Re: finally: new authors

On Monday, May 5, 2003, at 09:06 AM, Gabor Hojtsy wrote:

Sooo ;)

A person, thinking alone, in a room, and being brilliant with only *one* line or character is very hard to measure. We cannot measure it well by code, by lines, by characters, or even by their popularity among others.... no system has been proposed for such a case, be they doc, bugs, dev, notes, etc..... :-(

Well, nobody said that the human voting will be based on 'popularity among others'.

I was concerned that it would not be "well informed" voting, contrasted with the "buddy system" or, to use USA jargon, "the good old boys network". (Suffice to say that I am not happy with my current (US) electoral system for these reasons)

At least I never thought. IMHO everyone who votes has something in mind about the guys contributions (which is something you may call popularity), well, yes. Philip had some suggestion to make this more objective, and some derivative of this idea also appears in your list below: to ask a 'phpdoc resume' from candidates.

Ways of seeing and learning things beyond the day-to-day commit emails, yes, I like the idea. But: If, for example, there was an annual vote on an "employee of the year", most employees I know would suddenly become much more attentive to their work, and try to impress the voters. ;-)

OK. Can you afford to contribute time to code such a system (set of filters and stuff)?

To compare lines and changed characters? Yes, I have written much of this code already. To compare the *value* of the changes? No. :-)

Well, the latter would need artifical intelligence probably ;)

Or human intelligence.... to discern the value of a one-character commit.

Your new examples (though trimmed) are good, and have altered my perspective.

I am happy to hear that ;)

Reward below. :-)

How would you (and others) feel about the following:

Well, my first thought was that this is overly complicated, and would give a high burned on guys voting, which would drive them away from their real work.

1. Authors voting *must* read all commits.

This is something you cannot check I guess. While your starting point was that we need some system which is automated because people cannot be relied on. There are many commits at busy times a day... I read most of the commits anyway to some extent, but I don't think that this is doable.

So, the few "amazing one line commits" rely on human intelligence, and yet, humans rarely read them. It might be a problem, it might not be.

2. Of the 20(?) fixed authors, at least half (10) must be changed per year (to ensure new credit, stimulate keeping a position of credit, and avoid author overload). If there are ten listed, 5 must be changed.

How can you 'stimulate keeping a position of credit' while you enforce 10 members to go away every year, just because this is the rule?

If the want the credit, they *must* keep contributing, more than others. Trust me, the manual isn't exactly laden with people over-contributing.

If you have the top 20 listed the first year [let's say that this is measured by some meaningful way, not changed chars, so Hartmut won't be #1 forever ;)], do you think that will change that much after a year? Well, maybe we had a quite different group a year before, but key figures were much the same as today.

I guess I don't see it that way, because I've been inactive for so long. :-) My "top five" authors and editors has changed little, the top 20 changes almost monthly. Example: Sometime Jesus works on notes, sometimes he hits docs. Sometimes Damien changes hundreds of US pages, other times he........ has a life? :-)

3. Depending on the systems available, the voters *should* look at changed lines/chars, XML changes, section changes, etc. They should (by charter) look at as many pieces of information as they have available to them. It is up to the candidates to provide any enhanced information beyond the current systems.

...So what if we employ Philip's idea of a 'phpdoc resume' which would some up all the stuff done by someone? Then voters would have a short summarized list of what someone done (from his perspective), and why he thinks he deserves getting on the list.

Sounds good to me.

4. Voting should be sealed and private, to eliminate peer pressure, until votes are final.

I really don't understand this. Do you say voting should be private forever, so we should never know who voted for it, and who voted against is? Then a system restricting voting to those who are allowed (authentication, etc). is needed.

It was an idea to make sure voting wasn't based on pressure. In the US, who we vote for, as individuals, is sealed. What those people vote for is not sealed. I would say that this wasn't a big deal if there were 10 people.... but there seem to be much more.

5. Discussion of their expected votes, and actual votes should be public.

Well, here it would turn out eventually who voted for and against ;)

Yes, but *after* the votes are cast.

6. A majority of the voters should be allowed to disqualify any other person from voting.

Well, we are getting to closer with every letter to my original suggestion :)

You proposed a system (IIRC) where *any* voting member could cancel a vote, or candidate. To prevent turf wars, it made more sense to me that *half* (or more) of the whole group would have to agree.

8. Non-voters should not be put on the updated authors list, without explaining extreme difficulty (to prune inactive authors).

This is just a question of who is qualified for voting. Let's imagine we use the system you propose. What if someone is disqualified from voting? His contribution should not be disqualified :) Just to show the weakness of this point.

Non-voters was mean for pruning of inactives only. :-)

ROTFL.... what make/model? :-)

Nothing expensive. I need a stable and modern one, but I have no big budget for this. There are some 900-1000$+VAT modoles in Hungary, which seem to be ok. The very low cost models here are Gericom (from Austria) and Portocom (Hungarian brand, assembled with parts from Taiwan :).

Cheap stuff, my business is doing good. Is there a good money-transfer system to there?

Heh, there probably is ;) Should I treat this seriosly?? Unbelievable...

I've pulled in well over $100K on just PHP work. If I can't give back in doc work, or code work, because I'm too busy, the *least* I can do is make sure the people working on it have decent hardware. Hardware is cheap, human time is not. Ronald Chmara Ronin Professional Consulting LLC 520-326-6109 "Throughout human history, we have been dependent on machines to survive. Fate, it seems, is not without a sense of irony." --Morpheus, "The Matrix"