Manual talk:Importing external content
Martiniturbide (talk) 20:19, 13 June 2014 (UTC) There is a dead link " http://toolserver.org/~diberri/cgi-bin/html2wiki/index.cgi" html2wiki. It is very bad to lose that site. I hope we find a replacement.
- The pbwiki backup format is so poorly defined that the best I can say is "the script works on all my pbwiki data". It's a quick'n'dirty revamp of someone else's quick'n'dirty script.
- pbwiki is defunct now, and has been for a few years. Hopefully nobody needs to do this any more.
- Now that my pbwiki has been nicely imported in mediawiki, I don't intend to maintain the script... or having anything to do with pbwiki ever again.
So it's probably not high quality enough or useful enough for the manual, but I'm mentioning it here in the hope that nobody else has to waste as many hours as I just had to -- Treer (talk) 08:39, 27 March 2016 (UTC)
The link is dead.
Merging with Commons page
Commons:Convert_tables_and_charts_to_wiki_code_or_image_files is a far better written page and basically who's editing it there should come and edit it as part of the formal manual. It's long past time to identify major gaps in mediawiki importing.
Being able to import whole directory listings & keep them synchronized should be simpler given LDAP support & HTML support than it is. Seriously are we telling people to write cron scripts that go and see if directories have changed, when all they want is to embed a list of folders from some source in mediawiki? This should be in the standard build and support SMB, NFS, whatever type of folders. Use of mediawiki would skyrocket if it was easy to use as a front end for media libraries, things like Emby or Kodi.
The introduction text now mentions a workaround for this, which is to include HTML links so the browser can display those pages.
The live mirroring of GetWiki 1.0 etc was not supported because of load, but that's too bad, because it just meant that non-editable mirroring proliferated and spam sites benefitted. It's long past time to revisit putting nested/embedded live pages into basic mediawiki, in part because any intranet will want them, e.g. a support page for a product that has a parallel version for the customer support people internally listing things they can't tell the client about, then a deeper version for the developers on how to fix it. A workaround for this - importing XML regularly & then over-writing it with the smaller deeper more private wiki content (also in XML) is mentioned briefly.
The manual should have a standard tag for workarounds so we can easily find them all & update those that are no longer relevant. It's really too bad we're not using Semantic Bundle yet because that's the perfect way to do this... any other suggestions?