Release the Edits

Posted in

#1 by yorhel
2020-02-29 at 09:17
As part of a larger campaign to revisit and completely ruin... uhm, I mean, improve every aspect of the site, I've just uploaded a new edit form for release entries.

There's a bunch of changes compared to the old form. For the most part these changes are pretty minor - various fields have a more compact and/or less messy input method and the overall form organisation is structured to be a little more logical (hopefully). I got rid of the tabs, not because I didn't like them, but because the new form is just more compact than the old one and splitting it up into multiple tabs didn't seem very helpful.

The external links section in particular has been completely revamped, it's simpler in the sense that you can now just paste the URL and the form will figure out what to do with it. It's less convenient in the sense that you can't manually extract parts of the URL anymore (did anyone ever like that?) and the supported URL formats are also not as discoverable as before. I'll see if I can improve that if there's demand. The URL detection patterns probably still need a little tuning, so if you find a link that should be recognized but isn't, please let me know.

As with every update ever: The initial version sucks and I'm sure I broke something, I'll be fixing stuff gradually as problems get reported. The old edit form is still available in case something doesn't work. (And I still need to update d3 to reflect the new form structure. Ugh...)

Happy editing!Last modified on 2020-02-29 at 09:17
#2 by naiohoras
2020-02-29 at 09:23
it looks nice, I will mess around with it later and see if there are any bug or inconvenience.Last modified on 2020-02-29 at 09:32
#3 by eacil
2020-02-29 at 10:02
Can we have "both" as default for producers instead of "publisher", as it should be the norm?

I think it was already a problem before but the platforms are not correctly sorted, and I am not talking about the four shortcuts at the beginning. Dreamcast after the P, Nintendo after the S, etc.
A little bar after the shortcut would be helpful to differentiate the shortcuts from the alphabetical list, with Japanese and English as shortcuts for the languages.
And while we are at it, those two drop-down lists are very long, forcing you to scroll down and up the page. Wouldn't it be better to have a scroll bar inside the list like for the resolution field? Or three columns?
#4 by shining17
2020-02-29 at 10:24
Err, why removed the list of engines? I think it's better with the old drop-down list as I'm afraid with the new blank field of engine people will use different names for the same engine (i.e. Ren'py or simply Renpy). Or maybe I'm overthinking things...
#5 by yorhel
2020-02-29 at 10:48
Can we have "both" as default for producers instead of "publisher", as it should be the norm?
You're right, done.

And while we are at it, those two drop-down lists are very long
Yeah, I had been wondering what to do with that, but was curious about the feedback. One possible solution came to me when implementing the engine field - we can use an autocompleted input field to add languages and platforms. Downside is that it'll require switching from mouse to keyboard. :( Your suggestions may work, too.

I'm afraid with the new blank field of engine people will use different names for the same engine
Have you tried it? The engine field has instant autocompletion. If that turns out to be insufficient I can make it a little harder/more annoying to enter a not-in-the-list value.
#6 by beliar
2020-02-29 at 11:10
While we are at it, could I request a feature that I've been wanting for a very long time. We live in an age where VN releases get new languages added to them all the time. The current method of duplicating releases is extremely cumbersome and we eventually end up with something like this: Fault Milestone One. As you are remaking the edit form, maybe it will be possible to make it so new languages could be added to the same release, while also reflecting that the new language was added at a different date than the initial release.
#7 by ecchihieronymus
2020-02-29 at 17:02
Oooh, I like it. Only added two or three releases but this is nice, I like the OS selection. It was confusing to not have Windows in the top left section of that field. I didn't have a problem with inserting links but this is just as good/better.
#8 by lucumo
2020-02-29 at 17:21
Urgh, I don't like it...at all. Now it's way more messy. Instead of simply selecting platforms, one has to go through lists which go way too far down. What we had previously was really good. Tabs ensured that one could mentally check off a tab as being done and if one wanted a compact form, there was always the "all items" tab. I don't see why these changes were necessary...except for making creating releases take more time. More clicks for no gain.

/edit:

- the float chart-y input method isn't good; before, we could take a quick look and see "media", "quantity" and "medium", aha
- way too much red which is just distracting
- having an external links list overview made a lot of sense...also what was being required; now one would have to test the links...and well, there are a lot of Japanese shops out there while VNDB lists only a very small amount -> so not having a list overview of what is supported isn't good
- apart from the advantages of tabs mentioned above, more information is good and those information being split into tabs that make sense is also good; if anything, the tabs could have been more fleshed out with more information, rather than doing the opposite and removing information and/or making additional information suddenly appear
- the key to good usabilty is speed, and having as much information in a well-formatted manner available as possibleLast modified on 2020-02-29 at 18:09
#9 by yorhel
2020-03-01 at 11:02
@beliar: That affects a little more than just the edit form, but sure, such features should be much easier to implement now that I've streamlined the edit form code. Need to work out the proper data model and its implications first, though.

