Topic: General bug report thread - bugs here, bugs there, bugs everywhere, come here and report your bugs

Posted under Site Bug Reports & Feature Requests

It's very likely just the elasticsearch query timing out, the maximum time for that hasn't changed (it's 3 seconds for members, 6 for privileged, and 9 for contributor+)
That happens from time to time 🤷‍♂️

donovan_dmc said:
It's very likely just the elasticsearch query timing out, the maximum time for that hasn't changed (it's 3 seconds for members, 6 for privileged, and 9 for contributor+)
That happens from time to time 🤷‍♂️

Well, that's a little bit annoying. Could more search time be a donor perk or something?

peacethroughpower said:
Well, that's a little bit annoying. Could more search time be a donor perk or something?

That's not really how e621 works
earlopain said he'll be looking into it more when he gets back from vacation

Not a bug, just a nuisance, but could it be shown that a user does not allow dmails without having to send a dmail first?

I'll gladly be corrected if I'm wrong, but updated_at doesn't seem to work for me as a sort order under tag aliases, per the help page. I'll use the below search as an example, where items 0-3 appear to be in order however item 4 (id:50422) is not ordered properly by updated_at. It seems that the sorting method is actually by created_at.

https://e621.net/tag_aliases.json?page=a50340;search[order]=updated_at

As a side note it would be nice if updated_at was available as a sorting option for the regular tags list, as well.

azerai~ said:
I'll gladly be corrected if I'm wrong, but updated_at doesn't seem to work for me as a sort order under tag aliases, per the help page. I'll use the below search as an example, where items 0-3 appear to be in order however item 4 (id:50422) is not ordered properly by updated_at. It seems that the sorting method is actually by created_at.

https://e621.net/tag_aliases.json?page=a50340;search[order]=updated_at

As a side note it would be nice if updated_at was available as a sorting option for the regular tags list, as well.

First of all, are you aware that there is a typo in your link?
It should be https://e621.net/tag_aliases.json?page=a50340&search[order]=updated_at

But besides that, it's just not going to work with that page format.
It'll probably be ordered by ID regardless of what you do with it.

What exactly are you trying to accomplish here?

Exactly what Cinder said, your link is busted, and wither way you can't use order with sequential pagination (a/b), it will always be ordered by id - it has to be to go to that page

donovan_dmc said:
Exactly what Cinder said, your link is busted, and wither way you can't use order with sequential pagination (a/b), it will always be ordered by id - it has to be to go to that page

I see, wasn't aware sequential pagination was only available to be ordered by id, thanks.

cinder said:
First of all, are you aware that there is a typo in your link?
It should be https://e621.net/tag_aliases.json?page=a50340&search[order]=updated_at

But besides that, it's just not going to work with that page format.
It'll probably be ordered by ID regardless of what you do with it.

What exactly are you trying to accomplish here?

Thanks for the tip, in some other links I've used the semicolon worked just fine. ("page=3;limit=5" absolutely works, for example.)
My aim was to get a list of tag aliases (and tags, in a perfect world, too) ordered by the updated_at field. In the case of tag BURs, I don't see how I could keep an offline list of e621's tags updated without regularly pulling a fresh copy from the API, page by page, every so often. Periodically checking the API ordered by updated_at would be much simpler and less resource intensive.

Given that my original question has been answered, though, and that this no longer falls under the category of a possible bug, feel free to ignore.

azerai~ said:
I don't see how I could keep an offline list of e621's tags updated without regularly pulling a fresh copy from the API, page by page, every so often. Periodically checking the API ordered by updated_at would be much simpler and less resource intensive.

There's always the database exports.

azerai~ said:
My aim was to get a list of tag aliases (and tags, in a perfect world, too) ordered by the updated_at field. In the case of tag BURs, I don't see how I could keep an offline list of e621's tags updated without regularly pulling a fresh copy from the API, page by page, every so often. Periodically checking the API ordered by updated_at would be much simpler and less resource intensive.

Given that my original question has been answered, though, and that this no longer falls under the category of a possible bug, feel free to ignore.

