Commons talk:Robert Lavinsky

Question[edit]

I'm looking at the pictures on irocks, and many of them are pictures of the same rock from different angles. Should we upload all pictures, or only one of each rock? --Cerebellum (talk) 21:36, 23 March 2010 (UTC)[reply]

I would do them all (usually 4 shots it seems) if we don't have much of whatever mineral it is already, otherwise one or two is good. Of course, there's no problem uploading all of them if you want. Isn't there a bot that will help with this? Rocket000 (talk) 21:54, 23 March 2010 (UTC)[reply]

Please wait with that work. It can be, that the most pictures will be uploaded from mindat, because most of them are double uploaded on mindat and irocks.com. So it's better to wait, until the bot's work is finished. Then we can compare the pictures and upload the missing from irocks.com.

@Rocket000: I don't know, if a bot can do this, because the description is on another page than the picture with big resolution and the description is for more than one picture. Greetings -- Ra'ike T C 22:02, 23 March 2010 (UTC)[reply]

Thanks for the quick response, folks! I'm sorry, I thought that the bot had already run. I'll wait before starting. --Cerebellum (talk) 22:12, 23 March 2010 (UTC)[reply]

I believe that some of the work could be automated since there is a database behind irocks.com. Do you know if we could have access to this DB, or just get a dump as a CSV file? — Xavier, 23:43, 23 March 2010 (UTC)[reply]

PS: I have asked on the dedicated project page — Xavier, 00:39, 24 March 2010 (UTC)[reply]

Bot work[edit]

I would not do this (stupid!) job manually, but use a bot for that. As far as I have seen it, it should be possible to grab all images from irocks.com as I did it with mindat.org. To generate an image fingerprint and to compare the images by fingerprint is an easy job then. I will care about this when the upload for the mindat.org is done. --Reinhard Kraasch (talk) 08:26, 1 April 2010 (UTC)[reply]

Agreed with the fact that this a job for a bot. Yet, there is a simpler solution than comparing image fingerprints since the URLs of images that are both on irocks and mindat are easily identifiable. See the section "Comparison between pictures from mindat and irocks.com" on Commons:Batch uploading/Minerals from Rob Lavinsky on mindat. — Xavier, 22:01, 1 April 2010 (UTC)[reply]

Well, the problem is to find and grab the images that are not in mindat - they have names like "den07-03d.jpg", "edd103c-vlt-11a.jpg"... But anyway, I will investigate in this when I'm finished with the upload. -- Reinhard Kraasch (talk) 22:12, 1 April 2010 (UTC)[reply]

I'm not sure I understand the issue. What I mean is that you can safely assume that (the few) pics whose URL is NOT of the form "*/mdpics/*" can be uploaded here (e.g. http://www.irocks.com/db_pics/pics/t06-72a.jpg is OK). Downloading all images from irocks + mindat and comparing their fingerprint seems overkill to me. And AFAIK, though I doubt there will be many of them, their is already an automatic duplicate detection integrated in Mediawiki. My suggestion: download the whole irocks site except the images in the http://www.irocks.com/db_pics/mdpics/ directory, and that should be enough to retrieve non-mindat pictures. — Xavier, 20:29, 2 April 2010 (UTC)[reply]

You are of course right regarding how to find out duplicates between mindat.org and irocks.com. The problem I see now (I have meanwhile downloaded the entire irocks.com website) is how to find out which images are mineral images at all (see e.g. [1] - none of the images on this page shows a mineral at all!) and how to match the image descriptions with the images (see e.g. [2] - the arrangement of description and image is very variable and by far not as canononical as on mindat.org). --Reinhard Kraasch (talk) 10:10, 12 April 2010 (UTC)[reply]

Sorry, but there are mineral pictures, when you scroll down ;-) (for example: "Garnet pseudomorph after Limestone Fossil", "MADGARN-01 - Andradite var. Demantoid garnet" and so on). But I think, it's much easier to find the same pictures in the sub pages of irocks.com - By-Species Links (for example: Andradite page 2 "MADGARN-01 - Andradite var. Demantoid garnet"). greetings -- Ra'ike T C 10:37, 12 April 2010 (UTC)[reply]

@Reinhard: I may be wrong but apparently, images of minerals are all located in the "/db_pics/" directory on that server, so here again you can use their source URL to eliminate images of non-minerals. Unfortunately, http://www.irocks.com/demant/mad2.jpg would not match but I doubt there are many images like this.

