V2 feedback needed: custom character sheets

I guess a nice thing about my DACA having expired is I have a lot of time on my hands and have made a TON of v2 progress. I've about 70% of the way through, functionally speaking, though the remaining 30% will likely be the most intensive. That said, one of the features on my list now is character sheets, and I need advice/feedback from the community.

The current forums and custom character sheets work off a system called BBCode. BBCode is OLD, from before the turn of the century. Before Javascript could do much, it let people "code" without actually having to write code. As GP is now over 12 years old, it was the right choice at the time. In v2, I've moved the forums to a WYSIWYG called TipTap. BBCode is stored as it's written into the database, and then converted to HTML when shown to the user. With TipTap, you don't write any code yourself; you use the toolbars to change what you want to live. So the two systems don't mix.

Most of v2 is meant to be a carry over of the current system into a modern mechanism, so it can be updated and maintained more easily. One big change I was thinking of making was to build a form builder for custom character sheets, similar to how Google Forms works. You could add a component (d20 stats), drag it to be where you want, and build a sheet out of blocks. I could then make all (or most) of the current character sheets out of that tool, and all new custom character sheets could be built that way. Some pros and cons of the new system:

Pros:
- Building and sharing custom character sheets becomes REALLY easy (share a link to a custom sheet template instead of copy/pasting a BB template)
- Building new "official" character sheets becomes easier

Cons:
- "Official" character sheets won't be as visually appealing (I've tried hard to make the official ones as close to RL sheets as I could, including graphics; that would be much harder under the new system)
- I'd have to build specialized tooling that would allow you to view old custom character sheets, but they would not be editable; new custom sheets would have to be created for them

I'm hoping for feedback on what people think. This is a bit of a crossroads that didn't occur to me during early designing. If I were to keep to the current templating system, I'd have to maintain both the old BBCode alongside the new TipTap editor, and the site would be running on two different tools. The custom editor would just be a text area where you enter BBCode manually (no toolbar like the current site). Any other pros/cons folks can think of to help me decide? Any strong opinions on which way to go? I know custom sheets are a necessity, due to how limited the current system is.

I like the idea of the new template system, but I think I'd have to see it before I could be sure.

As 'used to' the old system as we are, moving forward into the new TipTap based format is the correct step, I feel.

Many of us have been around BBCode long enough that we can use it but the number of folks casually familiar enough with it to be functionally proficient is declining rapidly and a more modern WYSIWIG UI is probably the way to go.
Those of us who use the Custom Sheets put in the time to make them before and, odds are, we'll put in the time to recreate them again.

I don't like visual programming or forms/blocks based stuff, so I doubt I will like TipTap. Also I'm sorry for the touchy topic, but the first thing I see on their site is "AI-native tools", and that makes me want to do nothing with it.

I get the pros of making it easier for people who don't like coding, but my personal opinion is I'd rather stick with BBCode. Or I guess something else that's newer but allows actual coding.

Yah, I'm miles away from their "AI" stuff. That's the marketing side. I use the standard WYSIWYG without touching any of their AI content. I admit I use Claude in developing GP v2, largely because it helps me do a lot of the conversion in a lot less time than it would take to do manually. I'm not really happy with the direction of AI either.

Unfortunately, there's really no in between of a WYSIWYG and freeform that I can think of. I'm open to suggestions.

Unfortunately I don't really have suggestions, forum development is not in my expertise. Maybe if keeping BBCode would be strenuous to maintain, then perhaps the new editor can have a raw HTML option instead? Users who don't want to code can use the visual mode, and those who do can use HTML (and perhaps add features which would be difficult to add in the visual mode).

Unfortunately that's not really viable. It would mean maintaining two different, unrelated sets of code. It could easily lead to a lack of parity, and require more work to implement stuff. I am definitely not going to implement any raw HTML system. It's a GIANT security hole, and honestly, even some of the CSS stuff Adam added is problematic and is going to be removed in the v2. The more advanced CSS and JS are getting, things like that are only going to be possible on larger sites with security teams behind them.

And it's not that BBCode is hard to maintain (it is, but it's acceptable), it's that it's outdated. There's only so much you can do with it without it becoming coding, and that's not something most users want to do. It limits how much people are capable or willing to do, specially in a world where we can have a button that will insert a note block, allow you to add users validated by the backend, and then add functionality on top (I was able to build that in TipTap without too much hassle). How often does someone break BBCode because they forgot to close a tag, or closed nested tags in the wrong order? It's just not user friendly.

