VAFA KHALIGHI writes: > Hi, I am building a Persian format of LaTeX and going to upload that to > CTAN, the format name is farsilatex. I just wanted first to contact you guys > to see if there is anything that I should consider? that's a difficult question you pose, especially if (as I suspect) nobody on this list really knows what the Persian typesetting needs are. As a general advice I would say that building a separate format should be the last resort taken if any other option fails, i.e., - building a babel package - building a general package to add to a document (assuming that babel is too much targeted for "latin based languages") > this format will be the modified version of latex2e format that is suitable > to meet Persian typesetting needs. what are those typesetting needs? writing a short paper on requirements and outling what isn't being managable with the current standard setup should be the first step in my opinion a) because that would enable others to point you at the right resources or possibilities (assuming some exist) b) would help to understand us what would we need to support in a future kernel to allow such support (assuming that within the current framework some things are not possible) I would also suggest to take that paper then to a wider audience than this list (for example c.t.t) > If there is no comment, then my assumption will be that there is no problem > and I continue the project. well, my main concern is that a separate format has a number of drawbacks, to name a few - maintenance is problematical, ie any fix/changes to the official version needs to be separately managed - user base will be small, ie an addon package to the standard can normally be obtained and installed by everybody (if not already part of some distribution) for a different format this is not the case normally - certain functionality as offered, for example, by babel would missing, eg the ability to easily switch between languages in a document there is nothing really stopping you from going ahead in the direction you intended (and it may turn out that it is in fact the only decent solution eventually) but I would start not with it for the reasons above. Instead document what the needs are, which of them you find/think are not doable within std LaTeX and then spawn discussion on that regards frank