Commons talk:File types

This talk page is automatically archived by MiszaBot. Any sections older than 180 days are automatically archived to Commons talk:File types/Archive 2. Sections without timestamps are not archived.

Archives 2005-06-30 — 2008-10-01 2008-11-18 — now

Tag with sample for sound files

Need help with this. Please insert sample here

WebM

When is WebM going to be supported? The difference between WebM and Theora is basically the difference between VP3 and VP8. WebM can challenge H.264 whereas Theora can't, and because of that eventually WebM will become the open source standard. "VP8 has a better quality per bit rate than Theora." -Mike Shaver, Mozilla VP Engineering.--Brian Dell (talk) 23:24, 16 April 2011 (UTC)

I second that, although I don't know where to ask for more info. Beuc (talk) 13:19, 2 October 2011 (UTC)

bugzilla:30653 → bugzilla:27699. --AVRS (talk) 14:36, 2 October 2011 (UTC)

+1 - KTucker (talk) 00:22, 11 November 2011 (UTC)

Is there a reason .webm / VP8 are not allowable upload formats? - KTucker (talk) 22:21, 14 January 2012 (UTC)

What is a reasonable way to upload a wikiversity pdf fullscreen presentation + source ?

i've started creating a Wikiversity course wikiversity:Special relativity and steps towards general relativity. For people interested in the (plans for) different types of "learning resources" at Wikiversity, see wikiversity:Help:TYPES. In particular, Wikiversity is not supposed to compete with Wikipedia articles or Wikibooks books - see e.g. wikiversity:Help:Article.

The first half of the material for the course consists of:

• original source:
  • latex .tex file (my work), 110kb of text
  • xfig .fig source files (my work), typically 1-2kb of text (tiny!)
  • one-line script (my work) for creating a partial or full screen viewable .pdf file, for a lecture-type live presentation by a teacher
• non-original source:
  • WMCommons image files, mostly .svg, some other formats
  • a few en.Wikipedia static image files
  • one WMCommons animated .gif file
• intermediate files:
  • Commons (+w:) files, mostly converted locally to .eps postscript files
  • .eps files rendered from .fig source files
• final:
  • two pdf files (one for special relativity = SR, one for a few steps towards general relativity = GR)

So my question is, what is the best way to upload the original material and some of the rendered material in a way that makes it relatively wikified - other Wikimedians can edit, check the edit history, discuss on the talk page, etc. - but also not too difficult to recreate the final "usable" pdf files, i.e. what sort of script should i provide and how should it be uploaded? My tentative answers are:

• The .fig files are in principle renderable output as svg, so probably uploading an svg rendered output version + the source .fig text together should be easy - just take a long time because there are many :). Some of them could potentially be used in Wikipedia articles - where they were absent (where i was unhappy with the actual figure, the question of the preferred image to use would have to be discussed on a case-by-case basis at those WP article talk pages). So, upload to Commons?
• The latex source file could be uploaded verbatim as a slightly big text file - 110kb is not so big. So, upload it to Commons as verbatim wiki page content along with the rendered pdf files?
• Uploading the two rendered pdf files to Commons once all the source files are in place should presumably be acceptable on the grounds that it's used in another WMF project, and it's a format that IMHO is optimal for physics/maths type Wikiversity lecture material.
• The script is probably the biggest question. For security reasons, my guess is that a script should certainly not be uploaded as an executable file. The script is presently just a commented-out line in the .tex file, a chain of latex etc commands which can be easily copied/pasted to the command line. However, this assumes that all the image files have been converted to .eps files and placed in (or softlinked to) an appropriate local directory. A minimal script should offer the options of downloading image files from Commons/WP and/or producing .eps files from the .fig files, while not encouraging excessive use of bandwidth (e.g. downloading 40 URLs each time the script is executed without any download delay and repeatedly executing the script after every minor change). My feeling is that what needs to be done in the script is something like the question of a package management system for an operating system, e.g. apt in debian, so it would make sense to have a fairly intelligent, well-developed system and not just a plain shell script. On the other hand, maybe a plain script is enough to get started, and later on a more scalable system could be developed if this approach starts being seriously used in Wikiversity. People interested in improving the package would keep their own copy of a script and comment in/out files that they know have been modified. Each editor would have to be careful not to accidentally include a non-freely licensed image.
• intermediate files: Uploading intermediate files - especially in .eps format - to Commons sounds like a bad idea. On the other hand, reconverting files locally requires some more complexity in the script - i.e. possibly a download plus a conversion rather than just a download. My guess is that these should not be uploaded.

i'd rather get some feedback on this before starting any uploads. Boud (talk) 11:11, 29 June 2011 (UTC) (minor correction: "rendered" was incorrect Boud (talk) 09:29, 6 July 2011 (UTC))

i guess i can follow the User:Nichalp/Upload script example for uploading a script as a subpage. Any more alternative recommendations? Boud (talk) 13:20, 5 July 2011 (UTC)

