Rough Outline

[Gazook89]

What are some goals of this project? Since this is still in the very very early stages I think it's a good idea to actually list them out. It'll help any contributors that want to offer help by directing efforts. And we can be sure that any effort in one direction doesn't hamper efforts in others, such as decisions about core libraries etc.

So I'll start with just some thoughts I've had about it. I'm not yet advocating or pushing for any specifics on these, it's just my thoughts written down. I'll write until I run out of steam.

1. character sheets can look like many different things, even more so than the average TTRPG rulebook page. Any tool used to create character sheets should aim for offering maximum flexibility, or taking a fairly restrained/opinionated approach. Restrictions can be based on what the technology-stack allows, but also just on what we think a character sheet should look like. In either case, restrictions should have a stated reason.
2. Character sheets can be very complex. Often there are many graphics involved-- text frames, icons, themed fonts, and flowing layouts.
3. Sheets have numerous smaller text frames or inputs compared to book pages. Many of the text boxes do not need line wrapping or columns, though some probably do.
4. Good sheets likely group those smaller text frames into large frames for visual clarity.
5. There can be overlapping elements.
6. Character stats are commonly linked to other stats-- the input of one can influence the display of another.

At the start of this, I was skeptical that there was much utility in a new tool and thought that Homebrewery itself would be a good enough tool. Now I understand some limitations of HB for this:

1. Because HB uses one big editor input (the codemirror box), it naturally seems geared more towards long-form content like you would find in a book. A character sheet, being a form, should have a more natural building experience that matches the experience of filling out a character sheet.
2. The unique complexity of the page layout makes a compelling case for drag and drop editing. HB could possibly even benefit from d'n'd, but implementing it adds a layer of complexity which is a bigger risk when you are working with homebrew a hundred pages long.....character sheets don't have that level of risk, since they are only a page or two long.

Do other tools already have this locked in? Should we even take a step into this?

1. There are many graphic editors out there: Adobe, Affinity, Excel as examples of paid options, but also pixlr.com, photopea, Google Sheets, ....?
2. Roll20 and FoundryVTT and likely other VTTs have sheet building abilities

With the above in mind,

1. is the best underlying technology the CSS Grid layout? More structured layouts reduces 'freedom'. But many graphic designers espouse the benefits of using a hidden grid structure in layouts anyway. And the grid layout is easier to work with by making alignment between items easier. Right now, we are using react-grid-layout library--- I have looked at gridstack.js and it might be a better fit because it allows nested grids and moving items between grids.
2. ....HTML Canvas? Maximum freedom in terms of layout, graphics, etc. But maybe more difficult to work with when laying out text frames or other elements. Makes me wonder why someone wouldn't just switch to a more capable offline editor.
3. ....SVG?

I haven't touched on other capabilities, like the import/exports, saving, pdf generation, etc.

[5e-Cleric]

I'll tackle or comment on these points as i go, let me first say i agree with most of what is stated.

4. Good sheets likely group those smaller text frames into large frames for visual clarity.

I have looked at gridstack.js and it might be a better fit because it allows nested grids and moving items between grids.

1. Large frames can be easily dealt with new components or even custom components, i don't think most users will create hundreds of sheets, and almost every sheet is going to be wildly different, we could benefit of nesting components in the drag and drop, and if you really thing grid-stack is better for this, we can go that way, but that is going to be a very big change that will nullify most of the current work, we need to agree on this, and someone needs to take full responsibility and likely a due date so we don't overwrite all of this stuff.

5. There can be overlapping elements.

2. I don't think overlapping makes sense in a character sheet, i have never seen it, nesting yes, but overlapping? nevertheless, i think leaving the door open for overlapping can be a bit of a headache, but i guess we could set collisions conditionally based on user input? with a disclaimer to tell the user what it's getting into.

6. Character stats are commonly linked to other stats-- the input of one can influence the display of another.

3. My current idea for this tool is that of a PDF maker tool, creating a character sheet layout, not a character builder, if we do that, which we could, and users want to build D&D characters, we are going to need either:
   • Users imputing all their class data, which is a pain, or,
   • Official data input from roll20 or D&D Beyond, which i don't think we can get without paying a lot of money

We can't forget that more than 53% of all games played globally are D&D (according to roll20 data).

Do other tools already have this locked in? Should we even take a step into this?

4. The only real character sheet builder i have found is that of roll20, and it allows for one input only, HTML and CSS coded elsewhere, i sincerely believe there is a place for this tool in the WWW.

As for the rest, i reject SVG for this, as it is not its purpose or intention of the format. Canvas on the other hand could very well be an option, but would also require lots of unnecessary refactoring, and to be honest, i have never worked with it.