Nicholas Clark nick at
Tue May 8 21:19:42 BST 2007

On Sun, May 06, 2007 at 03:09:10PM -0700, Marvin Humphrey wrote:
> On May 6, 2007, at 1:30 PM, Nicholas Clark wrote:
> >This happens to be exactly the reason why I didn't enjoy doing the  
> >grant work
> >for TPF, and hence would now not accept money to work on Perl full  
> >time.
> I think we could keep you happy on that task.  It would just take the  
> right project manager.

I'm really not convinced. For the reasons below.

> It's unlikely in the extreme that such a person would be Dr. Nick's  
> technical superior, but that wouldn't matter so long as they were  
> $adequate.  Goals would be planned cooperatively.  Target dates for  

Goals are the key part here. Goals made for the sake of making goals are
worthless. And no project manager is going to be able to set technical
goals, as distinct from prioritising supplied goals.

> completion would be set and reviews would be conducted regularly to  
> evaluate execution.  The social price of failure would be ratcheted  
> up intentionally.

Why is that motivation? I see no carrots here, only sticks.

> Touching base with management daily would be mandatory, but it would  
> be best if you weren't flying solo 100% of the time and reporting to  
> only one person.  Ideally you'd have co-workers who would depend on  
> you to complete your tasks in a timely manner, but due to the nature  
> of your responsibilities in the Perl world, that would probably be  
> hard to swing.

The problem with Perl 5 now appears to be that it's not broken enough.
There's nothing major and bounded left to fix, hence no entity (individual,
firm, or other), has caused any input. [Hence no goals]

"major and bounded" means tasks that initially appear to need anything
between 2 weeks and 3 months to complete. There are many smaller bugs,
but it would be no fun *just* fixing the 1500 RT tickets currently open.

> In lieu of co-workers, we could have you prepare a short weekly 
> presentation for a small group.  That's a good exercise in its own  
> right for a variety off reasons, but the main idea would be to graft  
> human faces onto the vast amorphous blob of users who have benefitted  
> from your enormous volunteer contributions[1].

I'm failing to understand why repeatedly writing presentations is fun.
It appears to get in the way of solving problems or writing code.
Pride may go before a fall, but I'm now sufficiently comfortable about giving
presentations that I feel I can pretty much write the slides, edit them until
I am happy, then present it from them without any actual practice sessions.
Or alternatively, [with practise beforehand] give a talk that keep people's
attention for at least 10 minutes without any slides.

> Would such a regime have made your TPF time more enjoyable?

No. Because it needs co-workers. The lack of anyone to bounce ideas off is
stifling. Neither $job[-1] nor $job[-2] really ended up having any business
needs best served by tweaking the perl core, or, really, anything on CPAN.

Perl 5 is insufficiently buggy. Maybe we should give it an imaginary broken
arm to make us perform better in comparison with the competition :-)

Nicholas Clark

More information about the mailing list