Commons:WikiProject Tree of Life/Archive 2008

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

Taxonavigation sources[edit]

The taxonomy/systematic sources are seriously wrecked. They violate the SOPs of the wikipedia taxon projects. Way. Looking at Buteo gives me phyiscal pain. From the standpoint of one who in fact has a peer-reviewed taxonomic publication on birds, that page is was a vile and outrageous lie.

For birds:

  • English Wikipedia/IUCN Redlist/BirdLife factsheets or
  • HBW, whatever was updated last.
  • ITIS may be accetable for a few species.
  • Wikispecies and other Sibley-Ahlquist taxonomy sources are not allowed. That unfortunately means that the spiffy new taxonavigation is dead wrong.

For non-avian sauropsids, ITIS is at present the source to use IIRC. They are working to supercede that.

For fish, FishBase (natch)

For spiders, you need to parse the English Wikipedia, if you do not want to parse the World Spider Catalog (which is strongly advised against; I watched it being done once. It is a HUUUUGE work). Use no other source or you will mess up everything.

For invertebrates, ITIS is not acceptable. It usually contains 10-25% of the actual taxa only. Wikispecies is possibly the least bad default source. Markku Savela's Lepidoptera pages are not complete, but the best one-shot source there is. Note his disclaimer!

You may try to use Mikko Haaramo's Tree of Life e.g. for amphibians.

It seems that this was automated work. I hope it can be corrected as automatically. For birds, the taxonoic/systematic standards of Wikimedia Commons have dropped to a hitherto unprecedented low, I am sorry to say but it is true. It is only good that the "Old World warblers" have not been "treated" yet... Dysmorodrepanis 00:32, 17 February 2008 (UTC)[reply]

This is not good. ITIS is everywhere in birds. This means that much of the work conducted over the last years, trying to bring avian taxonavigation to contemporary scientific standards, has been laid to waste in a matter of 6 weeks or so.
The widespread application of a grotesquely incorrect phenetic taxonomy defies belief. Can this be reverted? Better to have no taxonavigation at all than one that doesn't fly (pun intended). Dysmorodrepanis 03:50, 17 February 2008 (UTC)[reply]
I wonder if the Sibley-Ahlquist_taxonomy page might be changed to help correct this? I've been adding Sibley-Ahlquist because of the guidance therein. Is it a matter of making pages for the ones listed above and deprecating Sibley-Ahlquist? Walter Siegmund (talk) 20:54, 17 February 2008 (UTC)[reply]
The Sibley-Ahlquist taxonomy was the original basis for Wikispecies, and also - apparently - for the Japanese user who added the specieslist-pages. The Commons content was originally organized according to it at least in part, but apparently never consistently so and in a rather lacklustre fashion. I got rid of the grossest mistakes; the Commons phylogeny does not use it at all. You can simply use the Commons categorization as a default for the most part, it is the best quick fix in a pinch. If there is discrepancy, you may can checking the English 'pedia if there are references newer than say 2004. If not, try an IUCN redlist search here - they have every species of bird except those that were described in the last 12 months. If the English 'pedia does not say otherwise, it uses this as a source in fact.
Nobody actually seems to want to touch the S/A pages. On en: too. The problem is that whoever added it was very enthusiastic, my god it has DNA and all. Problem is, "DNA" is not enough; newer studies use molecular data too and in a better way.
The Sibley system is the HMS Vanguard of the phenetics era, so to say - it is monumental and far outshines everything that had been built earlier with the same "technology". It is indeed a splendid effort and one simply has to appreciate it as such.
The major problem: the technology in itself is deeply flawed. Only 5 years after it was finished, it was completely superceded. Only 1 1/2 of Sibley/Ahlquist's main proposals are accepted as valid today:
  • The Galloanseres ("Galloanseri" in the original), placed at the very start of neognaths
  • The Passerida basic subdivisions, although the actual situation is a bit more complex than the neat 3-way split they proposed. Nonetheless, the 3 superfamilies will remain at least the main groups of Passerida.
The other key proposals are essentially refuted, sometimes vehemently so (the "Ciconiiformes" a la S/A are probably one of the 2 greatest fallacies of avian taxonomy ever)
The "" illustrate the problem best. In the Sibley system, they are the counterpart to the Passerida. In reality however, they roughly correspond to the songbirds that evolved before the last common ancestor of the Passerida. These early songbirds - 2 superfamilies and a huge number of minor lineages - are not closer relatives among each other than to the Passerida. The lyrebirds, according to S/A, are closer to crows than to sparrows. But in reality crows are closer to sparrows than either is to lyrebirds.
This is the typical problem of DNA-DNA hybridization, and phenetics in general: they discover groups well... too well in many cases. Basal radiations following an evolutionary innovation are typically recovered as apparent natural groups in phenetic analyses, because they share so many ancestral traits. But what counts is key innovations, not the lack thereof. Dysmorodrepanis 22:25, 17 February 2008 (UTC)[reply]
Hello, I am also partially responsible for the use of Sibley.
I already asked Dysmorodrepanis what classification should be followed for bird. He said that for bird, we cannot follow a single source/classification. Personnally, I find it sad, as it means that there is a risk that everyone will mix differently.
For, Ciconiiformes, Dysmorodrepanis don't worry, I reverted my modification (Only french and japanise wikipedia follow Sibley).
Cheers Liné1 06:52, 21 February 2008 (UTC)[reply]