As for how to match the image descriptions with the images, since the pages seem to be generated by a program, it would not be too hard to parse them. However, this is not the right way to do this. As I already said in the previous section, those pages are generated with data from a database. Instead of downloading the whole site and awkwardly analyze the content of those pages, we should rather work directly on the database (or on a copy of it, like a dump file) and all this would be way much easier. And this would avoid embarrassing comment like the one below... — Xavier, 20:45, 19 April 2010 (UTC)[reply]

Using the list [3] as proposed by Ra'ike eliminates indeed most of the non-mineral images (there are still some of them, e.g. collection labels if the minerals are from collections), and it also eases the parsing of the descriptions. In fact I have mostly finished this job, as well as downloading most of the images. (It is always hard to anticipate how to minimize the hassle for the material donor. So far my intention was to do the job as far as possible without the donor's assistance to accomplish this, and not to try to get the material directly and raw from Rob Lavinsky, since this would mean probably a lot of extra work for him we could easily spare him.) Regarding the issue below: We will have to reconsider the entire approach as soon as the issue is settled, I guess. --Reinhard Kraasch (talk) 08:06, 20 April 2010 (UTC)[reply]

Please see the note I added at http://commons.wikimedia.org/wiki/Commons_talk:Batch_uploading/Minerals_from_Rob_Lavinsky_on_mindat. These photos were downloaded without permission from mindat.org. Some of the descriptions you have used are not the original Rob Lavinsky descriptions, are copyright of mindat.org and you do not have permission to use them. If you want to use bots to retrieve information from mindat.org please do the correct thing and discuss this first with the webmaster. Last month I spent a long time trying to work out why six out of our top ten heaviest users (according to awstats) were from .dip.t-dialin.net - nearly causing me to block the entire ISP. It seems that the wikimedia user in question has downloaded not just the Rob Lavinsky photos, but ALL 290,000 mindat.org photos (the majority of which are copyrighted). This behaviour is totally unprofressional. While I have no problem with the use of Rob's photos, and I would have been happy to have helped in organizing an efficient way to transfer them to Wikimedia, I'm thoroughly disgusted with the way that this has been dealt with. --Jolyon Ralph (talk) 22:43, 17 April 2010 (UTC)[reply]

Hello Jolyon Ralph, I answered on Commons talk:Batch uploading/Minerals from Rob Lavinsky on mindat. greetings -- Ra'ike T C 09:43, 19 April 2010 (UTC)[reply]

Just so you all know, I've been contacted now directly and everything is resolved - the images and descriptions can stay and everything can continue as planned! --Jolyon Ralph (talk) 14:03, 21 April 2010 (UTC)[reply]

License[edit]

Under which Licence we have to license the pictures? CC-by-sa or GNU-FDL, or both? --Der Messer (talk) 10:33, 5 April 2010 (UTC)[reply]

I am confused.[edit]

Can we work on uploading now, or is the bot not done yet? --MithrandirAgain (talk) 02:59, 7 April 2010 (UTC)[reply]

Per Commons:Batch uploading/Minerals from Rob Lavinsky on mindat, I think there was a last minute misunderstading that delayed the bot upload. It should be resolved soon, hopefully. You have the details of the steps of the process at the bottom of the page. Jean-Fred (talk) 10:22, 7 April 2010 (UTC)[reply]

Okay, I see now. Thank you! --MithrandirAgain (talk) 23:47, 7 April 2010 (UTC)[reply]

Another mindat photo collection?[edit]

I've been going through the Lavinsky photos, posting them to individual articles & lists: I illustrated en:List of minerals A-B (complete), for example.

