feat: multiply conv reward with priority for higher reward

[ishowvel]

Resolves #175

[ishowvel]

/help

[ubiquity-os-beta[bot]]

Available Commands

Command	Description	Example
/help	List all available commands.	/help
/allow	Allows the user to modify the given label.	/allow @user1 label
/query	Returns the user's wallet, access, and multiplier information.	/query @UbiquityOS
/start	Assign yourself to the issue.	/start
/stop	Unassign yourself from the issue.	/stop
/wallet	Register your wallet address for payments.	/wallet ubq.eth

[ishowvel]

heres the qa

[ishowvel]

@gentlementlegen please review this :pray:

[ubiquity-os-beta[bot]]

@ishowvel, this task has been idle for a while. Please provide an update.

[gentlementlegen]

@ishowvel You should add Jest tests to your pull-request.

Here is my latest QA:

The decimals should not look like this. Also, I am confused about the total here. Can you walk me through the calculations?

[ishowvel]

After the total formatted reward is calculated, the priority level is multiplied to the reward for that specific comment, this all happens in the formatting evaluation. Then after the reward is passed to content evaluation to calc and multiply it with the relevance for that specific comment.

The only change here is adding a multiplication of the priority level to the formatting reward calculation

[ishowvel]

the total itself is not multiplied but the specification and comment incentives are

[ishowvel]

for jest testing, no mock result has ever contained the "priority" key. Can you help me find a way to make the mock results for this? all results starting at the formatting module would need to be changed

[0x4007]

Just copy the priority level names? Our config is public you can see all of them .ubiquity-os repo

[ishowvel]

Just copy the priority level names? Our config is public you can see all of them .ubiquity-os repo

There are 100's and 100's of results

[ishowvel]

https://github.com/ubiquity-os-marketplace/text-conversation-rewards/blob/development/tests%2F__mocks__%2Fresults%2Fformatting-evaluator-results.json

[0x4007]

There's five priority levels

[ishowvel]

I am talking about the results, each result object would need to be edited to have an extra priority key and the total and reward values would need to be edited

[ishowvel]

 [ 818.482 WXDAI ] 

Contributions Overview

View	Contribution	Count	Reward
Issue	Task	1	400
Issue	Specification	1	15.93
Issue	Comment	9	112.864
Review	Comment	25	289.688