Categories and/or galleries[edit]

It seems that there is no consensus since 25 months or so. The amount of content has shot through the roof during that time.

From my experience, I noticed that with the phylogenetic tree's leaves being categories, the tree becomes very unfriendly to use if you want to do good science. Categories approaching 100s of items render intelligent content recovery impossible. Probably about 60% of bird content can be at least tentatively IDd to subspecies for example.

On the other hand, there are purposes for which categories to collect a species' (or actually a lineage's least inclusive taxon's with at least "a few" items of content) content are, or seem, indisposable.

So we NEED gallery pages for species and we NEED categories for all taxa and if by any means possible also for species.

I found, rather by accident, Category:Elephants. Now, nobody would seriously propose that for deletion "because it's Category:Elephantidae, dammit!"), no?

This was interesting, because I have played around with categorizing birds by common name on en: following a similar approach. We had all these "game birds" and "seabirds" stuff that we tried to suppress, but with articles being given common-name titles routinely, you have no chance ;-)
It's actually rather useful; see for example here (note that page sorting is incomplete. But you get the idea.).

Category:Birds of prey (it's in a dismal state at present) should according to Commons SOP do what exactly? Redirect because it's a vernacular term? Well where exactly should it redirect to, hmmm?

You are presently trying to map a common-names category onto the phylogenetic tree. Consequences are redirects like Category:Eagles, Category:Butterflies, or Vulture, which Stewie Griffin - if he were a taxonomist - would describe as "vile and odious lies". Every professional taxonomist would agree that bijective mapping of folk taxonomy to scientific taxonomy on a large scale is impossible, pointless, misguided and should best not be attempted at all. It only leads to errors such like that "warblers" thing (<- If you don't know what I'm talking about, read it! It's an eye-opener.)

This led me to the conclusion: why not try to have both cakes and eat them? I have prepared a short summary and am presently assembling a more detailed discussion. I think it might actually work.

I semi-officially propose you guys read the summary or simply take a look at the examples. If you think that this approach would be entirely incapable to deal with a specific taxon, your can leave a note below. Otherwise, you may want to refrain from saying what has been said over and over again, at least until I have finished assembling the detailed rationale. ;-)

I have also created Category:Tanagers and adjusted Category:Thraupidae accordingly so you can compare the content (The changes don't amount to much and can be easily reverted). It demonstrates a major advantage over everything (I guess) hitherto proposed. In the finalized version, cat:Tanagers would contain common-name species subcategories only. Thraupidae would be deleted without replacement (see en:Tanager as for why...). The 3 photos in cat:Thraupidae would also be moved.

Simply take any favorite taxa of yours and think the trees through that would result under this approach. It is very likely that you'll find it to work like a charm.

(If you think it sucks - well so does the status quo, and it has been sucking for more than 2 years now with no end in sight and the complications increasing in an accelerating rate due to growth of content.)

Note that the proposal at first glance seems a radical break with almost every solution or fix suggested so far. This is actually not correct; the only really important change is what happens to species categories.

Note also that this is based on my personal experience with the Mediawiki engine as a classification tool on en: and Commons, and not on a priori theoretical desiderata. It is based on what I found to work and more importantly what I found not to work, across a wide range of taxa, while giving them the most accurate and complete categorization possible.

