On Wed, Oct 5, 2011 at 9:41 AM, Martin Langhoff <[email protected]> wrote: > On Wed, Oct 5, 2011 at 1:12 AM, Chris Leonard <[email protected]> > wrote: >> L10n of long-form content is an area where I think there are some >> excellent bits and pieces, but I'm not convinced that there is a >> really nice end-to-end solution yet. I think very highly of >> FLOSSManuals as a book publishing platform, but was less than >> impressed with it's L10n workflow. > > AFAIK FLOSSManuals software infra is being worked on, and that > involves (at least in part) Douglas Bagnall who used to work with OLPC > on the School Server.
By no means should my *critique* of FLOSSManuals L10n be taken for *criticism* of FLOSSManuals, As both a community and a tool, they have been very good to Sugar / OLPC. The collaboration with them has been valuable and productive and perhaps we can build on that relationship to work with them to improve their L10n tools as they do such an excellent job in other important areas of long-form content creation and sharing. I admit that I do need to look at their tools with a fresh eye, as it has been some time since I have thoroughly investigated the L10n workflow. As for the use of FLOSSManuals as the source and L10n workflow for the Help Activity, sadly, I think the more pressing challenge may be the lack of documentation of the process employed by Seth and his collaborators in the original Help Activity creation from FLOSSManual sources. I'm simply not sure that I would know exactly how to go about taking a refreshed set of FLOSSManual chapters and convert it into a revised Help Activity. Better documentation of documentation creation is a very "meta" concept, but I know from my time in a heavily regulated industry (pharma) that SOP #1 is always the SOP on creating SOPs. >> I am happy that several of the >> Sugar OLPC books have been translated, but these have been time-focused >> efforts requiring a lot of coordination and not amenable to the slower >> accumulation of collaborative work that characterizes Poolte L10n >> work. > > Well... for translating something long-form, I do think you need an > initial focussed translation effort, perhaps followed by piecemeal > maintenance. Prose style and cohesion do need that initial translation > to be done by one person or a small well coordinated group. I completely agree that translation sprints are an important aspect of content creation, but have some lingering concerns about the "ultra-marathon" of maintaining L10n over time and versions. Another non-technical aspect is the difficulty in creating such focused periods of effort to tackle a large project within multiple language communities. We do fairly well with gathering small quanta of discontinuous effort that accumulate over time in Pootle (it is a bit like the advantages that developers see in the distributed nature of git as a source code repo), but it is a much greater challenge to organize concerted efforts by a language team. In any event, I am happy just to be having a conversation about the challenges and to see people thinking carefully about the i18n / L10n process. There may be no single ideal solution simply waiting for us, but it is only by considering the issues and options that will we move towards better and more reproducible processes. Warmest Regards, cjl _______________________________________________ Devel mailing list [email protected] http://lists.laptop.org/listinfo/devel