Anyway. I came across another sizeable photo collxn from mindat here at Commons, but neglected to make note of it, and now I can't find it. Help? TIA, PDTillman (talk) 17:02, 9 April 2010 (UTC)[reply]

Hello PDTillman, there are many other free pictures on mindat from other user, licensed under CC-BY or PD. Please have a look onto my list of user categories. Perhaps you can find the picture there. Greetings -- Ra'ike T C 19:14, 9 April 2010 (UTC)[reply]

Wow. Thanks for the list, and the prompt reply! Cheers -- PDTillman (talk) 21:45, 9 April 2010 (UTC)[reply]

Mass deletion[edit]

I've been posting pictures from the Robert Lavinsky donation to the English Wikipedia, and was surprised to discover that almost all of these files had been deleted.

I guess, had I read the plan more carefully, I might have realized these were test images, subject to deletion -- but it would have been nice to make this clear on, for example, the "Category:Images by Rob Lavinsky" page.

Oh, well, I guess I'll put them back once the "real" upload is done. Still, my Wikipedia volunteer time isn't unlimited. Could you notify us when you think the category is stable? Do you plan more mass deletions?

Thanks, PDTillman (talk) 21:13, 15 April 2010 (UTC)[reply]

I'm sorry for this hassle - the upload was not planned as a test upload, but unfortunately User:Docu has spoilt my plans by communicating his wishes for modification by far too late (please see Commons:Batch uploading/Minerals from Rob Lavinsky on mindat/Old issues). Most of the files will be uploaded under the same name pretty soon, I hope. --Reinhard Kraasch (talk) 21:41, 15 April 2010 (UTC)[reply]

Good work![edit]

Congratulations and many thanks to Reinhard Kraasch and everyone else involved in shepherding the Lavinsky donation along the rocky road to Commons. There are some spectacular photos there. Nice work, guys! Cheers, PDTillman (talk) 20:30, 27 April 2010 (UTC), Consulting Geologist, Arizona and New Mexico (USA)[reply]

Parent categories[edit]

I've been adding some parent categories manually, for example adding Manhattan to Category:Minerals of 162nd Street and Broadway ...

But it occurred to me, this might be a task suitable for automation. Reinhard, have you considered this? TIA, PDTillman (talk) 00:34, 8 May 2010 (UTC)[reply]

What the bot could do is: filter out the county name for the US categories (as far as a county name is given), and then set the appropiate county category. Everything else is hard to figure out for the bot (in this case: is "Manhattan, New York" the real category, or rather "Manhattan Island" or "New York City" or whatever ...) Then: Lots of geographical names are not unique (think of "Georgia"), and one would have to parse the entire hierarchy to find out what is really meant. Unfortunately, the hierarchy with this mindat locality info is not canonical. Will say: You cannot rely on that it is the full order <MineralLocality>, <City>, <CountyOrDistrict>, <StateOrProvince>, <Country>. Quite often there have been used very unspecific locality names like e.g. "Freiberg District" (which is apparently not de:Kreis Freiberg, but also not de:Kreis Mittelsachsen or de:Freiberg) or "Zillertal valley" instead of the correct geographical entities. It is hard to sort this out manually, even more for a bot. --Reinhard Kraasch (talk) 09:20, 8 May 2010 (UTC)[reply]

Thanks. It does seem like it would be worth setting the county categories for the US files with the bot. Would you want to do a test set, and I'll check it?

For others: I put "1" through "A" of Category:Minerals of the United States into parent categories, somewhat haphazardly (as I'm feeling my way into it) and yes, there are judgment calls.

I'll probably start working on Category:Mines in the United States, cross-categorizing these into regional mineral categories (since they are almost all the Lavinsky mineral fotos), which also requires judgment and some knowledge of the topic.