It does not require an all-out massive effort to implement. That what would need to be changed can be changed gradually, as part of routine maintenance work.

I do not propose this as a system to be implemented wholesale, but as a working standard for further changes, making the content under the scope of WP:ToL by and by consolidate into an increasingly usable structure rather than deteriorating due to guerilla edit wars as is the case now.

(It will also give any non-Project people as might object to the ToL way of categorization some work on which their time is better spent ;-) )

Thanks for your consideration.

Dysmorodrepanis 04:48, 18 February 2008 (UTC)[reply]

OK, so looking over my collected issues I realized that it was basically the same-old-same-old. That what I linked above shall entirely suffice as basis for discussion I think - any details and rationales can be resolved in the discussion

Again - the key point of my proposal is that everything stays as it is, except that every species page gets a species category also, though not with the scientific but with the most widespread English common name. This latter then gets tied into a phenetic tree of life that does not use scientific nomenclature and to which non-English users are given access via redirects-from-pages.

The underlying rationale - and that, I think, really has not yet been brought up - is twofold:

  1. The present redundancy is painful and cruft. Both will eventually cause the present system to collapse; there is presently already not even a rule-of-thumb anymore, with whole sections of the phylogenetic tree designed after contradicting approaches.
  2. The present system in deed if not in word actively denies the overwhelming majority of users access to the category system to locate ToL content. By wholesale redirection of established common-name categories, it does so in a preposterous and entirely unscientific manner, that moreover constitutes an in-your-face violation of Wikimedia's underlying principles, namely veracity and accessibility. To the average user it's like we tell them: "either take Taxonomy 401 or piss off! Oh and by the way, butterflies ARE moths, condors ARE Old World vultures, and eagles ARE hawks." This is intolerable.

My proposal resolves each of these issues in a way that would seem to benefit all, and be compatible with the desires of (almost) all.

Dysmorodrepanis 17:10, 18 February 2008 (UTC)[reply]

You've definitely put a lot of time into this, which I admire. I must disagree, however, with anything to do with common name categories for species. There have been too many edit conflicts over "what is the most common name" as it differs greatly for many disciplines (particularly, botanical common names are a mess) that I think you'd be creating many, many more problems than this might solve. --Rkitko 04:17, 20 February 2008 (UTC)[reply]
I too admire the amount of work you put into it. As Commons should be multilanguage, i dont think article or category anmes with vernacular names should ge in here. Teun Spaans 21:32, 25 February 2008 (UTC)[reply]
Just to say my word, I am also against vernacular names for categories. And I prefer to have a category for each species, because it makes Mayflower a really useful tool to find uncategorized images. And please everybody stop to delete what the others have done. This seems to me the first principle ! Frédéric (talk) 20:16, 23 June 2008 (UTC)[reply]

Another spin of the wheel[edit]

Someone suggested (but I can't find their message) that instead of removing categories from images (once they have been found places in a gallery), why not just add a sort field to their category. For instance a bird that is uploaded and categorized as [[Category:Passer domesticus]], can be changed to [[Category:Passer domesticus| ]]. So all galleried images show at the top of the corresponding category. We can then also add a marker to the category, that shows where the new, unsorted, images start - eg add Image:Unsorted1.svg to the category with [[Category:Passer domesticus|*]]. For an example see Category:Passer_domesticus.

Now when people add new images to Category:Passer_domesticus they automatically appear separated from the processed images, after the 'unsorted.svg' divider (ie towards the bottom of the category). This way someone browsing the category first sees images which have had their identity confirmed (by whoever is updating the gallery) before reaching the possibly wrongly categorized new images. And someone maintaining the gallery can easy find recent images added to the category.

The use of a space ' ' and asterisk '*' for sorting might not be the best, as they are used for other sorting purposes too, but two other non-alphabetic characters would do (assuming images have names only starting with alpha-numeric characters, I'm not sure how well this works with other non-latin character sets).

A bot could be used to maintain the category by updating the sort order of any (already categorized) image found in the corresponding gallery.