The only other thought I have is to have a template builder, that you write in bbcode, and that gets turned into an HTML character sheet? I wonder if I can build something like that in JSON instead? So instead of maintaining two sets (BB and Prose), sheets are converted into HTML templates which are then filled out, like our standard sheets?

I like what FlyingSucculent is suggesting here. As someone who used to build websites for a living waaaay back in the old days, I have an inherent distrust of WYSIWYG editors. The amount of garbage code that would be added was frustrating when one would have to go in and edit them.

I realize that a lot of time has passed. WYSIWYG editors have probably improved, and bandwidth constraints of the old days are no longer a concern. Still, I like the flexibility to edit code when I can, even when its BBCode.

I find it surprising that the character sheets will be less visually appealing with the new software. I am more of a function over form guy, but I don't know the extent of the drop off in visual appeal you're referring to. A screenshot of before and after (if possible) would definitely help understand the difference.

Btw, what happened to the "edit" function on posts in this forum? It appears to be missing over here.

I see the Edit button myself, so not sure. The problem with allowing HTML is that it opens the site up for injection-style attacks (even the tricks we use to do fancy things now with forum posts are dangerous). While I'm sure that no GP member would do anything, those kinds of attacks ARE on the rise.

And anything like that leaves GP open to legal liability. Unless we want to move GP to a paid model where I can make enough money to be able to afford something like that, security is always first.

Fair enough on the HTML. The suggestion was mostly inspired by online games which let users to modify HTML of their profiles and forum posts, but I admittedly have no idea about the security measures behind it. :'D Or if there are any, in cases with small teams - or even big ones, sometimes.

I still would like BBCode to stay, but if I am in the minority, then I guess I will begrudgingly deal with a visual editor. I just don't like these, I like being able to type everything on the keyboard, not click buttons. ;...; (Also most of visual editors are a real pain on mobile, so I guess that's maybe another con? Depends on how it'll look, I suppose.)

PS: semi-unrelated, using the formatting buttons on GP doesn't work for me like third of the time, and I still can't figure out why. Hopefully the new one doesn't break it completely, that'd be awkward. XD

(I don't see Edit either, probably permission thing?)

I'll look into the edit issue. It's likely a broken permission. You do bring up a good point about editors and mobile though. TipTap does allow for keyboard shortcuts, which we can all learn over time.

As for formatting, going to a WYSIWYG allows for more control over formatting controls. I want to say i share your concern over WYSIWYGs. I spent days doing research on modern WYSIWYGs vs BB Editors that still exist (nothing's been made new in a long time, the rest are old and just maintained). I opted for TipTap after a bunch of research and experimenting.

I do also like this! If it's possible.

I see a lot of advantages to TipTap and WYSIWYG for forum posts. It's more approachable, avoids BBCode headaches, and gives you much more flexibility to add functionality.

I'm more hesitant about using the same approach for character sheets, though. A sheet is more like a web UI than a document, and I'm concerned that a form builder could make more interesting, automated or visually distinctive sheets difficult or impossible.

I like your idea of moving toward a JSON-based system. If it provides enough building blocks—fields, checkboxes, lists, tables, images, sections, calculated fields, layout options, etc.—I think it could work very well. I'd just want to see a moderately complicated existing custom sheet recreated before recommending it.

One other suggestion: don't have "special" built-in character sheets at all. Instead, make the official sheets simply premade custom sheets built using the same system available to everyone else. That would make it easier to do "dogfooding" and the distinction between "official" and "custom" much smaller. Ideally we could create, modify, and share all sheets ourselves using the same tools.

I also don't have permission to edit my posts in this topic.

Here's a YouTube video showing how TipTap works on v2.

Looks great for posts! 👏

You've hit one of my main pros to the templating system: I build the "official" sheets via the templating system, and someone can always copy/paste it into their own template, or create something from scratch (be that a build-a-form or JSON).

That also hits the con though, where visual fidelity would drop. For example, the 7 Seas character sheet uses one of their graphics and CSS positioned buttons, which would be much harder under either new mechanic.

It's a trade of ease vs uniqueness. I could always do the official ones custom made. They'd be prettier, more functionality, etc, and have another mechanism available for custom sheets, which would definitely be a step above what we have now, but would be less visually distinct. However, I do point out that it being code driven (be it bbcode or JSON) limits how many people build sheets. A drag-and-drop system would allow more people to work with it, and if the drag-and-drop builds a JSON (which is the sensible result, I think), and people could then edit that JSON, that kinda covers both sides?

