ShaydDeGrai

This is a very good question and I don't think it's ever too early to address it because how you organize your reference sets can really impact how well the entire system functions in the end. Years ago, I worked on a finger print matching system that faced many of the same issues. Our goal/expectation wasn't to find an exact match, at that time machine based matches weren't admissible in court but were used to create a short list of possible suspects for a human expert to consider matching by hand. We wanted to be able to reduce the search space from tens of thousands to hundreds as quickly as possible, so that when we did our second, more computationally expensive pass we could produce a list of our top ten or so candidates before all the suspects die of old age. (I just love it when, on CSI and such, they have a cheap laptop that can search the entire DoD, INS and FBI databases for a partial print and not only come up with an exact match in under a second, they render a spiffy real-time display of all 20 million comparisons just for good measure...) Anyway, the way we chose to simplify our first pass candidate reduction was to use borrow a idea from Information Theory called the Hamming Distance. The concept is named for Richard Hemming and to grossly over simplify things what it represents is, given two strings of equal length, the number of characters in the string that you need to change to make the strings identical. So, for example, the Hamming Distance (HD) between "CAR" and "CAT" is 1 (change R to T and you're done) Likewise the difference between "CAT" and "BAT" is one but between "BAT" and "CAR" is 2. If every string in the database is considered a node in a graph, then nodes with a Hamming Distance of 1 are considered adjacent, 2 would be two nodes removed, 3 becomes 3 nodes removed, etc. So if I were to organize my three word vocabulary at this point it would look like: CAR --- CAT --- BAT. If I were to add a 4th word, "BAR" it would be adjacent to both CAR and BAT but two steps removed from CAT. As you can probably guess, building up this graph takes time and ultimately means comparing each new entry in the database to every other entry in the database. This is the trade off, a (potentially large) one time expense in exchange for significant speed ups at run time. What we did with the finger print database was to come up with a binary string of "features" where individual bits indicated "has this feature or not" and small groups of bits represented binary enumerated "buckets" of ranges for scalar values (ie. "00" means none; "01" means 1-3; "11" means 4 or more - note that the binary patterns are Hamming Codes themselves, there is no "10" because it would have the same HD as "01" when compared to the other bit patterns in the set.) This simplified encoding allowed us to give every reference a (in our case 128bit) signature. Comparing signatures was in assembly code and fast (XOR the two binary values, bit shift through the result counting all the 1's that fall off the end, the total number of 1's is the HD. Patterns that map to the same signature string have a HD of 0 and, assuming you came up with a good signature scheme should be very similar to one another, Those with a HD of 1 only differ by one attribute and might be worth checking if the distance 0 search comes up sparse or empty, etc. The signature strings became our hash keys, and a B-Tree structure was used to deal with hash collisions (because in our case, we WANTED hash collisions, if things mapped to the same point, it meant we were getting close). Each node in the B-Tree had a list of all of its adjacent (HD=1) nodes to assist in a secondary search if need be. Of course all this is predicated on coming up with a good "signature" string. The first temptation is to copy something like Bricklink's taxonomy; I think they have about 200 different brick families ("Brick,Modified","Technic, Axle", "Animal, Body Part", etc.) The problem with this approach is that it assumes you know something about the context of the part your trying to identify, which might not be true. Going back to fingerprints for a minute, we expressly DIDN'T use the fact that some prints on file came from known felons and other came from trusted people holding security clearances because if someone just hands you a print and says "match this" without telling you whether it came from a crime scene or the White House, initially limiting the search to one demographic or another might actually hamper the overall search effort. The trick to coming up with a viable signature is to focus on testable and countable attributes. Figure out, based on your image processing efforts, what you can learn about any given piece easily. For example, does it have studs? If so, how many and do they form a row, a rectangular grid or an irregular pattern? Does it have studs on more than one face? Are the studs hollow? It it wider than it is tall? Does it have flat sides? etc. One of the early explorations into Artificial Intelligence was a simple game game called Animal. The idea was that the system would build up a decision tree, trying to guess an animal based on a series of questions and asking for more differentiating questions each time it guessed wrong. Over time, it was theorized, it would become an "expert" system. The dialog went something like this: PLEASE THINK OF AN ANIMAL AND I WILL TRY TO GUESS IT. DOES IT SWIM? >no (human input) IS IT A BIRD? >no I'M STUMPED, WHAT WAS IT? >a cow PLEASE ENTER A QUESTION THAT WOULD DIFFERENTIATE A COW FROM A BIRD: >does it give milk? (...) The next time the system ran, and the user said the animal didn't swim, rather than guess bird, the system would ask if it gave milk. If so, it would guess cow, otherwise it would guess bird. This is the sort exercise I think you need to consider when coming up with the "signature" for classifying your reference cases; broad yes/no questions based on easy out-of-context observations and measures that rapidly reduce the search space. If the cross sectional area if greater than the height on a part with studs, it's probably a plate rather than a brick. If the volume of the convex hull significantly less than the that of the bounding box, chances are good that you're dealing with a slope or a bow. If the studs are hollow and the side walls have holes in them, it's probably in the technic family. If the piece barely has any recognizable aspects, its probably either a minifigure accessory or a Bionicle part. Figure out what questions to ask for any given part you hope to recognize and encode the answers to those questions into the "universal" signature string. Again the finished signature doesn't have to be _unique_ it just needs to map similar parts to the same base index and, more importantly, take clearly dissimilar parts out of the competitive range as quickly as possible. Again, I hope this helps and I apologize to any non-techno-geeks whose heads may have recently exploded from this discussion ;->

