Everpresto

Automating the document signing process, cause you've got better things to do.

Why do reference id attributes on models not have a ref option?

If the ref option is provided. The reference id will not be loaded with with the document. The tradeoff to not using ref is that mongoose's populate function won't work. See Automattic/mongoose#3020

Why are models being required by import instead of mongoose.model in other model files?

The is a dependency loading issue. We might want to use ModelB in a static function of ModelA, now ModelA depends on ModelB to be loaded first; but what if ModelA is loaded before ModelB? This would throw a MissingSchema error. This type of cross dependency can get very complicated very fast - so when using models in other models, it's best to import from the direct source.

Why are we using Model.find + Model.save in order to update vs. Model.findOneAndUpdate in some places?

Because Model.save is a Mongoose method, which means Mongoose Document pre/post hooks will be run every time save is called (this comes in especially handy in Template.js). Model.findOneAndUpdate and Model.update are actually MongoDB methods, so calling these will not invoke Mongoose Document hooks such as pre('save'). See Automattic/mongoose#2672

**Why are we using res.status(200).json({}) instead of res.status(204).end() when successfully closing a delete response?

This is because we're using the fetch. In fetch, data must be converted to JSON using response.json(). However, if the response is empty, response.json() throws an error. Therefore it's better to close the response with an empty object instead. response.json() also fails when the status code is 204, this may be because fetch interperates 204 as an empty response. Therefore, we have to make due by using 200 instead. This is a necessary step to normalize data because there are some instances where we want to send back data after a succesful delete.

Provide feedback