[HCoop-Discuss] On organizing people to get work done

[Adam Chlipala]

Michael Potter wrote:
> I personally didn't reply because when I've disagreed with the party 
> line in the past, I've been shouted down and silenced, and I know the 
> party line around here is your paid staff.

I don't remember any party line.  Anyone is welcome to disagree with 
anyone else.  I think you spent a lot of bandwidth arguing for a 
position that wasn't shared by many other members; everyone needs to 
know how to realize when enough is enough, and his message has already 
been conveyed as best as is possible.  But, please, don't hold back from 
sharing your opinion on any issues relevant to the running of HCoop.

> I don't like how having 
> employees will create a class of member who is bound to serve us (as 
> opposed to the respect we give volunteers who are doing us a favor).
>   

So you're saying you don't trust people to set monetary values on their 
own efforts?  If you _do_ trust in that, then there is no worry that 
anyone will be "serving us" without adequate compensation.  Or are you 
somehow opposed to the idea of work for pay in the abstract?

> Also, I think paid staff will involve a rate increase, and I personally 
> don't want to pay extra for that.
>   

There are two main categories of benefit we'd see from any kind of 
improvement and quantity and reliability of admin effort.

First, we have emergency response, which has the diabolical property 
that you don't realize how bad you're lacking for it until you have a 
big disaster.  I submit that you may be mentally overvaluing the quality 
of HCoop services that you're receiving now for the current price, 
because we haven't happened to have a big data loss or hardware failure 
disaster yet.  To me, $10/mo. vs. $5/mo. is worthwhile for the peace of 
mind it brings.

We also have situations like the present poor performance of many of the 
network services we offer, including e-mail and database access.  These 
are annoying, but things "mostly work" anyway.  I can understand that we 
have a wide spectrum of stances on this among our member base: some 
prefer minimizing dues and are OK with slight flakiness, some are 
begging for a chance to pay a little extra for significantly better 
service.  I'm in the camp of being willing to pay several times more for 
a variety of different sorts of improvements to service.

"As we speak," there is a thread going on hcoop-help, where a member 
wants to run a Django web site.  The easiest way to get that going is 
with FastCGI or a similar scheme, but we don't have support for that.  
We currently only have one official "admin," as the rest have resigned 
over the last few months.  New functionality gets implemented when 
someone who has been granted root access thinks it will be fun.  Many 
members, I'm sure, would love to be able to throw money at us and get 
things done, but we just don't have the social infrastructure to support 
that kind of thing.

Another great example of this kind of problem is the idea of a 
non-AFS'd, normal UNIX machine that we would provide for the use of 
members who don't want to learn to deal with Kerberos, et al., just to 
run simple services.  This would take a significant amount of effort, 
but it could still be planned and executed in a week by a half-time 
admin, I claim.  We've been batting the idea around for more than a year 
with zero progress.  There are many other extensions to our offerings 
that I know would be seriously valuable to current members and would be 
seriously enticing to prospective members.  It's a question of what 
price each member wants to place on that kind of benefit.