Honestly, I have very little interest in historic prices, only in historic discounts, so whenever I look at what you display, I have to do the mental math to translate from price to discount, which I would rather avoid.

If you wish to avoid information overload, how about replacing historic prices with historic discounts? I don't know about others, but I would prefer that.

Not a deal breaker, I can continue doing the math if necessary.

Yes, I feel you ^^. It's quite a different craft, filled with sharp aesthetic concerns, ergonomic and readability issues, compared to the more common (invisible) aspects of design that a programmer takes pleasure with ^^ ! And reaching a merely decent result in those (mostly) visual arrangements will unfortunately emphasizes even more their defaults : one millimeter gaps anywhere can hurt a great work more than any perfectible (but still consistent) design choice at a more general level. Which probably explains why automated predefined mockups are that popular... and interfaces build through mish-mashes of boxes (options, etc.) seemingly placed at random far more forgiving !
I'm no visual designer myself, far from it actually, but I spent years ago some time in a training course in order to learn professional proofreading methods (mostly catered to book edition), helped in various small projects, and will tell you this : even the most simple concept benefits a lot from polishing basic details, for those contribute enormously to any user's first impression.
Fortunately, meticulous people like you are well equiped to deal with this aspect of any project ^^ : as long as feedbacks keep bringing you "food for thought" and that you aren't in a hurry, you'll probably come up with a clear design. What you've already reached is already fairly good... As a front page, Gamesieve looks much cleaner, less congested and more readable that Gog's main page to me, even though I still prefer the latter's colours...

In the end, as some of the first comments noticed, your long term goals (both the amount of functionalities and your investment in advertising those) will probably weight more anyway : any concrete reason to use Gog's interface is one less opportunity to launch yours and vice versa. For experimented users, even a small amount of very specific possibilities will be enough to tickle their fancies from time to time : it's what Gogdb does (well) already. But for the others, you'd probably need to add more links to Gog's common functionnalities.
Let's take a (deliberately) extreme example : if you added (near the Search panel for instance) a drop down menu allowing one to reach his daily destinations on Gog (forums, Blog, wishlist, etc.), he could even launch Gamesieve everytime without bothering to browse the former's main page. Of course, it means a complete front-end, but anything less means that Gamesieve will remain a occasionnal tool... and must be planned as such.
By the way, as someone previously said, it would be interesting to ensure that people can find your front-end and forum thread from somewhere else than itself ^^. Don't hesitate to mention it in well related topics, like you did during that NSFW one. You're working on a service better than Gog's in some regards, so it's only fair to type a line or two about it whenever it feels appropriate...

Oh, indeed, I didn't notice... and yes, I prefer it's darker version, but that's completely subjective. As long as you don't plan in adding numerous options, I'd say the manual toggle seems a good idea - two dots for that being enough, perhaps at the far right of the Search panel for example. But, at least on the paper, your current version seems more elegent to me. The question is "How many people will realise its existence ?".

Oh yes, indeed. Being able to save money within the Euro zone, just by buying games from a different place, seems so counter-intuitive that I forgot about that.

Yes, font metrics can be cruel masters ^^... Both situations look now properly fixed : it's a clear improvement ! I agree that the spacement around those filters is debatable - not only before the sub-filters, but in between categories, plus between any category's title and its filters as well, as the difference feels slightly too forced to me - and I'm still not convinced both by those arrows on the left - even when checking them on a smaller screen -, plus the weirdness of check boxes on the left and check marks on the right of the said filters - why not only check boxes all the way ? -, but I guess you've made your experiments ^^...
The weirdest (non alignment) issue remains, in my opinion, the positioning of the total amount of games presented, immediately under Gamesieve's presentation, when you obviously spent so much time ensuring that all other elements are seperated by a minimum of free space. I'd just place it on the same line as "Sort by" - therefore aligning "Showing X-Y" with "Sort by", which feels consistent, even with the applied filters under it.
I didn't notice any other immediate issue. The few other whacky details I keep noticing are usually related to rare situations. For example, Type "Shaq Fu" and then zoom to simulate a smaller screen... The "First time on sale..." label will eventually mess up with your whole arrangement. Another situation with RoboCop or Northgard : since the "verified" and "total" ratings are similar, they'll be regrouped on a single line... under their usual (top right) position. Ensuring the opposite happens (regrouping "total" on "verified"'s line) would avoid that.

