Hi Jonathon Blake and all, > F Wolff wrote: > >> It might mean that OOo will have to do basic sentence division, simply to >> see if it roughly correlates with the language boundaries. > > Taking your example, the "What is an inyanga?" should be tossed to > both the Zulu grammar checker, and the English grammar checker. with > each grammr checker treating the material in the other language as an > error, but the material in the language as correct. > > The Zulu grammar checker will break up inyanga, to verify that > everything is correct. It should throw the English back as a > potential error. > > The English grammar checker should flag inyanga as a potential error. > > This is where Joan's suggestion that each grammar checker have its own > color is a very good suggestion. In this instance, the entire > sentence, or paragrpah, is flagged as a potential error --- unknown > words. The underlining would indicate that the Zulu was flagged as > incorrect by the English grammar checker, and the English was flagged > as incorrect by the Zulu grammar checker.
I don't think this having different underlinings is a good idea from the UI point of view. ^^° How much different colors do you want? 5 different colors for 5 different grammarcheckers? And where will you stop this? And what if two of them choose to find the same text part incorrect? Should it have double underlines in different colors?? Also I think the current UI is not up to the task. Please have a look at the possible font effects in the "Format/Character" dialog. Asking for different kind of underlines for the same text at the same time is too much. Since there will already be one color for the spellchecker and at least one for grammar checking (maybe two if we like to show different levels of errors or style errors in a differnt way). That make already 3 colors. Do you really want to add an additional one maybe even two for each grammar checker? I think this will result in an overloaded UI that is going make more users uneasy about grammar checking than it is helpfull. I still appreciate comments from some more people here though. >> indicate the start and end positions of the percieved erroneous part, rather >> than the sentence. > > +1 Sure, only the wrong part (or may parts) of the sentence should be marked. > The frequency of multi-lingual word usage in a sentence corresponds to: > i) the number of polyglots who are expected to read the document; > ii) the type of document; > iii) when in the polyglot fad the document was written; > iv) the subject matter; > v) the subculture; > > In some instances, people won't even realize that the words are from > different languages. [This is how Spanglish originated.] I'm not > sure that having both an English, and Spanish grammar checker will > help, when people write Spanglish. Agreed. I would also guess that about at most 3-5% of documents are multi lingual. Still, from an engineering point of view we have to take care of them as well. Regards, Thomas --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
