Commons talk:Quality images candidates/Archive 24
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 20 | ← | Archive 22 | Archive 23 | Archive 24 | Archive 25 | Archive 26 | → | Archive 30 |
QI and Structured data
Hi, I just want to share with you an exemple of querry about the QI assessment, the exemple below shows the Qualities Images depicting Hummingbirds, of course are shown only the files that have the adequate values within depicts (P180) and Commons quality assessment (P6731):
#shows files that depict Hummingbirds and with the Quality Image assessment
#defaultView:ImageGrid
SELECT ?file ?image
WITH
{
SELECT ?item
WHERE
{
SERVICE <https://query.wikidata.org/sparql>
{
?item wdt:P171/wdt:P171* wd:Q43624.
}
}
} AS %get_items
WHERE
{
INCLUDE %get_items
?file wdt:P180 ?item .
?file schema:contentUrl ?url .
?file wdt:P6731 wd:Q63348069 .
BIND(IRI(CONCAT("http://commons.wikimedia.org/wiki/Special:FilePath/", wikibase:decodeUri(SUBSTR(STR(?url),53)))) AS ?image)
}
Regards, Christian Ferrer (talk) 19:17, 8 October 2020 (UTC)
:-)
https://twitter.com/Lunarbaboon/status/1313588903014735880 --Smial (talk) 19:37, 22 October 2020 (UTC)
- Nice! Yes, at least for our QI reviews here these are very reasonable suggestions. --Aristeas (talk) 11:16, 13 November 2020 (UTC)
- +1 ... and a few more --Kritzolina (talk) 13:01, 13 November 2020 (UTC)
Bot?
Is there an issue with QICbot? I see it ran today, but I also see Gzen92 adding tags manually. Did it miss a day? — Rhododendrites talk | 13:54, 16 November 2020 (UTC)
- Hello. For 13 November, the bot did not put the model (too fast I was wrong about for cette image). Gzen92 [discuter] 14:05, 16 November 2020 (UTC)
- No problem. Thanks for doing the bot's work. — Rhododendrites talk | 14:11, 16 November 2020 (UTC)
Bot didn't finished today
The bot created the archive page Commons:Quality images candidates/Archives December 04 2020, but didn't added the templates and didn't inform the nominators. I don't know what to do. (I've added the qi template at my own image manually. Some procedure for the other images?) --XRay 💬 08:55, 4 December 2020 (UTC)
- I'll add the qi template. --XRay 💬 08:59, 4 December 2020 (UTC)
- Done. --XRay 💬 09:07, 4 December 2020 (UTC)
deadlines
We have formulated two deadlines in the rules for consensual review. One is that the final decision may be registered at the earliest 48 hours after the last vote has been cast. The second is that if there has been no decision, an affected image may be archived at the earliest eight days after the transfer to CR. Again today, although a photograph was (correctly regarding the 48h-rule) rejected 48.5 hours after the last vote, the picture in question was in fact only available for discussion for less than five days. I think it is illogical and incomprehensible that undecided reviews have to remain standing for eight days, while others, even with an extremely close result (only one vote majority), are closed after only five days. We have no lack of storage space and I strongly advocate that every picture in CR can be discussed for at least eight days. It simply has to be guaranteed that at least one complete weekend is within the voting period, so that contributors who may not be able to do so during the working week can vote. -- Smial (talk) 22:31, 2 December 2020 (UTC) Translated with www.DeepL.com/Translator (free version)
- Now Augustgeyler has reverted my change in the case in question, which was intended to prevent the bot from archiving it prematurely, and which I have explained in detail and put up for discussion here, without comment. I would have hoped that such decisions would be discussed together beforehand in a collaborative project. Very disappointing. It would at least have been more decent to leave the decision on a photo one dislikes to a neutral person. --Smial (talk) 07:29, 3 December 2020 (UTC)
- @Smial: I did so, to follow the guidelines. The 8-days deadline for undecided is logical, because undecided means there are as many opponents as there are supporters. This proves there is now majority for one side. To give those candidates an 8 day chance is a good thing, because otherwise there would be simply NO decision. But in cases of nominations where a majority of at least one vote came out, it is a clear thing, these CRs should end after 48h with no new votes. I can not be blamed for all the others who did not take their 48 hour chance to take part in that particular decision. My own candidates were rejected ore promoted that way in the past. So I can understand your feelings. It can be disappointing seeing no one taking the time to take part in that decision. But what disappoints me here is, that you reverted my edits twice while they where made in accordance with the guidelines. So I am happy seeing you starting this debate here. And I would like to see if the guidelines can be changed at the end. But as long is this did not happen, please do not revert correct edits which I made at the end of such deadlines. --Augustgeyler (talk) 16:24, 6 December 2020 (UTC)
- I have not accused you of not following the guidelines. I criticize the literal interpretation of the guidelines, which in my opinion contradicts the "spirit" of the "consensual(!) review" procedure. I have justified my reinstatement of "QICtotal" in detail, both in the editing comment and here on this page, to gain time for discussion. You bluntly reset this attempt to initiate a discussion only a few hours later in the middle of the night without giving reasons and without responding to the need for discussion. Such behavior is unfriendly and is often the beginning of an editing war, which I of course did not get involved in. But since the QICbot has "done" the job for you, you have already "won". You can be proud to have stalled a discussion before anyone else could react. --Smial (talk) 13:08, 7 December 2020 (UTC) Translated with www.DeepL.com/Translator (free version)
- I just can wonder about what your are writing. I just followed the guidelines and your are telling something about wining and a bot who might be doing my job... – I am very sorry that I did not found your discussion here earlier. But I hardly disagree with you about your procedure. It was no good way to act against the guidelines by reverting my edit before the rules could be changed after a discussion. That is as if you would park your car in a parking ban and telling others complaining about it, that you have started a legislative initiative for abolishing parking bans and you are going to ignore them from now on. Sorry: please make your discussions here but respect the guidelines as long as they are not changed. --Augustgeyler (talk) 15:44, 7 December 2020 (UTC)
- It is a pity that you do not want to or cannot understand my point. I plead here for fairness and for equal chances for all candidates stranded on CR. You prefer to insist on a literal interpretation of a sentence in the rules to defend an action that has a "G'schmäckle" (taste), because by archiving the photo in such a hurry you have enforced your first judgment on the photo in question. That is not the fine way. -- Smial (talk) 13:51, 9 December 2020 (UTC) Translated with www.DeepL.com/Translator (free version)
- I just can wonder about what your are writing. I just followed the guidelines and your are telling something about wining and a bot who might be doing my job... – I am very sorry that I did not found your discussion here earlier. But I hardly disagree with you about your procedure. It was no good way to act against the guidelines by reverting my edit before the rules could be changed after a discussion. That is as if you would park your car in a parking ban and telling others complaining about it, that you have started a legislative initiative for abolishing parking bans and you are going to ignore them from now on. Sorry: please make your discussions here but respect the guidelines as long as they are not changed. --Augustgeyler (talk) 15:44, 7 December 2020 (UTC)
- I have not accused you of not following the guidelines. I criticize the literal interpretation of the guidelines, which in my opinion contradicts the "spirit" of the "consensual(!) review" procedure. I have justified my reinstatement of "QICtotal" in detail, both in the editing comment and here on this page, to gain time for discussion. You bluntly reset this attempt to initiate a discussion only a few hours later in the middle of the night without giving reasons and without responding to the need for discussion. Such behavior is unfriendly and is often the beginning of an editing war, which I of course did not get involved in. But since the QICbot has "done" the job for you, you have already "won". You can be proud to have stalled a discussion before anyone else could react. --Smial (talk) 13:08, 7 December 2020 (UTC) Translated with www.DeepL.com/Translator (free version)
- @Smial: I did so, to follow the guidelines. The 8-days deadline for undecided is logical, because undecided means there are as many opponents as there are supporters. This proves there is now majority for one side. To give those candidates an 8 day chance is a good thing, because otherwise there would be simply NO decision. But in cases of nominations where a majority of at least one vote came out, it is a clear thing, these CRs should end after 48h with no new votes. I can not be blamed for all the others who did not take their 48 hour chance to take part in that particular decision. My own candidates were rejected ore promoted that way in the past. So I can understand your feelings. It can be disappointing seeing no one taking the time to take part in that decision. But what disappoints me here is, that you reverted my edits twice while they where made in accordance with the guidelines. So I am happy seeing you starting this debate here. And I would like to see if the guidelines can be changed at the end. But as long is this did not happen, please do not revert correct edits which I made at the end of such deadlines. --Augustgeyler (talk) 16:24, 6 December 2020 (UTC)
Proposal: Commons_talk:Quality_images_candidates#Grace_period_in_consensual_review
Use of 16-bit tif format to preserve depth of colors
Hello all, I would like clarification on whether .TIF submissions are welcome.
