I rolled back the latest rapid security update and reset a lot of stuff. Cannot get it to work. Everywhere else it’s great, but until I can beta the Artemis app (@hariette) these features would be welcome.
I'm going to be expanding coverage on a rolling basis through a series of updates. If there is a specific magazine where you keep running into this problem, please let me know.
Thanks. I've released a small hotfix that extends coverage further. As for stuff showing up in the "random threads" sidebar, I believe this area is a lost cause, since it populates content regardless of whether it is being blocked or not. I suggest just disabling it via the General > Hide sidebar elements feature.
There's a cosmetic issue I've been having with the script that this update doesn't seem to have fixed: with the script enabled, even with all its features turned off, certain divs (which as far as I've found are all children of main, with both the section class and the content id) are being given the display:unset style. This 1) removes the background and 2) conflicts with the styling for .section to somehow create two bars along the left side of the element. Examples are on the "add new thread" page, the login page, and in almost every tab of the "magazine panel" moderation tool.
I tried looking for the issue in the script and couldn't find anything that jumped out at me (I... also don't know javascript, so it was a bit of a shot in the dark), but I'm as close to certain that it's this script as it's possible to be. I'm able to replicate the issue consistently across three different devices (Ubuntu/Linux Mint laptops and Android phone), running the script via the Tampermonkey extension in Firefox. (I saw that this had already been reported awhile back specifically for the login page and that you hadn't been able to replicate it, so I'm trying to provide any device-specific information I can think of that might be relevant).
Anyway, I was pushed to finally bring up the issue today because although previously the issue has just been a minor visual annoyance, I just changed the background on one of my magazines to an image and now, with the script removing the nice solid background from behind the formatting icons on the new thread page, I can barely see them. :(
Thanks for the report. I did a quick test now and was still unable to reproduce it. I believe you when you say it is happening consistently. The soonest I'll be able to investigate this further is this weekend. In the meantime, it would be helpful if you could submit a formal bug report on GitHub and include direct URLs, screenshots, and the specific element ids/classes in question. Thank you.
Truth be told, I am not sure what turbo mode is intended to be. The setting is there, but it has no description. I turned it on and observed no change. Is it different from infinite scrolling mode?
Thank you, I looked into it. Here is a reply I made in another thread:
By design, Greasemonkey scripts necessarily trigger on page load. However, KES initializes its own mutation observer and watches for events that change the DOM, but does so in a granular fashion so that only those of relevance (changes to the thread index and comments) are propagated. It looks like when turbo mode is enabled, the entire DOM up to the HTML tag is replaced at once, so our more granular approach is missing this change. We need to watch for basically all mutations now and filter them accordingly.
After making the necessary changes, it seems to be working again as intended, but some additional refactoring is needed, and I haven't checked the implication of this on specific scripts/add-ons yet. It looks like it should be easy enough, since the majority of functionality I looked at was working again when using a more permissive observer.
I haven't added these changes into the release version of KES, and given that turbo mode is listed as experimental, I don't think we can support it immediately over a more traditional page load structure, but at least I am aware of the issue now and can start refactoring it. Hopefully it is not problematic.
Great to hear! I'm assuming Turbo mode compatibility will be announced when it's released. Is there a experimental KES version that does have Turbo mode support?
There is a testing branch that can be used by intrepid souls who don't mind random breakage, but compatibility with Turbo Mode has not been added in there yet. I just workshopped the change a bit locally and haven't pushed any changes. I will ping you again when the prototype is working with a link to the testing branch.
Thank you for waiting. This feature is now available in the testing branch and should apply by default if Turbo Mode is enabled. I had to rejigger a few things to make the functionality comply with both normal and turbo modes, but KES should apply its settings correctly in all of the expected scenarios:
On initial page load
When navigating between pages
When loading threads/comments via infinite scrolling
Let me know if you see something out of the ordinary.
Note that the testing branch may introduce breaking changes from time to time, so after testing this feature, you can go back to the stable branch. The finalized version will make its way into version 3.2.0.
I disabled everything and tried it with/without and saw no observable difference on the login page. Can you indicate what anomaly you are seeing in the above screenshot? Also, could you please click the clipboard icon on the bottom right of the KES menu and paste the contents of the clipboard to me in a PM so I can try to reproduce this?
It's the background behind the credentials that's messing up, see those little grey bars on the left.
I disabled the other userscripts I had installed and still had this problem but now I went through the KES settings and configured stuff and the issue is now gone.
If it occurs sporadically, it might be something failing to load on kbin's side, I have seen visual bugs like this before that are unrelated to KES. Check the dev console when refreshing the page and see if there are any warnings or errors. I will investigate regardless.
Getting back to you, this issue has been extensively described and reported by another user on this ticket. The bug is now fixed and will make its way into the next release (3.2.0).
kbin Enhancement Suite
Hot