I can't say that I've ever regretted buying a particular set. I've certainly regretted NOT buying certain sets at certain times; usually when I returned to a given store intent on correcting my oversight only to discover that the store was sold out or the sale had ended and the price had gone up. Like the boy in the Emerald Knight video (elsewhere in this forum) I can relate to saving up for that really big kit only to discover that, by the time I could afford the set, it had been discontinued and all the stores were out of stock. I can understand looking back on many of the older set and thinking "what was so special about that?" Put the old Yellow Castle next to Helm's Deep or Hogwarts and some would say it's like comparing cave paintings to fine art. Still, when _I_ look back at my old Samsonite LEGO kits from 40 years ago, I don't see a dated toy or a novelty collectable, I see a reminder of how I grew up and the circumstances I had to deal with to get where I am. I grew up poor in an immigrant family where my mom stayed home with the kids and my dad worked construction by day and went to school part time at night. I subsisted on the most disgusting meals the public school breakfast/lunch program could provide, and even when writing my "wish list" to Santa I was reminded not to "sound greedy" (code-speak from my mother to keep the list short and try not to go over $10 total) LEGO kits, any LEGO kits, were a luxury and I knew it. If I got one as a gift, it meant the family would be eating pasta instead of meat. If I bought one for myself, it wasn't a casual treat, it was six months of delivering papers or shoveling snow or raking leaves for the neighbors. Every kit I got meant that I either worked for it, or somebody sacrificed something to be able to give it to me. So no, I don't regret having any them; they may not look like much, but they mean a lot to me. Today, LEGO has become my vice of choice. Maybe it's my take on a mid-life crisis ;-> I probably spend more in a month or two now than I did in my first 10 years combined and while all kits are definitely not created equal, I can't say that even in this period of increased acquisition I really regret any particular purchase, at worst, I just break the model down and use it for parts. It's shocking to realize how things add up from time to time (as my wife once pointed out - "Do you realize that for what you spent on LEGO last year, we could have spent a week at a spa in Hawaii?") but the same could be said for any given spending category (dining out/Scotch/wine/automobile expenses/etc), for what we spent on groceries and heat last year I could buy a new car - it's just a question of priorities. The only difference is that LEGO introduces the added complication of long term storage that you don't encounter with luxury vacations or fine dining so one's spending habits become more visible over time. As for seller's remorse, I must have the ultimate case of Sellers-remorse-a-phobia. I'm so convinced that if I sell any part of my collection I'll regret it, that I never sell anything. I'll occasionally give stuff away, but in those cases I take consolation in the notion that I have shared the joy of LEGO with others.

I didn't realize the invites were coupled with spending habits. Part of me now feels a little uncomfortable that I've been invited two years in a row. I knew I was "avidly embracing my hobby" but I didn't realize that, living in the 10th largest city in the US, I was consistently outspending 4.5 million of my neighbors. I am so glad my wife never reads this forum... I feel like I should go buy her some jewelry or something to balance things out.

I got my invite yesterday. I didn't go last year as I was traveling but I'm thinking of checking it out this year. I don't really know what to expect as I usually avoid retail outlets like the plague on (actual) Black Friday, I'm more a Cyber Monday sort of guy, but this might be fun.

I mean no disrespect to the people of China in general, but frankly, given the reputation China has for disregard of intellectual property, lax quality control in products for export, "questionable" (am I being too kind?) human rights stances, and environmental policies that make the biggest polluters in the U.S. look like tree-hugging, Luddite flower-children, I think I'd be quite happy paying a little more for a product NOT made in China. The rational side of me knows that TLG takes the quality of their product quite seriously and that with over a billion people in the world's third biggest economic power, it is unfair to stereotype the actions and output of the whole country based on a few anecdotal news stories here and there. Still, the irrational side of me flashes back to the realization that some LEGO parts come from China every time I see a news story about some heavy metal poisoning case because the paint on this toy had Lead or Cadmium in it, or the plastic for that doll tested positive for Arsenic. I don't even have kids but I'm surprised (even assuming LEGO parts are pure and clean) that such stories haven't turned in to public relations nightmares for them the way Mattel, Hasbro and Parker Brothers have been dogged off and on over the years due to Chinese manufacturing "issues". I don't know if the rumors are true or just wishful thinking on someone's part, but I as I said, I'd be willing to pay a little more to help TLG pay to upgrade their facilities elsewhere in the world for the luxury of not buying a Chinese export.