State:
- We require full spatial resolution (pixels) when available
- Lossy compression is necessary to bring a reasonable image size (bits) with visual degradation that's imperceptible to the human visual system.
- JPEG and WEBPG are the only lossy formats supported on Wikimedia Commons at the moment. (and typically on web browsers.) These formats only support 8-bit color precision (where each color channel is a value between 0 and 255).
Problem:
Wide color gamut is a thing. Current cameras and imaging software deal with higher precision: high dynamic range (+/- luminosity), wide color gamut (+/- chroma), and larger bit depth (12-16 bits source, 32-bit processing)
A color profile such as en:sRGB is used to encode images, so a value of 0 green in sRGB will lie on the top-left side of the sRGB triangle shown to the right (middle). Anything "greener" is clipped to that value, like highlights.
When using JPEG with the sRGB profile (its default use), 8-bits color values (0-255) are scaled to encompass the range of values present in the sRGB triangle.
Nowadays, there are color profiles which encompass many more values. The largest color space in the graph above shows en:ProPhoto RGB which goes so far as to allow imaginary values (wasteful but necessary to capture parts of the curve). My personal goto is Linear Rec. 2020 standard (shown below), which encompasses a wide range of real values (and is used internally by some development softwares like darktable).
JPEG and most codecs are perfectly capable of working with any given color profile, but they are limited to 8-bit values still. While the difference between 42 and 43 is minimal when the range of value represents a relatively small subset of colors, it becomes a big jump when 0-255 is supposed to represent every possible shade of red/blue/green. This creates en:Colour banding
Need:
There are several standardized lossy image codecs that solve the bit depth issue by allowing 16-bit images (0-65536), namely JPEG 2000, JPEG XS, JPEG XT, JPEG XL. None of them are supported by Wikimedia Commons or even common browsers such as Firefox. The future of JPEG XL in Firefox looks promising [1], but we are not there yet. (AVIF may get there sooner and it supports 12-bits or 4096 values which is good too.)
We have the higher resolution images but no suitable format to store it. The only current solution is the .TIF format which applies only lossless compression (hence the resulting file is large, needs to be downloaded, and loads slowly on the web). Wikimedia Commons describes TIFF as an archival format that should not be used for images intended to be displayed (Commons:File_types#TIFF). However, Wikimedia Commons handles TIFF well, including the embedded color profiles, and it creates full-resolution JPEG previews that should be used for most viewing purpose.
I think that TIFF images with 16-bit color resolution should be welcomed when the colors would otherwise be clipped (until a better technical solution is in place), for the same reason we encourage full pixel resolution. It may not make sense for most viewers now, but we provide the fullest resolution so that the end-user can do whatever they would like with the images now or in the future, and the Wikimedia servers provide thumbnails of any desired size (from article thumbnails to full pixel resolution).
-
The Wikimedia-generated JPEG preview looks fine for most viewing purposes. The TIF file provides some more color depth when it's necessary (eg red)
-
This is a disaster in 8-bit/JPEG, the TIF file is the only way to get proper colors
-
This type of blue isn't handled in sRGB so the image is saved with Rec. 2020 profile again. As expected, the Wikimedia preview looks quite bad with en:Colour banding because 8-bit isn't enough to capture this wide range of color.
-
The same picture saved in JPEG with sRGB profile. 8-bit is sufficient with this color profile, but this shade of blue isn't part of the sRGB profile so the color differs.
--Trougnouf (talk) 14:42, 4 December 2020 (UTC)
In the meantime I'm at a loss. The hardware and software tools I use often go beyond the old sRGB profile and I don't want to knowingly degrade the colors of my pictures to conform to what was deemed necessary in 1996. I also don't want to hold off uploading pictures to Wikimedia Commons for a year while I wait for modern standards to be implemented. I know that I won't go back to uploading my old pictures again in the future when I can't even catch up with the current ones.
TIF works well since nothing is lost, though the preview can (currently) be ugly, but that'll likely change in the future when the previews will be processed with a proper 16-bit lossy format. I guess the best compromise for now would be to upload two versions of my pictures, Rec. 2020 16-bit TIF and sRGB 8-bit JPEG, and explain that in the description / other versions (like in the train picture pair above), though that's a lot of work. --Trougnouf (talk) 21:57, 5 December 2020 (UTC)
- Sorry, ich kann das nur mit Google übersetzen und verstehe es vielleicht falsch. Warum unterstützen wir nicht DNG? --Ralf Roletschek 01:06, 6 December 2020 (UTC)
- DNG contains the raw sensor data, it needs to be processed by a software (such as Lightroom, darktable, etc, even viewers) which applies various modules (s.a. demosaic, white balance, denoising, crop, contrast, base curve / filmic, ...) before displaying it or exporting it to a "flat" format like JPEG/TIFF/PNG/... The DNG file is never modified directly, instead the editor creates a "sidecar" file which contains a list of instructions needed to get the desired result. Using DNG is thus not suitable here because the image either hasn't gone through any of the required processing, or if the "sidecar" file is included then the result would differ with every viewer/editor/browser (if it's compatible at all, typically it's not). A flat file format is needed to ensure that the image looks the same as intended no matter what software the user uses. --Trougnouf (talk) 09:34, 6 December 2020 (UTC)
- Vor allem aber ist DNG mit Patenten beladen, also kein freies Format und damit von vorherein ausgeschlossen. DNG ist aber nicht das einzige RAW-Format und für ein freies RAW-Format ließe sich die Frage stellen. --C.Suthorn (talk) 16:58, 6 December 2020 (UTC)
- DNG beinhaltet die Daten der Entwicklung, ist nicht nur RAW. Wohin es führt, auf sogenannten freien Formaten zu beharren, zeigt die Verbreitung von Videos auf Commons. Inhalte sollen frei sein, Dateiformate sind egal --Ralf Roletschek 19:01, 6 December 2020 (UTC)
- I have never analyzed in detail the DNG files that fall out of my Pentax cameras, but I am pretty sure that the "development data" they contain is as proprietary as that in the original Pentax PEF format. --Smial (talk) 12:49, 7 December 2020 (UTC)
- I don't know of an editor that writes development data directly in the DNG file. Which one are you referring to? --Trougnouf (talk) 08:32, 7 December 2020 (UTC)
- Photoshop --Ralf Roletschek 12:01, 7 December 2020 (UTC)
- I wasn't aware that some editors could write directly to the RAW/DNG file, online search results are unclear about Photoshop's behavior but it looks like it may indeed be possible to append the sidecar directly to the RAW/DNG file [2][3]. Either way it would be a list of instructions that's specific to Photoshop, so the original pixel values are never overwritten and any other software would likely not interpret these instructions, reverting to the original raw image. (DNG is a standard for raw image pixels/metadata but Photoshop instructions are not standardized.) --Trougnouf (talk) 12:23, 7 December 2020 (UTC)