And for clarity, I am NOT proposing using TipTap for character sheet generation. It would be an entirely separate system.

I think an advanced/editor mode might be that good middle ground you mentioned. The drag-and-drop builder could cover the common cases and make sheet creation accessible to everyone, while an advanced mode could expose the underlying JSON (or whatever representation you end up using) for people who want more control. Maybe they enable a special flag in their profiles, to access that JSON... We could use it for special graphics, positioned buttons, margins, etc. I still think we should be able to create the 7 Seas character sheet, even if only in this advanced format.

That way you're not forcing everyone to deal with code, but you're also not putting an artificial ceiling on what a custom sheet can do. Even if only a small percentage of users ever touch the advanced mode, I think having that escape hatch would be valuable.

Is ease of editing that important for character sheets? Users who make them are usually the ones who don't mind a bit of coding and formatting elbow grease. I doubt the number of people who actually want to make templates but don't because they don't know BBCode is very high. But I suppose we don't really have statistics for that...

Still, is it important enough for everyone to be able to make a sheet template? Even if someone can't build a template for a unique sheet they want, people here are helpful; someone usually offers to help. Accessibility to general public is good, but I dunno about sacrificing advanced customisation for it.

The building a JSON idea does seem like a decent compromise. Time to brush up on JavaScript, I guess!
Is the purple colour in quotes customizable? Colours this bright might be a bit much to look at a lot while reading text.

Just to clarify, I meant reading while editing; I imagine the actual quote doesn't look like that. XD

There's still A LOT of design work to be done. It won't be that bright purple.

That's good to know, thanks. (I'd vote for the same colour as current GP buttons. Maybe a bit darker.)

I'm not so concerned about visual fidelity as I am about functionality. I'm in the camp where I have existed with bbcode long enough that I'm not hopeless with it, but I can't make it sing and I build very visually simple custom sheets that work fine for me. I have used WYSIWYG editors, but not to build forms that need to be able to grow and change with time (some games get more complex as you level up, and it would be nice if character sheets could grow with them.) If there's a way to import bbcode sheets (and the library of custom sheets we have) functionally, if not visually, into the new system it would be good.

I'm concerned about anything that breaks reverse compatibility on the user side, so anything that breaks BBCode is worrisome.

I'm concerned about not being able to copy/paste a post as plain text for saving off-line. WYSIWIG tends to be corrupted when copying between different locales because the formatting standards are not implemented consistently between different software and site.

I'm concerned about sheet editing because I often want to do fine adjustments like changing the positioning/alignment of the character portrait depending on how the text flows around the page or what picture dimensions I have. I also had a case where I edited older sheets because I came up with a QoL feature that I wanted to add to it.

Apologies if these impressions of potential changes are too superficial.

There's no good choice. BBCode is an old and outdated system that has very little in the way of modern support. I couldn't find any actively managed projects that support BBCode. And I don't have the time or inclination to build and support a project like that myself. Truth is, we need to move forward, or keep the site as it is now, which works, but cannot grow.
This is a good point I can look at how it's working. I know currently folks often write posts elsewhere because GP will randomly not save. The nice thing about the new site is it's locally stateful, which means if a post fails to send, I can rig up code to save it local (to the device). So if a post fails, you can try again with no problem, and the server crashing won't crash your local.
So converting told sheets into any new system will be difficult. But I always intended on giving people fine control as much as possible.
Valid concerns, and I'm open to it. I'm just not sure of the solutions yet. I feel like a builder + JSON is the right step for custom sheets. Adam's BBCode additions were a great hack to add versatility to custom sheets, but it's inherently limited without exposing security holes. I understand the concerns over the new editor, but I'm hoping we can address them in a way that makes folks happy.

A quick test, the new editor doesn't copy into raw text, but it does copy into a rich text editor (LibreOffice Writer for the test).

Anything that makes the site more user-friendly for the average, non-specialist user is a positive for me.

I'm somewhat surprised by the claim that BBCode is outdated. It seems to be a de facto standard for what 'a forum' in the normal sense looks like on the Internet. SJGames uses BBCode. The Onyx Path uses BBCode. SufficientVelocity uses BBCode. Myth-Weavers uses BBCode too (despite having a WYSYWIG and HTML option too). It seems like any forum dedicated to RPGs that I remember visiting uses BBCode.

Now, I'm not a webcoder, so maybe I am missing something about how all those places approach to use of it, but it seems fora built on it are very much still active.

