The first two image links in the spoilerbox of my Timasomo update don't work (as in aren't clickable or turn the text into a link), but others do.

Bug?

I noticed an odd behavior, and I can't tell if it's an intended feature, or if it's a mistake. I selected some text in a comment. I then scrolled it offscreen as I read more comments. Eventually, I started to reply to a comment further down on the page. When I pressed the "Reply" button, it used the selected text from an entirely different comment as a quote block in the reply.

I think it makes sense to pull selected text into a reply, if the selected text is part of the comment you're replying to. I was surprised to see it happen when replying to a different comment. Is this intentional or a bug?

To simplify the title:

(Formatted text, no space. (Behind the arrows.) While it's expected for quote blocks to not separate with one 'line' between them, it's definitely not expected for the block to be the same size, even w/o text.)

(Formatted text, with a space. Added this one in to contrast with Preformatted w/ space and because it separates the code blocks.)

>
>
>

(Preformatted/Codeblock text, no space. The "quote trigger" arrow in the middle is the bug, since if you look at the "view markdown" option of this post you'll realize that arrow shouldn't be there.)

 >

 >

(Preformatted/Codeblock text, with a space. This is how I personally fix the bug, if it is that. You can also fix it by typing space into the phantom arrow.)

Now with text inside the quote blocks, for comparison. (And because quoteblocks have to quote something.):

qwerty

asdf

(Formatted text, no space. Here the block expands normally for the text.)

qwerty

asdf

(Formatted text, with a space.)

>qwerty
>
>asdf

(Preformatted/Codeblock text, no space. The arrow in the middle I never typed in is still there.)

 >qwerty

 >asdf

(Preformatted/Codeblock text, with a space.)

The CSS selectors that determine whether or not to apply the display: none rule to comment text sections use the pseudo-class rule :not(:target). While this is great for keeping a comment in a non-collapsed state, it's a bit too effective as it prevents user-initiated collapsing of the comment. This can be problematic when you visit a direct link to a comment with an extensive reply tree beneath it and want to collapse it so that you can view the surrounding reply trees. As it stands, you can't do this without needing to either a) collapse the parent (prevents viewing sibling comments), b) collapse the children (requires collapsing potentially multiple child comments), or c) remove the fragment portion of the URL (requires reloading the page and possibly losing your place on the page).

Reproducing should be as simple as clicking Link in a comment's header, then trying to collapse the comment after being redirected.

Usually bugs, glitches, and exploits degrade the experience of a game, but occasionally they can actually work to a game's benefit. In some cases, they can become significant enough to become part of a game's identity. In others, they make a broken game worth playing in the first place. Even without such legendary status, a given wrinkle in a game might simply make it more enjoyable or entertaining, or perhaps open up unexpected modes or paths of play.

What are some examples of these, and how did they improve their associated games?

I just noticed today that in ~music, the "topic-info-source" metadata isn't visible in listings; it shows the author name instead. Clicking through to the post it's clear that it's been scraped, it just doesn't get a site name or favicon.

eg: Youtube link on ~movies versus Youtube link on ~music

Is this intentional? It sorta makes it look like everything on ~music is a text post.

I might've missed something, but why did the Financials block disappear from the main page? And the numbers on the financials page are zero.

Also, while we're at it, there should probably be a “Financials” link in the footer.

I've been noticing a couple of weird formatting issues with tags on topics. For some posts, the first tag is shown before the metadata, then the rest of the tags are on a new line below. This isn't always the case, and so far I can't find any rhyme or reason for whether it does or not. It's not like some CSS float issue or anything, it's actually creating two <ul>'s. — example

The second issue I've noticed is sometimes the dangling part of a letter like lowercase "g" will be cut off by the footer. It seems like this is caused by .label {line-height: 1.2;}, but it's also inconsistent, and I haven't figured out why. In this example the "g" in "climate change" is not affected.

Maybe it's just me, I dunno. Firefox 70 on linux. I'd create a gitlab issue, except I am very lazy.

UPDATE:

It is actually NOT broken, but it seems that a blank line is required after the <summary> tags for it to work.

• Therefore, this does not work:

   > Interstellar - just watch it
  
   I like this movie a lot. BUT, it bothers me that
  
   <details>
   <summary>spoiler</summary>
   a story so grounded in hard science fiction chose an unjustified book
   metaphor to represent a dimension that would be, for us, absolutely
   unknowable. I much prefer the choice made by Kubrick in *2001, A Space
   Odyssey*: to represent one abstraction with another, preserving the
   alien character of an inhuman reality.
   </details>
  

• But this does:

    > Interstellar - just watch it
  
    I like this movie a lot. BUT, it bothers me that
  
    <details>
    <summary>spoiler</summary>
  
    a story so grounded in hard science fiction chose an unjustified book
    metaphor to represent a dimension that would be, for us, absolutely
    unknowable. I much prefer the choice made by Kubrick in *2001, A Space
    Odyssey*: to represent one abstraction with another, preserving the
    alien character of an inhuman reality.
    </details>
  

When I try to access tildes.net over an IPv6-enabled network, the connection simply times out.