- I understand raw files as "digital negative". A software that edits around in it and changes something would be deleted from my computer faster than I could say "WTF?". --Smial (talk) 12:49, 7 December 2020 (UTC)
- DNG beinhaltet die Rohdaten (die auch nicht angerührt werden) sowie die Metadaten und Entwicklungsdaten. Man kann das extern in XMP speichern aber das widerspricht dem Konzept. Die Rohdaten kann man nicht verändern, nur mit brutalen Methoden (HEX-Editor), Photoshop bzw. DNG bieten dafür kein Werkzeug. --Ralf Roletschek 17:16, 7 December 2020 (UTC)
- Right. I think it would be beneficial for Wikimedia Commons to allow raw/DNG files, but that's a different use-case (not meant for direct human consumption, good archival method to get better results in the future, open to the users' creativity). There is a discussion on [4] --Trougnouf (talk) 17:29, 7 December 2020 (UTC)
- DNG beinhaltet die Rohdaten (die auch nicht angerührt werden) sowie die Metadaten und Entwicklungsdaten. Man kann das extern in XMP speichern aber das widerspricht dem Konzept. Die Rohdaten kann man nicht verändern, nur mit brutalen Methoden (HEX-Editor), Photoshop bzw. DNG bieten dafür kein Werkzeug. --Ralf Roletschek 17:16, 7 December 2020 (UTC)
- I understand raw files as "digital negative". A software that edits around in it and changes something would be deleted from my computer faster than I could say "WTF?". --Smial (talk) 12:49, 7 December 2020 (UTC)
- I wasn't aware that some editors could write directly to the RAW/DNG file, online search results are unclear about Photoshop's behavior but it looks like it may indeed be possible to append the sidecar directly to the RAW/DNG file [2][3]. Either way it would be a list of instructions that's specific to Photoshop, so the original pixel values are never overwritten and any other software would likely not interpret these instructions, reverting to the original raw image. (DNG is a standard for raw image pixels/metadata but Photoshop instructions are not standardized.) --Trougnouf (talk) 12:23, 7 December 2020 (UTC)
- Photoshop --Ralf Roletschek 12:01, 7 December 2020 (UTC)
- DNG beinhaltet die Daten der Entwicklung, ist nicht nur RAW. Wohin es führt, auf sogenannten freien Formaten zu beharren, zeigt die Verbreitung von Videos auf Commons. Inhalte sollen frei sein, Dateiformate sind egal --Ralf Roletschek 19:01, 6 December 2020 (UTC)
- A tiff file whould be downloaded to be evaluated?, for example mediawiki has problems to render the thumbnail image in this image but I want to place this image as a quality image candidate, is it ok? --Wilfredor (talk) 16:31, 6 December 2020 (UTC)
- The file is 1.6GB large. It will probably fail to render a thumb in any file format. --C.Suthorn (talk) 17:32, 6 December 2020 (UTC)
- Yes, but This image can be evaluated on QIC ? --Wilfredor (talk) 23:13, 6 December 2020 (UTC)
- I think it would be reasonable for the reviewer to expect a JPEG version if Wikimedia can't generate a JPEG preview (your example) or the generated preview is degraded beyond QI standards (the two dark examples above). In that case I don't know if a QI label should/can go to the JPEG upload, original TIFF, or both. Note that TIFF can (and should) apply (gzip-based) lossless compression, the file you uploaded takes up 476 MB when it's saved in 16-bit with highest compression level [5]. --Trougnouf (talk) 08:47, 7 December 2020 (UTC)
Verteilung Dateiformate auf Commons: Jpeg 84%, PNG 4 %, Tiff 2 %, Svg 2 %, PDF 2 %, Gif 0,2 %, Webm 0,1 % --C.Suthorn (talk) 17:32, 6 December 2020 (UTC)
- @Trougnouf: Add (gzip-based) lossless compression to the TIFF file will add an additional decompression process on the server and that will make it more difficult to generate thumbnails. --Wilfredor (talk) 00:56, 9 December 2020 (UTC)
- @Wilfredor: I think it actually makes it easier on the server; this type of compression is light and ubiquitous (it's even part of the PNG coding scheme and some filesystems like Btrfs compress everything on-the-fly), and the reduced transfer time (just from hard disk to RAM/cpu) would more than make up for the decompression complexity (this is one reason it works so well in filesystems). One great thing about gzip compression is that we can pick the most complex compression level and it changes nothing on the decoding end, so the complexity burden is on the image creator (us) who uses the highest level of compression. --Trougnouf (talk) 08:16, 9 December 2020 (UTC)
- @Trougnouf: Add (gzip-based) lossless compression to the TIFF file will add an additional decompression process on the server and that will make it more difficult to generate thumbnails. --Wilfredor (talk) 00:56, 9 December 2020 (UTC)
What about using 16-bit-PNG? It is better supported, also has 16bit/channel? --C.Suthorn (talk) 10:37, 7 December 2020 (UTC)
- PNG works as well as TIF, and the current standard allows EXIF metadata and color profile (I wasn't aware of that and EXIF is not yet widely supported), but it's a lossless format like TIF so the resulting file size is roughly the same (tested with the bicycle trail image and Wilfredor's image, PNG brings only 0.6 to 0.7% bitrate reduction over compressed TIF).
PNG images can be viewed directly in most web browsers (instead of downloaded) so that's a significant benefit.
If there is a consensus to accept 16-bit PNG while we wait for a better standard then I would be happy to upload wide gamut images in PNG instead of TIF.
@Ermell: your opinion would be valued since you requested JPEGs in the reviews
--Trougnouf (talk) 11:06, 7 December 2020 (UTC)- PNG ist nicht für Fotos geeignet. --Ralf Roletschek 00:19, 9 December 2020 (UTC)
- I think png is the best option even for film scanning? --Wilfredor (talk) 00:56, 9 December 2020 (UTC)
- Wikipedia verwendet bei der Generierung von Thumbs eine Optimierung, die bei Jpgs für Fotos optimal ist und bei Pngs für Grafiken. Das ist ein bekanntes Problem und per Software lösbar. Grundsätzlich sind PNG, DNG, TIFF identisch, was die Bildqualität betrifft. --C.Suthorn (talk) 03:23, 9 December 2020 (UTC)
- Die Dateigröße ist aber sehr unterschiedlich. TIF ist höchst kompatibel und deshalb auch heute noch das favorisierte Archivformat. DNG kann weitaus mehr Informationen speichern als TIF und wird in den Behörden (allerdings viel langsamer als erwartet) mit der Zeit zum Standard erhoben. Es gibt keinen Grund für die Existenz von PNG, dieses Format spielt absolut keine Rolle und ist überflüssig. --Ralf Roletschek 15:10, 9 December 2020 (UTC)
- Wikipedia verwendet bei der Generierung von Thumbs eine Optimierung, die bei Jpgs für Fotos optimal ist und bei Pngs für Grafiken. Das ist ein bekanntes Problem und per Software lösbar. Grundsätzlich sind PNG, DNG, TIFF identisch, was die Bildqualität betrifft. --C.Suthorn (talk) 03:23, 9 December 2020 (UTC)
- I think png is the best option even for film scanning? --Wilfredor (talk) 00:56, 9 December 2020 (UTC)
- PNG ist nicht für Fotos geeignet. --Ralf Roletschek 00:19, 9 December 2020 (UTC)
voting at CR
I have noticed that votes recently are often and increasingly being cast on CR without any justifying comment. Please remember that this is not just a vote, but a review. After all, the comments should always help the photographer to recognise shortcomings and to correct them in the future. Even with a "pro", it is always useful for others to know what the motivations are. --Smial (talk) 10:46, 22 December 2020 (UTC)
Cat pictures, etc.