I agree with animperfectpatsy – DB exports would probably be the best solution here.
They are only updated daily, though, so if you need more up to date information you would need to query the API.

What exactly are you trying to make here? Why do you need to keep a local list of aliases?

animperfectpatsy said:
There's always the database exports.

cinder said:
I agree with animperfectpatsy – DB exports would probably be the best solution here.
They are only updated daily, though, so if you need more up to date information you would need to query the API.

What exactly are you trying to make here? Why do you need to keep a local list of aliases?

Oh, I didn't know about the database exports, that may sufficient. Thanks.

The bot I'm developing allows people to watch/block tags but, in the interest of not letting users just put whatever gibberish, mistyped tags, or aliases they want to into their blocked/watched list, I query their inputs against a local list of e621 tags/aliases before actually accepting their input. The local list, as opposed to simply querying the API, is specifically to prevent too many API requests at once. This has only presented a problem when a tag has been changed/renamed on e621 after it's already been added to the local list, as checking for the latest tags via the API every hour or so doesn't seem to show tags that were recently updated, only created.

Thanks again for the input.

scth said:
Found something odd with the type metatag. Negatives don't seem to work correctly when combined. It seems that only the last one is taken into account.
For instance, animated -type:gif -type:webm comes up with a page full of gif files. animated -type:webm -type:gif comes up with a page of webm files.

yeah, that's just how metadata searches work, you can only do one search per category. luckily in this situation the webm tag is automatically applied to all .webm posts, so you can use -webm in place of -type:webm.

scth said:
Found something odd with the type metatag. Negatives don't seem to work correctly when combined. It seems that only the last one is taken into account.
For instance, animated -type:gif -type:webm comes up with a page full of gif files. animated -type:webm -type:gif comes up with a page of webm files.

To add on to what sipothac said, meta tags like type indeed can only be used once, and the rightmost will always win due to the parsing order being left to right. If you want to exclude all "animated" content, -animated will typically exclude what you'd consider animated, though may not include all gif posts. -webm -type:gif -animated_png should cover almost everything that can be "animated" in the basic sense of having multiple frames.

scth said:
Found something odd with the type metatag. Negatives don't seem to work correctly when combined. It seems that only the last one is taken into account.
For instance, animated -type:gif -type:webm comes up with a page full of gif files. animated -type:webm -type:gif comes up with a page of webm files.

You don't have to search for type:webm, I reckon there is a meta tag called webm that is either added automatically to webm posts or Pup adds it hinself. Try animated -type:gif -webm

Anyone notice the blacklist isnt working properly? I have images showing up in the posts that are listed in my blacklist. Yes, I'm logged in. The blacklist was working before and its only been relatively recently that its been borked for me.

zaiaku said:
Anyone notice the blacklist isnt working properly? I have images showing up in the posts that are listed in my blacklist. Yes, I'm logged in. The blacklist was working before and its only been relatively recently that its been borked for me.

Click "Re-enable all".

snpthecat said:
Approved unimplicate does not change anything

The implication is still there for all intents and purposes, it's mention in the wiki pages is the only thing removed

Each implication stores all its descendants for performance and for some reason those didn't get updated correctly resulting in both gen 2 and 3 being applied. I will fix this up when I got some time.

Should be fixed now. There were 42 tags that were implying things they shouldn't. Here's the list:

