabnet developers, and developers of Turbine, and at one time was even official bugtracking system of ASF. Developers of Turbine had an idea to reorganize framework with the use of service architecture based on Avalon containers. The new version was to be Turbine 3. And they decided to use the Scarab as a testing ground. As a result, year after year, activity decreased in Turbine, because although the framework was not bad, but but not so good as many modern frameworks. Ultimately, people involved in the development of Turbine 3.0 lost interest and to the turbine and to Scarab (For obvious reasons, ASF abandoned him and he too became Collabnet not needed). But the legacy remains.
Ultimately, by the forces of community the best solutions lay in the basis Turbine 2.4 (which has not yet released). But Scarab was forgotten.
As a result, the architectural Scarab now in a landfill like good ideas. And only through community he is still alive, but all new features are made without understanding the concept of Turbine 3 (because it was not defined), and creates even more junk.
If it made even a dozen such changes, the stability, of an already not high, to be lost.

Scarab quite unique software. At least for me there is no free analogs was found. He focuses not on tracking bugs, but on tracking issues. Problems can be either on demand or request. Adjusted absolutely everything. Scarab allows separate all space by modules and customize (types of problems, fields in problems), each module is a unique way. This enables it to use in a wide range of areas: bugtracking, helpdesk management applications, CRM and so on. And most importantly that combine all of this simultaneously in a single system. All who use Scarab very happy with him, except that he does not develop.

I do not wish that the system would have died. And I shall make every effort to this. What I intend to do:

Priorities (release 1.0)
* To port Scarab at Turbine 2.4. This will lead to the destruction depending on Torque
* Clear debris and a refactoring most functionaly overloaded classes.
* Communicate features which was recently implemented to its logical conclusion Scarab
* Finally there old task to enable relevant search, which is provided by Lucene
* Define XML-RPC API for integration with external programs.
* Change UI. Current critically perceived by users.

Secondary objectives (release 1.1):
* Change ORM. If by that time will be ready more or less acceptable OXM (Object XML mapper), then ported to the XML: DB or XQuery.
* Realise integration with Jabber. Make Jabber tool for discussing applications with simulationosly logging conference talk as comments in issue.

This is the main directions of my activities after passing the diploma.

No comments: