On 30/11/08 22:54, Bill McCarthy wrote: > Hello Vim Developers, > > I've written on this subject before. IT ISN'T A CRITICISM - > just a suggestion. > > There have been several updates to the runtime files, > including eval.txt, options.txt and spell.txt. The first > two are already in subversion. > > However the "new" files are a mixture of new material not in > the "old" files and older stuff than in the FTP files. > > ============================================================ > Whenever there is a need to update a runtime file by > patches, the patch should be built between the latest > CVS/SVN file and the latest version of the file. That > latest version of the file should be immediately posted to > the FTP site (including the DOS version). > ============================================================ > > This way we wouldn't have the latest FTP file showing a > missing argument in glob() and '100 as the default 'vi' - > but the released patch seen in CVS/SVN and this list shows > the extra argument in glob() but only '20 as the default > 'vi'. >
It is debatable whether the rare runtime file patches included in patches/7.2/7.2.* should build on each other (as they have until now) or on the latest runtime files including changes not published as separate patches. However IMHO the FTP, CVS and rsync repositories (at least) and the SVN one (as soon as it is updated, but it is known that it lags behind, sometimes by as much as a week) ought to reflect ASAP (preferably on the same day) the "runtime changes published as patches", because it is distressing to see patch apply a runtime file patch, only to see it negated by rsync a moment later. Alternately, there could be "some official place" where to check when the rsync/FTP/CVS repositories were last updated (and what id the lowest-numbered patch, if any, which modified runtime files afterwards), in order to avoid undoing the latest 7.2.* runtime patches by syncing with an out-of-date runtime set. (Currently, since I don't know when the rsync repository is upgraded, I sync with it quite frequently; the advantage is that it deletes spurious files added by patch in the case of rejects, but the disadvantage is that it often undoes the latest published patch(es) to runtime files. AFAIK, glob() has exactly one argument (according to eval.txt, 2008-Nov-02) and the 'viminfo' default starts (in 'nocompatible' mode) with '100,<50,s10,h (according to options.txt, 2008-Nov-10), both of these being part of the "current" set on the rsync repository. Slightly (but not wholly) OT: Bill, I know that your extra float functions' help has of course long since been erased by syncing with the "official" doc/eval.txt. Maybe it would be worth while to publish it separately, and with a different filename, as something which could be dropped into $VIM/vimfiles/doc/ in order to avoid conflicts in the future. Best regards, Tony. -- If you're right 90% of the time, why quibble about the remaining 3%? --~--~---------~--~----~------------~-------~--~----~ You received this message from the "vim_dev" maillist. For more information, visit http://www.vim.org/maillist.php -~----------~----~----~----~------~----~------~--~---
