Dear Axelle,
Hello,
sorry for not replying to your mail. I am really busy for more than a month now and I am looking forward to July to finish the most urgent projects and probably do some vacation. If you could sent me the changes, I will review and add them to the distribution.
Gantt charts
I really thought a lot about gantt charts recently. Although I cannot explain my hesitation precisely, I somehow feel that gantt charts are not what I would like to support in streber. I learned the hard way, that implementing a feature is much easier than keeping it free of errors during the future development. In this way, streber is a personal project for which I have not enough resources of managing features I do not need urgently. I did this once with mysqli-support and it turned out to be the worst mistake in streber's development (beside choosing php as programming language). So if you
really need gantt charts, maybe parallel development would be a good idea. I am perfectly aware of that forking is a very hard decision but I just do not have enough energy to implement AND organize a complete project management system. Esp. if I do not see the personal benefits of its features. My core wishes for the next versions are following:
- Clean up Back end. Rewrite database abstraction layer, drop mysqli support, provide flexible filters and an API for extensions.
- Clean up the integration of ajax functions.
- Further tune the interface for managing tickets, documentation and efforts.
- Implement very basic functions for handling invoices.
Learning for the past
I am really hesitant with new features until the back end gets more stable. In some areas the system already got too complex. Instead of implementing new features I am thinking about making radical changes to the system:
Probably replace/extend folders with tags
Folders are excellent for documentation and small projects, but they soon get overly complex and clutter the actual work tasks. Learning from
Trak might be a good thing here.
Reduce the right system
The user right systems is way too complex. This is not only a problem to maintain but also to learn. I am thinking about reducing the pub levels to "Published", "Normal" and "Private". To differenciate weather items are writable or not is not necessary if anything can be redone. (I already hear the screams).
I know that this text will probably not help to increase streber's look as a serious business application. I am really sorry for that. I wish I would have the resources to put more energy into the project. Well, sometimes open source has its drawbacks.
yours sincerely
pixtur