Conversation Incentives

Comment	Formatting	Relevance	Reward
In the v1 of the Ubiquibot, when a result gets evaluated, a reca…	15.93 content: content: p: score: 1 elementCount: 6 a: score: 1 elementCount: 1 ul: score: 1 elementCount: 1 li: score: 1 elementCount: 3 result: 11 regex: wordCount: 98 wordValue: 0.1 result: 4.93	1	15.93
This needs https://github.com/ubiquity-os/conversation-rewards/p…	5.11 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 35 wordValue: 0.2 result: 4.11	0.8	4.288
To me 1 is the most straightforward to do for few reasons:- th…	19.95 content: content: p: score: 1 elementCount: 5 ul: score: 1 elementCount: 1 li: score: 1 elementCount: 3 result: 9 regex: wordCount: 111 wordValue: 0.2 result: 10.95	0.8	17.76
I think each plugin should output JSON not html as it is not rel…	17.29 content: content: p: score: 1 elementCount: 3 code: score: 1 elementCount: 3 result: 6 regex: wordCount: 115 wordValue: 0.2 result: 11.29	0.8	15.032
If you want to manipulate and convey data, HTML really is not ma…	17.85 content: content: p: score: 1 elementCount: 4 a: score: 1 elementCount: 1 result: 5 regex: wordCount: 134 wordValue: 0.2 result: 12.85	0.8	15.28
But then how do we consider the formatting of that output?Prac…	19.87 content: content: p: score: 1 elementCount: 5 result: 5 regex: wordCount: 159 wordValue: 0.2 result: 14.87	0.8	16.896
This can work, but we skyrocket coupling and to me defeat purpos…	7.03 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 55 wordValue: 0.2 result: 6.03	0.8	5.824
@pavlovcik To mitigate that that's why inside the comment reward…	10.73 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 85 wordValue: 0.2 result: 8.73	0.8	8.984
I realized that to carry this task properly we need to handle fl…	9.31 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 2 result: 3 regex: wordCount: 58 wordValue: 0.2 result: 6.31	0.8	8.048
Agreed, I think currently there are 3 possible things to annotat…	21.94 content: content: p: score: 1 elementCount: 5 ul: score: 1 elementCount: 1 li: score: 1 elementCount: 3 code: score: 1 elementCount: 7 result: 16 regex: wordCount: 54 wordValue: 0.2 result: 5.94	0.8	20.752
Resolves #5 Depends on #8 Reviews looks gigantic but it's on…	3 content: content: p: score: 1 elementCount: 3 result: 3 regex: wordCount: 19 wordValue: 0 result: 0	0.7	0
I think in this case it makes sense because this function will r…	12.49 content: content: p: score: 1 elementCount: 2 code: score: 1 elementCount: 4 result: 6 regex: wordCount: 60 wordValue: 0.2 result: 6.49	0.7	21.086
It is supposed to represent a comment. Would you prefer a descri…	4.11 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 16 wordValue: 0.2 result: 2.11	0.7	6.954
True, just picked this up from the previous codebase. https://rp…	3.44 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 19 wordValue: 0.2 result: 2.44	0.7	5.416
Seems convoluted to me and I don't know what problem that solved…	7.6 content: content: p: score: 1 elementCount: 2 code: score: 1 elementCount: 1 result: 3 regex: wordCount: 40 wordValue: 0.2 result: 4.6	0.7	12.44
Changed `COMMENT` to `COMMENTED`, added jsdoc de…	6.98 content: content: p: score: 1 elementCount: 2 code: score: 1 elementCount: 2 result: 4 regex: wordCount: 24 wordValue: 0.2 result: 2.98	0.7	12.172
Simply makes the docs linking the code within the comment, for c…	8.37 content: content: p: score: 1 elementCount: 3 result: 3 regex: wordCount: 48 wordValue: 0.2 result: 5.37	0.7	13.518
In the case of a string it would coerce the value and you would …	4.09 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 25 wordValue: 0.2 result: 3.09	0.7	6.326
`observe` would help to chose what we want to include an…	7.7 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 a: score: 1 elementCount: 1 result: 3 regex: wordCount: 41 wordValue: 0.2 result: 4.7	0.7	12.58
Not very useful indeed. I am working on the tests on a separate …	6.31 content: content: p: score: 1 elementCount: 2 result: 2 regex: wordCount: 37 wordValue: 0.2 result: 4.31	0.7	10.034
See https://github.com/ubiquity-os/conversation-rewards/pull/14 …	3.22 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 17 wordValue: 0.2 result: 2.22	0.7	5.108
Good point, also symbol should not be DAI but ETH isn't it? …	7.89 content: content: p: score: 1 elementCount: 2 code: score: 1 elementCount: 1 result: 3 regex: wordCount: 43 wordValue: 0.2 result: 4.89	0.7	12.846
Had changed it everywhere else but here, now it should be fine	2.65 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 12 wordValue: 0.2 result: 1.65	0.7	4.31
This matches any whitespace character (spaces, tabs, line breaks…	4.91 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 33 wordValue: 0.2 result: 3.91	0.7	7.474
By definition `dump` was just outputting results to stdo…	9.22 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 68 wordValue: 0.2 result: 7.22	0.7	14.108
Un-nested them one level. I think that the `buildXrow` s…	5.09 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 25 wordValue: 0.2 result: 3.09	0.7	8.326
I assumed BigInt only handles integers, and we manipulate a lot …	3.98 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 24 wordValue: 0.2 result: 2.98	0.7	6.172
There is one permit generated per participating user, last run:…	4.44 content: content: p: score: 1 elementCount: 2 result: 2 regex: wordCount: 19 wordValue: 0.2 result: 2.44	0.7	7.416
I made it according to the previous version:- Issue Specificat…	12.8 content: content: p: score: 1 elementCount: 5 ul: score: 1 elementCount: 1 li: score: 1 elementCount: 2 result: 8 regex: wordCount: 42 wordValue: 0.2 result: 4.8	0.7	22.72
Wouldn't that lead to precision loss in JS and after the divisio…	5.6 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 30 wordValue: 0.2 result: 3.6	0.7	9.04
That is correct. I understand now why you're confused:The valu…	21.95 content: content: p: score: 1 elementCount: 8 code: score: 1 elementCount: 7 result: 15 regex: wordCount: 65 wordValue: 0.2 result: 6.95	0.7	39.73
I don't think is it relevant in this scenario because we are not…	6.37 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 48 wordValue: 0.2 result: 5.37	0.7	9.518
@whilefoo you are correct, these were wrongly used. I corrected …	16.31 content: content: p: score: 1 elementCount: 5 code: score: 1 elementCount: 5 result: 10 regex: wordCount: 58 wordValue: 0.2 result: 6.31	0.7	28.834
Example of successful comment posting with results:https://git…	4.11 content: content: p: score: 1 elementCount: 2 result: 2 regex: wordCount: 16 wordValue: 0.2 result: 2.11	0.7	6.954
@0x4007 I believe that if I had all the tests in this PR it will…	4.29 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 27 wordValue: 0.2 result: 3.29	0.7	6.606

