Option to avoid duplicates

[azumukupoe]

Is it possible?

[Revadike]

Could you please explain what you mean?

[azumukupoe]

I don't need to see more than one "swi-block boxed" for each game

[Revadike]

I don't need to see more than one "swi-block boxed" for each game

Define "game". The script can only see links and images. It doesn't know if they belong together.

[azumukupoe]

I don't need to see more than one "swi-block boxed" for each game

Define "game". The script can only see links and images. It doesn't know if they belong together.

Sorry, I meant appid

[Revadike]

so, per page, u only want it to show icons ONCE per appid?

[azumukupoe]

Maybe per frame?

[Revadike]

Maybe per frame?

frame?

[azumukupoe]

Maybe per frame?

frame?

<iframe>?

[AxJ11]

Is it possible to make the script only show up icons once within the same block/div section?

https://i.imgur.com/Mn8fMy3.png
https://i.imgur.com/MUU0p1T.png

[Revadike]

Probably, but this likely doesn't apply to every case. And how many blocks higher up do u want to go when selecting this group. It's something hard to implement globally for every case.

[AxJ11]

Yes, I think it's would be a huge work for the whole process. Maybe deal one specific URL category at a time? I mean to customize the rule for each specific URL(giveaway All/Wishlist/.., Entered/Won/..., inside each giveaway and discussion post. Because the page structure of each URL may not change drastically very often, maybe it's possible to make a specific rule for each URL category?

For example, for "inside each giveaway," there are always three sections( to detect):

1. div class="featured__container"
2. div class="page__description"
3. div data-comment-id="56449198" class="comment esgst-es-page-1" => ESGST
   div data-comment-id="56449198" class="comment" => Original

Can we use these "class" and "data-comment-id" for the detection boundary?

And I think there are similar div structures to capture just like those mentioned above for other URL categories, eg. giveaway All/Wishlist/.., Entered/Won/... and discussion post.

https://i.imgur.com/MmOHWl8.png
https://i.imgur.com/WahrLT6.png

[Revadike]

I'm trying to keep it a global solution. If I start filtering and supporting different sites individually, it will be too hard to maintain. Besides, most individual sites have their own scripts/addons out there.

[AxJ11]

Sure, of course. It's probably the best to work with your own workflow and workaround.

[AxJ11]

Or you can think it's a backup workaround since the Steamgifts.com is probably the only website with this problem because of ESGST.

[AxJ11]

Another idea about the solution is to have an option to show the tag only when the HTML value or content contains the string of the "game title" or the word "Steam." Does it make sense?

Also, it's probably great to have a system to let users decide what webpages/websites to customize the condition of the way when the tag shows up, e.g. the detection range or by game title rather than the appid in the URL. Maybe this could reduce your workload.

[AxJ11]

...how many blocks higher up do u want to go when selecting this group. ...

Maybe one or two tiers up? Or have a system let users decide how far is the selecting range per webpage respectively.
I also found the bundle page of Barter.vg has this problem too. And the issue fits the one-tier up selection.
https://i.imgur.com/Wg1MlXa.png

[Luckz]

There are a lot of sites where "invisible marking" can happen, meaning due to the page design, a SWI checkmark and such will be added to the DOM but not visible to the user. Just showing the first match per app ID is no solution if for a particular site it's the second or third match that they would conveniently see.
On the other hand I agree with the benefit of this suggestion, seeing several SWI annotations for each game can definitely be annoying on some websites.

One idea would be to kind of pollute the user settings with per-site nth element preferences or selector limitations / custom regex.

[AxJ11]

Not quite sure what you meant but thanks for the input. Luckily, at least on ESGST end, the duplicated tags on Steamgifts.com would probably be solved by editing the URL of the "Game Categories" tag because the author added the feature this weekend. Hope this change would be live in 1~2 days.

[AxJ11]

Finally, the duplicated thing is reduced to the minimum, err....almost, on Steamgitfs.com because of the recent update of the ESGST. Now, can we have a system to let the user choose and save which HTML tag, class, id not showing the icon for a specific url or web title?

The last mile...almost done!
https://i.imgur.com/97lzzdh.png

[Luckz]

Not quite sure what you meant

Assume you're on a website.
Let this website have an IMAGE for each game, a LINK to it on the image, and ANOTHER LINK for example on the price field. That game now exists three times.
Depending on web design / CSS, or even your resolution/device, SWI-added icons might be visible on all three, or might be visible on only 0/1/2 of them.
It's thus individual to each website that SWI can modify which HTML element should be annotated (for the very same game).

[AxJ11]

@Luckz Like your explanation and solution. Good job!!!

BTW, an update a long time ago on the Steamgifts.com giveaway page, finally no more duplicate icons:
https://i.imgur.com/VagLAsK.png
It took me some time to find out that the SWI icons can only be modified through the "GM_addStyle" method. Don't know why. 🤔

[bestpika]

It may be possible to check if the parent element is the same one.
However, if the pages are too complex, it is very difficult to check.

[Revadike]

It may be possible to check if the parent element is the same one.
However, if the pages are too complex, it is very difficult to check.

This made me think. How high should the parent node be, since eventually they're all under the same root node. But perhaps we can look at subtrees that exclusively contain the same appid and avoid those occasions from happening somehow.

So, when you add a new SWI-block, you recursively keep checking the parent node for SWI-block children, and if they all have the same appid, make sure only one is displayed.

Now, poses a new set of issues:

• How do you check if you have not added all SWI-blocks yet?
• Which SWI-block do you want to display and which to hide?
• How badly will this impact performance?
• How will it work for every kind of web page.

Yeah... that's a lot of difficult issues to tackle.