Another thing that's come up is whether it's legitimate to vote down a nomination because you're bored with the motif, or perhaps that portraits of someone's cat are equivalent to selfies and therefore off-topic for Commons (I think that's a difficult argument to make, though). Should we adopt some clear standard of how many photos of a particular motif are too many, should we just judge every photo individually in terms of how good it is, or does someone want to offer another guideline? I'm not accustomed to thinking of QIC as a place where only a limited number of otherwise acceptable photos per motif are passed, but such a guideline could be adopted if there's a consensus for it. -- Ikan Kekek (talk) 22:38, 23 December 2020 (UTC)
- From guidelines: only nominate your best work. Every nominator should hence choose only one frame per set and nominate it. IMO it's just common sense. —kallerna (talk) 23:11, 23 December 2020 (UTC)
- I see under "Number of nominations": "Carefully select your best images to nominate. No more than five images per day can be added by a single nominator." But I don't think we're supposed to judge nominations against others by the same photographer, are we? How would we apply this? For great photographers, should we decline images that are merely very good? -- Ikan Kekek (talk) 00:00, 24 December 2020 (UTC)
- Some people upload only their best work to Commons, one frame per set, and then of course all their photos could be QI (or FP). But other upload almost every single frame they shoot, and then it's another case. —kallerna (talk) 07:10, 24 December 2020 (UTC)
- I see under "Number of nominations": "Carefully select your best images to nominate. No more than five images per day can be added by a single nominator." But I don't think we're supposed to judge nominations against others by the same photographer, are we? How would we apply this? For great photographers, should we decline images that are merely very good? -- Ikan Kekek (talk) 00:00, 24 December 2020 (UTC)
- @ Ikan I think here it's certainly not about the servers, but much more about the usability of categories; especially for those less or not familiar with Commons, Wikipedia etc. --A.Savin 00:08, 24 December 2020 (UTC)
- I'm probably on the lenient side in terms of how much of FP standards should carry into QIC, but for me an image is an automatic pass if I cannot identify any clear flaws with it (and "being boring" is not a clear flaw). -- King of ♥ ♦ ♣ ♠ 03:31, 24 December 2020 (UTC)
- To A.Savin: The categories are Category:Indoor cats, Category:Lying cats (sideways) and Category:British Shorthair cats. What's the issue in terms of usability? -- Ikan Kekek (talk) 04:21, 24 December 2020 (UTC)
- There is also problem with flooding categories. Slightly off-topic, but just couple weeks ago I tried to find a proper picture of Rialto Bridge in Venice. There are hundreds of photos of the bridge in Commons, but not one decent version. It took me lots of time to scroll trough all the photos - from a common subject like the Rialto Bridge, we have too many low-quality shots. Maybe on my next Venice-trip I manage to take my DSLR with me... —kallerna (talk) 07:10, 24 December 2020 (UTC)
- To A.Savin: The categories are Category:Indoor cats, Category:Lying cats (sideways) and Category:British Shorthair cats. What's the issue in terms of usability? -- Ikan Kekek (talk) 04:21, 24 December 2020 (UTC)
- SDC will(can) be much much more powerfull than cats... in the extand that you don't need cats..., exemple, here are all the qualities images of arch brigdes in Italy, of course the relevance of certain results may be questionable, but not more than if someone put a dog image within a bird category...):
#shows files that depict arch bridges in Italy and with the Quality Image assessment
#defaultView:ImageGrid
SELECT ?file ?image
WITH
{
SELECT ?item
WHERE
{
SERVICE <https://query.wikidata.org/sparql>
{
?item wdt:P31 wd:Q158438 .
?item wdt:P17 wd:Q38 .
}
}
} AS %get_items
WHERE
{
INCLUDE %get_items
?file wdt:P180 ?item .
?file schema:contentUrl ?url .
?file wdt:P6731 wd:Q63348069 .
BIND(IRI(CONCAT("http://commons.wikimedia.org/wiki/Special:FilePath/", wikibase:decodeUri(SUBSTR(STR(?url),53)))) AS ?image)
}
Christian Ferrer (talk) 12:04, 24 December 2020 (UTC)
- Kallerna, I think the big problem in finding quality photos in different categories is that the search for FPs, QIs and VIs usually doesn't work. That's not an issue of quantity. But even if overall quantity is an issue, that's separate from the quantity of QIs. -- Ikan Kekek (talk) 12:36, 24 December 2020 (UTC)
- Definitely, it just came into my mind, off-topic as I said. Sometimes I just think we should focus more on the overall quality than quantity. —kallerna (talk) 13:02, 24 December 2020 (UTC)
- And thanks @Christian Ferrer: now I found what I was looking for! —kallerna (talk) 13:06, 24 December 2020 (UTC)
- Definitely, it just came into my mind, off-topic as I said. Sometimes I just think we should focus more on the overall quality than quantity. —kallerna (talk) 13:02, 24 December 2020 (UTC)
Different standards for older photos: Yes or no?
When I vote on nominations here, I try to judge them according to what I understand to be the QI criteria and generally accepted standards of the community. However, I'm confused about what to do with nominations of digital photos from, say, 2007 that are smaller and less technically accomplished than we normally expect from photos taken in the last 4 years or so. In the past, someone pointed me to a discussion in which it was agreed upon that photos should be judged according to the standards used in the year they were taken, so that if a photo taken in 2007 presumably would have passed QIC if nominated in 2007, we should promote it now. But I'm now seeing other users disagreeing with this and advocating an alternative position, namely that all photos should be judged using the same standards, regardless of when they were taken.
I have no stake in advocating either of these positions, but I would hope we could come to some form of coherent consensus, so that even though we will continue to disagree on the margins, we could agree at least on the basis we're using to judge each nomination. -- Ikan Kekek (talk) 16:16, 23 December 2020 (UTC)
- My point in the discussion was that already in 2007 the "criteria" was the same - actually I feel that the QI criteria has lowered. Nowadays people nominate near duplicates, rainy day photos, photos with no clear subject, photos of information boards etc. Of course cameras have evolved, but the difference is not that huge. It would be far too difficult to have different standards for "older" photos - all photos should be judged using the same standards. And as I have said before, I definitely feel that we should raise the bar. People are nominating whatever just to get as many QIs as possible (quality or quantity?). —kallerna (talk) 23:08, 23 December 2020 (UTC)
- I take it, you were judging photos at QIC in 2007 and found that the standards were higher at the time? I'd certainly welcome a broader discussion of standards. I'm happy to do my best to apply any standard that's agreed upon. I don't see what's wrong with a rainy day photo, though. -- Ikan Kekek (talk) 00:02, 24 December 2020 (UTC)
- How should we judge then? A table of technical standards for each year since QI's establishment? --A.Savin 00:04, 24 December 2020 (UTC)
- Of course when you get that detailed, it does become absurd. What I've been doing is allowing for smaller pictures in earlier years and allowing for a somewhat greater degree of unsharpness and noise, but as I said, I'll judge things according to whatever standards a consensus supports. Of course if I start rejecting like 75% or even 50% of older photos that are nominated, I'll get a lot of flack for that as unfair. -- Ikan Kekek (talk) 04:24, 24 December 2020 (UTC)
- One thing I have always insisted on at both QIC and FPC, is that pictures should be judged not on their sharpness, but on "true resolution". So a 100 MP panorama could get away with a little blur at FPC (and a lot of blur at QIC), but a 4 MP image will be held to the highest standard at QIC (and likely be disqualified from FPC for easy subjects). My standard is 100% perfect edge to edge sharpness at 2 MP. Older cameras are capable of that, and modern 24 MP APS-C DSLRs with kit lenses should be judged by the same standard as well (rather than expecting great results at full resolution). -- King of ♥ ♦ ♣ ♠ 04:58, 24 December 2020 (UTC)
- That makes a lot of sense. -- Ikan Kekek (talk) 05:49, 24 December 2020 (UTC)
- KoH’s suggestion seems very reasobable to me. --Aristeas (talk) 16:31, 24 December 2020 (UTC)
- "Higher standards back in 2007", that's a bold statement, —kallerna can you please proof that? and please, don't send me another image of Diliff which were in a completely different league, we are not interested in that bookend but in the middle or rather in the low bookend. All you or anybody can do to form an opinion is just looking into e.g. Commons:Quality_images_candidates/Archives_August_2007
- It's just human, wenn you look into images and 95% have more than 10 MPx and even 50% more than 20 MPx that you will perceive a drop of quality when looking into a 5 MPx image, no matter what leve of sharpness. On the other way around if 80% of the images have 10 MPx or less then a 5 MPx image is just one more off the shelf.
- Agree though that all images should be sharp at 2 MPx but whenever we rise that standard, if we succeeed somewhen to do so, that shouldn't be done retroactively for older images, which are very useful in most cases for Wikipedia and with the devopment tools today even in JPEG format they can be perked up to a QI today --Poco a poco (talk) 13:51, 25 December 2020 (UTC)
- Well, going through the August 2007 archives IMO proofs my point: overall the nominations are composition- and subjectwise better. —kallerna (talk) 21:07, 25 December 2020 (UTC)
- KoH’s suggestion seems very reasobable to me. --Aristeas (talk) 16:31, 24 December 2020 (UTC)
Requirements
For a QI page some requirements have to be fulfilled, for example a description. IMO it's time to add a requirement for structured data. Some items should be mandantory, some recommended. For example depicts (P180), creator (P170), copyright status (P6216), copyright license (P275) and source of file (P7482) should be mandantory. --XRay 💬 08:30, 4 January 2021 (UTC)
- Das soll gefälligst ein Bot erledigen. Irgendwann gehts hier nur noch um Drumherum und das Foto ist nicht mehr relevant. Mir wurden früher Bilder abgelehnt, die nicht die Beschreibung im Dateinamen hatten... --Ralf Roletschek 11:17, 4 January 2021 (UTC)
- Oppose If you want to massively restrict participation in QIC and limit it to just a handful of contributors, that seems to me to be the most effective way.
