> On 03 Sep 15, at 07:35, Nick Burch <apa...@gagravarr.org> wrote: > > On Thu, 3 Sep 2015, John D. Ament wrote: >> I'm adding some shepherd notes for ODF Toolkit. I think my main concern is >> lack of mentor participation on the project. Rob Weir (I swear every time I >> type his name I type the d) has taken a mentor role on the project, but >> doesn't seem to be in podlings.xml. I'm not too worried about this, but can >> we add him to podlings.xml (if so, I'll add him)? > > If Rob is happy to take on the formal mentor role, I'd be in favour of adding > him in as an official mentor > > > In terms of the codebase, I've heard rumours that there's a huge pile of > fixes an enhancements sitting in a non-ASF git repo, waiting for the sign-off > to bring it back. Bringing that in would probably help, but there is the > issue that all of this code was developed externally and not here, which > doesn't help with community development. > > > If there's some energy for community development, I would suggest people put > together a list of "easier" things that could be fixed / contributed. Then, > importantly, existing committers should resist the temptation to do them! If > there's a list of easier tasks, then new community members (of which we've > had the odd volunteer over the last year) can have something to get started > with, to get them into the project and the codebase. > > If there's coding energy, I'd suggest working on an implementation of the > Apache POI Spreadsheet APIs. Currently, the spreadsheet components are by far > the most popular areas of POI, and there's a fairly general UserModel SS > interface across .xls and .xlsx. While it won't be easy to implement all of > that for ODF, there's probably enough in common between all 3 spreadsheet > formats that the key parts could be done. With that in place, it becomes a 2 > line change for existing POI users to try out ODFToolkit. That could then > help drive interest and users. There are also a few new POI committers who > might (I haven't asked them!) be interested in helping out a little bit with > this too. > > Nick
Nick, et al., I think looking at the history of the ODF Toolkit would be useful. From what I recall—I was there at its birth, but that was ages ago and I was peripheral to the main event—it began pretty much as a Sun/IBM endeavour and was always meant to be small, focused, and actually effective. I tried to grow the community then, almost a decade ago, but there was little interest in anything but organic growth. In that case that meant growth resulting from specific corporate interests. Outside of the main players, there seemed to be some interest in the ODF Toolkit, but the clamour was not deafening. And as marketing it was deprecated, even when "marketing" meant developer outreach, via the usual routes, it was never likely that there would be clamour. However, the project was also initiated at a time when the ODF had some glister, though its primary strategic goal hd probably been achieved, so there was value there. (ODF has actually continued to grow in awareness and even has gained prominence, which is good; but that prominence has not, I think, meant that there are more developers assigned to working on it, at least not in public spaces. Microsoft’s latest office products do have decent support. This is great, for all, including MSFT.) The granting of the code to Apache by Oracle preserved the essence of the project and probably gave it a boost. But with flagging corporate interest over the years the project has essentially become, AFICT, a good idea maintained by independents or abundantly generous savants, like Rob, who see the abstract value in it. What this means, if I read it correctly, is that if we want to preserve and even advance the project, we really need to put some effort into marketing it. That would entail, at a minimum: * representing the code and what’s there in a way that makes it sound relevant * representing the project at the ODF Plugfests (surely, this has been done? I last went in 2012 or 13, in Brussels. No mention of it then, if I recall. Yes, the toolkit project is not about the format spec. per se, but the gathering is composed of ODF stakeholders.) * Other, Java-related events? Like the one Oracle puts on? Presuming that would be relevant, that is. Alternatively, one could append it to a likely other project. I think now AOO is a better choice. best louis