Very nice! Really a top notch job! I'd been thinking about tackling Meduseld and I don't know if I should feel inspired or intimidated by this (in any case I'm clearly going to need more bricks...). Congratulations on a great design.

On a slight tangent, I recall one project I worked on that was supposed to use neural nets to recognize and classify US, British and Soviet tanks based on satellite and/or spy plane footage. We built the network, came up with all sorts of exhaustive training sets (with neural nets you try to get the computer to figure things out by example "this is a T-62, this is a Chariot, this is an Abrams, etc. and hope the network figures out what differences one from another when shown photos from all different angles, perspectives and levels of obscurity). We felt pretty good about the system in the lab, it had a 99% accuracy for identifying "ours" versus "theirs" and about a 60% accuracy on the actual model of tank. Then we fielded it and it failed horribly. In analyzing what went wrong we realized that all of our training photos for "our" stuff were professional, staged photos given to us by proud defense contractors and most of our Soviet samples were from field footage, spy cameras or frame captures from newsreels. When we correlated image quality with output we found that we had invented a multi-million dollar photo critic. If it was a nice clear, patriotic looking photo the system declared it to be a US tank; if it was a good quality photo but the sky looked overcast, it assumed it was a British tank, and if the photo was grainy, blurry or washed out, it got classified as a Soviet tank. My cat, who refused to sit still for the photo, was classified as a T-62. While Neural Nets are an interesting thing to play with, I'd recommend sticking with traditional image processing techniques for this project ;->

There are actually two type of "rotation" to consider, one where the information content of the image hasn't changed and one where the underlying image conveys a fundamentally different story. Let's say I put a 2x2 plat on the table and photograph it from directly above, I can see the studs and the edge of the piece, for argument sake, lets say the edges are exactly parallel to the edges of the photo itself now I spin the piece on the table but still photograph it from directly above, I get an image where the edges aren't parallel to the edges of the photo any more, but if I printed out both photos on acetate and helf it up to the light I could rotate the two printout and get the image of the plate to align. The important part of the image (the information) is the same, it's just the orientation within the sampling area that was impacted by the rotation. This sort of rotation is easy to detect and compensate for when using FFTs. If you look at the Magnitude of the transform of a box (our piece in this case) and apply a threshhold filter, you'll get a blurry plus (+) sign image that eventually reduces down (depending on where you set the threshold) to a dotted line along the dominant axis of the box. So if your box is taller than it is wide and perfectly aligned with the edge of the frame you should see a vertical dotted line appear in the reduced plot. If the box is rotated, this dotted base line will also be rotated by the same amount with respect to the center of the image. From there you can calculate the angle, apply the opposite rotation matrix to the image and an inverse FFT will realign the original photo. Similarly, you can use a phase correlation filter to deal with issues of magnification or translation (say your photo is off center and magnified slightly). The strongest point of the phase correlation should get you the offset you need for a translation matrix operation and the signal to noise ratio in getting that strong point can clue you into difference of scale. The harder case is when the camera has moved such that the different rotation has a different perspective of the object (and potentially different information). When this happens you're often better off looking for features of an image rather than trying to match the whole thing at once. If I look down at the corner of a cube, for example, the two dimensional projection of its outline is a hexagon; If I look at one face dead on, I see a square. Without knowing that I stared with a cube, how do I know if I'm supposed to see a hexagon a square or something else entirely. Since we know we're looking for LEGO parts and they, as a general rule of thumb have studs and tubes (which are round, and round things are easy to spot in FFT projects even when skewed by perspective) I might try to write a "stud-finder" module and see how that does with arbitrary points of view in a few test cases. Once you know where the studs are you can figure out a) likely parts to compare to, and b) scale and perspective skew (because the studs are always circles of know size) to either massage the source image to make it more like your references, or generate new references that recreate the camera positioning used on the physical model.