I see two other possible workarounds : either imagine a small "Titles only" check box after "Find", restricting results for that one case. Or - but it would imply a bit more work and I'm not sure the result would be more elegant - come up with a small sub-title (on one line) to distinguish the different (sorts of) results provided. "By title", "By genre", "By developper", etc.
Now, that said, the current situation isn't bad at all and I understand your idea of incorporating all kinds of possible queries into one search engine. It's a solid concept... It just feels weird indeed to end up with so many results while, basically, just searching for a precise title - which, of course, seems the most basic possible use of that engine. I simply don't like this kind of approximation since it often leads to the mess one can witness on websites notorious for their horrible search engines (like Amazon). Type a word and, in addition to a few precise answers, you'll almost always obtain dozens of other elements devoid of any apparent relation with your query.

The use of any website's bottom part - even in other modern media, up to books - to store additional optional information is well documented, throughout marketing researchs correlating eye movement and reading habits with interface organizations. On most websites, one can be therefore sure to find some important (legal, etc.) information there and privacy disclaimers would feel right there, with any FAQ, About, Report sections or Forum link.
The same logic applies to any informational (sub-)panel, including research results, and explains why the "Purchase" button usually ends up in their bottom right section ^^. Gog's newer Thumbnail function, in its catalog, also demonstrates that... and is quite well done in itself. It is absolutely useless in my opinion (lol), but well done nonetheless.

Same opinion as Mr. Kgnao : As long as possible (base) price changes are taken in account - otherwise, obviously, the discount percentage would be misleading -, the discount information would be enough.

It's the base price changes that make the discount information alone not enough.
But when you factor in the different currencies, that's one more thing. Granted, GOG hasn't updated the exchange rates in a very, very long time, but when they do, that changes the actual prices in various currencies even if the discounts remain the same, so the discount information once again doesn't tell you how much you'd have paid and how the current price compares.

Hence my concern above ("As long as...). Of course, this solution would require proper algebra to take in account those changes and display meaningful discount percentages. Which, on the paper, doesn't involve more than a very easy formula combined with proper tracked values... But, of course, you're right : that part would be mandatory for the result to be really helpful.
It's not as if, like Mr. Kgnao said, this is a important detail anyway. I'm merely giving my opinion, as someone who doesn't really care about historic prices : simply knowing whether or not the current discount leads to a better deal (all prices taken in account) would be enough for me. But the other solution's fine enough too as far as I'm concerned...

Oh, another suggestion, while I'm thinking about it : how about mirroring to the right the standards you applied to the left, with the filters (for "Available now" and "Coming soon" games) and the arrow to remove out of view any uneeded category.
A few filters or checkboxes would save you the need to display lenghty informations after your titles - "(editions, expansions, modded games, goodie packs, demos and bundles)". Even "(games)", after "Available now", looks less elegant in my opinion.
Bonus effect : since, on the long run, you may want to display more information (other categories related to Gog, news about your project, etc.), you'd have space at will to do so...
Keep experimenting ^^ !

Is there a way to inplant a feature that shows what engine the game use ?

" this game using Unity "
" this game using UE4 "

etc

Maybe if the GOG DB had that info, but unfortunately this won't do because the Galaxy/GOG API isn't that open.

Thank you (also for the examples). I'd hoped to simply be able to re-use the data model for wishlists, owned games and hide lists, but I'll leave the "mix" designation for a further stage of development, then.

When I get around to this eventually, it's going to be hard to make this perform well - I think I'll probably have to end up with two "views"; the "mix" with comments, and the regular main page where you can filter by a list, but not see the comments.

There indeed hasn't been an exchange rate update since Sep 2023, but just from the countries I'm tracking, almost every month there are a handful of products which get their local prices updated for 2-5 countries at a time. Those adjustments can be quite large, too.

I find this perspective interesting, as it's completely not what I personally look at. That means I need to come to grips with the underlying viewpoint first, before I can determine if it's something which can fit in my vision for the site (and then if it's technically feasible / would improve things / ...) I also wouldn't immediately know what the "correct" percentage is to display for products with (local) base price changes.