 [ 2073.928 WXDAI ] 

Contributions Overview

View	Contribution	Count	Reward
Issue	Task	1	400
Issue	Specification	1	63.72
Issue	Comment	9	451.456
Review	Comment	25	1158.752

Conversation Incentives

Comment	Formatting	Relevance	Priority	Reward
In the v1 of the Ubiquibot, when a result gets evaluated, a reca…	15.93 content: content: p: score: 1 elementCount: 6 a: score: 1 elementCount: 1 ul: score: 1 elementCount: 1 li: score: 1 elementCount: 3 result: 11 regex: wordCount: 98 wordValue: 0.1 result: 4.93	1	4	63.72
This needs https://github.com/ubiquity-os/conversation-rewards/p…	5.11 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 35 wordValue: 0.2 result: 4.11	0.8	4	17.152
To me 1 is the most straightforward to do for few reasons:- th…	19.95 content: content: p: score: 1 elementCount: 5 ul: score: 1 elementCount: 1 li: score: 1 elementCount: 3 result: 9 regex: wordCount: 111 wordValue: 0.2 result: 10.95	0.8	4	71.04
I think each plugin should output JSON not html as it is not rel…	17.29 content: content: p: score: 1 elementCount: 3 code: score: 1 elementCount: 3 result: 6 regex: wordCount: 115 wordValue: 0.2 result: 11.29	0.8	4	60.128
If you want to manipulate and convey data, HTML really is not ma…	17.85 content: content: p: score: 1 elementCount: 4 a: score: 1 elementCount: 1 result: 5 regex: wordCount: 134 wordValue: 0.2 result: 12.85	0.8	4	61.12
But then how do we consider the formatting of that output?Prac…	19.87 content: content: p: score: 1 elementCount: 5 result: 5 regex: wordCount: 159 wordValue: 0.2 result: 14.87	0.8	4	67.584
This can work, but we skyrocket coupling and to me defeat purpos…	7.03 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 55 wordValue: 0.2 result: 6.03	0.8	4	23.296
@pavlovcik To mitigate that that's why inside the comment reward…	10.73 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 85 wordValue: 0.2 result: 8.73	0.8	4	35.936
I realized that to carry this task properly we need to handle fl…	9.31 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 2 result: 3 regex: wordCount: 58 wordValue: 0.2 result: 6.31	0.8	4	32.192
Agreed, I think currently there are 3 possible things to annotat…	21.94 content: content: p: score: 1 elementCount: 5 ul: score: 1 elementCount: 1 li: score: 1 elementCount: 3 code: score: 1 elementCount: 7 result: 16 regex: wordCount: 54 wordValue: 0.2 result: 5.94	0.8	4	83.008
Resolves #5 Depends on #8 Reviews looks gigantic but it's on…	3 content: content: p: score: 1 elementCount: 3 result: 3 regex: wordCount: 19 wordValue: 0 result: 0	0.7	4	0
I think in this case it makes sense because this function will r…	12.49 content: content: p: score: 1 elementCount: 2 code: score: 1 elementCount: 4 result: 6 regex: wordCount: 60 wordValue: 0.2 result: 6.49	0.7	4	84.344
It is supposed to represent a comment. Would you prefer a descri…	4.11 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 16 wordValue: 0.2 result: 2.11	0.7	4	27.816
True, just picked this up from the previous codebase. https://rp…	3.44 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 19 wordValue: 0.2 result: 2.44	0.7	4	21.664
Seems convoluted to me and I don't know what problem that solved…	7.6 content: content: p: score: 1 elementCount: 2 code: score: 1 elementCount: 1 result: 3 regex: wordCount: 40 wordValue: 0.2 result: 4.6	0.7	4	49.76
Changed `COMMENT` to `COMMENTED`, added jsdoc de…	6.98 content: content: p: score: 1 elementCount: 2 code: score: 1 elementCount: 2 result: 4 regex: wordCount: 24 wordValue: 0.2 result: 2.98	0.7	4	48.688