The state of QIC is already pretty bad, essentially it's always the same relatively small group of contributors who fill the candidate list and an even smaller group who then also give ratings. This is very much at odds with the original idea of using QIC to improve the average quality for photos in Wikipedia on as broad a basis as possible. Instead, it has largely become a competition of know-it-alls with strong narcissistic tendencies. Also, some people seem to be only interested in collecting as many medals as possible and submit large amounts of really insignificant snapshots and technically borderline images, probably in the hope that not too many of them will end up on CR.
This already quite small group of contributors would be further reduced by the required condition, which in my opinion would make the point of QIC completely absurd. --Smial (talk) 11:45, 4 January 2021 (UTC)Translated with www.DeepL.com/Translator (free version) - A bot can't do the work for some SDC items, but I agree with Smial. It may be too much for the nominators. --XRay 💬 07:09, 5 January 2021 (UTC)
Which are the best? What's "best work"?
@Kallerna: often quotes "only nominate your best work". The full sentence in the guidelines is "Carefully select your best images to nominate." There are different opinions and an interpretation is possible. IMO the sentence means only images without errors like unsharpness or similar issues. That's because QIC is not VIC or FPC. So similar images can be nominated. That is also the path that we have been treading for years. But Kallerna's interpretation may be right too. Should the sentence be improved? Should the statement be clarified? --XRay 💬 08:24, 4 January 2021 (UTC)
- Comment This requirement can be deleted completely in my intemperate opinion. --Smial (talk) 11:48, 4 January 2021 (UTC)
- +1 I agree. We do not need the sentence in the guidelines at QIC. --XRay 💬 07:08, 5 January 2021 (UTC)
- We now have two suggestions or requests for changes or clarification regarding the rules. See above under "deadlines". Are there other people interested in this topic? --Smial (talk) 15:41, 5 January 2021 (UTC)
- I'll try to make proposals. --XRay 💬 05:36, 6 January 2021 (UTC)
- We now have two suggestions or requests for changes or clarification regarding the rules. See above under "deadlines". Are there other people interested in this topic? --Smial (talk) 15:41, 5 January 2021 (UTC)
- +1 I agree. We do not need the sentence in the guidelines at QIC. --XRay 💬 07:08, 5 January 2021 (UTC)
Proposal: Commons_talk:Quality_images_candidates#Removing_a_sentence_at_number_of_nominations
Category:Quality images and archives pages
Hello,
First of all excuse me for my english, this is essentially a machine translation.
Following Commons talk:Quality images candidates/Archive 24#Bot? and Commons talk:Quality images candidates/Archive 24#Bot didn't finished today, I looked into the problem.
In the archive pages (until December 31, 2020), I noticed some quirks of the bot:
- In Commons:Quality images candidates/Archives August 19 2013, there were images in fifteen copies (for example File:Claudine Glot, Rencontres de l'imaginaire de Brocéliande 2013, Paimpont, France.jpg)! (I have deduplicated all the archive pages);
- In Commons:Quality images candidates/Archives October 08 2013, Commons:Quality images candidates/Archives October 11 2015 et Commons:Quality images candidates/Archives October 15 2015, il y a File:Brimham Rocks MMB 19.jpg (same date and time);
Also some user quirks:
- File:2011 Dubrownik, Widok na morze z tarasu widokowego (03).jpg promoted in Commons:Quality images candidates/Archives December 14 2014 but refused after in Commons:Quality images candidates/Archives December 22 2014;
- File:Nottingham railway station MMB 40.jpg offered five times and promoted twice;
- others of the images are proposed and promoted several times, record for File:Eglise Saint-Cyr-et-Sainte-Julitte de Canac 01.jpg, five times!
So following a comparison of the category and the archive pages:
- I have a list of 1 300 images present in the archives and not present in Category:Quality images. The first in my list for example: File:12-12-15 Mahnmal Köttingen 04.jpg, the template Template:QualityImage has disappeared and File:14-09-02-oslo-RalfR-442.jpg where the bot did not go;
- And the invese, a list of 1900 images present in Category:Quality images and not present in the archives. The first of my list for example: File:-1360 Schreiberfigur anagoria.JPG, add the Template:QualityImage without reasons;
I will correct this over the next few days, thank you for your indulgence in case of error.
Gzen92 [discuter] 15:43, 6 January 2021 (UTC)
- The last example: Some users just add an image to a quality image category, but never send it to QIC. I remove such images from the categories, if there is no archive page and QIC bot never updated the page and it looks like not fulfilling the guidelines. This may be not OK, if the bot failed to create the archive page and failed adding the template, and the template was added manually, but the user has forgotten to create the archive page manually. If, if, if,... --XRay 💬 16:54, 6 January 2021 (UTC)
- I understand the problem. For this file, File:-1360 Schreiberfigur anagoria.JPG, I watched l'historique de QIC, no trace of the file in the diff I looked at, no trace of User:Anagoria. Could we offer these files to QIC (but 1900 is a lot), and for refusals, we remove the template (and categories) on the files ? Gzen92 [discuter] 08:40, 7 January 2021 (UTC)
- Note : 58 files QI of User:Anagoria without vote... Gzen92 [discuter] 13:27, 7 January 2021 (UTC)
- Done First case, in the archives and no model (after manual checks, 629 files), see pages list. Gzen92 [discuter] 11:22, 12 January 2021 (UTC)
- Note : 58 files QI of User:Anagoria without vote... Gzen92 [discuter] 13:27, 7 January 2021 (UTC)
- I understand the problem. For this file, File:-1360 Schreiberfigur anagoria.JPG, I watched l'historique de QIC, no trace of the file in the diff I looked at, no trace of User:Anagoria. Could we offer these files to QIC (but 1900 is a lot), and for refusals, we remove the template (and categories) on the files ? Gzen92 [discuter] 08:40, 7 January 2021 (UTC)
- Second case (part 1 of 2) : 374 images with Template:QualityImage and not in the archives ("Commons:Quality_images_candidates/Archives_xxx"). But images in "Commons:Quality images/Subject/xxx" or "Commons:Quality images/Technical/xxx". I created the page Commons:Quality images candidates/candidate list/January 21 2021 (1) to vote, all pictures should be IQ. Gzen92 [discuter] 11:03, 21 January 2021 (UTC)
- Second case (part 2 of 2) : 622 images with Template:QualityImage, not in the archives ("Commons:Quality_images_candidates/Archives_xxx") and not in "Commons:Quality images/Subject/xxx" or "Commons:Quality images/Technical/xxx". I created the page Commons:Quality images candidates/candidate list/January 21 2021 (2) to vote, probably few QI.
- I'm in charge of applying the results accordingly but i need your promotions. But there is no QI vote helper, I do not know how to do. Put them on the main voting page? Gzen92 [discuter] 15:12, 21 January 2021 (UTC)
- Just a hint, your bot is adding the QI template to files that have never been QI but happen to have the name of a file that had been awarded the QI stamp. Let me elaborate. A picture with the title aaa.jpg becomes QI, I move it to aab.jpg, then I upload a new image with the name aaa.jpg which has never been QI. You bot removes the QI from the file that was QI and gives it to the file which wasn't. I believe that you shouldn't rely on files names but rather on the object ID, which should be unique. Poco a poco (talk) 20:34, 21 January 2021 (UTC)
- Thank you I understand. It's twisted ! Gzen92 [discuter] 07:55, 22 January 2021 (UTC)
- For Commons:Quality images candidates/candidate list/January 21 2021 (1) 357 images have QICbot in the history with adding of the model -> ok. 17 do not have it -> I remove Template:QualityImage and categories QI. Gzen92 [discuter] 10:14, 28 January 2021 (UTC)
- For Commons:Quality images candidates/candidate list/January 21 2021 (2) 51 images have QICbot in the history with adding of the model -> ok. 571 do not have it -> I remove Template:QualityImage and categories QI. Gzen92 [discuter] 15:11, 28 January 2021 (UTC)
- Thank you I understand. It's twisted ! Gzen92 [discuter] 07:55, 22 January 2021 (UTC)
- Just a hint, your bot is adding the QI template to files that have never been QI but happen to have the name of a file that had been awarded the QI stamp. Let me elaborate. A picture with the title aaa.jpg becomes QI, I move it to aab.jpg, then I upload a new image with the name aaa.jpg which has never been QI. You bot removes the QI from the file that was QI and gives it to the file which wasn't. I believe that you shouldn't rely on files names but rather on the object ID, which should be unique. Poco a poco (talk) 20:34, 21 January 2021 (UTC)
- Please be careful when you remove QI badges. For example for File:USA_Old_State_House_1_MA.jpg you can see the positive decision here: Commons:Quality_images_candidates/Archives_January_2007 --Dschwen (talk) 17:00, 28 January 2021 (UTC)
- Hello, my excuses, you are right. I was very meticulous but there was a failure. Sincerely, Gzen92 [discuter] 07:34, 29 January 2021 (UTC)
New user.