Thoughts? (excluding those that dismiss the need for categories ;-) --Tony Wills 12:18, 1 March 2008 (UTC)[reply]

I know that from somewhere :) A first attempt was made in this discussion but don't take my arguments against the proposal too serious. That was before I tried to start LifeBot, and before all other discussions, and I've mellowed somewhat. So, let's look at a few arguments:
  • The scheme gets unpracticable whenever there are more than 200 pics in the category. With species, there would rarely be a problem. However, apply it to higher taxa, and you'll not only need to browse (and here a lot of bandwidth is taken because the thumbs have to be loaded) but also due to a MediaWiki bug, some galleries and subcategories will also be on the 'next' page only because of the pics.
  • You'll need a delimiting picture.
That said, I'm not against it at species level, and will look forward to any small scale implementation you'll try. I have the suspicion there will be gotchas but let's see. --Ayacop 18:27, 28 May 2008 (UTC)[reply]

Small milestone[edit]

Not on our favorite subject of galleries/categories :-), but with my latest rounds of uploads I topped a thousand, with User:Stan Shebs/Plant taxa showing I've uploaded pictures of 1,029 plant taxa to date (not counting rose breeds). And I've still got massive backlog... Stan Shebs 15:39, 14 March 2008 (UTC)[reply]

Cheers Stan! That's impressive - many thanks! Pudding4brains 16:29, 14 March 2008 (UTC)[reply]

Extinct/fossil major mess[edit]

The scheme grossly violates Categories#Over-categorization. "Fossil" is for practical purposes a subcategory of "extinct".

See Category:Fossil Neuroptera and Category:Extinct Neuroptera for the redundancy (up to threefold) this creates. Things like Category:Fossil Ichthyornithiformes and Category:Fossil Archaeopterygiformes are completely pointless, since the taxa do not contain non-fossil members. And these cases stretch through all the taxa involved. While some sort of categorization is warranted, the present approach is unworkable. The redundant categories need to be deleted, not redirected; according to SOP they should never have been created in the first place. Dysmorodrepanis 16:32, 27 April 2008 (UTC)[reply]

More nonsense: Category:Meiolaniidae. So a fossil of a Ninjemys qualifies as fossil Ninjemys, but a drawing of the same species qualifies as non-fossil Ninjemys? OK, but it would have to be "Ninjemys fossils" then. But that is also bad, for Image:Diatryma Gigantea -cast skeleton.jpg is NOT a fossil, as its title says - the closest it comes to being a fossil is a Carboniferous tree... there is not a single scrap of fossil bone in this thing, no matter how much it looks like a fossil. It is a modern-age replica. Dysmorodrepanis 10:27, 7 May 2008 (UTC)[reply]
I will empty all these categories as I come across them, so they can be deleted. The mess it has made in birds alone is simply unbvelievable. The categories must NOT be made redirects; redirecting categories is a bad thing as they are still active and contant can be placed there, which effectively hides it from view. A bot can slate the emptied categories for deletion, as it is being done on the English Wikipedia. Dysmorodrepanis 10:33, 7 May 2008 (UTC)[reply]

So fix it. I've restructured the Ichthyornithifornes tree, and nominated the Fossil Ichthyornithiformes subcategories for deletion as redundant. Hesperian 11:31, 7 May 2008 (UTC)[reply]

Plants/Flora of categories and whether species should associate to them?[edit]

Hi TOL users, I have posted these two threads at the Village Pump: Commons:Village pump#Plants of vs Flora of and Commons:Village pump#Should species be categorized to area-specific categories?, which somewhat relates to the work here. -- Slaunger (talk) 23:22, 2 July 2008 (UTC)[reply]

non-TOL v. TOL; categories vs. galleries[edit]

Howdy. There has been frequent discussion regarding the TOL's stance that galleries take precedent over categories, whereas elsewhere on Commons: all media must be categorised; and galleries just highlight the selected best. I opened a discussion to try and settle, once-and-for-all, categories versus galleries. The discussion is here, and as I believe members of this project will be able to offer a counter-argument. Therefore, I would greatly appreciate your input. Cheers! --Bossi (talk • gallery • contrib) 23:07, 3 July 2008 (UTC)[reply]

Please be aware of this change [1] and be aware that reverting it would not be prudent. (see also this COM:AN discussion) Counterarguments are going to be somewhat difficult to come up with, I suspect, as categories are the Commonswide way of cataloging things. Force of numbers won't be a consideration, only strength of argument. ++Lar: t/c 13:20, 6 July 2008 (UTC)[reply]
Buffering, perhaps, Lar's justifiably forceful statement above, I'd like to point out that I'd be glad to work with any interested parties in finding a solution to efficiently tracking workflow, so long as the solution doesn't involve decatting images. :) For example, we could create a TOL review tag which gets automatically placed on species images and which can be updated as the images is reviewed. The need to review images and track progress is by no means unique to TOL, so figuring this out would be useful to everyone. Cheers.--Gmaxwell (talk) 18:58, 6 July 2008 (UTC)[reply]
I missed this poll. Some questions arise:
  1. Does a poll have priority over a vote?
  2. Why was only a small selection of the ToL arguments presented in the vote and another misrepresented?
  3. Why have ToL members not been informed of this poll?