I'll continue with my main effort, which is actually using the new photos in English wikipedia articles -- which is definitely more fun -- and do the categorization stuff at odd moments or on sick days. Too many of those lately (sigh). Best, PDTillman (talk) 18:41, 8 May 2010 (UTC)[reply]

I let the bot run over a few categories: http://commons.wikimedia.org/wiki/Special:Contributions/RKBot - please let me know what you think of these edits. Cheers Reinhard Kraasch (talk) 01:12, 9 May 2010 (UTC)[reply]

They all look good to me. No real problems noted. Run a bigger sample, or just turn it loose? Thanks, PDTillman (talk) 05:06, 9 May 2010 (UTC)[reply]

As a side note, I wonder if it would be better to rename cats such as Category:Minerals of Alexander County to "Minerals of Alexander County, North Carolina"? -- which would be more intelligible to someone finding it in a list.

This would only be worth doing if it could be totally mechanized, and arguably is superfluous, since we are planning to add Category:Alexander County, North Carolina anyway. Thoughts? PDTillman (talk) 05:14, 9 May 2010 (UTC)[reply]

I gave this job to the CommonsDelinker bot, see User:CommonsDelinker/commands --Reinhard Kraasch (talk) 17:16, 9 May 2010 (UTC)[reply]

Thanks for removing the empty cats too. Should I bring others I see to your attention, or tag them, or is there a cleanup routine you can run sometime? Cheers, PDTillman (talk) 18:28, 9 May 2010 (UTC)[reply]

If you move cats or empty them it would be a good idea to track this somewhere, since I will use the same cats for the next uploads, especially the irocks.com images, which use a different locality description scheme (or, to be exact: no scheme at all), but I am just about to match the irocks.com locality info to these mindat categories. --Reinhard Kraasch (talk) 18:33, 9 May 2010 (UTC)[reply]

Request to include size/scale for Lavinsky photo users[edit]

Lavinsky is pretty good about giving dimensions for his specimens. It's good practice to cut & paste this info when using his photos in an article. See, for example, here for a discussion of scale for mineral photos. Thanks, PDTillman (talk) 18:49, 15 May 2010 (UTC)[reply]

Duplicate cats[edit]

I noticed we have both Category:Ojuela Mine, Mexico and Category:Ojuela Mine, both for the same place, presumably a bot error. I don't know how to fix it, but xref'd both for a workaround. --PDTillman (talk) 22:07, 26 May 2010 (UTC)[reply]

Well, Category:Ojuela Mine, Mexico is from before the bot upload (has been created manually by Ra'ike), Category:Ojuela Mine has been created by the bot. There will be of course a lot of duplicates (especially of popular localities as Tsumeb), since I had (and have) no idea how identify the localities and to find out whether a appropiate category already exists. One also might think of aggregating localities (as e.g. Category:Minerals of Lavrion, Category:Esperanza Mine and Category:Km-3 Mine) But I fear that can only be done by somebody who really knows the locality or at least the literature mentioned in mindat.org (http://www.mindat.org/loc-18501.html) --Reinhard Kraasch (talk) 22:50, 26 May 2010 (UTC)[reply]

Well, that would be me (of the people active here) ;-[

Given my volunteer time-constraints, it might be best just to xref these for now (ie, put the dupes in each other's cats), and leave for some (hypothetical) detail-oriented volunteer to tidy up later? That's my proposal, anyway.

