Management of our information


#1

I am currently trying to find somewhere to deposit the Inkscape SVG files from the EMP–E poster. These are licensed as CC BY 4.0 and the information is embedded in each SVG file. My best option, in this case, is probably the openmod wiki.

But our information management system is frankly a shambles. As best as I can tell, the following applies:

Wiki

The openmod wiki should only accept Creative Commons CC-BY permissive license and CC0 public domain waiver files uploaded by the author (in contrast, Wikipedia has a complicated registration system in which the author can confirm their license by email and third party can undertake the upload). The upload menu currently offers a CC BY-SA (copyleft) option, but this is wrong in our case (probably a hangover from the Wikipedia origins of the underlying wiki software).

Most open formats are valid, including but not limited to (I would guess): JPG, PNG, PDF, SVG (including Inkscape), XCF (GIMP), WebM (video), ODP, ODS, ODT (the last three being LibreOffice). But proprietary formats are barred, including: DOC, DOCX, XLS, XLXS, PPT, PPTX (all Microsoft).

Revised versions of files can be uploaded to overwrite historical versions, with rollback also possible. A commit log is automatically maintained.

Forum

Our internet forum currently accepts only JPG, PNG, GIF (all raster graphics). That excludes PDF and SVG, which probably cannot be satisfactory rendered.

As best I can tell, the forum does not currently carry a Creative Commons BY license, therefore normal copyright applies to all text and content (unless graphics files themselves carry metadata to the contrary). That should be rectified, particularly given that all contributors can still be contacted. Indeed, text from the forum cannot be transferred to our wiki or to Wikipedia without rewriting (which must extend beyond “close paraphrasing”). (The same issue applies to traffic on the mailing list as it happens.)

Regarding files, I am not sure if file overwriting is possible on the forum, and, if so, whether a commit log is maintained. Does anyone know?

Google Docs

While convenient, the use of so-called editable wikiposts on the forum should, in most cases, supersede the use of google docs. Especially with the new polling functionality being established by Tom Brown.

If I am not mistaken, google docs are only viewable by those registered on the openmod wiki, while forum postings are publicly viewable. (The forum does not accept anonymous edits to my knowledge, but the Discourse server can be configured to allow unregistered editing.)

The licensing status of a google doc is rarely, if ever, specified and so defaults to standard copyright.

Other storage

Other file servers and websites are used for openmod files. For example, the neon site carries files (one is a zipped collection of PDFs) from the first Berlin workshop and the FIAS file server carries files from the Frankfurt workshop. These are subject to linkrot (as was recently seen in the case of neon, but quickly fixed, thanks Lion). Moreover presentations residing in zip files cannot be hyperlinked. Other examples of files on foreign severs may also exist. Files on foreign servers are difficult to archive and backup.

GitHub

Is this a useful place to work on shared documents, say Inkscape SVG (which is also natively text)? Would GitHub also be useful as a repository for graphics files, for example?

Does GitHub make sense as a PDF archive (I rather think not)?

GitHub allows an open license file to be added and automatically records it presence.

File metadata

Metadata should, in my opinion, be written into uploaded files wherever possible.

The cross-platform metadata utility of choice is exiftool. exiftool writes to PNG and PDF, but you need to know the relevant field names first. Alternatively, video metadata can be added, edited, and viewed with ffmpeg, but again a knowledge of the supported field names is required.

Solutions

Solutions are not so easy to come by. Possibilities include:

Technical

  • the openmod forum could be configured, if possible, to support more file types
  • the openmod could add a file server to its suite of internet services, to circumvent the restrictions imposed by the wiki and the forum
  • the openmod website could become a de facto file repository

Legal

  • the openmod wiki should be modified to remove the CC BY-SA upload option
  • the openmod forum should embed a CC BY license and a notice that all contributions are thus licensed

But ultimately, openmod needs a coordinated policy on the storage of openmod-related files, be they graphics, presentations, videos, and also work-in-progress documents, SVG, TeX, and similar files.

The forum licensing issue is particular topical. The lack of a suitable CC BY license means material cannot be transferred to our wiki or to Wikipedia, for example.


#2

Forum

Okay, I just found the license notice for the forum, buried here. And it is a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported (CC BY-NC-SA 3.0) License. Not what we want and incompatible with the openmod wiki, Wikipedia, and most other downstream use cases, including open access publishing. Moreover, there is no license confirmation process when uploading graphics files.

Should we contact our contributors and change the license to CC BY 4.0?

And should we make the copyright notice more visible?


#3

Hi Robbie, that licence wasn’t a conscious choice, it’s just the default. Your suggestions make sense.

Could you export the list of emails from the admin interface and give everyone a week to object to changing the licence?


#4

Hi Tom, all

All 31 contributors have been emailed with the proposal to change to a Creative Commons BY 4.0 International license and asked if they disagree. The email was titled “openmod forum: proposal to change to Creative Commons BY 4.0 license” and a copy was also sent to the openmod mailing list for general reference. The deadline for objections is Monday 23 May 2017 and I will be collating these. Further debate can also take place here as required.

