Thanks, Sourav for that crisp explanation.

I don't think so. I just re read the git-scm-tagging just to confirm your this statement a tag is a line segment of a branch so it has a defined start and a defined endpoint. It's just another way to name a commit, you can do git show <tag name> instead of commit hash and it will show you that commit. Which means it's not a line segment, it's a point that's all. it's one way to define 'stable' reproducible packages. I do agree with this one though. Let me know if I am correct. Tag is one of the lesser used features of git and requires a bit of understanding before one can get mastery over it

I don't agree on the first point. There are n number of Code Formatting tools available. If you have an automated solution which you know would do the job for you. Why bother? Though the intention should be to write code which is legible but making it a yes/no criteria, seems a little harsh. Coding Practicums, Yes. How do you write variable and function names, Design Patterns used. Are you using language specifics to get the best out of language is what I would look at.

Thanks manh.... Even I had the same theory. Plus mongodb is schemaless, so I profit from the fact that, if I change the schema, if shouldn't break the system. Profiling the metrics could be a key here.

Or probably hire someone/consultant with more experience of this kind. That is true, lack of experience is key here.I might have to go for a complete re write, to make it more scalable. Hiring someone that good wasn't in our favor when we started out. Maybe now we are expanding, we will give a complete thought to it, with someone senior.

Well that is a method I told as well, but if number of feature increase, it becomes cumbersome. You may be storing lot of variables unnecessarily .Don't you think?

Well that's the brute force solution!!! Still it does the job. Thanks!! Will try and implement this. I was thinking more like a state based solution. Like your program will have states, based on certain conditions and when those are fulfilled it enters a participating state and does the required action, in case it fails anywhere. All actions done up and until that point are reverted and system restored to previous condition, waiting for next input to come.

I already use it, but has some issues on what it marks as coding time. Need that to be more transparent.

Thanks @spsiddarthan I am already looking into RabbitMQ. It might take some time to evaluate and let you know the result but I would love to know other solutions as well.

Okay. So that is one good way of doing it. I am seriously looking at making something like this into an NPM/Yarn Package. Not really good open source auth packages out there.