Hello Brianna,
Here is some feedback regarding the Greenspun project, as seen by graphists from fr.wikipedia (which I will call fr.graphists in this email). I C/C this email to the Greenspun mailing-list, I guess other people could be interested.
Regarding the way illustrations would be done :
* The images should have a very high quality: they should be reviewed both by good graphists and by people who know the topic. Images should be assessed both regarding their accuracy, their use of colours, their possible internationalisation, etc. : fr.graphists said that payment should be done after quality has been acknowledged. High quality means using good sources: these sources should be sought beforehand and approved by people involved (zoologists for animal diagrams, for instance). This is a key point for making good diagrams. The source(s) should be quoted/acknowledged on the image description page.
* For each illustration, there should be a version with numbered labels and a blank version, on top of versions with English or other localised labels. The numbered version allows easy reuse on small Wikipedias (without graphists), the blank version allows interactive diagrams using ImageMap for instance.
* Each illustration should be translated in as many languages as possible. This part can be done by volunteers, but the image has to be advertised to them.
* Free software should be used if possible. By free software, they mean of course Inkscape. There are several reasons: first, it is free :-) so in the same spirit as Wikimedia projects; this software can be used by anyone without the need to buy a license; it produces better SVG code than OOo or Illustrator if you don't have the good plugin. Fr.graphists said that this may be a problem with professional graphists, who tend to use more professional tools.
* Maybe it goes without saying, but all illustrations should be in SVG format. PNG is much harder to translate / adapt. ANother good reason to use Inkscape.
* The selection process may be private (ie not open to anyone) as competition for money should be avoided. When images are produced, they should be made available to everybody for review, criticism, and inspiration !
* Some fr.graphists were anxious about a possible "en.wikipedia-centrism", meaning that some illustrations would be useful to the en.wikipedia project but maybe not to others. This can work the other way around too (Wikipedia vs. Wikibooks, etc.). This also means that they should be easy to translate (maybe problematic when jargon or technical language is used). Maybe illustrations for "core topics" should have higher priority as well...
Finally, they also said that it would have much more impact to teach graphists to use their tools (Inkscape / GIMP / etc.), to create tutorials, etc. than to pay for a few hundred pictures. This would be much more in the spirit of Wikimedia projects. People could be paid to teach new graphists, to write well-illustrated tutorials, etc., which should also potentially generate less frustration between paid and volunteer graphists. Another way of avoiding this "frustration" is to make sure the diagrams produced in the Greenspun project have a really high quality.
The graphists had some general suggestions for the project (which may or may not fit into the objectives) :
* create commonly accepted standards or recommandations for maps and diagrams. The French Graphic Lab already uses cartographic standards extensively (see [1])
* help User:Sting to finish his cartographic tutorials, and translate them (see [2])
* Make work together a macro-photographer and a graphist, to produce good quality diagrams of insects and this sort of stuff (see [3] for instance)
* Get programmers to create a cartographic software for Wikipedia, including standards, etc.
* Help the development of Inkscape, and the SVG rendering software in MediaWiki (patterns, masks, some issues with arrows, support for animated SVG...)
That's it, I think. You may read the full comments on [4]. Cheers, Rémi Kaupp User:Korrigan
[1] http://fr.wikipedia.org/wiki/Aide:Cartographie#Conventions [2] http://fr.wikipedia.org/wiki/Utilisateur:Sting/Brouillon/Carte_topographique... other pages from http://fr.wikipedia.org/wiki/Utilisateur:Sting/Brouillon [3] http://commons.wikimedia.org/wiki/Image:Sch%C3%A9ma_abeille-tag.svg [4] http://fr.wikipedia.org/wiki/Discussion_Wikip%C3%A9dia:Atelier_graphique/Ima...
2007/12/7, Rémi Kaupp kaupp.remi@gmail.com:
Hello Brianna,
Here is some feedback regarding the Greenspun project, as seen by graphists from fr.wikipedia (which I will call fr.graphists in this email). I C/C this email to the Greenspun mailing-list, I guess other people could be interested.