Tags

excess | missing | source_tag
{visual_novel} | {} | burner_(tail_tale)
{visual_novel} | {} | diggle_(tail_tale)
{east_asian_mythology} | {} | eastern_dragon_humanoid
{east_asian_mythology} | {} | eastern_dragon_taur
{visual_novel} | {} | eye_patch_(tail_tale)
{visual_novel} | {} | lick_zack
{visual_novel} | {} | mug_(tail_tale)
{visual_novel} | {} | nap_zack
{third-party_edit} | {} | photomorph
{portal_penis} | {} | portal_strapon
{visual_novel} | {} | run_turn
{visual_novel} | {} | tail_tale
{generation_2_pokemon} | {} | unown_!
{generation_2_pokemon} | {} | unown_?
{generation_2_pokemon} | {} | unown_a
{generation_2_pokemon} | {} | unown_b
{generation_2_pokemon} | {} | unown_c
{generation_2_pokemon} | {} | unown_d
{generation_2_pokemon} | {} | unown_e
{generation_2_pokemon} | {} | unown_f
{generation_2_pokemon} | {} | unown_g
{generation_2_pokemon} | {} | unown_h
{generation_2_pokemon} | {} | unown_i
{generation_2_pokemon} | {} | unown_j
{generation_2_pokemon} | {} | unown_k
{generation_2_pokemon} | {} | unown_l
{generation_2_pokemon} | {} | unown_m
{generation_2_pokemon} | {} | unown_n
{generation_2_pokemon} | {} | unown_o
{generation_2_pokemon} | {} | unown_p
{generation_2_pokemon} | {} | unown_q
{generation_2_pokemon} | {} | unown_r
{generation_2_pokemon} | {} | unown_s
{generation_2_pokemon} | {} | unown_t
{generation_2_pokemon} | {} | unown_u
{generation_2_pokemon} | {} | unown_v
{generation_2_pokemon} | {} | unown_w
{generation_2_pokemon} | {} | unown_x
{generation_2_pokemon} | {} | unown_y
{generation_2_pokemon} | {} | unown_z
{visual_novel} | {} | yellow_(tail_tale)
{visual_novel} | {} | zile_(tail_tale)

And here's the SQL I used to find these because I spend way to much time on it to just keep to myself:

SQL
select
  array(select unnest(a3.descendants) except select unnest(a3.consequents)) present,
  array(select unnest(a3.consequents) except select unnest(a3.descendants)) missing,
  a3.source_tag
from (
  select array_agg(distinct a1.consequent_name) consequents, array_agg(distinct a2.descendant) descendants, a2.source_tag from tag_implications a1
  inner join (
    select distinct descendant, antecedent_name source_tag from (
      select id, antecedent_name, consequent_name, UNNEST(descendant_names) descendant, status from tag_implications
    ) descendants
    where status = 'active'
  ) a2 on a1.antecedent_name = a2.descendant or a1.antecedent_name = a2.source_tag
  where a1.status = 'active'
  group by a2.source_tag
) a3
where
  array_length(array(select unnest(a3.descendants) except select unnest(a3.consequents)), 1) > 0
  or array_length(array(select unnest(a3.consequents) except select unnest(a3.descendants)), 1) > 0

I am unable to update my Blacklist. If I try to use the popout one from the Posts menu bar, it just throws an error: "Failed to update blacklist.". Attempting to do so from my account settings page displays the message "Email has already been taken." and doesn't save the changes I made to the blacklist. I am on the Mobile layout, if that helps.
Not sure if anyone else is experiencing this?

pat180 said:
I am unable to update my Blacklist. If I try to use the popout one from the Posts menu bar, it just throws an error: "Failed to update blacklist.". Attempting to do so from my account settings page displays the message "Email has already been taken." and doesn't save the changes I made to the blacklist. I am on the Mobile layout, if that helps.
Not sure if anyone else is experiencing this?

forum #367524

pat180 said:
I am unable to update my Blacklist. If I try to use the popout one from the Posts menu bar, it just throws an error: "Failed to update blacklist.". Attempting to do so from my account settings page displays the message "Email has already been taken." and doesn't save the changes I made to the blacklist. I am on the Mobile layout, if that helps.
Not sure if anyone else is experiencing this?

Yes, many people.

