Wikidata talk:WikiProject Chemistry/References
Latest comment: 11 years ago by Snipre in topic GESTIS
GESTIS edit
German Wikipedia heavily relies on the hazard substance database http://gestis.itrust.de/ as it aggregates and gives a credibly opinion when multiple assessments exist. w:de:Vorlage:GESTIS creates automatic links as the database has it's own internal ZVG number and can also be searched by CAS which makes the point "the url might change" invalid. Also of course in this case we have to rely on up-to-date toxicological data instead of dangerous outdated information printed in analog media. I have also never heard of http://sicherheitsNet.de nor seen this in the wild on German Wikipedia. I therefore strongly 
Oppose the proposition to avoid or disallow trustworthy digital references. Matthias M. (talk) 18:34, 16 February 2013 (UTC)
- If I suggested to avoid internet database it's because of a previous experience with the database ESIS which changed their system and all links (although we used a model) were broken. It is possible to change the model and the input data in the articles but this needs a lot of effort. Then I am a little afraid when I see that NIST webbook claims that it can ask for a payment to access its database in the future. and finally if we take a look at the references of gestis or NIST wekbook most of those databases relies to books or other databases so why not used directly the primary sources ? Snipre (talk) 19:41, 16 February 2013 (UTC)
- Secondary sources are reviewed primary sources, see en:WP:PSTS. --Leyo 19:46, 16 February 2013 (UTC)
- Ok, I was not accurate: Gestis is not doing the review but refers to other documents which did already the review (but primary sources in sciences are mainly scientific articles which are already reviewed so for me we don't need to consider thoses sources in the same way as other primary sources). Then for some data like SGH definition, there is no need of review, because it's a definition. I am not against Gestis or other internet databases, but I prefer references which are available in every good library in order to be sure that in the future everybody can check them. Gestis is free and maintained now, what's about the future I don't know. But this should not be a reason of a long discussion because wikidata can handle different values for a same property, the only thing to consider is to avoid to depend too heavily on internet databases. Snipre (talk) 20:27, 16 February 2013 (UTC)
- @Matthias M. http://sicherheitsNet.de and all german referneces I proposed are the references of Gestis so if gestis is using them, I think we can use them (look just once at the reference page of a datasheet of Gestis). Gestis is mainly a compilation of other secondary sources which are very specific. I think it is a better way to work by using specific compiled references instead of general references like Gestis: there is a reduced risk of propagation of errors (and I already found some error in gestis). Snipre (talk) 20:39, 16 February 2013 (UTC)
- Ok, I was not accurate: Gestis is not doing the review but refers to other documents which did already the review (but primary sources in sciences are mainly scientific articles which are already reviewed so for me we don't need to consider thoses sources in the same way as other primary sources). Then for some data like SGH definition, there is no need of review, because it's a definition. I am not against Gestis or other internet databases, but I prefer references which are available in every good library in order to be sure that in the future everybody can check them. Gestis is free and maintained now, what's about the future I don't know. But this should not be a reason of a long discussion because wikidata can handle different values for a same property, the only thing to consider is to avoid to depend too heavily on internet databases. Snipre (talk) 20:27, 16 February 2013 (UTC)
- Secondary sources are reviewed primary sources, see en:WP:PSTS. --Leyo 19:46, 16 February 2013 (UTC)