search

dovy / Pedigree-Viewer

Real-Time Collaboration is open sourcing the effort they made to create a fully html5 pedigree viewer. This works on most modern browsers and features panning, zooming, and view modifications.
GNU Affero General Public License v3.0
23 stars 11 forks source link

Tree is not search engine friendly #2

Closed stuporglue closed 11 years ago

stuporglue commented 11 years ago

Pedigree should be printed out as static content in a

dovy commented 11 years ago

That's fine if it's one GEDCOM per page load, but what about dynamic content? Now you're going to basically output the entire gedcom to the HTML dom before making the JS tree. Is that within the scope of a pedigree viewer?

stuporglue commented 11 years ago

It would be a single GEDCOM, and it probably needs to be the whole gedcom too -- at least names, dates and places of individuals. Each person wouldn't need to be in it's own div, or even it's own

necessarily. If they were all in a single div, the load on the DOM wouldn't be that great.

So, the DOM would contain this text summary of the entire GEDCOM file in

Users aren't going to see this content (unless they have JavaScript disabled) it's mainly for Search Engines with a secondary goal of making it somewhat accessible to screen readers.

I care about search engines because when I put up my family history website, it's with a goal of sharing my info and making it findable to others.

dovy commented 11 years ago

Hmm, you're thinking very GEDCOM based. What about those who use a database? Or the results from an API?

Just saying if you have like a GEDCOM file of some 3000 names, that's going to take a while to generate a file to utilize. Also then using json to parse that file and get the proper nodes created will be a memory beast. This may be a side step into your PHP gedcom parser. I'd suggest a cache of json and a cache of html for including in the page. Not sure the best approach here.

stuporglue commented 11 years ago

My goal is that this will be something that non-programmers who are minimally technically savvy can use. Ideally if you're technical enough to find this project via Google, then you would be technical enough to use it. So, the non-programmer genealogist who has a website should be able to unzip the project, plop in a GEDCOM and have it "just work".

The more technical can modify makeJson.php to connect to an API, a database, or whatever they want. Or they can modify the call to drawChart() in index.php to call a static file, a ruby, python or .NET script.

GEDCOM has its flaws, but it's the lowest common denominator.

As an aside, the PHP gedcom parser is pretty fast. If you can get me a big gedcom file, I bet the times would be acceptable, even in a 3000 name file.

stuporglue commented 11 years ago

"Is that within the scope of a pedigree viewer?"

Upon further review, probably not. Print the info you want on your own page. Just use this for tree display