Simply makes the docs linking the code within the comment, for c…	8.37 content: content: p: score: 1 elementCount: 3 result: 3 regex: wordCount: 48 wordValue: 0.2 result: 5.37	0.7	4	54.072
In the case of a string it would coerce the value and you would …	4.09 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 25 wordValue: 0.2 result: 3.09	0.7	4	25.304
`observe` would help to chose what we want to include an…	7.7 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 a: score: 1 elementCount: 1 result: 3 regex: wordCount: 41 wordValue: 0.2 result: 4.7	0.7	4	50.32
Not very useful indeed. I am working on the tests on a separate …	6.31 content: content: p: score: 1 elementCount: 2 result: 2 regex: wordCount: 37 wordValue: 0.2 result: 4.31	0.7	4	40.136
See https://github.com/ubiquity-os/conversation-rewards/pull/14 …	3.22 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 17 wordValue: 0.2 result: 2.22	0.7	4	20.432
Good point, also symbol should not be DAI but ETH isn't it? …	7.89 content: content: p: score: 1 elementCount: 2 code: score: 1 elementCount: 1 result: 3 regex: wordCount: 43 wordValue: 0.2 result: 4.89	0.7	4	51.384
Had changed it everywhere else but here, now it should be fine	2.65 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 12 wordValue: 0.2 result: 1.65	0.7	4	17.24
This matches any whitespace character (spaces, tabs, line breaks…	4.91 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 33 wordValue: 0.2 result: 3.91	0.7	4	29.896
By definition `dump` was just outputting results to stdo…	9.22 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 68 wordValue: 0.2 result: 7.22	0.7	4	56.432
Un-nested them one level. I think that the `buildXrow` s…	5.09 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 25 wordValue: 0.2 result: 3.09	0.7	4	33.304
I assumed BigInt only handles integers, and we manipulate a lot …	3.98 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 24 wordValue: 0.2 result: 2.98	0.7	4	24.688
There is one permit generated per participating user, last run:…	4.44 content: content: p: score: 1 elementCount: 2 result: 2 regex: wordCount: 19 wordValue: 0.2 result: 2.44	0.7	4	29.664
I made it according to the previous version:- Issue Specificat…	12.8 content: content: p: score: 1 elementCount: 5 ul: score: 1 elementCount: 1 li: score: 1 elementCount: 2 result: 8 regex: wordCount: 42 wordValue: 0.2 result: 4.8	0.7	4	90.88
Wouldn't that lead to precision loss in JS and after the divisio…	5.6 content: content: p: score: 1 elementCount: 1 code: score: 1 elementCount: 1 result: 2 regex: wordCount: 30 wordValue: 0.2 result: 3.6	0.7	4	36.16
That is correct. I understand now why you're confused:The valu…	21.95 content: content: p: score: 1 elementCount: 8 code: score: 1 elementCount: 7 result: 15 regex: wordCount: 65 wordValue: 0.2 result: 6.95	0.7	4	158.92
I don't think is it relevant in this scenario because we are not…	6.37 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 48 wordValue: 0.2 result: 5.37	0.7	4	38.072
@whilefoo you are correct, these were wrongly used. I corrected …	16.31 content: content: p: score: 1 elementCount: 5 code: score: 1 elementCount: 5 result: 10 regex: wordCount: 58 wordValue: 0.2 result: 6.31	0.7	4	115.336
Example of successful comment posting with results:https://git…	4.11 content: content: p: score: 1 elementCount: 2 result: 2 regex: wordCount: 16 wordValue: 0.2 result: 2.11	0.7	4	27.816
@0x4007 I believe that if I had all the tests in this PR it will…	4.29 content: content: p: score: 1 elementCount: 1 result: 1 regex: wordCount: 27 wordValue: 0.2 result: 3.29	0.7	4	26.424

[ishowvel]

@gentlementlegen are these the desired outputs?