Anyway, as for the bug I came here to report. Since Earlopain mentioned fixing lingering glitches with the visual_novel tag implications, I've been having another look at cleaning it up with an eye to its eventual nuking. It came to my attention that the search display at the top of posts doesn't expand to hold very long searches, and the clickable search link overflows into the post itself. This is particularly annoying when you have navigation at the bottom of the post enabled, because then the search link renders the playback controls of video posts inaccessible. Sample link:

https://e621.net/posts/3912337?q=visual_novel+-yaminabe_aries+-boyfriend_to_death+-kanon_%28visual_novel%29+-repeat_%28visual_novel%29+-clannad+-angels_with_scaly_wings+-morenatsu+-echo_project+-doki_doki_literature_club%21+-katawa_shoujo+-nekojishi+-stardust_kids+-purrfect_apawcalypse+-great_troubles+-i_promise+-password_%28visual_novel%29+-tennis_ace+-knights_college+-the_human_heart_%28game%29+-extracurricular_activities+-burrows_%28captain_nikko%29+-kemo_coliseum+-far_beyond_the_world_%28series%29+-temptations_ballad_%28visual_novel%29+-dawn_chorus+-psychic_connections

So I'm getting a weird bug. Not everything I uploaded is showing up under User:Gyro. I can find the images through searching for them, and it does show I uploaded them in the history. My blacklist is empty save for a couple users so its not that.

This image for example doesnt show up in that search: https://e621.net/posts/4196582

User:Gyro Search:
https://cdn.discordapp.com/attachments/255506995879215104/1136307735210774648/image.png

A Search to demonstrate where it should sit:
https://cdn.discordapp.com/attachments/255506995879215104/1136308574633595010/image.png

And see I was the poster of it.
https://cdn.discordapp.com/attachments/255506995879215104/1136308051981389874/image.png

millcore said:
Post the bugs you've found here. Please be as specific and as constructive as possible; the more info about the issue you include the easier it will be for us to fix it.

Other forums you can check:
https://e621.net/forum_topics/25717 - List of changes that are NOT bugs
https://e621.net/forum_topics/25748 - Quick FAQ/Q&A thread
https://e621.net/forum_topics/25718 - For documentation problems
https://e621.net/forum_topics/25716 - The theme/aesthetic issues (please read the main post before using)

If you find other bugs floating around in the forum you can post a link to those forums here so they don't get lost.
If you already made another thread for a bug then please go edit [Bug] at the start of the title so it's easy to spot.

Hey! I have an issue where all of my favourites have disappeared :( I changed my password so it should be fine but I had a lotta good stuff in there so it’s irritating. Any way I can fix this?

I just got a question that, just in case I use the phone to see all the post. I got an issue that it is only watching animations, cause sometimes doesnt shows anything, just the background from the website, or sometimes the screen go green.
For example
https://e621.net/posts/4005902?q=dimwitdog
https://e621.net/posts/3150776?q=natural_habitok
This animation i was able to see it like previously but now I'm not able to see it anymore only if I use the pc.
Some videos I'm able to watch it but, yeah just wants to know whats happening.
Thanks for the help.

This has been an issue for a very long time, so it might already be known, but trying to go to a profile page of a user with a very large number of favorites (such as mine) sometimes fails with an error like:

An unexpected error occurred.

Log ID: bd5527b8-50b9-47e4-8db4-2054d2905200

crocogator said:
This has been an issue for a very long time, so it might already be known, but trying to go to a profile page of a user with a very large number of favorites (such as mine) sometimes fails with an error like:

An unexpected error occurred.

Log ID: bd5527b8-50b9-47e4-8db4-2054d2905200

Can confirm, had errors looking at your profile page before

crocogator said:
This has been an issue for a very long time, so it might already be known, but trying to go to a profile page of a user with a very large number of favorites (such as mine) sometimes fails with an error like:

An unexpected error occurred.

Log ID: bd5527b8-50b9-47e4-8db4-2054d2905200

No problem loading your profile for me with my longer timeout :P

But yeah, this is one of the numerous reasons people aren't allowed that many favorites anymore.

faucet said:
No problem loading your profile for me with my longer timeout :P

But yeah, this is one of the numerous reasons people aren't allowed that many favorites anymore.

That makes sense. It's not a huge deal, just a minor inconvenience.

faucet said:
No problem loading your profile for me with my longer timeout :P

But yeah, this is one of the numerous reasons people aren't allowed that many favorites anymore.

Huh, what's the favourites cap now?