So, if there's a product which in the USA has a base price of $10, was previously discounted to $2.99, and is now $2.97 (both 70%, but 1% better), and in Germany it was previously €9.99 with a 70% discount to €2.99, but has since had its price increased to €10.99, and now has a 71% discount to €3.19 - what should that game show? (I think that example shows why exact prices make it much easier for me to determine a single truth to display, but it's not a rhetorical question - I hope the answer will help me understand the viewpoint of wanting historical percentages.)

You call this an extreme example, but I can see the use of putting in a navigation menu to a couple of GOG pages, to really make gamesieve functional as an alternative homepage for GOG. I find it hard to determine which pages would be useful, though. E.g. personally I'd never pick the Blog (I've visited the thing thrice, I think, including just now, and every time I look at it, my eyes immediately glaze over the marketing layout and animations, and I move on.)

=> Are there people who'd seriously like a navigation menu for GOG on gamesieve? If so, could y'all give me a comprehensive list of which pages should be in it? If there's enough consensus, I can add this to the backlog.

(I also get your general point; I've been mulling over eiii's request from early in the thread for a news thread archive, since I'm likely to start gathering those anyway to more swiftly detect new releases.)

The only thing I find harder than graphical design, is marketing. I've been doing minor bits of self-promotion here and there, but each one feels like a major effort.

If people here are present in other communities where gamesieve might be well received, and would be willing to do some "organic" marketing for me, I'd be much obliged. :)

There is 2px extra whitespace underneath the title of each filter block, beyond just the spacing provided by the increased font size. Is that what you're talking about here? That's there generically (also underneath the titles in the "Available now" section on the right) primarily because when a title wraps to two lines, the subsequent content (filter or price) would be positioned closer to the second line than the second line to the first. I can't reduce the line-height of titles even further, since some fonts have long ascenders and descenders which could clash with a lower line-height. (For the filter blocks, it also prevents the top shadow (when scrolled down) from touching descenders on the title.

Ah, those crosses do look like checkmark crosses, indeed. (I couldn't use the cancel cross from the applied filters, due to - again - font metrics from that messing up my line heights.) But their functionality is very much different - "exclude" the filter (where the regular link is require), and when filters overlap (making AND-selection possible), that's something which is simply impossible with the OR-selection of multi-select checkboxed filters.

What I'm now curious about is: did you really not understand the "exclude" behaviour of those crosses? (I'd hoped the cross metaphor, together with the crossed-out hover effect would be much clearer than GOG's closed eye.) Or was this a more rhetorical sentence?

Yeah, I'm not happy with this myself. But I'm loathe to add yet another line before the actual results. Thanks for mentioning it, though - it's good to see which painpoints are shared.

Do you just mean that it isn't left-aligned with the title when it jumps to the next line? If that's not it, a screenshot would be welcome.
I do have a fix for this in mind, but it's touching the OS icons e.a. as well, so it needs a bit more time than I've been willing to give it.

That's a good one. Bottom aligning with the title was deliberate when the show button for included products was positioned there, but with the ratings top-aligning for the entire block makes more sense. ... it's annoyingly non-trivial, sadly, since lining out the baseline with the developer e.a. requires some tricky adjustments, but I'll try beating it into submission soon.

I already have the "search without applying defaults" checkbox there. Two is really the absolute maximum I feel I can get away with (I've prototyped this already), so I need to feel confident that the checkbox is really doing exactly what's desired, and that there's no way I can make the usecase work without asking the user to determine the desired behaviour.

Unlike others, probably, I am really only interested in percentages. I don't care whether the base price has changed in the meantime. I just want to know about the current and historical percentages, regardless of prices.

For the Unreal Engine specifically, many games using it mention this on the game page, so a search for "Unreal Engine" gives a starting point. (For Unity there are also some games like that, but a lot more noise due to the word being pretty generic.)

As dnovraD mentions, I lack (knowledge of) a good authoritative source which can tell me about this for each game. However, since I do have a starting point with the game page information, and feel it'd fit well with my overall goal for the site, it's conceivable that I'll at some point (far down the road) put in the manual work to annotate each game with this information. (With a very large bucket of "Unknown".)

Not sure about good sources where you could pull the info but both SteamDB and PCGamingWiki tracks the most popular engines.

Edit: Tired of fixing links, here they are:
https://www.pcgamingwiki.com/wiki/The_Witcher

https://steamdb.info/app/699170

As shown, for problematic URLs (containing colons), within the url= portion (not label), replace : with %3A -- Hopefully, that is of some assistance to you.