Whatever the answers are, it is my impression that due to the continued pressure of the categorist lobby enthusiasm for the project was low for the past year. I guess that with the result of this poll ToL is dead. I have seen that aboutv 5 people are working on categories now, but just for myself: categories are so inferior in organizing images when compared to pages, that i dotn wnat to invest time in it. And the result of the poll spells further trouble for the galleries and the organization of images. Teun Spaans 13:37, 6 September 2008 (UTC)[reply]

Which poll it was, where is the link to the page? @Teun Spaans : if you don't like categorization, don't do it, do galleries/pages, but i think it was a mistake since the beginning to make a work of uncategorization of images that where previously categorized at the species level by other contributors. It generated to many conflicts and many new contributors were aware that TOL was making "problems" and finally didn't join the project. Finally, i think that TOL pay the price of this decision (undoing the work of others) : new contributors will not join and current members will stop one day. You have to be more open and not so conservative : welcome everybody and let people do galleries/pages or categorization as they wish, finally more people will join this project again. Guérin Nicolas (messages) 11:55, 10 September 2008 (UTC)[reply]
I think that you have to see it as teamwork. Ignorants like me can try to sort some of the million incoming images and drop them in the right area or category. Specialists can then try to harvest and place them properly on the right place in the right galleries with the right text. --Foroa (talk) 12:11, 10 September 2008 (UTC)[reply]
Actually, the presence of the categories creates a lot of extra work. But as this has often said before, this seems to be a lost fight. I will resume work when decent tools become available. Teun Spaans 20:12, 8 November 2008 (UTC)[reply]
Me too. I find it impossible to work with this mess as often as I would like. Galleries are a legitimate feature, only they are a leaf (in phylogenetics lingo) and not a branch (like categories). We want to use them for leaf taxa. What is so hard to understand about that?
I will sort content that users like Foroa harvest, but as a resource where it matters most, at the endpoints, Commons has lost its credit for me. Dysmorodrepanis (talk) 02:11, 28 February 2009 (UTC)[reply]

Tree of Life or Confusion of Life?[edit]

I don't know if there already was a discussion about it, but I don't think the changes User:CarolSpears is currently doing on the TOL-Categories are very useful. As far as I see he's trying to merge different systematic approaches into the category-tree. For instance Category:Polemoniaceae is now included in Category:Ericales according to the AGP II system and in Category:Solanales according to the Cronquist-System. The problems can also be seen further down in the tree: Category:Cuscuta is included in Category:Convolvulaceae (AGP II) as well as in Category:Cuscutaceae (family level according Cronquist, subfamily-level according to AGP II). Next problem: there are at least 3 different approaches for a subfamily classification within Solanaceae (none of which is included in the AGP II system!) - and right now it is not clear which one CarolSpears is following in Category:Solanaceae. Additional I think a Category:Solanaceae_incertae_sedis is very superfluously ...

If you want to show different systematic approaches don't mix them up in one category-tree. Maybe create an alternative tree for Cronquist-classification. But right now it's a total mess! --Carstor (talk) 12:19, 18 August 2008 (UTC)[reply]