@lucumo: I like feedback, I really do, but I feel like you're grasping at straws and nitpicking about minor things that I already admitted weren't fully fleshed out yet. I find it hard to imagine how you can defend the old form when it either had such obvious deficits in some aspects and is pretty much the same as the new form in many others.

Languages/platforms: The most important aspect is being able to quickly see which options are selected. The old form sucked bad at this, especially the platforms. I often missed a checkbox in that unorganized sea of checkboxes. Now you can instantly see what is selected. The available options are only relevant when you're changing that particular field, so an additional click to avoid throwing unecessary info at you seems justified to me.

The complaint about adding more clicks seems weird. Only a few field need one extra click, which you already save by not having to switch tabs. The new form is for the most part number-of-clicks-compatible with the old one.

"Way too much red" when, in comparison to the old form, only three red (*)'s were added. Granted, it'd function quite alright being just white as well, I suppose, so I may have gone a little overboard. And form errors are more red now, yes, but those are *supposed* to stand out like a sore thumb. That's the point of error messages.

In terms of information: You, as a big fan of information density, should love this change. The new form shows exactly the same information as the old one and takes less space to do so. The list of supported shops is still there. Granted, it lacks supported URL formats, but I mentioned that as a possible deficit and was curious about feedback to see how much demand it has - not only the amount of information but also the relevance of it is important, and this is my attempt to optimize for it based on feedback rather than guesswork.

The media inputs in both forms is almost exactly the same, so I don't understand your argument there.

(As for "why was this change necessary": It wasn't, strictly speaking, but I'm working on rewriting most parts of the site in order to get rid of 12 years of technical debt, and in the process of doing so I'm trying to look for ways to improve a few things while I'm at it. I was never a fan of the chaotic and organically-grown release edit form, so here's one such an attempt.)Last modified on 2020-03-01 at 11:49
#10 by lucumo
2020-03-01 at 12:51
What were the obvious deficits in the old form? I didn't notice any. And as mentioned, the tabs gave the opportunity to add more information over time.

Really? The old platform overview was perfect. As someone who basically only added old games (PC88, PC98, MSX etc), it was very easy and quick to do. Not sure how missing a checkbox is the fault of the layout, as it was alphabetically organized and it had the system icons.

Switching tabs was optional though, thanks to the "all items" one for the people that preferred it like that. And as mentioned, the tabs served a purpose in mentally checking off a part of the creation process. So trading a useful click for a pointless one is...not ideal.

It's about visual clarity. Unless it's extremely important, it's mental clutter. At least I wouldn't categorize "Fields marked with (*) may cause other fields to become (un)available depending on the selection." as extremely important, same with "no language selected", as it's natural to add the language.

Huh? But you did the exact opposite by taking lots of useful information and hiding them behind drop-down menus or getting rid of them altogether.

The question is...why make it float chart-y when one could also see all the items from the start? Your brain already processes what's next, even if you aren't focused on it yet. However, the float chart design works contrary to that and as such hampers speed (especially for people for add stuff the first time).
#11 by savagetiger
2020-03-01 at 13:05
Why is media hidden by default?
And it doesn't show the Cero rating letter anymore, which was helpful as a shortcut because the letter on the logo is way bigger than the number.Last modified on 2020-03-01 at 13:33
#12 by skorpiondeath
2020-03-01 at 14:17
@yorhel: release is not recognizing this getchu link format when I try to add it: link (NSFW)
#13 by yorhel
2020-03-01 at 17:27
release is not recognizing this getchu link format
Fixed.

Why is media hidden by default?
Hmmmmmm that's an interesting way of thinking about it. I didn't so much hide it by default as change the input method from a dropdown to an "add media" link. But now that you mention it, maybe changing the paradigm from "list of 0 or more media" to "list of 1 or more media where the first one may be left empty" is indeed more intuitive. I'll go and play with that.

And it doesn't show the Cero rating letter anymore
Will re-add.

@lucumo: If you were able to instantly spot which of the many platform checkboxes were enabled, then good for you. I sure wasn't. :P I'm not sure what you mean with "float chart design"; the new form doesn't change the order in which you can enter stuff in any way, nor does it even suggest a particular way of working. The most significant change is the switch from "long list of checkboxes" to "dropdown selection". If that's what you mean, then should I instead have gone the other way and expanded all previous dropdown boxes into lists of radio buttons? That doesn't seem like an improvement in my book, for the same reason as stated before: It makes it much harder to tell at a glance which values have been selected.

