[Hpr] Show flow rate

Ken Fallon ken.fallon at gmail.com
Mon Apr 22 12:56:42 PDT 2013


On Mon, 22 Apr 2013 14:46:45 -0400
epicanis+hpr at dogphilosophy.net wrote:

> The direction this is going worries me a bit. The focus lately has
> been entirely on "shows on hand" with less attention to "show
> flow" (except as something to make "shows on hand" as big as
> possible).
Your show has been uploaded on 2013-03-26 and should be released this
week. That's a average wait time of a month, not ideal but in that time
we have had a lot of new and returning hosts plus a 4 slot backlog from
the previous queuing system. 

We have been running the first in first out system for 2 years and we
are barely into four months of the "longest release date" method.
Which still had carry over from the previous method. So let's give it a
chance and see how it works when we are in "famine" mode.
> 
> > Well I do [care when shows are released], because this network will
> > cease to exist if shows do not continue to come in. Right now, New
> > or Older hosts looking at the Calendar might not contribute because
> > it looks like we are swimming in shows. We are not. We still need
> > 166 shows to fill to the end of the year.
> 
> This implies that in mid-November or so nobody should bother doing
> any more shows, because we'll have all the shows we need to fill to
> the end of the year...obviously that's not the intention. I think the
> "number of shows to fill the whole year" count is a distraction at
> best (the year end is something of an arbitrary distinction).
But will we have enough shows by mid-November ? The spring is by far
the most prolific and we're living on that for the rest of the year. 

During the summer there are no shows submitted. None. Weeks go by
without shows going into the queue except from the fallback hosts who
respond to the call. I would need to spend a lot of time proving that,
but giving the file naming having now upload date, it should be easier
to track and graph wait times.

> 
> I also think it would be potentially fatal to HPR if 166 people
> showed up tomorrow to fill all of those slots. If they did, anyone
> (other than first-time contributors) who submitted a show would be
> waiting until next year for the show to be useful.
I would hope that the regular contributors would be happy that there
are new hosts on board. However once they post their show, it's back to
the end of the line for them and they get the same crack of the whip as
anyone else.

> That would kill the reinforcement that can make contributing to HPR a
> habit for at least some of us.  (It's like sending a check to a
> charity who "really needs your donation!" but then the check doesn't
> get cashed for half a year.) I suspect contributions would crash as a
> result. 
If your show is in for six months, then yes fair point but a months
delay is not a lot. I could be wrong here, perhaps people expect their
shows to be out sooner. If that's the case then someone needs to crack
the nut of supply versus demand.
> 
> This might be a factor in the boom-and-bust cycles HPR seems to go
> through sometimes.
Might be, but if you continue to upload the check will be cashed
sooner or later. 
> 
> > The logical thing to do would be to remove the files from the
> > Calendar page. I have just now done this for all the shows. The
> > page now focuses on hosts only.
> > 
> > Waitttttttt.
> > 
> > The only reason that the files were ever shown is to allow the
> > hosts to know that they have been processed. To that end I put the
> > file list on the ftp server in a file called queue.txt. That way
> > the hosts get to know where the shows are.
> > 
> > Is that sufficient or should we continue to shoot ourselves in the
> > foot by listing all the shows ?
> 
> I like the change overall (the files list made the page look a lot
> fuller than it really was, though I DO miss seeing what topics were
> pending), but I would take issue with the "shoot ourselves in the
> foot" comment. It just doesn't feel right to me to hide HPR's backlog
> as a psychological trick to get new first-time contributors (the
> latency due to the backlog doesn't apply to new contributors, so the
> only element this alleviates is the lack of urgency - recurring
> contributors are affected by both).
> 
> The host list alone indicates that there are at least three weeks
> worth of shows in the queue so it's not really hiding the backlog,
> just understating it (there's 2 full months worth of shows ready
> right now - Ahuka, Ken Fallon, and Klaatu have put in some serious
> labor!) . Anyone currently in the queue who wants to contribute more
> has to get over the fact that whatever work they put into it won't be
> useful to HPR for about a month or more.

No anyone who has *multiple* shows in the queue. And they already know
that. Ahuka has said that he expects to take off during the summer and
has released the shows for that purpose. Klaatu has released shows in
batches, that's what he does. Those shows of mine are a year and a half
old and if the go out next week or in three months isn't going to make
much of a difference. The release order is actually right there on the
page "Last release time of show hosts:". Your up next, then Mike then
sigflup. Once you release your show, the clock starts ticking for you
again. If you submit a show the day after or a month after, your place
will be kept in the queue. 

This benefits new hosts but it also benefits old hosts that submitted
and didn't bother again. And it benefits HPR as the mix of hosts is
greatly increased. The point is it's not just you that is waiting to
get posted, it's everyone. All of us that are waiting behind you
right now in the queue. But so long as the rules are followed I am
happy waiting. On a very important side note, and if anyone at anytime
feels the rules are not been followed, well shout, right here on this
list. 

> 
> I would suggest that HPR's need is NOT "166 more shows this year" but
> rather "4-5 shows per week for the next 34 weeks and beyond", and I
> think the distinction is important. Having a big backlog that can
> supply 4-5 shows per week for a while is one way of doing this, but
> at a certain point it just turns into podstipation inhibiting the
> flow of new submissions.
I agree. I can change that but how many weeks do you pick ? I would say
a year because our most bountiful time is the spring.

> I'm not entirely sure what the actual solution would be to getting a
> more sustainable balance between the submission rate and the
> publication rate. Although whenever this topic comes up several of us
> suggest shoving out the shows faster in one way or another, Ken's got
> me questioning whether that's really the right approach.
Everyone assumes that releasing more would be a good idea. However if
we started releasing on Saturday and Sunday it _would_ lead to people
unsubscribing. We release a massive amount of content as it is and it's
the thing I hear most from people who have subscribed is that they cant
keep up and feel guilty having missed shows. 

> 
> What would motivate you all to contribute more regularly?
Contribute your shows and let the queue do what the queue does. I
personally go through spurts where I'll record several shows and upload
them then not bother for a while again. This has the happy side effect
that I have forgotten my show and can hear it again as a hpr listener,
spotting the mistakes, and assumptions that I made in the fist show.

Ken.





More information about the Hpr mailing list