Watsit

Privileged

snpthecat said:
) apparently marks the end of a url, even if it is part of the url, so you get links that are only halfway finished like
https://en.m.wikipedia.org/wiki/Bracket_(mathematics)

Limitations of auto-deducing URLs without explicit delimiters. Sometimes they should be part of the URL, and other times they're not supposed to be. Funny enough, I'm having a hard time making an example where the closing parenthesis is not made part of the URL (like https://e621.net/posts/4248464). Other characters, like ! are also routinely mistaken as part of a URL when they're not supposed to be, forcing me to add odd spaces between URLs and various punctuation markers. I kinda wish there was a way to specify a URL with explicit delimiters, like <https://e621.net/posts/4248464> or "foo":<https://e621.net/posts/4248464> or something, so the parser knows exactly where the URL begins and ends without having to guess whether a given character is part of the URL or not.

watsit said:
Limitations of auto-deducing URLs without explicit delimiters. Sometimes they should be part of the URL, and other times they're not supposed to be. Funny enough, I'm having a hard time making an example where the closing parenthesis is not made part of the URL (like https://e621.net/posts/4248464). Other characters, like ! are also routinely mistaken as part of a URL when they're not supposed to be, forcing me to add odd spaces between URLs and various punctuation markers. I kinda wish there was a way to specify a URL with explicit delimiters, like <https://e621.net/posts/4248464> or "foo":<https://e621.net/posts/4248464> or something, so the parser knows exactly where the URL begins and ends without having to guess whether a given character is part of the URL or not.

There is for bare urls

donovan_dmc said:
There is for bare urls

https://en.m.wikipedia.org/wiki/Bracket_(mathematics)?

Huh, thank you. Also apparently ) and ! don't get included as part of the link unless there's a non-space character after it (without <>). Well this is good to know.

It would be nice if it were mentioned in the help: dtext wiki
Also in the dtext wiki the escaping dtext and block formatting anchor link doesn't work because of _

Updated

Crossposting from the re621 thread here since it's a problem with e621 itself, not re621:

faucet said:

lox said:
I found a bug with blacklisting, it has something to do with posts having a child-parent relationship. Here is an example: search id:3954312, blacklisted gynomorph, in the search results the picture is shown, but when you go to it - it is blacklisted.

Actually this issue goes beyond re621, and seems to be a fault with e621 itself. The API isn't responding with the correct tag list, with the only general tag being tearing_own_clothing. re621 can't blacklist it if the e621 API isn't reporting the correct tags.

"tags": {
    "general": [
      "tearing_own_clothing"
    ],
...
Full response
GET https://e621.net/posts.json?tags=id%3A3954312

{"posts":[{"id":3954312,"created_at":"2023-03-26T11:47:30.116-04:00","updated_at":"2023-08-28T16:28:32.486-04:00","file":{"width":3000,"height":2734,"ext":"png","size":4321341,"md5":"cff01b00b5a9186028a01d57ff9fc404","url":"https://static1.e621.net/data/cf/f0/cff01b00b5a9186028a01d57ff9fc404.png"},"preview":{"width":150,"height":136,"url":"https://static1.e621.net/data/preview/cf/f0/cff01b00b5a9186028a01d57ff9fc404.jpg"},"sample":{"has":true,"height":774,"width":850,"url":"https://static1.e621.net/data/sample/cf/f0/cff01b00b5a9186028a01d57ff9fc404.jpg","alternates":{}},"score":{"up":658,"down":-11,"total":647},"tags":{"general":["tearing_own_clothing"],"species":["canid","canine","mammal","raccoon_dog","tanuki"],"character":["michiru_kagemori","nazuna_hiwatashi"],"copyright":["brand_new_animal","studio_trigger"],"artist":["plundered"],"invalid":[],"lore":[],"meta":["absurd_res","english_text","hi_res"]},"locked_tags":[],"change_seq":50221454,"flags":{"pending":false,"flagged":false,"note_locked":false,"status_locked":false,"rating_locked":false,"deleted":false},"rating":"e","fav_count":1494,"sources":["https://twitter.com/PPlundered/status/1640016688048373760","https://www.furaffinity.net/view/51521682/"],"pools":[],"relationships":{"parent_id":null,"has_children":true,"has_active_children":true,"children":[3954310]},"approver_id":95927,"uploader_id":174929,"description":"Be careful when checking your social media in public, you never know WHAT will happen!\r\n\r\nWinner of the 2nd Gaytron poll, thank you for the support!\r\n\r\nhttps://patreon.com/Plundered","comment_count":4,"is_favorited":false,"has_notes":false,"duration":null}]}

