Posted under General
Could weibo be a sample prone source? I was never able to get fully going on this source and so I'm not sure whether my own posts could have been sampled (post #2459751). Since there is a watermark I suspect it was.
Would be really good if this source had some automation!
Read the Help:Image Samples wiki and realized that I didn't mark posts tagged with MD5 Mismatch with resized as applicable when processing the images... (I can't imagine many upscaled from official sources).
Would it be beneficial for me to go back and tag all of those then before moving on to the other sites...?
Also, how should images that are the same size but different filesizes be handled...?
Also, changed Deviantart thumbnail to Deviantart sample for standardization purposes.
Updated
BrokenEagle98 said:
Read the Help:Image Samples wiki and realized that I didn't mark posts tagged with MD5 Mismatch with resized as applicable when processing the images... (I can't imagine many upscaled from official sources).
Would it be beneficial for me to go back and tag all of those then before moving on to the other sites...?
Also, how should images that are the same size but different filesizes be handled...?
I think it would definitely help. Hopefully it doesn't take too long though, but you can prioritize whatever you think is most important on your slate first.
We still have a LOT of images to clean through regarding twitter samples, anyway.
Should the child post be tagged duplicate? I could understand why not if this tag should not be used on status:deleted posts. My practice with ArtStation was simply to parent the /original/ and tag the antecedent as duplicate. I would understand however if this practice should be saved for Twitter/Pixiv mixups and sampled images should rather be deleted.
Also. Why not imply all these *_sample tags to something image_sample or something else? It would enable non-builder/non-gold members to help out more easily by being able to formulate broader searches and hell it would help us since I've been wishing it was there. As an aesthetic note such a tag's wiki could replace the help page you created.
Edit!: /original/, not /large/, totally changes the meaning conveyed
Updated
+1 to a parent tag for all image samples. I'll go ahead and create a BUR for this in a new topic.
See topic #13629.
Updated
If I understand correctly from glancing at this thread, does this mean that the URL of the Twitter pictures that should be uploaded should be like in this format:
https://pbs.twimg.com/media/image.jpg:orig
That is, if the image filesize is different from what has been uploaded?
Also, since downloading the image to the computer and then uploading it here doesn't work if the file name ends with '':orig'' or '':large'', is it possible to just copy the URL to the uploading page to have the site download it? Because I tried checking the source data, and it says: ''Error: Couldn't get status ID from URL'' so I'm not sure if the file would still upload properly.
Use the bookmarklet to upload images from Twitter... it'll automatically pull in the largest image size.
If you can't get the bookmarklet to work, then just use the image link as you showed from the uploads page, e.g. https://pbs.twimg.com/media/image.jpg:orig
As far as the error you're getting, that's because the image URL is independent of the Twitter post URL, so there's no way to derive or find the original Twitter Post ID with just the image URL. That doesn't mean that uploading won't work... just that the commentary or artist data won't be auto-populated, which is what "Fetch Source Data" attempts to do.
Okay, thanks.
Also, I'm pretty sure the answer to this is yes, but just to be sure:
If the image filesize from ''.jpg:orig'' is smaller than the one from ''.jpg'' (regular file extension), I should still upload the one from ''.jpg:orig'' anyway, since it's the original image file, right?
Mostly finished with Deviantart, but I've encountered a couple of examples that I don't know what to do with.
For the following, the picture links still work, but the post page...
The first could be tagged with bad ID since the image is no longer publicly available, or it could be tagged MD5 mismatch since it's been replaced with a revision.
The second could also be tagged with bad ID since the image is no longer publicly available.
Or the could be just left as is...
Hmm... this is similar to how pixiv allows revisions, doesn't it? I'd tag the first as md5 mismatch.
Second, definitely bad ID. bad deviantart id, more specifically...