I suppose something like Markdown/Markup of some sort is a reasonably lively contender against BBCode, though usually outside the normal forum ecosystem and more in places like Discord and Reddit (and the latter seems to have had issues with WYSYWIGifications lately).

I took a look at all of those to remind myself, but they're all ancient (web speaking). None of them uses any modern technology I can see. SufficientVelocity and Myth-Weavers are in fact written in phpBB, which while still an active project, is very out dated. Like I said, I've had no luck finding a bbcode editor written in any modern technologies.
I'm not denying folks still use it, but that doesn't make it less old/outdated. It was definitely the defacto standard for forums for, I'd estimate, 20 years. All the sites you mentioned are largely the same site now that they were when they were created, and in some cases, even without a fresh coat of paint. I see barely any Javascript on any of them, next to no interactive functionality. GP is well more than a forum. The tools and features take it beyond that. I started GP as basically just a forum, and it was lost in the noise.
Markdown is too limited and not expandable to be used. It's not viable for a forum beyond basic formatting, unless you're willing to break standards, at which point it's not Markdown anyway.

For forum post editor, is it plausible to have something like "Paste as BBcode" option where it would detect BBcode tags from raw text and do the formatting based on it? Maybe not all, just common stuff like [b] or [i]? Sort of treat them as special symbols?

I predict that the new editor will make my mobile posting difficult, since I usually just type out the tags while writing and then copy-paste into the editor. That won't work as well with a visual one. :( It's not that big of a deal, but it would be a nice feature. I dread to imagine having to select dialogue on mobile to bold it. XD

I am putting serious consideration into the mobile concern. But in general, the idea was to not copy over the bbcode parser, because again, I'd have to maintaining multiple separate systems and keeping them in sync.

Maybe some sort of markdown then (not that it's probably much better maintenance-wise)? Although it really isn't that much of a pain, just annoying when you have to select text to format it. But, well, I guess it is a matter of habit. XD

So how's this for a starting point (and boy, do I mean starting point). This JSON:

{
	version: 1,
	elements: [
		{
			type: "section",
			content: [
				{
					type: "text_input",
					name: "name",
					label_pos: "top",
					label: "Name",
				},
				{
					type: "text_input",
					name: "class",
					label_pos: "before",
					label: "Class",
				},
			],
		},
	],
}

results in this form:



Initially, I'll allow a bunch of helper classes and a white list of style keys to allow more fine control.

So to clarify, each of these content types would be a button we could push to add it, and then we could also code it directly if we wanted to?

Still figuring out how the UI part of it will work, but yes, you'll have a UI to use and a code pane. We could always release with just the code part (and docs on all the functionality) wihle we build out the UI.

Looks good to me!
Will sheets still support having editable sections ([abilities] tag)? It's one of my favourite things about custom sheets, being able to just edit one section and not the whole sheet. :D

I don't think I follow. Still figuring it out, but custom sheets will now live at the same "level" as pre-built sheets. When a sheet is created, it'll start as v1. You can then select it as a character type to create to fill out. If you edit the sheet itself, it'll become v2, etc. So if you move stuff around, it'll continue to work (you'll just need to update any created characters to v2). But I'm building blocks for things like a "repeater", which functions like the dnd 5e skill section: some inputs + a "add skill" button, which i don't think can be done with the current custom setup?

I meant more the specific functionality of the [ability] tag, being able to edit a specific section of an already existing sheet (not the template, a sheet) by clicking the quill icon

Like, for example, with the current 13th Age sheet If I want to edit a section in it, I have to press Edit which allows to edit every section at once. But with a custom sheet, I can just press the quill icon next to the section and only edit said section without editing the whole sheet.

I'd like a class/element/block which can function like that, if possible?

"Add skill" one is good, I like it. But it might get pretty long for inventories, so just a text field which can be quickly edited would be handy. :3 I think so, at least. I know not everyone uses the quill button.

Oh, interesting, no, didn't think of that. It's doable, but given this site is API driven and thus will be a lot faster to edit, I wonder if that will be as necessary? The way it's being built, it should "hot swap" between edit and display modes.

I guess we'll see in practice then! Good to know it's doable though, maybe if there is demand it can be an add-on later on. :D

If it's not a big hassle, I would like to have that feature too - in my character sheets, the "ability" section functions a blank space to jot down notes with unlimited size.

If 'add skill/trait/whatever' becomes a thing, one thing I'm wondering about is calculating point totals. Right now I have a manually-made sheet that does that for skills, but does not allow adding more lines (since it would need to adjust the formula). I wonder if a formula that goes 'sum all across [category]'s [field]' would become possible instead.

