SHAPEWAYS: The Third Mile
Steve Cole writes:
As we moved forward from 11 July toward 1 August, we began to explore new ideas and to learn some new lessons. Some of those lessons involved flying at full speed into the side of a granite mountain, but that's what makes life interesting.
An ongoing lesson is that small items take more work
than large ones because they keep failing to pass the Shapeways tests
for successful printing due to small features. Even after passing the
automatic checks, they fail the first time someone tries to order a
set of fighters or gunboats. This led to some changes to how we do
things. For one, more time had to be allowed for small units. They are
easy to build, taking work equal to the bridge of a typical cruiser.
The problem is getting them to pass the tests, as the sculptor has to
make change after change, trying this and experimenting with that,
uploading them again and again to run the tests. This rescheduled many
releases (including some carriers that cannot be released without the
fighters). We thought that the Fed F18, Klingon ZY, and a Kzinti
fighter would take about the time of one ship (total), but they
actually take the same time as six ships (total). They had to be
rescheduled to a point after the Klingon B10 and are now scheduled for
1 October. That included some fighters which were in fact already done
and which already passed, but we needed the three key fighters to be
done first in case others had to be changed.
Another conversation resolved
around "Aztek" panels, the raised panels on a starship designed
to add the appearance of depth and detail. Lots of lines of starships
do these, and some of ours had them. After much discussion, it was
decided that Jean and I (SVC) will decide what gets Aztekked and what
does not. In some cases we may well do both. The 3125-scale ships for
the Lyrans will come in Lyran (with Aztek) and LDR (no Aztek)
versions. The Klingons will get a D7B and C8B (no Aztek) and a D7K and
C8K (with Aztek).
Someone asked if we could do ships in a
different scale, one he regularly buys. We can (legally) do anything.
It depends on time available and potential sales. Probably for the
rest of 2017 we need to focus on the three scales (3788, 3125, Omni)
we are doing now. The problem is that a single model can be done in
3788 and easily converted to 3125, but that's about the limit either
direction. If we wanted to do a version of a ship in, say, 1/7000, it
would mean a sculptor starting over from nothing, then probably having
a lot of trouble getting the details to print at the smaller size. If
we wanted to do, say, 1/2000, we could scale up the 3125 but even then
the sculptor would have to add details and make other changes, not
quite a new ship, but still a lot of work. To launch a new scale would
require making at least four ships and probably more, and that's a
lot for sculptors to invest if it turns out we only sell a few copies.
So such experimental new scales will happen, but with limited numbers
in carefully selected scale-markets, and sometime next year.
During the process, we received the first and probably
only shipment of the last two metal ships (in both scales), the
Klingon and Romulan heavy war destroyers. The start-up cost of these
ships had increased by 300% over the previous five years while sales
went down because we had already done all of the good sellers. The
result is that these ships will never turn a profit and should have
gone straight to video, I mean Shapeways, except that when they
started we hadn't thought of going to Shapeways. Which is a key
point. It can easily take five months to get a ship into the warehouse
in metal; it only takes a month (often a week) to get a ship into
Shapeways
One thing that happened
was that the BBS and my email program started crashing because there
were too many graphics being juggled. We had to take a day to clean
out the ones that were outdated.
Will McCammon had
done the first scouts (F5S, D6S, G1S) and determined that Shapeways
cannot handle the traditional sensor dishes that Star Fleet Universe ships and art
have. They just cannot build the thin edges of a hollow dish. So
sensor dishes are now a bowl with a flat surface and substantial peg
in the middle, said bowl being merged well into the ship's hull to
reduce the changes of failing the auto-checks. This is important
because the players want (and the sculptors want to get paid for)
scout variants of the warships, and having a standard sensor dish that
actually works will help things.
Before we reached the second release group on 11 July we had
decided that the sculptors would need to get their ships to Jean on 25
July rather than 31 July. That would give her a comfortable six days
to upload about 30 ships.
Well, we thought we had
a good plan for the production of a reasonable number of ships (30),
but enthusiasm got the better of the sculptors and ourselves. Will
McCammon was scheduled to do the "missing" Klingon C8
dreadnought for 1/3125, but we soon decided to also sell the same ship
in 1/3788 since he already had it done. (All ships are done in 1/3788
first and then scaled up for 1/3125.) Later when the Aztek
conversation happened, it was decided to have him do the C8K in both
scales. He also did the Federation gunboat as scheduled. Matthew
Lawson was scheduled to do four Omega ships and in fact sent two extra
Trobrins. Steve Zamboni was scheduled to do eight ships but in fact
sent 22, although we only uploaded 18 of them due to time limits. Gary
Pollock was scheduled to convert two of his Hydran fusion designs to
the hellbore variant, which would have been four ships counting both
scales. He actually sent six. Chris Nasipack was scheduled to do three
Lyrans in two scales (six total items) but ended up doing five Lyrans
in two scales plus an Aztek alternate for a total of 15. While more
ships is good (it pushed the store past 100 items for sale, a place we
did not expect to reach until October), uploading 55 ships instead of
30 meant some long days for Jean. These got longer as the inevitable
happened and some ships failed and had to be uploaded a second time.
Looking back, 55 ships was probably more than you needed to pick
from, but what's done is done.
It may be 10% extra work for a
sculptor to convert a 1/3788 ship into 1/3125 or to convert a
destroyer into a scout; it is 100% more work for Jean to upload the
extra ship. The company needs Jean (and for that matter, me) working
on many things, not just Shapeways, and the two of us were effectively
AWOL for a month as we became the Prisoners of Zenda (I mean, of
Shapeways).
The plan for the
fourth batch is for each sculptor to do an easy build (a variant or
something mostly done), then a priority ship the line needs (starting
with Will McCammon's B10), then to send in whatever he can. We will
be sure to get the first two into the store (which will probably be
more than 20 ships) but the "extra" ships sent after that
will be checked by Will when he isn't doing his own ships. Then they
will be reviewed and posted by me (Steve Cole) when I am not working
on my own projects (counter sheet reprints, Captain's Log #53, Scenario Log #2,
A Call to Arms: Star Fleet Book #2, Federation Admiral, and others). And finally they will
be added to the store for the first of September when Jean isn't
pushing to finish the GURPS Prime Directive revision and editing Traveller Prime Directive. That
should mean a less stressful August for everyone on the team, and will
still get you plenty of ships to pick from.
The road got bumpy at the end. The Hydran gunboats were delayed when the sculptor had to fix some issues with
previous files and ran out of time. The Kzinti dreadnought almost
slipped to the next month when we had to get it re-scaled as the
original 2500 resin and metal ship was too big. Steve Zamboni
completed it just in the nick of time.
But when it
was all over, the store had passed 100 ships in a mere 50 days. Jean
and I were already at work creating a new pipeline system designed to
get the priority ships done first and to better manage the time we
invest in the project.
posted by Federation Commander at 2:06 AM
<< Home