[X] My Assistant

Based on the work of Qserty I have enhanced the script to be more userfrendly with the bookmarks.

[github.com] https://github.com/Meldo-Megimi/EH-Page-Scr...robbler.user.js

Optimizing performance does not mean getting your page 0.1 sec faster, it means the server won't crash in 6 months because of the RAM, CPU, or bandwidth usage...

Thanks to everyone trying to answer all those stupid comments, and to those who actually worked to provide an alternative.

Pagination is dead, I share the frustration, there should have been something put in place instead, even if not as good.

I am no web developer, but I made some changes to the monkey script, so I will share it as well for whoever want to use/improve it.

- I overwrite firstGID or lastGID with the url parameter next or prev if available.
=> For example if the first result is GID 5000 and the last 6000, but prev is 4500, the result is covering 4500-6000, and not 5000-6000, there was simply no result in the 4500-5000 range, but the search do include that range.

- I'm inspecting the Prev and Next div, to check if they contain a SPAN instead of a link A.
=> If Next is grey and unclickable, it means there are no further result, so even if the last result is GID 450, the range 450-1 is covered otherwise "Next" would be available.

I admit it doesn't change much but I feel better when the green range is accurate, when you press next it starts exactly from where it ended (IMG:[invalid] style_emoticons/default/smile.gif)

- I added 2 buttons Save and Load, I do not think they will work with Chrome as I use unsafeWindow for the button functions and it seems not compatible with Chrome from what I read on the web.

During a search, save will store the current GID in the browser localstorage with the search as the key.
If you do the exact same search later, load will bring you back exactly where you were (it also store if you were going forward or backward)

Recreating a pagination is impossible, you can try to estimate the number of pages based on your current page, like 25 results for each 10k GID range, you will only create a mess, missing potential results between pages because the next 10k GID range had in fact 60 results and you had no way to predict that.

Thus that new GID navigation bar is the best way to navigate currently.

(IMG:[i.imgur.com] https://i.imgur.com/eDitjFQ.png)

[pastebin.com] https://pastebin.com/pWrA6ihD
In case I'm not allowed to post an url: pastebin.com/pWrA6ihD

The new page selector is a huge downgrade I can't keep track how far into a search I'm in now hopefully this is reverted or made better.

Good job making using the site more effort than it's worth. How am I supposed to navigate without page numbers?

This is such an obviously user-unfriendly mess that countless people who have never even looked at these forums are posting to express their disapproval, and the owner in their arrogance just thinks it's brigadeering? Nobody sent me here, I took one look at the change and found my way here by myself because I could only assume it was broken. But to see it was intentional is just baffling.

You can sit there thinking you've got it all worked out but nobody owes you web traffic. If you make something that was once simple and convenient a hassle to navigate then eventually they just won't bother.

Based on the work of Qserty I have enhanced the script to be more userfrendly with the bookmarks.

[github.com] https://github.com/Meldo-Megimi/EH-Page-Scr...robbler.user.js

Awesome, finally someone created a repo for this script. I hope that one will be the one, that receives PRs and stuff, combined with all the great ideas, without need for scrolling the whole thread and selecting the needed version and easily updatable.

It just feels bad that such a QOL function simply just disappears, but if this is a server load issue, then there is really no other way to deal with this issue other than increasing server capacity.

If users are that concerned then maybe we should all find some ways to monetize the QOL function for the site.

Do anything but kowtowing to shit bird Irodori or fakku that is.

Good job making using the site more effort than it's worth. How am I supposed to navigate without page numbers?

This is such an obviously user-unfriendly mess that countless people who have never even looked at these forums are posting to express their disapproval, and the owner in their arrogance just thinks it's brigadeering? Nobody sent me here, I took one look at the change and found my way here by myself because I could only assume it was broken. But to see it was intentional is just baffling.

You can sit there thinking you've got it all worked out but nobody owes you web traffic. If you make something that was once simple and convenient a hassle to navigate then eventually they just won't bother.

"dude just remember that you were on the gallery number 38236712 and use the script some people made that is only on the forums that make you select by hand" btw if you don't think this is the best change ever you are obviously a troll brigading

You know, back then when the site gonna shutting down, all of us said
"This is 'Hentai Library of Alexandria' we can't lose it. We got many valuable archives here."

And now the library comeback, but then you took off the label of every books to make this library going on.

Question is, library still there, but who gonna use it efficiently except the librarian themselves? What's the point of finding books harder in the library? To just keep it or to use it?

This post has been edited by epa: Nov 4 2022, 07:52

I still think my "by month" idea would work without substantially impacting performance.

Might we be able to do "pagination" by month? So the server would keep a static list of which GID is the first (or last) for each month of each year, and then just add unconditional jumps to the next and previous number in that list to a requested set of results? (setting ?prev=(that id))

That'd greatly reduce the amount of time I'd spend clicking and might be a suitable compromise, right?