I just added a formula system yesterday! It's rudimentary, but I'm sure we can expand it to fit needs.

Also, each block will be versioned, so if we add new features or change how a block works into the future, old sheets won't break.

Here's a doc outlining all the features of the current sheet builder. Feel free to give feedback. It's definitely not a fixed or exhaustive list, just the starting point. I know it's a bit technical. https://docs.google.com/document/d/1JWv1JewvUBHOh-Th19iBIvyWBT20I1G2124Qb2VwcaE/edit?usp=sharing

Ooh, I like what I'm seeing!

Are you going to add any node types for image elements, or would they be limited to background-image? I like adding full size avatars to my sheets, thus something to replace [img] on it would be nice (plus the usual avatars). :3

And maybe if you are going to be changing how avatars are added, sheets can support having multiple avatar images? So player can choose between them when posting (like different expressions, or different art for characters who shapeshift).

It looks interesting, but it's outside my expertise so I can only sort of parse what I'm looking at.

Yup, can do that.
That's an interesting one that would be a bit more difficult under the current structure. I'll have to defer that one to a later update.

I am not a coder, but one of the things that I love most about running games on GP is the ability to store groups of NPC avatars and easily post as them. In BBC I think it is the [npc] tag. Perhaps that touches on the multipe avatar images, but is this a feature possible to include? Even outside of character sheets, the GM having access to a toolbox of NPC avatars is wonderful.

Totally agree, C1NDER. Even if that easy-post-as-NPC feature isn't live out of the gate, I would consider it a core requirement -- it's a *big* differentiator vs. other PbP sites.

Took a cursory look at the document. Particularly like section 7.3 - Operators

Adding my name to C1NDER and Harrigan regarding the post-as-NPC feature.

If sheets could contain a set of avatar + name pairs you can post as, it would work for both making NPC lists and characters with multiple avatars. Just saying. :P

Speaking of, how would posting as a character work with these new sheets? In the example in the document the character name is not a class of its own, just a label-input group, so is it going to be a separate field automatically added for all sheets? Or is there going to be a class for it? Or would it depend on the input name? Same question for the avatars.

Avatars are separate, and name and notes are required fields in any character sheet. If you remove them, it won't save. This means you can have name anywhere in the sheet (doesn't have to be right at the beginning), as long as it exists.

As for the "post as NPC", do we mean as a NPC that doesn't have to have a char sheet? In the past, I considered NPC sheets, simple, low field count sheets that had basic info and an avatar. That's something someone could design (stat fields, maybe a skill field, notes) + avatars. As a note, I've upped the character avatar count to 5. Still figuring out how to do the posting part, but yah.

As for posting as NPCs, both full-sheet and no-sheet options are of use, because not all NPCs on whose behalf one posts have sheets. (Also, sometimes one posts on the behalf of a PC using a sheetless avatar, e.g. when a PC has many disguises or shapeshifts improvisationally.)

Ok, maybe we can do something where you can upload NPC avatars and then just post as a name. But I'll probably have to limit it, to avoid people spamming avatars and to not overuse space. What's a reasonable limit? 20?

I'm pretty sure people have way more than sheetless 20 NPCs. I nominally have 20 NPCs-who-will-eventually-get-sheets in my campaign, but I'm trying to run an academic setting centred on a specific class.

Also, can we retain externally-hosted sheetless NPC portraits like we have now?

For me, something as "simple" as this element would suffice:To render it, my suggestion would be to show a new drop-down, under "Post as", if there are more than 1 avatar for this character

Hm... I guess space is cheap, we can let it be unlimited, and if it becomes a problem, solve it then.

Just as a point of reference, I've run multiple games that had 30+ NPCs who were just a name and a pic.

An elegant solution :) I think that works well.

As Harrigan indicated, I too VERY often use NPCs without sheets. I keep NPC stats and stuff separate, so on my GM sheet I'll just have lists of 'NPC avatar + name'. It's really great for style and helping the players quickly identify with a visual who is talking in dialogue.

Keleth - if it is helpful to see an example in action, HERE is a post where I used multiple NPC avatars from the GM sheet for a back-and-forth dialogue including multiple NPCs in a single post.

And HERE is an example where a set up multiple avatars for use on a PC, as the character would change physically depending on how he channeled his abilities (toward the bottom of the sheet).

Mind, one of those is 403, though perhaps that can be circumvented by a server owner.