I see two options:
  1. We mandate that our category system use a particular system e.g. APG-II for flowering plants
  2. We split Category:Ericales in Category:Ericales sensu APG-II, Category:Ericales sensu Cronquist, etc.
Hesperian 13:27, 18 August 2008 (UTC)[reply]
I guess option 1 would be the best, using one system - alternative systems can be shown by using the Taxonavigation-template. But you will always have to mix different systems: Infra-familar classification isn't covered by AGP-II (and as far as I know neither by Cronquist or any other high-level classification, but correct me if I'm wrong) as well as everything above order-level. So adding a taxonavigation-template to subfamilies, genera, species etc. and stating it's AGP-II or whatever is just wrong. I.e. for Solanaceae you can follow the infrafamiliar classifications of either William D'Arcy, Armando Hunziker or Richard Olmstead - at least the last two are still being used, even if the phylogenetic classification sensu Olmstead seems to be more and more preferred. --Carstor (talk) 13:47, 18 August 2008 (UTC)[reply]
I agree with all that. The only problem is when there are competing taxonomies. I don't think it would be proper for us to endorse a particular one. Hesperian 14:04, 18 August 2008 (UTC)[reply]
I don't see a problem by just using one system for the categories. Alternative taxonomies can be mentioned in the article-text just as it is done right now, there's no need for neglecting their existence. But currently there are 10 families in Category:Solanales which isn't correct for any of the systematic concepts you could choose. --Carstor (talk) 14:42, 18 August 2008 (UTC)[reply]
I don't see a problem, either. It's not like the Commons is a scholarly reference. The first herbarium I worked in was arranged in the Engler-Prantl sequence, despite the fact that no one felt it was an adequate representation of anything: it was simply the arrangement that was used, for (at that point) purely historical reasons.--Curtis Clark (talk) 16:51, 18 August 2008 (UTC)[reply]
  1. I like the idea of categories specifying the classification.
    Why: A taxon name does not mean anything without the systematic involves (I introduced the {{Taxonavigation}} parameter classification for that + I always add the comment <!-- for YYY classification--> after a category).
    Small proposition: Perhaps for simplification, wikicommons could have a default classification (APG II for plants) that would not need classification precision. That way we would have Category:Ericales (meaning APG II) and Category:Ericales (Cronquist)
    Advantages: Keeping cat name for APG II with give us less work. The parentheses in the Cronquist taxon name will allow us to hide the parentheses in the Taxonavigation (It is already provided by the parameter classification=)
    Problem1: How will we navigate between the cat of the same taxon in different classifications ?
    Problem2: english/french... sites should probabily link to the cat without parentheses. So the cat without parentheses should propose link to the alternate/old classifications.
  2. I don't understand why we precise the Strassburger classification. It is a pure german classif (I am myself 80% german ;-)) that mimics APG II.
  3. I would prefer that we keep at least one classical classif (like Cronquist). The Category:Ericales (Cronquist) would allow us to have both tree in parallele.
  4. I also think that we should stick to pure APG II classif and avoid post APG II taxon (I mean those created by {{APWebsite}}). We should nicely wait for APG III without jumping on all the taxon propositions made by biologists.
Cheers Liné1 (talk) 07:13, 7 September 2008 (UTC)[reply]

category redirects need undoing[edit]

I noticed this morning that RocketBot (talk · contribs) is picking up on files from categories with {{Seecat}} and moving them, often moving things out of a species category and into a genus or higher. I'm guessing this is a relic of the former categorization setup for TOL images (add to galleries and remove cats), but it's causing difficulty because when people use hotcat to categorize, the js suggests specific categories which are then botted to a less appropriate higher category. So, we should try to check species categories when adding them to see if it's a redirect, and undo the redirect if it's there. Admins and rollbackers might want to keep an eye on Rocketbot's contribs to catch this problem... it seems to address the category redirects every other day or so. --SB_Johnny talk 09:25, 23 August 2008 (UTC)[reply]

Warning: Post APG II invasion[edit]

Someone creates a lot of categories corresponding to orders and families not following APG II. Most of those taxon come from {{APWebsite}} which is a cool web site, except that it is only proposing modifications to APG II. This website is not recognized by the community, also it creates and suppresses taxon all the time, because it is a systematic in preparation. I try to systematically set the {{tl|Taxonavigation}} parameter classification=APWebsite on those categories to precise that their classification is not APGII, but post APGII. here is a list of problematic taxa (old list, because you know how bad the seach engine is ;-)).