As I mentioned in the announcement, I have nothing against tabs and I'll re-add tabs if the form grows again in size. But at the current size I don't see the point. And again: what information did I remove, apart from the supported URL formats?

I'll be unreddening that (*), I agree it doesn't deserve that much attention.
#14 by lucumo
2020-03-01 at 18:41
Flowchart (whoops) design is when you click on something and based on what you selected, something else then appears. Granted, the options here are limited but it's the same design. I already mentioned it before and savagetiger did again for one example: Click on "add medium" --> *drop down menu of media appears* --> select medium --> *drop down menu of quantity appear* --> select appropriate number. Same with adding the "producer", where you select the name first and then developer/publisher which is also illogical. Before, you could select developer/publisher beforehand and then add the name. After all, the title is always mentioned before the name of the company/person/etc, in every medium even. Replacing all the dropdown menus would likely be even better but I would have to see it before I can say something definite about it. The gain wouldn't be as big though, as it's mostly shorter lists. Dropdown menus work better for things where numbers are involved though (and where logic is involved), so release date, age rating and resolution...and the quantity of media which is why the medium itself is ok in a dropdown menu as it makes it unified for that point.

The supported URL formats are what I talked about and those were pretty useful, so it's a shame you forcefully compressed them. As mentioned before, there are a lot of Japanese shops out there and trial & error is meh, especially if one is familiar with those.

Thank you. But again, I would still suggest going in the other direction and looking into expanding the amount of information/options. The more people are involved, the more information become available and since you are expanding VNDB, the whole system should become more complex over time. One could always add something like information about DRM for instance but that would be a giant mess.
#15 by alto
2020-03-01 at 18:56
From a quick look, all positive. Even if the platform selection is a bit long, I much prefer it. I often forgot to add a platform and finding the right one out of all the icons was unfun.

Feedback:
- The suggestion in #6 would be very nice for increasing site clarity.
- "list of 1 or more media where the first one may be left empty" is more intuitive to me
- Adding links to external sites says "Edit" after you paste a link, "Add" would make more sense to me. Even then, I wonder if automatically adding recognised links isn't better - in other sites/software with the same UI design I constantly see people think stuff is already added and they never click the Add/Edit/Confirm buttons unless adding multiple things (if it already works this way without clicking the current "Edit" button, ignore me).
- There's only one Animation field vs Story/Ero ones on the old form.
- Maybe you disagree but I'm a big fan of inline help for form fields. I think it'd be cool to have the info from d3 on the edit form somehow, like alt text on hover/click of a field name or a question mark in a circle with excepts from d3.Last modified on 2020-03-01 at 18:57
#16 by beliar
2020-03-01 at 19:29
Adding links to external sites says "Edit" after you paste a link, "Add" would make more sense to me.
Indeed, "Edit" confused me the first time I saw it. I didn't understand if I should press it or not.

There's only one Animation field vs Story/Ero ones on the old form
Err, I can see both fields.
#17 by alto
2020-03-01 at 20:02
Ahh, it only shows if the ero animation if rating is set to 18+. I'm not sure this is a safe assumption to make, AFAIK you can have what people would consider "ero" in 15+ rated games and maybe lower. At the least it makes the ero animation on r34979 uneditable now :P
#18 by ecchihieronymus
2020-03-01 at 21:18
#17

I'm going slightly off-topic with this but why isn't it rated 18+ to begin with, those sample pics on MG very much scream 18+ even though the description says 'Mild Nudity'. I don't perceive that as mild nudity.
#19 by beliar
2020-03-01 at 21:46
While you could have checked the edit history, the game has an official M rating by ESRB (MangaGamer were pretty proud of that fact). And official ratings supersede the unofficial designations.
#20 by butterflygrrl
2020-03-01 at 22:09
... which brings me back around again to wanting to argue for fixing or removing the age rating field because it's broken and constantly misused, but that's not related to the new edit thing.

When you paste in a URL from a random site the form removes the list of valid sites to show only 'Invalid or unrecognised URL'. It might be nice to still have the list of sites visible. It could also be convenient to have a link in there somewhere that leads to a fuller breakdown of all the URL formats. While that info doesn't need to be bulking up the page all the time, it's probably useful to have it accessible somewhere for someone who wants examples?
#21 by kumiko1
2020-03-03 at 11:14
Can we get support for EGS links to unreleased games, e.g. link
and maybe DLsite links to unreleased games as well, e.g. link
#22 by yorhel
2020-03-03 at 11:36
Less complaints and more feedback, great. :D

I've added an expandable list of supported URL formats now.

