The crypto academy season 3 is underway and we have several new changes in this season that reward the best and consistent users. To help make this more scalable, I had created the TagSearcher. Looking back at its initial version, I have added so many new features to make it more usable for the professors. However, there is always more room for improvement, right?

The Problem:

One of the highlights of this edition of cryptoacademy is that the best and the most consistent users get an extra bumper vote from @steemcurator01. To do this, the professors need to get the scores of every individual from every category for the whole period of season 3.

This is hard if all professors are not already maintaining a record of who they vote. It's hard because scrolling through 1500+ words of post to find the comment is just a lot of work when there are 100+ assignment submissions every single week per professor.

The Solution:

^{The new field in TagSearcher}

TagSearcher now has an additional field that displays the users score if they have been scored by the professor. Parsing the large comment for so many posts hinders performance. I have done as many optimizations as I can to have a smooth user experience. Hopefully, this helps mitigate the above mentioned problem.


^{Format criteria for scoring}

As discussed with @sapwood, I request all the professors to use the format as displayed in the above image when they post the final score. If a different format is used, then the phrase "incorrect format used" gets displayed in the UI. Hope this helps all the professors keep track of the scores on all the academy posts.

This is currently live on https://scribe.steemcryptic.me/tagsearcher

Improvements:

Currently, TagSearcher is entirely dependant on several API calls to the RPC node to get all the information. This is a network call and is the most expensive operation in teh entire process. To help mitigate this, I am building a generic data collector that suffices the needs for the academy posts. Apart from this, @steemchiller is also working on making my job a whole lot simpler with his SDS API. Both these require extensive testing. So, it might be a while before we get to see a very well-optimized version of TagSearcher.

Earlier today, I announced my entry to the steem witness family. https://steemit.com/hive-196725/@starlord28/in-retrospect-it-was-inevitable-or-starlord28-goes-witness
I would be humbled if you could vote for me as one of your 30 witnesses. This would not only motivate me to provide more services on teh technical front but in-turn, you would also be helping to secure the Steem blockchain further.
To vote me as a witness, you can simply head over to steemworld and vote for me in the witness section. You could also follow this link: https://steemlogin.com/sign/account-witness-vote?witness=starlord28&approve=1 and vote using your active key.

CC:
@steemitblog
@steemcurator01
@steemcurator02
@sapwood