In this particular case, the two cats are for exactly the same place, and so, ideally, would be consolidated into a single cat. --PDTillman (talk) 00:49, 27 May 2010 (UTC)[reply]

I guess, best is to wait until the upload finishes, and then leave the job to CommonsDelinker. However, for further uploads of mindat.org material the old category should not be deleted but left as a link to the correct category (or I would need some feedback to change the category name with my data). --Reinhard Kraasch (talk) 08:25, 27 May 2010 (UTC)[reply]

Hello, my manually created category of the Ojuela Mine, Mexico certainly can be replaced with the bot created category and other too, if they exist. It was a missunderstanding at the beginning of the upload process. greetings -- Ra'ike T C 10:26, 27 May 2010 (UTC)[reply]

I'm happy with either name. I think the convention is Whatever Mine, unless there are two different mines with the same name. Best, PDTillman (talk) 17:32, 27 May 2010 (UTC)[reply]

No, there's only one Ojuela Mine, I (and google) know. So it's no problem to say our CommonsDelinker the command {{move cat|Ojuela Mine, Mexico|Ojuela Mine}} ;-) -- Ra'ike T C 21:03, 27 May 2010 (UTC)[reply]

Quartz or topaz?[edit]

are duplicates of

I know nothing about gemstones, that is why I'm a little lost. Which title is correct?

More:

Please, help. :) Also, I really need help with Category:Topaz. I ran out of ideas for possible classification of the files. Already used colors, types and other sorts of things such as "bixbyite with topaz", but huge amount of uncategorized files still left (more than 200 or 300). Any suggestions? My brain is going to explode. -- deerstop. 02:26, 30 May 2010 (UTC)[reply]

As you see on Commons:Batch uploading/Minerals from Rob Lavinsky on irocks.com#Details the deletion of the image duplicates will be the next step in the upload process. There are several reasons for the duplicates, most of them have been created by uploading the same image from different sources (mindat.org and irocks.com), others because some files had already been uploaded manually before the bot started its work, and some (as these) are categorization errors. In any case you should not rely on the file name (what's in a name?) but only on the image description. --Reinhard Kraasch (talk) 10:04, 30 May 2010 (UTC)[reply]

So, do you suggest to let it go now or can I try to figure out the right title? Actually, after a bit of thinking I decided that duplicate files may have different names for a reason. I do not fully understand mineral hierarchy, that is why I thought that maybe topaz is a subspecies of quartz (like amethysts). :) But Wikipedia article says they are not related. Still, what if quartz and topaz are somehow connected? Or should I simply stick to image's description? I'd love to finish at least Topaz categorization since I've already spent so much effort. -- deerstop. 14:08, 30 May 2010 (UTC)[reply]

It seems as if Rob has put these quartz images (and several others) mistakenly in the category "topaz": http://www.irocks.com/render.html?species=Topaz&page=31, http://www.irocks.com/render.html?species=Topaz&page=32 - I hope there are not too many additional wrong categorized images. I fear I will have to check the mineral categories against the description to find further discrepancys. --Reinhard Kraasch (talk) 16:16, 30 May 2010 (UTC)[reply]

Okay, thank you, I will check the descriptions then. Frankly speaking, there were quite a lot of wrong categorized images. -- deerstop. 18:33, 30 May 2010 (UTC)[reply]

I have made up a maintenance list: Commons:Robert Lavinsky/Category missing in description --Reinhard Kraasch (talk) 20:42, 30 May 2010 (UTC)[reply]

Nice idea. -- deerstop. 21:04, 30 May 2010 (UTC)[reply]

I have split up this list into 8 smaller pages to ease editing. --Reinhard Kraasch (talk) 14:50, 1 June 2010 (UTC)[reply]

Thanks for the heads-up, Reinhard. I started fixing these, then realized there were a LOT of them.

Do you want us to post wrong ones as we find them, or just wait for the botwork? --PDTillman (talk) 23:33, 30 May 2010 (UTC)[reply]

In my opinion, we'd better post links to the files which are possibly OK. For example, at File:Antimony-Beryl-redberylharrismineutah.jpg red mineral at the picture is obviously beryl, but what's the yellow crystal? If it's an antimony, then the category is right. -- deerstop. 00:18, 31 May 2010 (UTC)[reply]

I guess most cases are just synonyms or varieties (as emerald or aquamarine for beryl). But I am no expert, so I need assistance to filter these out. With cases as File:Antimony-Beryl-redberylharrismineutah.jpg I have no idea what made Rob to categorize it as "antimony", maybe we should ask him. --Reinhard Kraasch (talk) 00:36, 31 May 2010 (UTC)[reply]

Hello, something about Morganite and Topaz: Morganite is a rose to pink colored variety of beryl and topaz is an independent mineral, but this irocks.com-category means "All Specimens with Topaz" and there are really two minerals to see in the first three pictures. Rose morganite and colorless topaz.