faucet said:
Crossposting from the re621 thread here since it's a problem with e621 itself, not re621:

Actually this issue goes beyond re621, and seems to be a fault with e621 itself. The API isn't responding with the correct tag list, with the only general tag being tearing_own_clothing. re621 can't blacklist it if the e621 API isn't reporting the correct tags.

"tags": {
    "general": [
      "tearing_own_clothing"
    ],
...
Full response
GET https://e621.net/posts.json?tags=id%3A3954312

{"posts":[{"id":3954312,"created_at":"2023-03-26T11:47:30.116-04:00","updated_at":"2023-08-28T16:28:32.486-04:00","file":{"width":3000,"height":2734,"ext":"png","size":4321341,"md5":"cff01b00b5a9186028a01d57ff9fc404","url":"https://static1.e621.net/data/cf/f0/cff01b00b5a9186028a01d57ff9fc404.png"},"preview":{"width":150,"height":136,"url":"https://static1.e621.net/data/preview/cf/f0/cff01b00b5a9186028a01d57ff9fc404.jpg"},"sample":{"has":true,"height":774,"width":850,"url":"https://static1.e621.net/data/sample/cf/f0/cff01b00b5a9186028a01d57ff9fc404.jpg","alternates":{}},"score":{"up":658,"down":-11,"total":647},"tags":{"general":["tearing_own_clothing"],"species":["canid","canine","mammal","raccoon_dog","tanuki"],"character":["michiru_kagemori","nazuna_hiwatashi"],"copyright":["brand_new_animal","studio_trigger"],"artist":["plundered"],"invalid":[],"lore":[],"meta":["absurd_res","english_text","hi_res"]},"locked_tags":[],"change_seq":50221454,"flags":{"pending":false,"flagged":false,"note_locked":false,"status_locked":false,"rating_locked":false,"deleted":false},"rating":"e","fav_count":1494,"sources":["https://twitter.com/PPlundered/status/1640016688048373760","https://www.furaffinity.net/view/51521682/"],"pools":[],"relationships":{"parent_id":null,"has_children":true,"has_active_children":true,"children":[3954310]},"approver_id":95927,"uploader_id":174929,"description":"Be careful when checking your social media in public, you never know WHAT will happen!\r\n\r\nWinner of the 2nd Gaytron poll, thank you for the support!\r\n\r\nhttps://patreon.com/Plundered","comment_count":4,"is_favorited":false,"has_notes":false,"duration":null}]}

probably related: tearing_own_clothing is on the post, but it's the only tag the api is not giving in the specific post's json.

cinnamoncrunch said:
probably related: tearing_own_clothing is on the post, but it's the only tag the api is not giving in the specific post's json.

Fixed that particular case. It has something to do with a programing mistake I made a while ago about how tag categories are getting cached. I will have to check for other tags where the same happens, I'm thinking there are probably a few more.

I already added a test so that this data doesn't get cached like that in the future here, I just didn't anticipate that lingering on for so long, the entries don't seem to be getting evicted like I thought. I don't see the point in doing validations against the live site when I can just add tests as needed with data that I actually control.

