I don't know how to implement a font substitution algorithm that makes sense. The problem:
fonts used in windows are not usually available on others systems. When you try to open on linux an abiword file that comes from windows, it will have "Times New Roman", "Arial", etc fonts, but the linux guy doesn't has these fonts. If we substitute these fonts for something else, we have to indicate it in the GUI, ie. if we change Arial for Helvetica, we have to show in the font family combo box the Helvetica name (as this is the font the abiword is using to display the text). This substitution should be a "only visual" one (only the string in the GUI is affected, but the doc is still saved with "Arial"), or a "hard" one (the file is saved with "Helvetica"). If the substitution is "hard", what if the user makes a silly change, saves the doc, and send it back to the windows user? The windows user will open it, and abiword will have to do another substitution from "Helvetica" to something else (supposing that the substituted font will be "Arial" is wrong, the windows user may very well have Helvetica installed). If the substitution is "only visual", say that the linux user adds a new paragraph in "Helvetica", it will be saved in the doc as a "Helvetica" paragraph, while the others paragraphs will be still "Arial" (and the GUI is saying that all them have exactly the same font). That's a real problem. Say a peer makes half an assignment in windows, using "Times New Roman", and that I finish it in my linux computer, using "Nimbus Roman" (that, AFAICS is the same font used by my windows peer). I send back him the assignment, and... surprise! half the assignment uses a font, and the other half uses a different font. That's not good. Anybody knows what should I do to solve this problem? Cheers, -- Joaqu�n Cuenca Abela [EMAIL PROTECTED]