Adding links to external sites says "Edit" after you paste a link
That's a bug. It should only say "edit" when you're pasting a different link for a site that is already listed (e.g. a different EGS link). In all other cases it should say "add" indeed. Fixed now.

I wonder if automatically adding recognised links isn't better
Good point, and something I kinda wanted to do, but it's tricky in the sense that it would not support editing a URL when it's pasted into the input box. I often lazily select only a part of the text I intend to copy, remember the part I forgot to select and then edit that in after pasting - it's faster than doing a perfect selection in many cases. Of course, that's a rather odd thing to do with URLs since you'll typically just copy the entire address bar, but still...
So the trade-off is in instant-adding and requiring perfect pastes vs. add-button and allowing editing. I'm still not sure I chose the right solution.

Maybe you disagree but I'm a big fan of inline help for form fields
So am I, and it's something I've been wanting to do for a looooong time indeed. But I can never figure out the best way to do that, either in terms of code/maintainability or in terms of UI. Granted, I haven't spent all that many braincells on it yet.

fixing or removing the age rating field because it's broken and constantly misused
I mean... yes, sure, I agree. But what I'm afraid of is that that makes us suddenly responsible for flagging VN contents. We're already having enough trouble with images. That's not an argument against doing it, just that it needs a better thought out proposal than the ones I've seen so far.

DLsite links to unreleased games as well
That URL should get recognized just fine.

Can we get support for EGS links to unreleased games
They have a different URL format for unreleased games? Wasn't aware of that. Does link work and does it point to the same entry?

EDIT: Which reminds me, I still had a TODO note to add more sites suggested in t12755. Maybe I can get to that now that I've made the external links management a lot easier for myself.Last modified on 2020-03-03 at 11:39
#23 by rampaa
2020-03-03 at 12:09
EDIT:
just that it needs a better thought out proposal than the ones I've seen so far
Some proposals then:

1) Mark r34979 as 18+ and declare that in-house ratings trumps official ones.
Advantages:
-It's the easiest solution.
-It doesn't result in redundant information for the majority of VNs

Disadvantage:
-It doesn't solve the problem with "All ages" titles not being actually all ages but rather non-ero
-It might not solve the problem with console games having 18+ rating without ero content.

2) Create a boolean checkbox for sexual scenes and only allow official ratings (or self-proclaimed ratings of developers) to be used.
Advantages:
-It's a pretty easy solution to implement.
-It doesn't need us to ignore official ratings.

Disadvantages:
-We will have to find all the instances where "All ages" has been used as "non-ero" and change their rating to Unknown.
-Since most of the "18+" games on Windows means ero, there will be a lot of redundancy.

3) Ditch the rating system all together and just have ero/non-ero flag.
Advantages:
-It's a pretty easy solution to implement.

Disadvantages:
-We will not care about official ratings.
-We will lose some nuance for ratings.

But what I'm afraid of is that that makes us suddenly responsible for flagging VN contents.
Are you hinting at the fact that tags should not actually be VN specific but release specific (or Edition specific, t13027.20, your last paragraph)? If that's what you mean, I don't see how not addressing the rating problem will change that fact.Last modified on 2020-03-03 at 12:29
#24 by kumiko1
2020-03-03 at 14:20
That URL should get recognized just fine.
Yes it gets recognised but the link to the page is broken, see r66506

Does link work and does it point to the same entry?
Yes it redirects you to the before_game page automatically if the game isn't released yet.

Also if you're adding more sites now there's no problem having lots of options, some ones you could consider include amazon, melonbooks.co.jp, twitter (Japanese VNs often have their own twitter account), suruga-ya, sofmap, toranoana, and traders.

Edit: incidentally, the before_game EGS page is where information on trials is, so although its not really that important you could make a way for trials to permalink to the before_game page.Last modified on 2020-03-03 at 14:27
#25 by tahu157
2020-03-04 at 06:01
One thing I've screwed up a few times when adding store links is putting the wrong version of the Getchu link. Usually I accidentally link to Getchu's physical games store rather than the downloads store for download editions of games. I'm probably not the only one who's done this too. When dealing with Getchu links, would it be possible to only accept dl.getchu.com links when the medium is set to Internet Download?


EDIT: Maybe also if the text "ダウンロード版" or "Download Edition" is detected in either of the title boxes, although that seems more prone to error.Last modified on 2020-03-04 at 06:04