[ 818.482 WXDAI ]

[ 2073.928 WXDAI ]

[gentlementlegen]

@ishowvel try with priority 3 as it is the most likely to give very long decimals if not properly handled. My comment was about the 0.3000000004 number in my screenshot.

Coming back to my calculation question, here is the detail:

I can't seem to understand how we get 11.264. 2 + 1.88 * 0.8 = 3.504 which would be the result on priority 1, like seen in the screenshot here

So how does priority 3 make it 11.264? Is that the expected result @0x4007 ?

[0x4007]

I guess 3x multiplier?

[gentlementlegen]

@0x4007 Sure but 3.504 * 3 = 10.512 so I don't understand where this result comes from.

[ishowvel]

https://github.com/ubiquity-os-marketplace/text-conversation-rewards/blob/f1e0ab4a3a3537a3bbd1cbed097e2c6cb0bf477d/src/parser/formatting-evaluator-module.ts#L72 cant seem to understand why theres a check for comment.score.reward as this check is bound to return always false, the first time comment.score.reward is given a value is in this very module

[ishowvel]

heres the new qa @gentlementlegen, the problem was not adding the priority mul at the right time, i should have added at the absolute end

[gentlementlegen]

https://github.com/ubiquity-os-marketplace/text-conversation-rewards/blob/f1e0ab4a3a3537a3bbd1cbed097e2c6cb0bf477d/src/parser/formatting-evaluator-module.ts#L72

cant seem to understand why theres a check for comment.score.reward as this check is bound to return always false, the first time comment.score.reward is given a value is in this very module

Because every module can be disabled individually, the type can technically be undefined (not within this module I agree) so this has to be done for TypeScript not to complain. I'll do a new QA as well.

[gentlementlegen]

New QA is ok: https://github.com/Meniole/text-conversation-rewards/issues/27#issuecomment-2456148025

Before merging, I would need that you add a Jest test with this feature thank you. Also, the tests are already failing: https://github.com/ubiquity-os-marketplace/text-conversation-rewards/actions/runs/11668599466

[ishowvel]

Done!

[ishowvel]

All tests that failed, failed because of not supporting this feature. So I just added support in all of them

[ishowvel]

@gentlementlegen can you please review these changes? Thank you!

[ishowvel]

Any updates?

[ishowvel]

@0x4007 can you review this as you have created the specification, gentlemengen may be busy

[gentlementlegen]

@ishowvel Sorry I have been extremely busy yes, I'll try to have a look today. In the meantime please fix the conflicts.

[ubiquity-os-beta[bot]]

@ishowvel, this task has been idle for a while. Please provide an update.

[ishowvel]

Currently out for a university, will work on this task in a few hours

[ishowvel]

idk why the tests keep failing even without the changes present in this pull request, i made a branch and force reset it to this repo's dev branch, heres the test

[gentlementlegen]

@ishowvel tests don't pass because the import types don't match, I think @octokit/rest is imported when it should be @octokit/core within the tests.

[ishowvel]

@gentlementlegen @octokit/core is imported, this has become a major blocker for me as i cant debug my other pr without working tests

[ishowvel]

tests even fail on a fresh clone of this repository

[gentlementlegen]

@ishowvel merge the development branch into this one.

[ishowvel]

Tests are passing 🎉 🎉 🎉

[ubiquity-os-beta[bot]]

@ishowvel, this task has been idle for a while. Please provide an update.

[ishowvel]

@gentlementlegen can you please review this

[ishowvel]

Do reminders exist for reviewers like the ones for assignee

[gentlementlegen]

@ishowvel I tried to run the plugin with your pull-request and this is what I got https://github.com/Meniole/text-conversation-rewards/actions/runs/11868228402/job/33077194863#step:2:173

Does it work on your side?

[ishowvel]

From what I have tried using @octokit/core makes the build work but makes the tests fail and @octokit/rest makes the build fail and test work

[ishowvel]

My code changes do not produce this bug as you can see here https://github.com/ubiquity-os-marketplace/text-conversation-rewards/actions/runs/11769854878 The latest merge is also failing because of the same reason

[gentlementlegen]

@ishowvel Okay let me try to fix the test and build then.

[gentlementlegen]

Seems to be ok in my latest QA: https://github.com/Meniole/text-conversation-rewards/issues/28