We should also make the license is visible on the foot of each page. Perhaps this is a configuration issue. I will see if this is feasible. Otherwise it could go on the opening page.


#5

Today is Monday 22nd May and no comments on the license issue.

As I discussed with Robbie face to face last week I am happy with your suggestion to change the license of the forum.
On the other hand I am very unhappy with changing it using the suggested procedure!
As far as I know all existing (forum) material can only be relicensed by the authors/ copyright owners. Use the “reply or die” method is not correct in my point of view.

I think we need an active statement (check the box that you read the agreement) of every single user. I know this will slow down the process, but openmod should take this issues serious.


#6

Ludwig is correct. The better process is to solicit consent. If no one suggests otherwise, I will reissue my original email on Wednesday 24 May 2017 with this new procedure and collect the responses personally. I don’t know much about setting up some kind of check-box arrangement (as Ludwig suggested) on this forum, Google Forms, or elsewhere. But someone else is welcome to do so prior to Wednesday.

Those who need to agree comprise: Tom Brown (already consented), Berit Mueller, Robbie Morrison (take this as consent), Stefan Pfenninger, Ludwig Hülk (consented in previous post), Bart Wiegmans, Tim Tröndle, Lion Hirth, Bryn Pickering (consented in following post), Magdalena Dorfner, Lukas Wienholt, Julian Krüger, Christian Brosig, Jonas Hörsch, Michail Vercoutter, PowerWatch1, Iain Staffell (already consented by email), Jarrad Wright, Martin Klein, Frauke Wiese, Wided Medjroubi, Mathis Buddeke, Editha Koetter, Giulia Garegnani, Fadi Bitar, Paula González, Niklas Wulff, Uwe Krien, Randy Long, Tue Vissing Jensen, and Daniel Huppmann.


#7

Not sure the easiest way to go about retrospectively getting permission, but you have mine!

For future sign ups, it is possible to make it more explicit by having a checkbox at signup which is required:

“field type” can be “confirmation” and “required at signup” can be checked. The wording of the field name/description will have to be up to someone who knows more about that sort of thing than me!

As to allowed file uploads, it is actually possible to open it up to any file extensions we so choose. That just requires a change in the admin panel. Each file has a unique hash, but there is no easy way of finding a list of them and I think unassigned files (e.g. attachment was updated, leaving the old file referenced) might get deleted after a time.


#8

Hi Bryn / good suggestions, how about:

Licence confirmation on registering

  • field type : confirmation
  • field name : “licence-consent”
  • field description : “I agree that my contributions will be published under a Creative Commons Attribution 4.0 International (CC BY 4.0) Licence”
  • required at signup : true

Upload file types

Should we add PDF documents, hence *.pdf? Does the software really delete unreferenced files after a time, that seems as if it should be a site policy and not hardcoded?


#9

That’s sorted:

You can’t ‘Create New Account’ without having checked the box. We should then also change the tos to match that. Are you able to do that @robbie.morrison?

I’ve added PDF and SVG as allowed upload extensions. Currently, the ‘delete orphaned links’ option is ON. I guess the reason is that the information is stored wherever @tom_brown is hosting it, meaning there is a limit on space. Also, (and this is the reason the option exists) there is the concern that it can be used to host illegal content. You can upload a file, cancel your post and still have a link to the file that doesn’t exist anywhere on the site. You can then share that link. I’m not sure what the best option is here!


#10

Thanks for taking care of this! Sorry, I’ve been a bit busy lately. There is currently 18 GB of space left on the server, so this shouldn’t be a problem. Technical details on the server can be found in my mail to the list on 27.01.2017, but basically it’s this:

https://www.hetzner.de/hosting/produkte_vserver/cx20

I have no opinion about the “orphaned links”.


#11

Hello @brynpickering, all

The Terms of Service section on licensing now refers to the Creative Commons Attribution 4.0 International License. The revised process for seeking consent for historical postings is covered in this mailing list thread. If existing participants object or fail to respond, then their historical postings will be footnoted with the old license in due course. As Bryn explained, people registering for the forum in the future will be required to explicitly consent to the new license (thanks).


#12

Hello @tom_brown, @brynpickering all

On the face of it, I don’t think it good practice to delete a file when all references to it in postings are removed. That can easily happen inadvertently. Regarding the malicious use issue, participants are required to register before they can edit and upload. Perhaps we should strengthen our screening procedure. Most people who wish to legitimately join this site will have some traceable background in energy modeling, one would imagine. Or is the admin overhead for checking credentials too high?

The ability to upload PDF and SVG files is a good move (thanks Bryn). But we should still think about a file repository where we can store lead up work for presentations, posters, and other documents. Any thoughts?

An allied question, does the Discourse software allow for new versions of an uploaded file to replace earlier versions (as Wikimedia Commons does)? I am reluctant to experiment on a live site.

By the way, I occasionally download an archive of the forum from time to time, as a last line of defense.