"New" user "commonists" sets up a user page on Jan 20. Knows a lot about templates and setting up user page archiving, has no contributions outside of commons, and immediately joins QIC to give uncommented pros and cons. Do we need a voting sock here? --Smial (talk) 10:42, 21 January 2021 (UTC)
- No. @Commonists: Any statement? --A.Savin 12:57, 21 January 2021 (UTC)
- @A.Savin: @Smial: I am a photographer and I have read the guidelines, did I do something wrong? If you want I will not vote anymore, sorry. I just wanted to be useful. Regards --Commonists 14:25, 21 January 2021 (UTC)
- Well, given the fact that by now you uploaded 0 (zero) own images and did 0 (zero) edits apart from QIC and your own userpage, you have no credibility as reviewer, that's for sure. --A.Savin 14:34, 21 January 2021 (UTC)
- Yes, you are actually right, I will upload photos and I will not vote anymore until I have a little more credibility. Excuse me again.Regards --Commonists 14:56, 21 January 2021 (UTC)
- @A.Savin: @Smial: As a newbie myself I'd like to object a bit here. Yes, there are rules to who is accepted to vote—for a good reason for sure—however I see "uncommented"/"Good QI" promotional votes from a lot of non-newbies, which are essentially cross-promoting the shots of the people, who just promoted their own candidates. Objectively if @Commonists: 's voting is coherent and reasonable I don't see why an opinion shouldn't be honored here and suggested that he/she commented more about the decision. Additionally unreasonable verdicts can be objected via discuss, even if this introduces more effort to the process. But in my opinion it's better to welcome new people to the circle and be appreciative than pushing them away immediately—especially if they didn't do something terribly wrong. To prove the point, even if Commonist would've technically acquired "credibility", constructive criticism about the commenting style could've been stated more constructively. I don't mean to step anyone's foot here, but to show a perspective from a "newbie" point of view. —Etaped (talk) 16:23, 27 January 2021 (UTC)
- Nice, but this doesn't change anything on the fact that someone with zero edits has no credibility as QIC/FPC reviewer. Anyone is free to contribute constructively and earn credibility step by step. For me it took three years from registration to first showing up at QIC. --A.Savin 18:45, 27 January 2021 (UTC)
- @Etaped: You are right in that even experienced contributors on QIC often only give a pro or con without comment. I have criticized this in the past, I think more than once, unfortunately without response. But A.Savin is also right: To be considered reputable on QIC, you have to have earned this reputation somehow. This is not achieved by voting without comment. Anyone who provides well-reasoned comments in the votes is welcome. --Smial (talk) 11:59, 28 January 2021 (UTC) Translated with www.DeepL.com/Translator (free version)
- @A.Savin: @Smial: As a newbie myself I'd like to object a bit here. Yes, there are rules to who is accepted to vote—for a good reason for sure—however I see "uncommented"/"Good QI" promotional votes from a lot of non-newbies, which are essentially cross-promoting the shots of the people, who just promoted their own candidates. Objectively if @Commonists: 's voting is coherent and reasonable I don't see why an opinion shouldn't be honored here and suggested that he/she commented more about the decision. Additionally unreasonable verdicts can be objected via discuss, even if this introduces more effort to the process. But in my opinion it's better to welcome new people to the circle and be appreciative than pushing them away immediately—especially if they didn't do something terribly wrong. To prove the point, even if Commonist would've technically acquired "credibility", constructive criticism about the commenting style could've been stated more constructively. I don't mean to step anyone's foot here, but to show a perspective from a "newbie" point of view. —Etaped (talk) 16:23, 27 January 2021 (UTC)
- Yes, you are actually right, I will upload photos and I will not vote anymore until I have a little more credibility. Excuse me again.Regards --Commonists 14:56, 21 January 2021 (UTC)
- Well, given the fact that by now you uploaded 0 (zero) own images and did 0 (zero) edits apart from QIC and your own userpage, you have no credibility as reviewer, that's for sure. --A.Savin 14:34, 21 January 2021 (UTC)
- @A.Savin: @Smial: I am a photographer and I have read the guidelines, did I do something wrong? If you want I will not vote anymore, sorry. I just wanted to be useful. Regards --Commonists 14:25, 21 January 2021 (UTC)
Pings
Is there any way we can make it so pings work on this page? I suspect the issue is how the page seems to treat signatures unusually, but I've never gotten a ping from QIC and regularly miss responses. — Rhododendrites talk | 21:21, 12 February 2021 (UTC)
Proposal to improve the illustration of the template {{Focus stacked image}}
➡ Please vote here. Thanks -- Basile Morin (talk) 04:50, 17 February 2021 (UTC)
"Double" Promotion - deadline of 2 days
The guidelines say that “if there are no objections within a period of 2 days (exactly 48 hours) from the first review, the image becomes promoted or fails according to the review it received.” I am currently experiencing the case where an image already set on promotion has received such a promotion a second time (seen here - File:Airbus Helicopters H135 HB-ZUE.jpg (nominated 2021-02-15)). However, the deadline of two days for the definitive promotion does not seem to run from the first promotion, but from the second. In my opinion, this is not in accordance with the guidelines. What is the reason for this? Chme82 (talk) 08:06, 19 February 2021 (UTC)
- Would it be possible that I receive feedback? Thank you Chme82 (talk) 09:08, 22 February 2021 (UTC)
There are must only be one active nomination of the image. If the image has two active nominations, the most recent one should be declined. --C messier (talk) 22:10, 25 February 2021 (UTC)
- The issue isn't two separate nominations, but one nomination with two people saying "support". The bot must look at the most recent timestamp. So yes, if a second person supports, I suppose you'll have to wait an extra couple days. I don't consider this to be a big problem, considering the great many areas where we could use technical help. Besides, I could imagine many other problems if a bot only looks for the first support. It's not like there's a limit on the total number of active nominations. — Rhododendrites talk | 23:13, 25 February 2021 (UTC)
- Thank you for your feedback. The issue is indeed a single nomination with more than one user saying "support" (seen here - Rescue helicopter Airbus Helicopters H135 T2+ HB-ZUE of Alpine Air Ambulance at its base at Birrfeld airfield in Switzerland (LSZF)). In this case I still don't understand why the QIBot must look at the most recent timestamp. As mentioned before the guidelines says that "if there are no objections within a period of 2 days (exactly 48 hours) from the first review, the image becomes promoted or fails according to the review it received." As a second "support" can not be considered as an objection regarding the first one it should also not cause a extension of the voting period. In my opinion it just doesn't make sense if an image which gets more support than another has to wait longer for promotion. As this really is just a little and rare issue we can leave it as it is. But if someone is going to refresh QIBot's tasks it should be taken into account. Chme82 (talk) 09:22, 1 March 2021 (UTC)
- Most likely explanation: limitations of what the bot can do. You could make a suggestion on the bot's talk page, but it happens infrequently enough and the consequences are small enough that you'll probably just have to wait an extra couple days if it happens again, sorry. — Rhododendrites talk | 16:19, 1 March 2021 (UTC)
- Thank you for your feedback. The issue is indeed a single nomination with more than one user saying "support" (seen here - Rescue helicopter Airbus Helicopters H135 T2+ HB-ZUE of Alpine Air Ambulance at its base at Birrfeld airfield in Switzerland (LSZF)). In this case I still don't understand why the QIBot must look at the most recent timestamp. As mentioned before the guidelines says that "if there are no objections within a period of 2 days (exactly 48 hours) from the first review, the image becomes promoted or fails according to the review it received." As a second "support" can not be considered as an objection regarding the first one it should also not cause a extension of the voting period. In my opinion it just doesn't make sense if an image which gets more support than another has to wait longer for promotion. As this really is just a little and rare issue we can leave it as it is. But if someone is going to refresh QIBot's tasks it should be taken into account. Chme82 (talk) 09:22, 1 March 2021 (UTC)
Media search filter
Hi, just a quick notification, a task in Phabricator have been created about the potential to add an additional filter about Quality/Valued/Featured in Special:MediaSearch which will surely become the default search engine in the long term. Christian Ferrer (talk) 19:58, 2 March 2021 (UTC)
Your process sucks
Apologies for the blunt heading, but as just a driveby comment, as someone who has passed by this area on occasion but isn't a regular, the way this process is set up is infuriatingly difficult to navigate. I have the "nominate this image as a QI" button, but it never works, and when I try to do it manually, inserting my signature as --~~~~
as seemingly suggested by the editnotice doesn't actually produce a plain signature either. Please put some minimal effort into making this system's UI a little less terrible. I shouldn't need to dig into whatever manual exists buried deep somewhere in the archives in order to be able to make a simple nomination. {{u|Sdkb}} talk 23:27, 4 March 2021 (UTC)
- Yep, it could be better (ditto VIC, the monthly photo challenge, etc.). There are lots of issues with structure and too few people with the technical ability to do fix them. Many of the people who developed the gadgets, scripts, bots, and processes we use are no longer around, so they degrade over time until someone steps up to take over (like GeneralNotability with VICbot recently). And wishlist items for these processes don't get much love, it seems.