The picture File:Antimony-Beryl-redberylharrismineutah.jpg is really a "red beryl" (also known as red emerald or old and unrequested Bixbit). It's color comes from a little bit of Manganese in the crystal structure, not from antimony. So I guess, that the matrix of this red beryll contains antimony. greetings -- Ra'ike T C 09:22, 31 May 2010 (UTC)[reply]

File:Quartz-Topaz-tz07a.jpg, File:Quartz-Topaz-tz07b.jpg and File:Quartz-Topaz-tz07c.jpg and File:Quartz-Topaz-tz07d.jpg should be renamed to "Quartz" or "Rose Quartz". -- deerstop. 12:52, 31 May 2010 (UTC)[reply]
@Deerstop: I think, rename to Quartz is ok, because rose quartz isn't enough. In this pictures are two kinds of quartz: Rose quartz and white quartz. greetings -- Ra'ike T C 22:12, 1 June 2010 (UTC)[reply]
Thank you, sorry for the late answer. -- deerstop. 00:22, 10 June 2010 (UTC)[reply]

Could you please make up a list of files that need to be renamed (or deleted)? We can collect these cases then and let the bot do the job. Proposal: Commons:Robert Lavinsky/Files to be deleted, Commons:Robert Lavinsky/Files to be renamed --Reinhard Kraasch (talk) 14:34, 1 June 2010 (UTC)[reply]

Misplaced specimens[edit]

Commons Category:Lake Superior Mine is referred to the Lake Superior Iron Mine. However, all the specimens but one are native copper. Obviously a mixup, --PDTillman (talk) 04:12, 30 August 2010 (UTC)[reply]

A "Stone information" template ?[edit]

I do not know anything about rocks, but from what I see here, a special infobox (something in the spirit of {{Painting}}) would fit these images very well. Apparently , it could have the following fields:

Type (the name of the rock, I don't know what the name of the field should be).
Discovery place
Size
Possibly something like current location.
Comments
Source

It would have the following advantages:

More multinlinguality
Better localisation
Smoother look

Given that the structure of the file description is standardized is the same for virtually of files in this category, a bot could do the conversion, provided that it is done before too many things are manually changed in file descriptions.--Zolo (talk) 07:59, 5 September 2010 (UTC)[reply]

Mistake by City[edit]

Hey, the Dreislar Mine was in Dreislar, Hallenberg not Winterberg.--Falkmart (talk) 17:30, 28 November 2012 (UTC)[reply]

Commons talk:Robert Lavinsky

Contents

Question[edit]

Bot work[edit]

License[edit]

I am confused.[edit]

Another mindat photo collection?[edit]

Mass deletion[edit]

Good work![edit]

Parent categories[edit]

Request to include size/scale for Lavinsky photo users[edit]

Duplicate cats[edit]

Quartz or topaz?[edit]

More questions[edit]

Misplaced specimens[edit]

A "Stone information" template ?[edit]

Mistake by City[edit]

Navigation menu

Commons talk:Robert Lavinsky

Question[edit]

Bot work[edit]

License[edit]

I am confused.[edit]

Another mindat photo collection?[edit]

Mass deletion[edit]

Good work![edit]

Parent categories[edit]

Request to include size/scale for Lavinsky photo users[edit]

Duplicate cats[edit]

Quartz or topaz?[edit]

More questions[edit]

Misplaced specimens[edit]

A "Stone information" template ?[edit]

Mistake by City[edit]

Navigation menu

Search