My question is: should we follow strictly APG II (and wait for future APG III) or follow any other phylogenetic classification ? Liné1 (talk) 12:06, 19 September 2008 (UTC)[reply]

We should be following the most up-to-date taxonomy, rather than stubbornly sticking to a static one. APG II is now five years old and known to be flawed in some areas. I think your "proposing modifications to APG II" reveals a misunderstanding of how systematics works. You don't need to ask the APG's permission to alter the taxonomy to accord with a newly revealed phylogeny. Hesperian 03:36, 20 September 2008 (UTC)[reply]
Follow the most up-to-date taxonomy ? Even if there are weekly, monthly updates ?
OK, but what if I don't follow the same modification of APGII as you do ?
If we do what you say, we have to provide for each Taxonavigation, the reference to the publication followed. Of course, this publication will not accessible on internet. So there will be no possibility to verify the information
Personnaly, I prefer a stable/verifiable/outdated information than a unstable/nonverifiable/uptodate information.
Liné1 (talk) 08:15, 22 September 2008 (UTC)[reply]
It doesn't bother you that "outdated" is essentially a euphemism for "wrong"? Hesperian 11:53, 22 September 2008 (UTC)[reply]
So you suppose that updated means "true" ? At least they are recognized by a community.
Seriously, if I see a newly created category named with a familly name almost familiar (for example: already existing in Takhtajan systematic) but not recognized by APGII:
  1. Do you find it normal to add taxonavigation=APGII in the {{Taxonavigation}} of this category ?
  2. How do I retrieve the genus list of this family ?
Liné1 (talk) 12:35, 22 September 2008 (UTC)[reply]

We need to stick to published and agreed-upon systems, even if somebody says they are "wrong". Commons categorization is really not the place to be doing original research, nor should we be taking a proposed change and enshrining it as new reality. Stan Shebs (talk) 13:14, 22 September 2008 (UTC)[reply]

It would be gilding the lily for me to claim that once a paper is published it is the new consensus, but it is even more inaccurate to claim that each new paper presents only a "proposed" change. A paper won't even see the light of day if it isn't seen as a step forward by a pack of feral peer-reviewers; so in general, newly published taxonomies are immediately accepted, and incorporated into the big picture. Hesperian 23:33, 22 September 2008 (UTC)[reply]
Yes and no. Yes, peer review results in survivors, and assuming a reference is peer-reviewed (as contrasted to things on Peter Stevens' web site that are not yet published), we can rely on its solidity, if not its validity. No, just because something is peer-reviewed, that doesn't make it necessarily either right, or accepted. This is especially true of publications based on molecular data sets that are at odds with either other molecular data sets or with morphology. Many of the APGII changes, and many since, have caused thoughtful taxonomists (a group in which I include myself) to say, "Oh, yeah, that makes sense." The reason I often mention the Scrophulariaceae is that I've known for years that it needed to be dismembered, and the dismembering left morphologically coherent units (Penstemon really is more like Plantago than it is like Scrophularia). But I know that there are some who disagree even with these more-or-less consensus changes, and I'm sure there are others that are even more controversial.
But all this is peripheral to the issue at Commons: here we're trying to make a system that people can use for finding resources, and I think that stability is more important here than in the Wikipedias, especially given that not all Wikipedias use the same classification, but all their users have to figure out Commons.--Curtis Clark (talk) 02:58, 23 September 2008 (UTC)[reply]

Although I agree that using up-to-date taxonomy in places such as taxoboxes is usually a good idea (with the caveats of original research and unintended POV-pushing), categories in Commons are to help people find items, not so much to represent current thought. Stability is important here. It's usually easier to figure out "Didn't that genus used to be in the ____aceae," that have to go searching to find out where it is now.--Curtis Clark (talk) 13:26, 22 September 2008 (UTC)[reply]

I think APG has been great for simplifying these discussion about which taxonomy to follow, but five years is a very long time in the context of the present post-paradigm-shift mop-up. Hesperian 23:33, 22 September 2008 (UTC)[reply]

Alas, the floor has not yet been fully mopped. Perhaps in two or three more years, things will be stable enough that new papers will refine rather than change classifications, but I'm not convinced we're there yet. (Again, I'm talking about Commons categories; I have no disagreement with incorporating the latest classifications in encyclopedia articles.)--Curtis Clark (talk) 02:58, 23 September 2008 (UTC)[reply]
I hereby bow to consensus (<bow/>). Hesperian 03:11, 23 September 2008 (UTC)[reply]
I agree with Liné1, Stan Shebs and Curtis Clark. To adopt an inchoate taxonomy is unlikely to be helpful. Walter Siegmund (talk) 04:29, 23 September 2008 (UTC)[reply]

proposal[edit]

Please have a look at my proposal on village pump: Commons:Village pump#Image redirect naming scheme for taxa. I think it is related to this wikiproject. --Slomox (talk) 14:57, 22 October 2008 (UTC)[reply]

Categorization issue/Structural Classification of Proteins[edit]

Hello, today I noted a quite unusual categorization attempt at Category:Structural Classification of Proteins (SCOP) (cf. category tree for its subcategories which names start with "Class: …" and "Fold: …"). This has been done by 76.88.72.233 (talk · contribs) and could be related to the work of Donabel SDSU (talk · contribs) who is only active a bit on en.wikipedia (in addition to Commons), from what I can see.