In other places this makes sense, I do something similar for dtext changes where I pull a bunch of data from the db and check for any changes that I didn't anticipate. I believe the term for that is acceptance testing.

the < prev and next > buttons show up on the post page for order: searches, this seems unintentional since they still function in ID order. (example)

sipothac said:
the < prev and next > buttons show up on the post page for order: searches, this seems unintentional since they still function in ID order. (example)

It is intentional.
Those buttons indeed only work if the search is ordered by the post ID.
That's what the default ordering is too.

cinder said:
It is intentional.
Those buttons indeed only work if the search is ordered by the post ID.
That's what the default ordering is too.

just to confirm, I'm saying that they currently are showing up on non-order:id_asc searches when, prior to a recent update, the search-seq-nav elements wouldn't have shown up at all on any post page if the search query contained an order: metatag.

Did the spectags: operator suddenly die? A search jumped from 20 pages results to 52 pages and I was trying to figure out why, then I noticed that it was ignoring the spectags.

kevsnowcat said:
Did the spectags: operator suddenly die? A search jumped from 20 pages results to 52 pages and I was trying to figure out why, then I noticed that it was ignoring the spectags.

Yep, look at the message 5 above yours
All <category>tags:<count> got borked

I'll check that out later when I got some time.

Edit: Fixed both the tag count searches and the next/prev links appearing when searching by order:X

Updated

I'm not sure what expected when adding a post to a pool, but when I tried about an hour ago, it worked, but there was no confirmation and the "Add to pool" box stayed open. That doesn't seem like it would be expected.

I use Brave and Opera GX on my phone and in both browsers the desktop site is activated for no reason, I have checked with other browsers like Firefox and Chrome and in those it shows me the mobile site, I don't know the reason or how to correct it

crocogator said:
I'm not sure what expected when adding a post to a pool, but when I tried about an hour ago, it worked, but there was no confirmation and the "Add to pool" box stayed open. That doesn't seem like it would be expected.

Oh yeah, no feedback when adding to a pool makes it feel awkward

crocogator said:
I'm not sure what expected when adding a post to a pool, but when I tried about an hour ago, it worked, but there was no confirmation and the "Add to pool" box stayed open. That doesn't seem like it would be expected.

It's supposed to refresh the page afterwards, but it hasn't worked for a long time, it relied on a hacky solution of just responding location.reload(); to the browser.

Ideally it should give a JSON response like pretty much every other route.

06000208 said:
On an artist page, specifically this one, it seems like the Group field is resulting in sanitized html rather than html that's part of the page:
...

It does indeed seem to be affecting all artist pages, most probably caused by 9e6ac15.

Honestly I think it's about time that the group column just gets nuked altogether, it's rarely ever used and 99% of existing usages are invalid. It's a relic from Danbooru where manga artists often belong to doujin circles, there's not really any analogue to this in the furry fandom.

alphamule

Privileged

I noticed that custom CSS isn't working and checked the Inspector and found this:
The resource from “https://e621.net/users/custom_style.css?md5=04da7b93af9603c0477b50d3917e7706” was blocked due to MIME type (“text/html”) mismatch (X-Content-Type-Options: nosniff). [Learn More]
It literally worked less than 12 hours ago. Since there was a site update, it's probably an unintended consequence of something changing.

Loading that URL in a new window gives "text/css is not a supported format for this page" which is DEFINITELY a bug.

alphamule said:
I noticed that custom CSS isn't working and checked the Inspector and found this:
The resource from “https://e621.net/users/custom_style.css?md5=04da7b93af9603c0477b50d3917e7706” was blocked due to MIME type (“text/html”) mismatch (X-Content-Type-Options: nosniff). [Learn More]
It literally worked less than 12 hours ago. Since there was a site update, it's probably an unintended consequence of something changing.

Loading that URL in a new window gives "text/css is not a supported format for this page" which is DEFINITELY a bug.

yeah, I just realized this too when I was trying to use the stylesheet to fix the .desc-popup box overfilling my screen and being unclosable.

it seems like old CSS stylesheets still work, like the one I currently have in use still is doing what it's supposed to be, but if I modify it at all the new one doesn't function... although that might just be because my old stylesheet is chached in my browser.

alphamule said:
I noticed that custom CSS isn't working and checked the Inspector and found this:
The resource from “https://e621.net/users/custom_style.css?md5=04da7b93af9603c0477b50d3917e7706” was blocked due to MIME type (“text/html”) mismatch (X-Content-Type-Options: nosniff). [Learn More]
It literally worked less than 12 hours ago. Since there was a site update, it's probably an unintended consequence of something changing.

