Documentation Issue: No automatic persistence in node.js

[tasinet]

Hello,

The project README mentions: "In a nodejs/node-webkit environment LokiJS also persists to disk whenever an insert, update or remove is performed.".

Is this accurate? I tried to look for this functionality in the source and it doesn't seem to be there - nor could I reproduce it with code. It looks like we have to explicitly call db.save() when we want an on-disk checkpoint created.

Am I missing something? If not, might be worth correcting.

[techfort]

You're absolutely correct in that the readme is wrong. It used be the case but we thought it better to let the user decide when to perform disk saves. Having said that, the first feedback we got is that many user may be attracted by this particular feature so my short term plan is to implement a .use(config) or configure(config) type-method that will allow the user to specify when to perform disk saves. This can be on particular events, but we are also researching a light and fast scheduling algorithm for scheduling jobs.
Would that suit your use-case?

[tasinet]

I recently discovered Loki, not quite sure what my use case is yet. In any case, I don't really care for that particular feature personally. My point was to get the README fixed ;)

I agree that letting the user select disk saves explicitly is the better route, as with large datasets the save op could get expensive - as you're writing the entire dataset each time.

[Ghirro]

I'd agree with tasinet here. Repeating a save operation at any useful point is going to dramatically increase overheads within the application, it makes far more sense to me to have the user explicitly call save().

One potential solution may be to have insert/update write a new record to disk separately (without the need for a huge file write) and then save effectively being a function that takes the new writes and combines them with the database. That way if your program crashes you'll be able to run a save call to amortize the dataset without loss.

[techfort]

We have been talking about incremental datasets, so I like your idea Ghirro. Meanwhile I corrected the readme so I'm going to close this issue, feel free to open a separate issue about incremental saves if you feel strongly about the feature, there's a very concrete chance this will be in the cards soon.