Could someone please have a look and see if categories like Category:Class: Small proteins need any maintenance? Well, I would rename this category back to just Category:Small proteins (which was blanked and is currently empty), but I'm not an expert on proteins. --:bdk: 21:56, 6 December 2008 (UTC)[reply]

Hi, I wanted to keep the category names similar to the way SCOP displays the protein lineage. I also needed to avoid putting protein structure images in an existing category. For example Category: Insulin already exists, so I created a subcategory named Category:Protein Domain: Insulin. If this a problem, please let me know and if there are any suggestions to categorize these protein images.
I also want to mention that I plan to upload around 50,000 protein images in Category:Structural Classification of Proteins (SCOP). So before I upload the images, I wanted to determine how to categorize them. I've been manually adding the categories for recent uploads to figure out a way to write a program to do this automatically. So any help with how I should name these categories will be appreciated too. Thanks! Donabel SDSU (talk) 23:37, 6 December 2008 (UTC)[reply]
The categories like family: <something> and class: <something> don't match our category naming conventions. You should just use the simple names and place them in the right category/categories. Do you have a list of categories you want to create somewhere online? Multichill (talk) 15:08, 7 December 2008 (UTC)[reply]


I originally planned to use the simple names until I came across protein images that have Category: Insulin. I did not want to put the protein images in the exisiting Insulin category because I want them as a subcategory for Category:Fold: Insulin-like. I just realized I can make the prefix to a postfix, so that it would look like this: Category: Insulin - Protein Domain. Is this acceptable? Donabel SDSU (talk) 16:26, 7 December 2008 (UTC)[reply]

You should only use the postfix if the category already exists. How much cats do you want to create exactly? What's the number of naming conflicts you're going to get?
The classification is three layers, right? You should tag every page with a standard template, something like:
{{SCOP
|class=
|fold=
|domain=
}}
Multichill (talk) 16:43, 7 December 2008 (UTC)[reply]
Yes, I agree that I should use the postfix for categories that already exist. I see that pagefromfile.py will not overwrite an existing page if I use -safe. That will help avoid putting images in an exiting category.
The classification has five layers -- Class, Fold, Superfamily, Family and Protein Domain. I only know of one naming conflict (with Insulin) and wanted to avoid any others. Right now, I don't know the exact number of cats I will create, but do know there will be a lot. For example, the Class: "All alpha proteins" has 258 Folds as shown in SCOP here. Another Class called "Alpha and beta proteins (a+b)" contains 334 Fold names as shown in SCOP here. So I anticipate even more categories because within each of these Folds are Superfamilies, Families and Protein Domain names. I do plan to upload each protein image with its SCOP lineage, so it would be displayed on the page as shown in this example. Notice the name for Fold = Superfamily = Family = Domain. I don't want to create categories for identical names, so I will go with the highest level (in this case, it is Fold) as the category page.
Now that I have a better idea on how to name the categories, is it possible to delete all of the contents in Category: Class: Designed proteins and Category: Class: Small proteins? I want to start over with these two rather than going in and manually editing the category names is in these two categories. Thanks again for the help and suggestions. Donabel SDSU (talk) 23:10, 7 December 2008 (UTC)[reply]
Donabel SDSU, maybe have a look at Template:Taxonavigation and develop a similar template Template:SCOP or Template:SCOP navigation first? Having a main template for all categories ready before starting with the creation of all the categories saves time :-) --:bdk: 01:06, 8 December 2008 (UTC)[reply]
Thanks for another great suggestion and all your help! Donabel SDSU (talk) 07:31, 9 December 2008 (UTC)[reply]
Retrieved from "https://commons.wikimedia.org/w/index.php?title=Commons:WikiProject_Tree_of_Life/Archive_2008&oldid=372756475"