Loading that URL in a new window gives "text/css is not a supported format for this page" which is DEFINITELY a bug.

sipothac said:
yeah, I just realized this too when I was trying to use the stylesheet to fix the .desc-popup box overfilling my screen and being unclosable.

it seems like old CSS stylesheets still work, like the one I currently have in use still is doing what it's supposed to be, but if I modify it at all the new one doesn't function... although that might just be because my old stylesheet is chached in my browser.

Sorry about that, there was an unrelated change and the test coverage for custom css was nonexistant until now.
I'm not sure how browser cache this but since it's wasn't a successful response I'm thinking they just don't? If your custom css still doesn't apply try making a change so that the browser goes to fetch it again.

I should have reported this ages ago but back when the site had a big update I got stuck with my account saying it has 3 unread messages

sorrowless said:
I should have reported this ages ago but back when the site had a big update I got stuck with my account saying it has 3 unread messages

First of all, that "big update" was three years ago...
But assuming that it's still an issue:

1. Have you checked your DMs for unread messages?
2. Have you tried clicking Mark all as read on the DMail page?

cinder said:
First of all, that "big update" was three years ago...
But assuming that it's still an issue:

1. Have you checked your DMs for unread messages?
2. Have you tried clicking Mark all as read on the DMail page?

I did in the past. If you click on the unread messages link your unread messages doesn’t appear with larger text. I tried going to my messages from my profile instead and now I see them. Either pages doesn’t have a “mark all as read” button, though as far as I can tell

sorrowless said:
I did in the past. If you click on the unread messages link your unread messages doesn’t appear with larger text. I tried going to my messages from my profile instead and now I see them. Either pages doesn’t have a “mark all as read” button, though as far as I can tell

https://e621.net/dmails has a "Mark all as read" button in the bar up top, the hamburger menu if on mobile

alphamule

Privileged

earlopain said:
Sorry about that, there was an unrelated change and the test coverage for custom css was nonexistant until now.
I'm not sure how browser cache this but since it's wasn't a successful response I'm thinking they just don't? If your custom css still doesn't apply try making a change so that the browser goes to fetch it again.

It's technically applicable using browser extensions/userscript, but yeah, it's nice to be done on site end. I'll test in a moment and edit with report.

Confirmed - it's working, now!

Minor display bug:

When viewing an image that's part of a set, esix displays a header showing the name of the set. If you mouseover that link, it will read something like "page 622/630". It should read "post 622/630"

As an example: https://e621.net/posts/3937097?q=set%3Atredfg543perfection - I *presume* others will see my set in the header, but I'm not certain that's the case.

edit: Taken by itself, I suppose "page" makes sense. I take issue with the fact that sets themselves are paginated though, so I see some ambiguity with listing post number as page number.

Updated

I'm fairly sure that a lot of raaggu's art has been mistakenly also tagged as kaneru's, because raaggu's Japanese nickname on Pixiv reads "irukaneru".

So I went to request a bulk update, assuming that this is the point of that feature, but apparently not? I can't be sure because there seems to be no documentation on BUR besides the help box in the BUR page, and all it lists is the syntax.

In any case, assuming mods have a way to mass-update posts, I suggest removing the kaneru tag from all posts which have both kaneru and raaggu.

onemoreanonymous said:
In any case, assuming mods have a way to mass-update posts, I suggest removing the kaneru tag from all posts which have both kaneru and raaggu.

You can't do that with a mass update.
I believe it is better achieved with a tag script.

onemoreanonymous said:
I noticed that several posts in pool #36111 which were deleted for being paid content have been released to the public by the author a few days ago. Are pics deleted by the mods still stored in the server somehow so that they can be undeleted, or do these need to be uploaded again now?

Do not upload deleted content. Deleted content can be undeleted. Even if there is a public version, that public version might not be the same quality as the deleted post, and if that's the case you can upload the publicly released vers