Regarding the way illustrations would be done :
- The images should have a very high quality: they should be reviewed both
by good graphists and by people who know the topic. Images should be assessed both regarding their accuracy, their use of colours, their possible internationalisation, etc. : fr.graphists said that payment should be done after quality has been acknowledged. High quality means using good sources: these sources should be sought beforehand and approved by people involved (zoologists for animal diagrams, for instance). This is a key point for making good diagrams. The source(s) should be quoted/acknowledged on the image description page.
Sure - we should force giving source in such images ;)
- For each illustration, there should be a version with numbered labels and
a blank version, on top of versions with English or other localised labels. The numbered version allows easy reuse on small Wikipedias (without graphists), the blank version allows interactive diagrams using ImageMap for instance.
Disagree. It's trivial to remove labels from labeled image. Going further labeled (with texts) images are much less usable than numbered. It's cause very low resolution of computer displays - text labels eat a lot of screen resolution but don't provide any information to image itself. It's why some people vote oppose on FPC - usable width of image on Wikipedia is up to maybe 600px, and average is about 300, so if image has 50% of width used for empty margins/text labels we loose 50% of effective resolution. Of caurse it doesn't affect printing where we have at least 10 larger resolution for images :) but it worth remembering that text labels are just sexi, but not good at all.
- Each illustration should be translated in as many languages as possible.
This part can be done by volunteers, but the image has to be advertised to them.
Hope you mean description of image (numbered labels in description page).
- Free software should be used if possible. By free software, they mean of
course Inkscape. There are several reasons: first, it is free :-) so in the same spirit as Wikimedia projects; this software can be used by anyone without the need to buy a license; it produces better SVG code than OOo or Illustrator if you don't have the good plugin. Fr.graphists said that this may be a problem with professional graphists, who tend to use more professional tools.
Yes you are 100% right, but I know some professionals that use Inkscape (eg/ Piom and Joystick from pl.wiki who both are professional graphics, but doesn't draw much for WM projects)
- Maybe it goes without saying, but all illustrations should be in SVG
format. PNG is much harder to translate / adapt. ANother good reason to use Inkscape.
Yes it's essential - if someone creates vector graphic and then uploads only PNG version it's against Wikithinking!
- The selection process may be private (ie not open to anyone) as
competition for money should be avoided. When images are produced, they should be made available to everybody for review, criticism, and inspiration !
Yes. Confidential making decision after public debate would avoin stresses :)
- Some fr.graphists were anxious about a possible "en.wikipedia-centrism",
meaning that some illustrations would be useful to the en.wikipedia project but maybe not to others. This can work the other way around too (Wikipedia vs. Wikibooks, etc.). This also means that they should be easy to translate (maybe problematic when jargon or technical language is used). Maybe illustrations for "core topics" should have higher priority as well...
There is always some kind of systemic bias.
Finally, they also said that it would have much more impact to teach graphists to use their tools (Inkscape / GIMP / etc.), to create tutorials, etc. than to pay for a few hundred pictures. This would be much more in the spirit of Wikimedia projects. People could be paid to teach new graphists, to write well-illustrated tutorials, etc., which should also potentially generate less frustration between paid and volunteer graphists. Another way of avoiding this "frustration" is to make sure the diagrams produced in the Greenspun project have a really high quality.
We think same on pl.wiki
The graphists had some general suggestions for the project (which may or may not fit into the objectives) :
- create commonly accepted standards or recommandations for maps and
diagrams. The French Graphic Lab already uses cartographic standards extensively (see [1])
- help User:Sting to finish his cartographic tutorials, and translate them
(see [2])
- Make work together a macro-photographer and a graphist, to produce good
quality diagrams of insects and this sort of stuff (see [3] for instance)
- Get programmers to create a cartographic software for Wikipedia, including
standards, etc.