- It's not like someone offers to fix it up and the regulars say "no, we love this!" :)
- As for the specific issues, the QInominator gadget requires two steps: nominate an image on the image page (it sounds like you're doing that) and then edit the most recent day at COM:QIC and click the green button at the top. Is that not appearing? I've been using it for a while, and it seems to work fine for me. The thing is, at some point it started also including some text along the lines of "I, the copyright holder, agree..." that has to be manually removed. That one is filed under inactive (or barely active) gadget developers.
- I don't know why signatures don't work as normal on this page. Perhaps as a kludge to standardize for the bot? Donno. It's annoying that it's impossible to generate a ping as such. It's been raised a few times, most recently by me just above (#Pings). — Rhododendrites talk | 23:59, 4 March 2021 (UTC)
- It was unclear what exactly I needed to do to get the nomination copied from the image page, but I think I was able to. When I clicked the green banner at the nomination page to paste, though, it didn't do anything. So idk. And to your more general point, yeah, I guess this is just part of the large pile of suboptimal technical Wikimedia processes; maybe someday the WMF will decide to provide enough support for them to be remedied. {{u|Sdkb}} talk 00:19, 5 March 2021 (UTC)
Invitation to participate in research to understand how you work with media
An invite to participate in a research, has been extended to media creators by MRaish (WMF). Take a look at WikiProject Photography. --Cart (talk) 17:31, 5 March 2021 (UTC)
Minimum requirement to evaluate a quality image
I think that users should have a minimum quantity of quality images to be able to evaluate quality images, I think that 100 could be a good number. In this way we force people to take time to learn how it works instead of going directly to evaluate, what do you think? --Wilfredor (talk) 20:47, 3 March 2021 (UTC)
- Not easy to check, I suppose. And we have users who review properly despite less than 100 own QI (e.g. Franz van Duns or LexKurochkin). --A.Savin 12:33, 4 March 2021 (UTC)
- Yes, it may be difficult to verify, the ideal would be a flag --Wilfredor (talk) 19:26, 4 March 2021 (UTC)
- Something that could encourage this is to have a counter of images reviewed per user, that would stimulate competition and experience. For example "Quality Image reviewed by XXX" --Wilfredor (talk) 21:00, 4 March 2021 (UTC)
- But we also demand that as many pictures be rated as nominated. --XRay 💬 12:40, 4 March 2021 (UTC)
- This is another problem, QICvote improves the ease of evaluating photos. Sometimes I see too many images to evaluate --Wilfredor (talk) 19:26, 4 March 2021 (UTC)
- We have CR to protect us from obvious errors in evaluation. If we make it harder to review images, QIC will probably fail because we have too few reviews as it is already. --MB-one (talk) 08:59, 7 March 2021 (UTC)
- I also consider a restriction in the way suggested by Wilfredor to be counterproductive. If we want to improve something, especially in CR, then I would suggest that every voter should be obliged to give at least a brief explanation. This can also simply be a reference to the reasoning of another voter. Not everyone necessarily has to write an extensive commentary, but pure voting without reasons doesn't really help anyone. --Smial (talk) 11:03, 7 March 2021 (UTC) Translated with www.DeepL.com/Translator (free version)
- We have CR to protect us from obvious errors in evaluation. If we make it harder to review images, QIC will probably fail because we have too few reviews as it is already. --MB-one (talk) 08:59, 7 March 2021 (UTC)
- Creating good quality image requires good quality photographic equipment. Judging the quality of an image does not. Both are also very different skills. For those reasons I very much oppose this suggestion. --Kritzolina (talk) 11:09, 7 March 2021 (UTC)
- Wilfredor, would you like me to be prevented from voting at QIC? I feel sure I have yet to nominate 100 photos at QIC, and all the photos I've nominated are not by me. -- Ikan Kekek (talk) 02:01, 9 March 2021 (UTC)
- Of course not :) ←, it is just a brain storm to improve the Process--Wilfredor (talk) 02:53, 9 March 2021 (UTC)
- Google. Translate: „Um ein Bild von guter Qualität zu erstellen, ist eine gute Fotoausrüstung erforderlich. “ Nein, für gute Fotos braucht man keine teure Technik, dazu habe ich schon mehrere Vorträge auf Wikimanias gehalten. Allerdings denken einige der Abstimmenden so, das stimmt. --Ralf Roletschek 11:50, 9 March 2021 (UTC)
- Ralf, wieviele Smartphonebilder kommen durch den Prozess? Das was du unter einer guten Fotoausrüstung verstehst unterscheidet sich möglicherweise von dem was ich meinte. --Kritzolina (talk) 20:17, 9 March 2021 (UTC)
- Das kannst du hier sehen:
- Telefone und Billigknipsen sind nicht für jedes Motiv geeignet aber sie können mittlerweile viel mehr als man allgemein annimmt. Sie sind schon allein aus optischen Gründen kein Ersatz für eine richtige Kamera aber sie nähern sich immer mehr an und können mittlerweile Dinge, die noch vor 5 Jahren undenkbar waren. Mein Fotolehrmeister brachte mir Ende der 70er bei, daß ein gutes Foto im Auge des Fotografen entsteht, nicht in teurer Kameratechnik. Ich arbeite heute damit, alle Kameras waren bereits mehrere Jahre alt, als ich sie gekauft habe. Nur die Objektive, irgendwie haben sich die Hersteller da mit der Kommastelle vertan. --Ralf Roletschek 11:00, 11 March 2021 (UTC)
- Der Markt für Wechselobjektive ist klein geworden, die Ansprüche höher. Wie sonst, außer im Preis, soll sich das ausdrücken? --Smial (talk) 11:58, 11 March 2021 (UTC)
- Ich muß allerdings zugeben, daß Zeiss-Objektive auch wirklich außergewöhnlich gut sind. --Ralf Roletschek 13:49, 11 March 2021 (UTC)
- Naja, die unterschiedlichen Mengen deiner QIs sprechen ja schon auch ein Sprache. Und wie du sagst, die Bedingungen dafür müssen stimmen. So hab ich mit meiner alten Kamera auch ein paar QIs von Lyriklesungen gemacht, aber deutlich mehr halt mir den Leihkameras von WMDE und jetzt mit meiner neuen. Lichtverhältnisse und Abstand zum Motiv sind da halt nicht immer ideal. --Kritzolina (talk) 12:13, 12 March 2021 (UTC)
- Der Markt für Wechselobjektive ist klein geworden, die Ansprüche höher. Wie sonst, außer im Preis, soll sich das ausdrücken? --Smial (talk) 11:58, 11 March 2021 (UTC)
- Ralf, wieviele Smartphonebilder kommen durch den Prozess? Das was du unter einer guten Fotoausrüstung verstehst unterscheidet sich möglicherweise von dem was ich meinte. --Kritzolina (talk) 20:17, 9 March 2021 (UTC)
- Wilfredor, would you like me to be prevented from voting at QIC? I feel sure I have yet to nominate 100 photos at QIC, and all the photos I've nominated are not by me. -- Ikan Kekek (talk) 02:01, 9 March 2021 (UTC)
- I strongly disagree. That seems too elitist (or at least non-inclusive) of a requirement, and Ikan Kekek is one of our most valued reviewers yet he doesn't upload images himself. A flag might be helpful (less intimidating to review knowing that it's likely to be checked over?) as long as it's voluntary. --Trougnouf (talk) 13:39, 12 March 2021 (UTC)
- Kind of you to say. -- Ikan Kekek (talk) 09:36, 13 March 2021 (UTC)
Proposals
@A.Savin: , @Agnes Monkelbaan: , @Andrew J.Kurbiko: , @Angry Red Hammer Guy: , @Anntinomy: , @Aristeas: , @Berthold Werner: , @Bgag: , @Cayambe: , @Ermell: , @Ezarate: , @Famberhorst: , @Granada: , @GyozaDumpling: , @Gzen92: , @Ikan Kekek: , @Jakubhal: , @Johann Jaritz: , @Kallerna: , @King of Hearts: , @Liberaler Humanist: , @M 93: , @MB-one: , @MIGORMCZ: , @Martinvl: , @Mdaniels5757: , @Michielverbeek: , @Milseburg: , @Moroder: , @Palauenc05: , @PantheraLeo1359531: , @Peulle: , @Pierre André Leclercq: , @Poco a poco: , @Podzemnik: , @Robert Flogaus-Faust: , @Scotch Mist: , @Sebring12Hrs: , @Smial: , @T.Bednarz: , @Tagooty: , @The Cosmonaut: , @Tomer T: , @Tournasol7: , @Trougnouf: , @Velvet: , @Vengolis: , @XRay: , @Саня Новиков: , @Ввласенко: , @LexKurochkin:
The proposals will be closed on 4th February 2021.