So far that list would have < 360 GID's in it. A lot less than the millions of galleries we are at now.

i'd be fine flipping through all the galleries in one month rather than across all of time. "next month/last month" buttons wouldn't be too hard if we just kept track of the first or last GID of a given month... right?

Is there something I'm overlooking here? Seems like a decent compromise to me.

BTW, when I say "pagination by month" I really mean that we keep the <n galleries per page> limit in place, but just add shortcuts to skip to the beginning/end of the month.

This post has been edited by dragontamer8740: Nov 4 2022, 07:51

I don't even know what GID is or how to use it as a replacement for page numbers.
If it Is like some sort of id every gallery has then what I need to do whenever I want to find a gallery among many is remembering the id number? Wouldn't that mean I knew exactly where the gallery is and had no need for searching?

Also why expunged and low-power tags searchs aren't included with normal non expunged/high power tags galleries when you check their checkbox but rather exclude the normal galleries?

This post has been edited by cuern1451: Nov 4 2022, 07:58

Haven't really posted here much, but have to add to this. As a normal user, this change feels completely unintuitive to use. When searching with tags, the Gallery ID does not match the number of searches in a page, which makes it very difficult to find specific galleries.

If I want to search 20 pages in for a set of searches, this could be 400 ID or 1000 ID, completely different based on the tag. Not to mention manually entering these searches from the browser link is also incredibly annoying. Hopefully this changes gets reverted or updated soon.

You know, back then when the site gonna shutting down, all of us said
"This is 'Hentai Library of Alexandria' we can't lose it. We got many valuable archives here."

And now the library comeback, but then you took of the label of every books to make this library going on.

Question is, library still there, but who gonna use it efficiently except the librarian themselves? What's the point of finding books harder in the library? To just keep it or to use it?

Speaking of that shutdown incident, let's not forget that it was a kneejerk reaction to fake news of a fake law.

And half the library is copystruck and has to be torrented.

How and what is a GID this forum is so confusing

Based on the work of Qserty I have enhanced the script to be more userfrendly with the bookmarks.

[github.com] https://github.com/Meldo-Megimi/EH-Page-Scr...robbler.user.js

I think we should make new thread dedicate to this script to bring something like page back (or better navigation), so we can do something like update or version control. And it might be harder for newcomer to this thread to find in this 40+ page (unless you click next 40 times, just like the new search engine does).

This post has been edited by epa: Nov 4 2022, 08:43

What's the logic behind having to go through ALL the pages you don't want to reach the one you want ?

Do you know what's the biggest existing resource in life ? Time.
Do you know what's the most important job in our history ? Librarians. Or in today's age any type of acts/job/software that sorts and classify files.
Because books would be lost if there was no one to sort them.
Because history and data would be lost if there was nothing to sort them

Have a nice day.

This post has been edited by funky360: Nov 4 2022, 08:48

Also, could I ask why downvoted tag search was removed? I've found it useful for tag cleanup before.

How was that useful for cleanup?

GUYS LOOK i was very quickly able to find my last post just by vaguely remembering what the last pages was when i posted

HOLY

Normal user that's also never posted before in the forums here, and I wanted to voice my extreme displeasure with the new updates like apparently a lot of other people. I had literally asked a bud if the lack of pagination was because of a bug or site maintenance. On some level, I feared it was a new layout decision, and I was shocked to see that's indeed the case. It really is god awful trying to navigate now, and I'm seriously considering using nhentai again or something even though I came here specifically to get away from their own search & viewing issues. E-H has been my go-to for years, and now it's just frustrating to even use/navigate, and I sincerely just hope these frankly horrible changes get rolled back because it is WAY worse than before. It should not be this hard to navigate a godd*mn hentai site.

Edit 11/5/2022: Just found out that there's a 5 keyword limit in the search now. So when i add a 6th, i get a message saying "Some of your keywords were not used. Try combining keywords with quotes. (Ignored x)" They fucked up the search too. No more simple plug-in and play. What in the actual fuck were these assbackwards retards thinking. Search is literally broken and a shell of its previous version, just like the site as it is now. Not to mention the loss of being able to see expunged galleries in my favorites so I can look for them elsewhere. Fuck everything about this update. Gonna have to start transferring my favorites over to nh.

This post has been edited by TheBrojangles: Nov 5 2022, 08:39

Any chance we can get a bit of a compromise in the form of something that keeps track of what page we're on, at least? It should still be possible to see how many pages back we are, right?

I not a UI designer but
such a mass Galleries just have next page buttom?
like dig a hole by teaspoon
just fix it by anything
plz

Bing, google and co do have page-numbers on their results..........

Can you imagine that happen? HAHA. user leaving the site. seriously The Old search engine is running out of RAM frequently. that are a old saying 'ether change the Software , or put a lot of money in hardware'

Or - hear me out - don't frick up the useability.
There are 2 fundamental problems with what you said:

1 - "change the software" should not mean "destroy usability and avoid the problem by getting people to leave"
2 - the removal of pages has no impact on overall performance cost.

The engine already has to scan all the results and order them. And it seems the problem was that for whatever stupid reason it is not caching anything.