- Help the development of Inkscape, and the SVG rendering software in
MediaWiki (patterns, masks, some issues with arrows, support for animated SVG...)
Would be great if WMF would donate Inkscape team with eg. symbolic 50k$
Cheers AJF/WarX
Hello Artur and others,
2007/12/7, Artur Fijałkowski wiki.warx@gmail.com:
- For each illustration, there should be a version with numbered labels
and
a blank version, on top of versions with English or other localised
labels.
The numbered version allows easy reuse on small Wikipedias (without graphists), the blank version allows interactive diagrams using ImageMap
for
instance.
Disagree. It's trivial to remove labels from labeled image. Going further labeled (with texts) images are much less usable than numbered. It's cause very low resolution of computer displays - text labels eat a lot of screen resolution but don't provide any information to image itself. It's why some people vote oppose on FPC - usable width of image on Wikipedia is up to maybe 600px, and average is about 300, so if image has 50% of width used for empty margins/text labels we loose 50% of effective resolution. Of caurse it doesn't affect printing where we have at least 10 larger resolution for images :) but it worth remembering that text labels are just sexi, but not good at all.
Regarding labels (i.e. not numbers), I tend to agree with you that the numbered version should be the default one. However, think at how the images are used beyond the Wikimedia projects. I personally have used images in Powerpoint presentations or with children, and numbered versions are not easy to use in this case. The same goes when printing the image, and you don't want to add manually the key. So I would think that numbered & blank versions are useful for Wikipedia projects, and labeled versions are useful for third-party re-use.
- Each illustration should be translated in as many languages as possible.
This part can be done by volunteers, but the image has to be advertised
to
them.
Hope you mean description of image (numbered labels in description page).
See above. But of course if the description page is translated, it is then easy to make translated labeled versions :-) so yes, translate first the description page.
Thanks for your comments, Rémi (Korrigan)
Hello everybody.. I'm Delaby Pierre, aka Walké. Firts sorry for my english, I'm french. I understand English thanks to Google translate... Second, happy christmas ! ( and other...) Now my presentation.. I'm a young french, who intended to study art, and wikigraphiste for nearly a year. With the help of other wikigraphists , I learned to master Inkscape, and I think I have reached a correct level. My opinion in (for ?) the project : I agree whit Korrigan for all the point, and I think the internationalisation of picture is critical.. Just a question : what about the time of the realisation of the request ? 15 days, it's short, no ? And If a take a request, but I have not finished in time, other participant can also make the request. But if I finished before any of them ( Say 3 days after), can i propose my work, or am I disqualified ? So, it's all for time... Yours, Walké
Here the text in french, say better. Bonjour tout le monde ! Je suis Delaby Pierre, aka Walké Escusez moi d'abord de mal parler anglais, je suis français.. Je comprend tout de même l'anglais, mais en partie grace à Google translate Ensuite, joyeux noel à tous, et compagnie... Maintenant la présentation. Je suis un jeune français, qui se destine à des études d'art, et wikigraphiste depuis bientôt un an. J'ai appris à maitriser Inkscape gràce à l'aide des autres wikigraphistes, et je pense avoir un niveau correct. Mon opinion sur le projet : je suis d'accord avec Korrigan pour tout, et j'insiste sur l'importance de l'internationalisation des images. Juste un point que j'aimerais éclaircir : le temps pour la réalisation d'une commande. 15 jours c'est pas un peu court ? Et si je prend en charge une commande et que je n'ai pas fini à temps, d'autre peuvent aussi la prendre. Mais si trois jours après j'ai fini, je peut soumettre mon travail, ou je suis discalifié ? C'est tout pour le moment... Amicalement, Walké
- Each illustration should be translated in as many languages as possible.
This part can be done by volunteers, but the image has to be advertised to them.
Very important but it shouldn't be part of the grading criteria. We shouldn't encourage artists to create versions outside of the languages they speak. They would resort to translation software/dictionaries which leaves room for error.
