Refactor Config: mdast element keys as HTML entities / ref

[0x4007]

• Use HTML tag names to unambiguously refer to the elements in a comment from the comment-incentives config.
• Reference: feat: default config #383 (comment)

[rndquu]

This issue adds unnecessary code complexity

So under the hood we should transform html's li to mdast's list and then assign node prices.

I don't really see the difference between a vs link or img vs image. To me the mdast syntax is more friendly to none tech users compared to the \w+ which is planned to be used for words count.

[0x4007]

1. I just reviewed the syntax on its readme. Theres still limitations with its implementation, in this context. Perhaps we will need more complex key names to encode all of this information. How do you propose to refer to element h2 in our config using mdast? Also notice that in communicating with you, I am able to unambiguously express which element I'm referring to.
2. mdast is intended to be used by remark which looks like it can solve our problems with plugins to work with the markdown source. It seems that we should include remark? I'm pretty certain that it can handle this conversion for us.

[rndquu]

How do you propose to refer to element h2 in our config using mdast?

We could use the same type as mdast already use:

{
  type: 'heading',
  depth: 2,
  price: 100
}

It seems that we should include remark?

remark is built on top of the micromark which we already use in the bot's code

So this task is about moving from mdast types to html types, right? In other words the incentive comments feature would work this way:

1. Get html nodes prices from the config
2. Convert markdown to html
3. Calculate payouts based on the parsed html tags

[0x4007]

It seems like a hack to try and serialize a JSON object and use it as a key for the config property inside of a .yml file. It's better to just use the html entity as the key name. e.g.

comment-incentives:
  code: 5
  img: 5
  h1: 1
  li: 0.5
  a: 0.5
  blockquote: 0

every-x-incentives:
  every-word: 0.1 # not sure what is the most expressive property name for the "every word" incentive
  # every-sentence: 0
  # every-paragraph: 0

• Get html nodes prices from the config
• Convert markdown to html
• Calculate payouts based on the parsed html tags

Seems correct.

---

I tried experimenting with the yml schema. It needs work but figured I would post here before sleeping. This gives us so many possibilities for customization. Note that these numbers are just for example purposes only:

incentives:
  comment:
    code: 5
    img: 5
    h1: 1
    li: 0.5
    a: 0.5
    blockquote: 0
    totals:
      word: 0.1 # not sure what is the most expressive property name for the "every word" incentive
      # sentence: 0
      # paragraph: 0
      # comment: 0
  reactions:
    received:
      thumbs-up: 1
      thumbs-down: -1
    sent:
      thumbs-up: 0.5
      thumbs-down: 0.5
      eyes: 0.5
  reviews:
    received:
      approved: 10
      rejected: -10
      left-comment: 0
    sent:
      approved: 10
      rejected: 10
      left-comment: 0

[rndquu]

The time label should be updated to at least <1 Day then

[whilefoo]

/start

[ubiquibot]

Deadline	Tue, 18 Jul 2023 11:23:26 GMT
Registered Wallet	0xA0B11F474d8ECE1205d38c66d5F2bE8917675d60
Payment Multiplier	1.00
Multiplier Reason
Total Bounty	450 USD

Tips:

• Use /wallet 0x4FDE...BA18 if you want to update your registered payment wallet address @user.
• Be sure to open a draft pull request as soon as possible to communicate updates on your progress.
• Be sure to provide timely updates to us when requested, or you will be automatically unassigned from the bounty.

[whilefoo]

Github API can return comments in the HTML form so I think it would be better to use that instead of using other tools for conversion?

[rndquu]

Github API can return comments in the HTML form so I think it would be better to use that instead of using other tools for conversion?

yes, it might be better

[whilefoo]

Is the scope of this issue also implementing incentive of comments in the PR related to the issue and reactions of the comments?

[rndquu]

Is the scope of this issue also implementing incentive of comments in the PR related to the issue and reactions of the comments?

I don't think so. The scope for this issue is refactoring mdast element names to html tags in the bot's config and making sure that nothing is broken. @pavlovcik pls confirm

[ubiquibot]

Do you have any updates @whilefoo? If you would like to release the bounty back to the DevPool, please comment /unassign
Last activity time: Thu Jul 20 2023 12:10:04 GMT+0000 (Coordinated Universal Time)

[whilefoo]

@ubiquibot waiting for clarification in #545

[ubiquibot]

Do you have any updates @whilefoo? If you would like to release the bounty back to the DevPool, please comment /unassign
Last activity time: Tue Jul 25 2023 08:45:17 GMT+0000 (Coordinated Universal Time)

[ubiquibot]

Do you have any updates @whilefoo? If you would like to release the bounty back to the DevPool, please comment /unassign
Last activity time: Sat Aug 05 2023 08:38:44 GMT+0000 (Coordinated Universal Time)

[ubiquibot]

@whilefoo - Releasing the bounty back to dev pool because the allocated duration already ended!
Last activity time: Tue Jul 25 2023 08:45:17 GMT+0000 (Coordinated Universal Time)

[ubiquibot]

[ CLAIM 450 WXDAI ]

0xA0B11F47...917675d60

If you enjoy the DevPool experience, please follow Ubiquity on GitHub and star this repo to show your support. It helps a lot!

[ubiquibot]

@whilefoo The time limit for this bounty is on Wed, 23 Aug 2023 15:00:32 UTC