>>>>> "Philippe" == Philippe Troin <[EMAIL PROTECTED]> writes:
Philippe> I'm just asking for some common sense. This package has 3 Philippe> different levels of details for the world coastline (low, Philippe> high and full). Are they really needed ? Can't just a low Philippe> can provided with some pointers to where to get the Philippe> additional data ? Yes! Why was this packaged in the first place? Why does this need to be a package? I use GMT myself (and have for years), but I would never DREAM of packaging the coastline data and submitting it to Debian. That makes absolutely no sense whatsoever. This data is already available on the Internet (and from 3 or 4 different ftp servers, IIRC). Wouldn't a better solution be to package a small script that downloads this data from the network and installs it into an appropriate place in the filesystem?? Thus, Debian provides automatic access to this data, without having to actually package the data itself and waste space in its archive duplicating data that is available elsewhere. Philippe> As a matter of facts, I'm kind of opposed to packaging Philippe> "pure data" (this encloses bible-kjv, Philippe> anarcho-syndicalism.deb, etc) because for "pure data", Philippe> packaging is minimal (just dump the file(s) into Philippe> /usr/share/doc and that's it). This doesn't apply to some Philippe> reasonably sized "pure data" like /usr/share/dict/words Philippe> etc... Exactly. If the data is available elsewhere on the Internet, IMHO it is better to package a mechanism for retrieving this data and installing it into its proper place in the filesystem. >>>>> "Chris" == Chris Lawrence <[EMAIL PROTECTED]> writes: Chris> Why not package "pure data" installers, like we do for Chris> non-free software (xanim-modules, realplayer, star-office, Chris> etc.), and provide the "pure data" on a dedicated "data Chris> server" (or data server network)? I think that this is a good idea. Philippe> Does this belong into -project ? I think so. Brian