- Free software should be used if possible. By free software, they mean of
course Inkscape. There are several reasons: first, it is free :-) so in the same spirit as Wikimedia projects; this software can be used by anyone without the need to buy a license; it produces better SVG code than OOo or Illustrator if you don't have the good plugin. Fr.graphists said that this may be a problem with professional graphists, who tend to use more professional tools.
I don't think the tools they use matter much as long as they submit clean code. If they do use something like Illustrator, the code easily be cleaned up, even automatically. Inkscape is arguably better than those "professional tools" when it comes to SVG. And hopefully this project attracts those professional graphists.
- Maybe it goes without saying, but all illustrations should be in SVG
format. PNG is much harder to translate / adapt. ANother good reason to use Inkscape.
Not specific to Inkscape, but yes, SVG should basically be the only format we suggest (except GIF for animation).
- Some fr.graphists were anxious about a possible "en.wikipedia-centrism",
meaning that some illustrations would be useful to the en.wikipedia project but maybe not to others. This can work the other way around too (Wikipedia vs. Wikibooks, etc.). This also means that they should be easy to translate (maybe problematic when jargon or technical language is used). Maybe illustrations for "core topics" should have higher priority as well...
I don't see that being an issue here. At least with the en part. There shouldn't be anything requested that's only useful in one language. Being Wikipedia-biased may be so. But it should be, as it's the most popular and has the greatest need for images. It's needs are probably also the furthest reaching, meaning a given image is likely to have a use somewhere else, too.
Finally, they also said that it would have much more impact to teach graphists to use their tools (Inkscape / GIMP / etc.), to create tutorials, etc. than to pay for a few hundred pictures. This would be much more in the spirit of Wikimedia projects. People could be paid to teach new graphists, to write well-illustrated tutorials, etc., which should also potentially generate less frustration between paid and volunteer graphists. Another way of avoiding this "frustration" is to make sure the diagrams produced in the Greenspun project have a really high quality.
Sounds nice, but that's not what this project's for. Hopefully it will encourage others to make some tutorials. OTOH, maybe a single tutorial request (SVG format) would be a good idea.
- create commonly accepted standards or recommandations for maps and
diagrams. The French Graphic Lab already uses cartographic standards extensively (see [1])
Yes, we should use multi-wiki standards wherever we can.
Some other good ideas in there.
-brad (rocket000)
Hello all,
Thanks, Remi, for posting this summary of the French discussion.
Regarding the way illustrations would be done :
- The images should have a very high quality: they should be reviewed both
by good graphists and by people who know the topic. Images should be assessed both regarding their accuracy, their use of colours, their possible internationalisation, etc. : fr.graphists said that payment should be done after quality has been acknowledged.
Yes, that will happen by the Review group.
High quality means using good sources: these sources should be sought beforehand and approved by people involved (zoologists for animal diagrams, for instance). This is a key point for making good diagrams. The source(s) should be quoted/acknowledged on the image description page.
I don't know about pre-approving the sources, but illustrators will be required to list the sources and references they use so that they can be verified.
- Some fr.graphists were anxious about a possible "en.wikipedia-centrism",
meaning that some illustrations would be useful to the en.wikipedia project but maybe not to others. This can work the other way around too (Wikipedia vs. Wikibooks, etc.). This also means that they should be easy to translate (maybe problematic when jargon or technical language is used). Maybe illustrations for "core topics" should have higher priority as well...
Yes, to avoid any kind of "centrism" I will finalise the request list myself. It won't just necessarily be whatever anybody suggests.
To have a look at the kind of requests that are being made so far (or to make a request yourself), please see http://meta.wikimedia.org/wiki/Philip_Greenspun_illustration_project/Request... . I will make comments so people can see what kind of requests are likely to make the final list and which are not.
Thankyou very much to the French graphists for all their thoughtful comments and to Remi for summarising them. Many of them I had already thought to do and the others I will keep in mind as the project goes on.
cheers, Brianna Project coordinator
greenspun-illustrations@lists.wikimedia.org