I am sorry but I find your post rather confusing and way to long (by the way, length of your post is inversely related to how many people will bother to read it). It is unclear if you are wandering about mechanics of uploading large number of files or how to handle different file types. If you are uploading large volume of images I would suggest using Commons:Tools/Commonist before looking into other tools. If you have very large volume (thousands) than a specialized bot might be able to help you, if you ask here. As for file types, you should first read Commons:File types for files supported by commons and other wikimedia projects. The file types you mentioned in your post do not seem to be covered in this list. I would suggest to convert latex to either wiki-text and use directly or change to PDF and upload to commons. I am unfamiliar with xfig but I assume you can export it to SVG. Hope this helps. --Jarekt (talk) 16:37, 5 July 2011 (UTC)

Sorry that you found my post confusing. i stated in the section title and in the first sentence that this relates to a Wikiversity course and linked to it. i guess i assumed that the reader would be familiar with the standard way of producing scientific (physics, astronomy, mathematics) documents, i.e. w:LaTeX, so i didn't think of stating overtly that the final files are produced in a chain from original source files and non-original source files, through intermediate files, out to final files in pdf form. It's somewhat analogous to wiki source text which is output to final html form, but a better analogy is the difference between a .tar.gz software source package and an executable binary created from the .tar.gz file. In the same way that it would be pointless to expect or hope users to edit the binary executable directly, it would be pointless to expect or hope Wikimedians to edit the two final pdf files for the course. Instead, the source files - maybe also intermediate files - should be available for editing, while the final pdf files should be available for people who just want to use them.

Yes, i did read the article which this talk page is attached to before adding my question. I didn't see it answering my overall question.

Regarding your specific responses:

original files - .fig files: It sounds like uploading the output .svg file along with the .fig source as wiki-text source associated with the file would be reasonable, so that people can easily edit the .fig source and then upload the new output .svg file, or alternatively edit the .svg. On a quick test i made, the .svg format file is 3 times the size of the .fig format file, because of all the xml-like tags. Xfig does not seem to import .svg, only export it, but if people edit the .svg version, a discussion of tools to use and preferred format could take place on the wikipage of the file being worked on.

original files - latex/full package/pdf: converting w:LaTeX source to mediawiki-text and then from mediawiki-text to a fullscreen presentation in a browser is an interesting idea. i suspect that the conversion would only be semi-automated - it would probably be easier to write directly in mediawiki-text, using internal and external wikilinks, <math>the math tag</math> for the main content of the what-was-LaTeX-source, and the [[File:filename.fmt|pixel width etc]] way to embed Commons figures (image files). This would need a mediawiki-text to fullscreen presentation convertor, and it seems from this D'Arcy Norman blog post that this 'Wikipedia Presentation' script can be installed for use with greasemonkey, and then be used for viewing a mediawiki-text page as a fullscreen presentation. Some problems: the script seems to be non-free, since no licence is stated, there seem to be some bugs, and it doesn't seem to have been updated for five years.

Do you or anyone know of a free-software mediawiki-text to fullscreen slide presentation converter that is actively maintained? Using a non-free script would defeat the whole point of contributing to WMF projects: we don't want someone to come along later and tell us we can no longer use/modify/redistribute the tools we use.

In any case, in the short term, uploading the latex source + a script (for creating the pdf file from all the source files) as verbatim content in wiki-text that is associated with an uploaded .pdf file is probably the best approach, along with a recommendation that the long-term format might better be mediawiki-text + a free-software converter to fullscreen slide presentation format. As long as the licensing is clear, anyone who has the time to do a conversion will be able to do it.

Boud (talk) 09:29, 6 July 2011 (UTC)

Please add new formats: KML, FLA, ODP & OO2

Please allow uploads of KML, FLA, and OO2 filetypes. --SJ+ 18:04, 9 August 2011 (UTC)

Comment added: https://bugzilla.wikimedia.org/show_bug.cgi?id=2089 --SJ+

What to compile FLA with, and what into? If it can be compiled into a free format accepted at Commons, but not using free software, check Commons:Restricted uploads. --AVRS (talk) 21:15, 9 August 2011 (UTC)

Epub?

What (if any) are the thoughts/plans on adding epub files? Jbarta (talk) 19:07, 14 September 2011 (UTC)

Where are the archives?

The first post on this page is 16 April 2011, and the last post in /Archive 1 is dated 1 October 2008. Where are all the posts in between? — Edokter (talk) — 12:51, 24 December 2011 (UTC)

• Fixed. The new archive page name had been set to “20” instead of “2”, and the link still pointed to “1”. Added a box with links to both pages. --AVRS (talk) 16:02, 25 December 2011 (UTC)