I didn’t even know you could sort by reactions 
2 Likes
Coming back to the title of this thread
" What’s a good bounty to implement Rengo?"
I wonder how you intend to measure “good”?
I think that offering $1000 certainly indicates a personal level of commitment to wanting the feature, that’s for sure.
OTOH, $1000 is possibly a mere scratch on what the actual cost of implementing it would be, at market rates. $1000 pays for barely a few of days of work from a developer.
I think if you had 10 such committed people (IE $10k) then you’d have enough to surely be starting to influence the decision… $1000 is really just a token of want, rather than a substantial payment, as far as software feature development goes…
2 Likes
What are the chances that software development is somewhat overpriced or inflated these days
2 Likes
I mean, what fields can you think of where people would be happy to work 3 months for $1000? Think GaJ makes a reasonable point. It is a great personal commitment but possibly not alone enough to really incentivise anoek (of course, all speculation until anoek wakes up haha)
2 Likes
What fields can you think of where you get paid $1000 for three days work! Imagine all the fields where you go to college and train for most of a decade like law or medicine, I don’t think you’d make near that much.
If the coder got a handful of sections of code done in three days it’d hardly be worth the money. If a coder got a huge amount done in three days, they’re probably copying large chunks from some existing work they’ve already developed and are editing it. In the second case you’re not really paying for new custom work as much as for the experience of the developer (possibly) assuming again it’s not just coming from something else.
It sounds quite overpriced to me. When I saw that it was costing more than $40k (nearly $50k) to design a website Polgote.com - a new platform to connect Go teachers and Go students I thought that was pretty crazy to be honest.
I guess I just don’t understand the scale of money, how much things are actually worth.
2 Likes
That works out pretty close to 80k annual salary before tax… Other than entry level unskilled labour, that sounds not that outlandishly high for most industries…? 
2 Likes
I think I’ll just opt out of this conversation and see what comes of it.
1 Like
$1000 for 3 days is less than $50 per hour.
That’s a lot for a waiter, but it is barely the average wage for a programmer - look it up.
Here’s a question for “you all” … how many hours work do you think it would be to “implement rengo”?
How much would you be paid for that many hours?
2 Likes
You’re kidding me, right?
My plumber charges these sort of rates, let alone my dentist … we were charged $700 the other day for a few hours visit by the plumber.
My friend reported $800 bill for an hour at the cardiologist…
3 Likes
I won’t start about salary per se for a given task. I simply consider this as a bonus incentive to implement rengo before other things.
7 Likes
Yeah Groin has a point. anoek is still receiving his regular salary, so the $1000 doesn’t need to fully pay for rengo, it just needs to justify bumping it up the priority list.
4 Likes
Just play rengo on KGS, it works well there!
1 Like
I think we’re not comparing the same things I suppose. If everything is ludicrously more expensive in another country, I suppose I can’t call the wages overly high, if that’s what they’re supposed to be compensating for. If $700 is a standard callout fee for a plumber without replacing a toilet or something requiring a large installation, that doesn’t seem right, but of course I don’t know. Same for the cardiologist.
2 Likes
But it’s not just $1000, since the regular income through site supporters doesn’t change: implementing Rengo is something on the to do list anyways, basically the 1000$ would be a bonus for changing the priority list of what to work on, right?
7 Likes
It’s a very generous proposal @Gozart, however it’s not a money thing.
Rengo hasn’t happened yet because it involves touching a lot of code and schemas that assumes exactly two players, as well as creating an entirely new flow for manual match making and automatching, and dealing with things like players dropping out / timing out nicely without trashing the game for everyone else. Additionally the current rating system won’t handle multiple players, so that implies implementing at least a secondary rating system that can handle multiple players, presumably using something akin to TrueSkill/OpenSkill or something of the sort.
That being said I have been slowly preparing to undertake it, but it has to happen when I have a fairly large block of time.
27 Likes
Since I brought up UserVoice earlier, I just wanted to note the Rengo was actually the most popularly voted item on that platform. I found this forums post noting that.
5 Likes
My opinion on the above:
- Automatching is useless. Too few people play rengo, setting up a match can only work if people arrange it in advance.
- dropping out/timing out: then the pair loses the match by time.
- rating: rengo matches should be unrated. Anyway too few rengos are played, ratings will never be accurate. To setup automatic handicap, consider the rating of the pair as the average of the ratings of its two members.
22 Likes
I agree, and in any case not all features need to be in place day 1. Having the ability to play Rengo would already be a great step forward, even without any matchmaking or ranking.
5 Likes
Completely agree. That’s a lot of useless features… and work. A solid manual matchmaker, and a good running unranked game, that’s all what is needed.
Rating will have to be for a whole pair so it will push to keep same team through games. I don’t feel very rational to rate individually results from a pair. All in all It’s so a strange idea that it never occur to me before. Just drop that.
To give opportunity to replace a player who wants to leave is another idea i didn’t think about it. At first stance it looks friendly but i am not sure It’s a good thing as it can be an incentive to not finish a game. That sort of thing almost never happen in face to face games.
Instead of an automatic pairing, what about a manual assisted pairing ?
From a list of candidates pairing can be suggested which favorize different settings (widest difference of level in a team, as less handicap as possible between teams etc… ) with a calculated handicap proposed by default (but can be modified too).
Players can agree or drop out at will until thr game start.
Something like that. Would be perfect if afterall noone as to be the manager of the pairing but all coming from agreements of each. (An option can be to have a manager who will fix all, assisted in background)
4 Likes