Running ping -4 tildes.net responds but ping -6 tildes.net times out. I've tested this on multiple networks (residential and commercial) but it doesn't work on any.

It looks like tildes.net has an AAAA record but the IP doesn't work.

Can anyone reproduce?

Hi, I'm new here. I use an extension for Firefox called Vim Vixen that allows the user to navigate webpages pretty effectively via the keyboard. One of its features allows the user to press the f key to bring up an overlay of keybindings over all links on the page, and pressing the keys associated with a link causes the browser to follow the link. I'm not intimately familiar with how the extension works, but with some basic Inspect Element-fu, I'm reasonably sure this feature is implemented with an iframe that is injected over the entire page.

On Tildes, the iframe is visible in the top left corner and shifts all the actual content over accordingly. Normally, it is invisible. This is the only time I've ever had a problem with the extension, which is why I'm posting here first rather than reporting an issue with the extension. If someone more knowledgeable is interested in creating a fresh Firefox profile and comparing how the extension works on Wikipedia and on Tildes, I'd appreciate it. The code for the iframe appears to be identical on both sites, and it's at the end of the page body.

https://tildes.net/~hobbies/bzx/what_are_your_hobbies_and_how_do_you_get_into_them#comment-2z4f

See for yourself. Text is squished to the left side of the page, one or two words per line. No idea if anyone else can see this but since it happens every time l visit that link l thought it was worth asking.

Going to poke @Deimos on this one l suppose.

Edit: Issue created.

Issue is occurring on iPhone 7 using Firefox mobile browser v15.1 but has occurred on previous versions of Firefox.

When I try to click the link to view comments, the post grows in size seemingly indefinitely (eventually either the link works and redirects or I get frustrated and scroll to the next post). This has happened a few times. As far as I can tell it is seemingly random.

https://m.imgur.com/8HHIvtl
https://m.imgur.com/rhiA30y

Sorry if this is the wrong place to post this. Still trying to get use to the proper place to put things. If you need any more details let me know.

Not sure if this is intentional or not, but I've noticed today that the first time I click on the new messages link at the top right, I cannot vote on any messages I read. I need to reload my inbox in order for the vote to stick.

Anyone else encountering this?

I'm using Safari 12.0.2 on macOS 10.14.2 (Mojave). The same issue also occurs on iOS 12.1.2 (using Safari).

When using 1Password to autofill with the browser extension on macOS or the "autofill" feature on iOS an error message pops up: username: String does not match expected pattern.

I have to either use the browser supplied autofill on macOS or manually copy/paste username and password into the corresponding fields. On iOS there's an autofill API which I have set to use 1Password in the browser, also causing the error

Edit: Video of the issue

My password is shorter than 8 characters. When I attempt to log in, I get a validation error telling me so.

Luckily, I'm signed in already on this browser. However, when I go to the change password page and attempt to make my password longer, I get a validation error telling me my old password is shorter than 8 characters, and it prevents submitting the form.

I have the "mark new comments" option enabled. Most of the time, it works, displaying "(X new)" under any topics which have had new comments posted since I last looked at it.

However, if the topic is old enough, this "(X new)" notification does not display. The topic is sorted to the top of my feed, because I'm using activity from all-time as my default sort. But it doesn't show "(X new)". I know it has new comments, or it wouldn't jump to the top of my feed, but I'm not seeing that notification.

I don't know how old is "old enough" but it's definitely longer than a month. The topics this is happening to are 3 or 6 months old, but it doesn't seem to happen for topics which are only a few weeks old.

This is a bit odd. Recently I've thought typing into tildes input boxes sometimes felt laggy but today I noticed my fan spinning up when I loaded the site (until today it's been so hot my fan was spinning all day). I checked Chrome's task manager and yes, tildes.net is using 100% cpu. If I wait 2-3 minutes, usage will drop back down to the expected 0-1% and remain there. This happens every pageload. If I load a page, wait for the cpu use to drop, then either reload that page or click an in-site link, the cpu ramps up again for 2-3 minutes.

Chrome Version 68.0.3440.106 (Official Build) (64-bit) on Debian testing, which is up to date as of midnight GMT 10/08/18. I even rebooted my machine just to be extra sure. Happens on all tildes pages. No other site that I regularly visit seems to do this.

Posted here rather than to the issue tracker for visibility, to see if this is affecting anyone else before raising an issue about it.