A very generous offer indeed! My own SLE and Magnus' WYSIFTW are indeed the most active projects, so that would be a good bet. Actually, for me the timing is just right, as I'll be working on a paper about this editor for a while, so it'd be cool to have someone(s) continue the project. If one of your researchers has a brilliant idea on how to do this right, that would obviously be really valuable too.
A lot of things Magnus mentioned apply to my project too: * Improving detection algorithms, i.e. better sentence-level editing (perhaps using an external language recognition library), better detection of other elements. Keep in mind that the editor excludes anything it doesn't 'understand', so this is a nice fallback, you don't have to write a complex parser that detects a lot of stuff at once. * Cross-browser/platform/device compatibility (think mobile, touchscreens, etc.) * Usability testing (the more the merrier!) * Verifying detection coverage (Which % of the wikitext is editable) and quality (Wikitext -> Adding markers -> MediaWiki parser -> Removing markings -> Wikitext??) Checking this on a large number of pages. * Test suites (again, the more the merrier, but only for parts of the code and interface that are considered stable!) * Lots of implementation details: embedding the (current) editor toolbar in the textboxes, making sure (a fair percentage of) gadgets still work with this, and handling unusual cases like edit conflicts, etc. Perhaps it'd be good to have a (video or IRC?) conversation with you, your developers, people from the Foundation, and people from the specific projects you want to contribute to. Again, really awesome that you guys want to work on this! :-) Best regards, Jan Paul On 19-Jan-2011, at 9:55, Magnus Manske wrote: > On Mon, Jan 17, 2011 at 1:47 PM, Panos Louridas <louri...@grnet.gr> wrote: >> Hi, >> >> At the Greek Research and Education Network (GRNET) we look at the >> possibility of contributing to the development of WYSIWYG editor support in >> Wikipedia. We understand that considerable work has already taken place in >> the area, e.g.: >> >> * http://www.mediawiki.org/wiki/WYSIFTW >> * https://svn.wikia-code.com/wikia/trunk/extensions/wikia/RTE/ >> * http://www.mediawiki.org/wiki/User:JanPaul123/Sentence-level_editing >> >> We therefore think that it will not be productive to reinvent the wheel over >> here. >> >> Our contribution can take the form of providing developers that will devote >> part (or all) of their time for some months in 2011. We welcome any comments >> and suggestions on how we could push this forward, and in particular: >> >> * Specific tasks / components that need to be designed, developed, >> optimized, etc., and estimates of effort and timeframe. > > Hi Panos, > > a very generous offer! One I would like to take you up on, for WYSIFTW > as you no doubt have guessed :-) > > WYSIFTW is approaching feature completeness, as far as wiki markup > parsing is concerned, and improves on usability as well. (just try the > new "floating context hover boxes", in lack of a better name, that I > added last night, wich come up when you hover over a template or a > references, for show/hide and rendered preview, and the new optional > rendering for templates as a key-value-pair table) > > For support later this year, tasks would include > * increase parsing performance (mostly post-parsing steps, focusing on > DOM lookup and manipulation) > * improve editing usability (cut/copy/paste, better specialised > dialogs for images, table/row/cell properties etc.) > * usability testing (I'm using up volunteers fast ;-) > * creating a test suite (to make sure that changes don't accidentally > break anything) > * general compatibility testing (find pages that parse/unparse > wrongly, and patch the code accordingly) > > I like the sentence-level editing function, but once I add > section-level editing to WYSIFTW, these two will start to converge. > I'm curious which of these will be more suited to small fixes and > adding single sentences/references etc. > > As for RTE, I know little about. Apparently, it is not suitable for > Wikipedia in its current form. From brief looks at CKeditor, it might > be quite some work to make it behave nicely around parsed wikitext, as > used on Wikipedia. > > Cheers, > Magnus > > _______________________________________________ > Wikitech-l mailing list > Wikitech-l@lists.wikimedia.org > https://lists.wikimedia.org/mailman/listinfo/wikitech-l _______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
