I learnt 3D modeling on my own in just 14 weeks, and earned an honorable mention and staff pick on Sketchfab. Sharing my journey and hope it's helpful to you :)

Let me know your thoughts!

Curious to hear your thoughts and feedbacks on the topic.

Yah, exactly how I feel about it. And then "save" takes at least two clicks, and you need to wait for it to load your boards, while "like" is just one... :sadpanda:

Another great resource: https://www.ibm.com/design/language/experience/data-visualization/chart-models/

If you have experience in Javascript, I think Javascript backend libraries could be the best place to start.

Personally I really like Parse.js and Backbone.js. It's very easy to learn and get started. Perfect for prototyping and validating your startup ideas. I got very excited when I found Parse.js and wrote an entire tutorial for it: https://code.tutsplus.com/series/building-your-blog-with-parsejs--cms-651

And this is an MVP stage product I made with Parse: http://premade.io/#/new

If you want to be more future-proof, Node.js could also be great. But fair warning, the MERN stack learning curve is steep.

Hi Eliot, I'm the author of the article there.

I think we are basically finding the same solution. The only difference is just I chose padding over margin. My reason there was margin only stack vertically and padding would make the framework more consistent in both axis.

That said, if margin worked for you, that's great, too. Glad we found similar solutions :)

Yup you are definitely right. We were testing no underline upfront, traditional underline upfront, and medium style underline upfront. If we set the no underline version as the baseline, traditional underline brings 21% uplift in CTR, but medium style underline only brings 11%. And that caused my confusion.

Yup. See the example in my comment above, should be pretty clear that it's an underline not a decoration tho.

Safari users aren't the majority of our user base.

It's the title in our data visualizations like the "Stephen Curry 2014-15 Playoff Shot Chart" in this example here: https://w.graphiq.com/w/315cxySqXzL

Sample size was a few thousand, large enough to call it significant. We decided to move forward with Medium style for now, but that's definitely a strange test result to me still.