My present engagement in the LibreOffice project suffers from lack of time. Thus, most of my ideas stay in a prolonged (but hopefully not eternal) draft stage, without having the time to dive in and to implement even a quick prototype. So you will mostly read "thoughts", and seldom see "results". Sorry. But that's the situation at the moment. The problem is that I tend to search for generic solutions too exhaustively before jumping into cold water and starting to implement.
Personally, I'm a "not-not". I'm not a developer, not a QA specialist, not a doc writer, not a marketer. To be honest, I'm not a specialist at all. I've been working in scientific research in my former life, but that's passed long, long ago.
On the other hand, I'm sort of "curious idealist". I love my family, people, in deed the whole world. (Ok, if they aren't nasty, of course). So that's all you need to know about me for the moment, I think :-)
As occasional contributor I've considered myself as "almost community member" of OOo for several years.
When TDF creation was announced shortly, I first was rather surprised, as with all its underhand planning and sudden presentation of facts it resembled more to a putsch than to a community driven action.
After some days, with all the publicly available statements now, my conviction was growing that it was a good action. Several of the Oracle staff tended to behave like pashas in those times in my eyes. I did not question their individual meritocratic legitimation, I just stumbled over their arrogant-feeling behavior. Thereby they showed - for me, an external, who just looked with his eyes and felt what he felt - that the action to bring forward an independent instance to manage this huge project is overdue.
Therefore, I decided to support the TDF.
With LibreOffice, my position is, that LibO is a necessary experiment, a proof of concept for the TDF and the ideas behind a better community integration. It is for me a prototype of what could be possible if OOo wasn't dominated by Oracle but set up as really independent project.
Therefore, the only possibility I saw (and still see) for me personally is to support TDF/LibO. Although I don't dare to believe that LibO will establish as an acceptable alternative to OOo, it will at least show what a community can achieve. It will show it to the world, but it will also show it like a mirror to the TDF founders, to the community. So after a year or two we will know if this experiment was a real improvement to OOo or not. But even if it shows that it was not, there could be many details and aspects from which we all can learn and which can then be transferred back to the OOo community. If it is willing to learn.
So, that's what I felt more than a year ago :-)
As I'm not native English, I'll write down most of my ideas in German. It's just quicker :-)
Konservativere Marketingstrategie (Marketing/QA)
Ich tendiere zu einer konservativen, QA-basierten Marketingstrategie. Im Moment geht's mir da möglicherweise noch etwas "zu subjektiv" zu.
- Die "empfohlene Version" (für private und Corporate Users) sollte möglichst fehlerfrei sein, neue Features sind in meinen Augen dafür erst mal eher nachrangig.
Einschätzung der momentanen Situation (Anfang-Mitte 2012)
- Die momentane Strategie erscheint mir persönlich etwas zu sehr "Banana-like": Zu wenig QA im Vorfeld, zu viel QA beim Endkunden, die Software wird zu früh empfohlen.
- QA intensivieren (Releases ankündigen, explizit zum Testen auffordern, "Sensoren" implementieren)
- Objektivere Kriterien für die "Empfehlung" bzw. die "Einordnung als stable" festlegen
Man könnte beispielsweise festlegen, dass eine Version (Codeline) erst dann als stable bezeichnet wird, wenn mindestens X % der Regressionstests erfolgreich absolviert wurden (oder, etwas differenzierter, zum Beispiel: 100% der Tests auf Blocker, 99,8% der Tests auf ernste Fehler und mindestens 95% der Tests auf triviale Fehler). Dazu wäre es notwendig, eine entsprechende Liste von Regressionstests zu erstellen und alle Testläufe müssten auf möglichst einfache Weise, um gut Community-Crowdsourcing betreiben zu können, dokumentiert werden können.
Personal ToDo list
Some ideas, what I could do if I had more time...
- [Community] complete this list of website form entries
- [Wiki] categorize uncategorized wiki pages
- [extensions&templates] help translating + enhancing extensions&templates site
- [documentation] continue contributing to User Guides
- [QA] Organize or help organizing a "Bug Review Week"
Finally, some crazy ideas (which probably never will get implemented, at least not by me, so they'd better be called dreams)
- community task manager
- I'd really like to have a public todo list, where every community member can enter tasks, wishes, ideas. Where items can be priorized, voted, tagged, organized (into subtasks etc.), and nicely listed. Status info (idea/planned/started), persons involved, links to deliverables etc. can be added. Similar to bugzilla, of course, but not developer/bug oriented but community oriented. TUTOS (www.tutos.org) is a web app coming close to these requirements - and probably, Drupal is another possibility.)
Ideas, Thoughts, Activities, Deliverables, Crystal Nuclei ...