Comment on xkcd #3090: Sail Physics
who@feddit.org 3 days ago
Any chance of getting the bot to use the source page as the post link, rather than burying it within the body? That would make it easier to view the whole comic, complete with hover text.
Bot author here - I thought that the current implementation was a big improvement because it meant you didn’t have to load up an external website but I should have known that not everyone would be happy!
Looking at the votes for the comments for and against this idea, it looks like if it went to a vote the current setup would win, but I’ll think about how it can be improved.
I’m not sure the current comment votes are representative, due to the biases that form as people engage (or don’t) in idea threads, but I acknowledge that it is possible that a proper vote would end up as you expect.
If directly linking to the full comic page is a no-go, how about putting the source link at very top, as the first line in the body? That would at least make it a little easier for desktop users to target the link they need for a single-screen view of the whole comic.
If you upload an image, the URL field is populated with the URL of the uploaded image, so there’s not really multiple fields like it appears.
I can definitely change the template, although I won’t edit the bot before the next comic which might be any second now.
If you upload an image, the URL field is populated with the URL of the uploaded image, so there’s not really multiple fields like it appears.
Sure enough; I just discovered this for myself when running some tests. I also noticed that Lemmy offers an Alt Text field, which it looks like the bot is already populating.
Test 1 : The URL field and the image attachment field were used; the latter overwrites whatever is placed in the former. I guess this might be modeled after Reddit.
Test 2: A direct link to the image at xkcd.com was placed in the URL field, and the source link placed at the top of the body. Result: This is similar to the bot’s current format, including the flaw that a desktop browser with strict privacy settings won’t show the comic image as part of the post when the thumbnail is clicked, because it’s an off-site image. Having the source link at the top of the body does at least make it a little more convenient to click through to xkcd.com’s single-page view.
Test 3: Only the image attachment field was used; the URL field was left blank. Result: This allows a desktop browser to show the comic image in-line when the thumbnail is clicked, even with strict browser privacy settings, because it’s an attached image hosted on the local Lemmy instance rather than an off-site link. The source link was again placed at the top of the body. Result: Best of the three tests, I think.
(Side note: I used m.xkcd.com links instead of plain xkcd.com links in these tests, just to see how the mobile site looks in different browsers. In practice, either ought to work.)
If you decide not to start populating the URL field with a link to the xkcd.com source page, I think the approach in Test 3 would be the next best thing. (At least, based on what I have found to be possible so far.)
Regarding the edit:
The mobile link might be better than the usual link, but I’d still prefer the one-stop-shop of this current method. Comic, hover text, and comments are all viewable in one place without the additional wait of switching to a browser app and loading the page. I wish the desktop experience mimicked this.
AlDente@sh.itjust.works 3 days ago
I actually like it better this way because the full comic can be viewed and zoomed in on in app (Voyager in my case). I’ve seen these posted with the source page as the post link and I can’t click on the image to zoom in without being pulled out of app with a wait for the web page to load. Also, when entering the post to see the comments, the thumbnail is so small when the source page is linked, making the comic unreadable. With an image alone, the full comic is viewable beside the comments. Is it different for you?
who@feddit.org 3 days ago
The thing is, with xkcd, the image alone is not the full comic. Part of the comic is the hover text, which you don’t get with the image alone.
Wxfisch@lemmy.world 3 days ago
The bot adds the alt text to the post, so no need to head to the site to see the entire comic.
aeharding@vger.social 3 days ago
Are you the bot author? If so have you considered using Lemmys alt text field directly? This would display the text with the image is some clients like voyager. It would also federate to Mastodon (post body doesn’t federate). You could also keep the text in the post body and do it both places :)
who@feddit.org 3 days ago
That splits the comic between two separate screens, and requires multiple steps (and page loads) to read it.
It’s not as good as linking to the whole comic in one place.
AlDente@sh.itjust.works 3 days ago
Ah, thanks for clarifying. I never knew about the hover text. I just viewed the source page on my desktop browser and now I see it. However, is this text even viewable on mobile, when you don’t have a cursor to hover?
who@feddit.org 3 days ago
The mobile browsers I’ve used let you long-press the image to see the hover text.
If yours doesn’t, you can always prepend
m.to the domain name, like this: m.xkcd.com/3090/