Grace period in consensual review
First of all: Fix of the link at Commons:Quality_images_candidates#Consensual_review_rules. Add following rule:
If there are no objections within a period of 2 days (exactly 48 hours) from the last review, the image becomes promoted or fails according to the review it received. If there is no decision, it can be closed after 8 days (exactly 192 hours) with no new review.
See Commons_talk:Quality_images_candidates#deadlines
Discussion
- Question I think this has always been handled this way? --A.Savin 13:17, 7 January 2021 (UTC)
- Yes, it is. But see the Commons_talk:Quality_images_candidates#deadlinesdiscussion above, it's not part of our guidelines. --XRay 💬 13:08, 8 January 2021 (UTC)
- My point was different. The 48-hour deadline is too short. Every photo that lands on CR should have an equally long voting period, and this period should include a weekend so that working people also have a chance to vote in any case. So I advocate decision and archiving for all only after eight days, not only for the undecided images. I think this is a requirement of fairness or equal treatment. --Smial (talk) 10:52, 21 January 2021 (UTC)Translated with www.DeepL.com/Translator (free version)
Voting
Pro
- Support --XRay 💬 05:50, 6 January 2021 (UTC)
Contra
- …
Result: IMO no clear result, only one vote. --XRay 💬 06:44, 6 February 2021 (UTC)
Removing a sentence at number of nominations
At Commons:Quality_images_candidates#Number_of_nominations remove the sentence
Carefully select your best images to nominate.
There are misunderstandings and the sentence is not necessary. Candidates should always fulfill the guidelines: Commons:Quality_images_candidates#For_nominators.
See Commons_talk:Quality_images_candidates#Which_are_the_best?_What's_"best_work"?
Discussion
- Comment If this will be removed, we should all take 1000 photos of same subject, upload them all to Commons and nominate 5 per day to QIC. Will be fun to promote them all. —kallerna (talk) 01:18, 8 January 2021 (UTC)
- Neutral it's ambiguous: some people select their best images to upload, so nominate all of them. someone else might take 1000 pictures of the same subject, so "best images" might mean 20 similar images out of the 1000. On the other hand, it's probably not detrimental. It seems like what people are more concerned with would be something like "try not to nominate too many very similar images" (regardless of whether we remove this line). I would support adding that. — Rhododendrites talk | 01:40, 8 January 2021 (UTC)
- Please do not consider "everything you shot" as a blanket. The guidelines must be met. --XRay 💬 08:47, 8 January 2021 (UTC)
- Well the guidelines are already so low, that with a decent camera it is hard not to succeed. Some of you definitely nominate whatever - and some of you also promote whatever. People think that I'm mean, but I just want to keep some standards - if all photos are "quality images", there are no quality images. —kallerna (talk) 08:58, 8 January 2021 (UTC)
- Regarding the guidelines, I agree with you, they should be up to date. Make a proposal! --XRay 💬 09:11, 8 January 2021 (UTC)
- My suggestion: Add a new discussion section and start collecting improvements for the guidelines. It's a good idea. For example there were several attempts to increase the resolution limit. May be it's time for improvements. --XRay 💬 10:53, 8 January 2021 (UTC)
- Comment I've read about similar images. IMO there is only one meaning of the sentence in the context of the section: Every nominator has to check the image himself for matching the guidelines before nominating. So the image should fulfill the guidelines. There is nothing about other images or similar subjects. If there was a rule for similar images, it should be made clear and the sentence should be replaced by an own section. --XRay 💬 07:04, 9 January 2021 (UTC)
- An alternative solution of removing the sentence may be replacing by something like Please check the compliance of your images with the guidelines before nominating. --XRay 💬 07:09, 9 January 2021 (UTC)
- How do you feel about nominating almost duplicates? IMO it is just common sense that you should not nominate those, even uploading several frames of the same subject is sometimes useless. Because people nowadays nominate whatever, we should add a sentence about similar images, common sense has been lost. —kallerna (talk) 07:17, 9 January 2021 (UTC)
- I don't want discuss about duplicates. The meaning of the sentence is misleading. IMO the sentence does not say anything about duplicates. If so in your opinion it should be clarified. IMO it is just a hint to follow the guidelines. --XRay 💬 09:59, 9 January 2021 (UTC)
Voting
Pro
- Support --XRay 💬 05:50, 6 January 2021 (UTC)
- Support --Trougnouf (talk) 09:45, 6 January 2021 (UTC)
- Support --Aristeas (talk) 13:37, 6 January 2021 (UTC)
- Support --Smial (talk) 11:26, 7 January 2021 (UTC)
- Support Poco a poco (talk) 11:28, 7 January 2021 (UTC) PD: To be honest, I don't understand why actually UPLOADING multiple similar images. The selection of the best images should be done when it comes to uploading to Commons.
- Support --Ralf Roletschek 07:27, 8 January 2021 (UTC)
- Support. I like the proposed alternate language. -- Ikan Kekek (talk) 13:20, 15 January 2021 (UTC)
- Support Thinking about this more, I support. "Your best work" is for FPC. That's where we can be picky about similar shots, too many images of one subject, etc. QI is just a marker of baseline quality. It's not our best; it's a standard. "Does this check all the boxes?" If so, there's no reason not to label it as such. More of a certification than an award. That said, I would still support discouraging near-duplicates along the lines of what I proposed previously. — Rhododendrites talk | 17:03, 24 January 2021 (UTC)
- Support Even though I am new here: Support per Rhododendrites. Commons allows the "other versions" classification. If there's a set of images of the same quality but same object, why not presenting all of them in QIC? Alternatively: Promote automatically all versions of the set, if one gets promoted? The review process should account for these "other versions" as well —Etaped (talk) 15:57, 27 January 2021 (UTC)
- Support --Commonists 13:48, 19 March 2021 (UTC)
Contra
- Oppose IMO nothing wrong about this sentence. It's a good advice, not a rule. I don't understand why some people deem necessary to nominate "just everything they ever shot". --A.Savin 13:17, 7 January 2021 (UTC)
- Oppose Yes, just nominate everything you shoot. And what is the point of QI again? —kallerna (talk) 01:14, 8 January 2021 (UTC)
- Oppose In my opinion there are already too many quality images that look almost identical. Selecting the best of them for nomination might be a very good idea. --Robert Flogaus-Faust (talk) 22:06, 8 January 2021 (UTC)
- Oppose This is a warning that could reduce the excessive number of nominations --Wilfredor (talk) 19:27, 4 March 2021 (UTC)
Result: Support 9 votes, Oppose 3 votes - Proposal accepted. --XRay 💬 06:44, 6 February 2021 (UTC)