I think one of the first questions you need to consider is whether you want to try to solve this problem in a spatial domain, a time domain or take a hybrid approach. A spatial approach is looking for artifacts in the image as YOU see the image; thing like matching an outline or a profile to count studs to get lengths (knowing the size of the stud in real life) the problem is that the more you move away from an orthographic (exact 90 degree shots from left, right, front, back, top and bottom) projection, things get messy. Add to that, lighting conditions, partially obscured parts (or self-obscuring, concave parts) and things can get ugly fast. A temporal domain starts with ugly math and non-obvious concepts but can render as moot a lot of issues like scale, lens distortion and perspective foreshortening. A hybrid approach, as the name implies, tries to find the best of both words Personally, I'd probably start with the temporal domain. Using LDraw wireframes, you can build up a database of ideal images of parts. I'd use either a Fourier Transform (floating point) or a Walsh transform (binary - less accurate but faster) on these images to get their temporal projections. (A book on Numerical recipes can be useful here but there are plenty of open source FFT code samples to draw from). When you take an image of a real part, do an edge detection, throw away the color information, and transform what's left (a good book on image processing can help here if you need algorithm ideas). Then perform a convolution on this reduced image with each of your reference images. Convolution in the temporal domain works a bit like waves in a ripple tank, the super-positioning of waves with different periods and frequencies tends to cancel things out and you're left with lots of noisy ripples, but if the waves match, the signal intensity doubles and you get one big wave. In essence this gives you a way to measure how similar two images are to one another and, by extension, which ideal image best matches the sample. Since, under this scheme you're never looking for an "exact" match, only a "best" one it's more forgiving of photographic artifacts. The down side, of course it that it's a lot of computation and a smarter algorithm would probably be to try and match features of the brick (studs, tubes, clips, etc) in isolation first, to narrow down the number of parts you need to compare it to (hint: round things are easier to track in a temporal domain, straight lines are more distinct in the spatial world) I hope I haven't made your eyes glaze over (I used to do this stuff for a living and studied image processing under the guy who cleaned up the Hubble telescope images back before the fixed the lens) I know I sometimes forget other people may have social lives and don't use terms like Convolution and Fourier Transform in casual conversation. Good luck with your project, it sounds like it could be fun.

I'm not saying that there shouldn't be low end entry sets or that every set needs to be "complete." My core points really boil down to three points of contention (four if you count battle pack marketing ideas but I feel we're very much in agreement on that one): 1) Not every set needs to be a UCS offering, there is a need for broad spectrum of sets in all price ranges, play and display expectations, and building ability level; the problem is that LOTR has a huge following among college students and adults (some of whom are even rumored to be female) and the Lego offerings to date _seem_ to be targeting pre-teen boys, a strong LEGO demographic for sure but in the greater Tolkien fan world, a pretty small minority (hopefully this will change with the release of the Hobbit films). TLG should _expand_ its offerings to have broader appeal beyond the pre-teen boy audience. 2) When I buy a mid- to high end kit, be it LOTR, the Hobbit, or whatever I want the price point to reflect the size and complexity of the build. If I spend $80 on a kit, I want, what in my mind - warped as it is - I have come to expect from an $80 kit, not eight $10 kits bundled into a common box. The Mines of Moria doesn't _feel_ like big kit despite being priced like one. It feels like a bunch of little builds - like building all the furniture and the fire truck that goes into the modular fire station, without actually getting the fire station. There are two ways to fix this: either break it up and market the pieces as small builds (Cave Troll with doorway and Orcs, Well with Hobbits and Skeleton, Sarcophagus with Gimli, skeleton and orcs, Wall segment with hero (Aragon/Boromir/Legolas) and more goblins and orcs, etc) for $10-20 each - collect them all and build the bigger scene yourself - buy all 6-8 kits at once on S@H and save 10%, etc., etc.; or build a more "complete" kit with a coherent, monolithic centerpiece to the build and set a different (probably higher) price point. 3) A number of the design choices they've made to date just seem hard to justify to me (okay, I've never been a fan of flick fire missiles - so shoot me with one - whatever...) For example, I MOD'ed the cart from Gandalf Arrives to make it big enough for Frodo to sit (technically stand, as his legs don't bend) beside Gandalf and so the rear tailgate of the cart could open. This involved replacing 6 parts and adding 4 additional parts. Even at S@H PAB prices (and giving myself a credit for the 6 parts I was no longer using) this was less than a dollar in parts to make it a much nicer scale and more playable set. True, a dollar in the context of a that kit's MSRP is a 10% price hike, but in the larger scope of things, the can of Coke on my desk right now cost $1.75 from the vending machine in the break room so it's not an unreasonable trade-off to address the issue. I mean no disrespect to the designers at TLG who are working hard to bring us these kits, but I look at that they've done for Star Wars, Harry Potter, Pirates of the Caribbean, and other licensed themes (where in some cases I like the kits better than the source material they're drawn from) and I just don't feel like they're bringing their best work to the table on these lines. There are a lot of (little) things that could have been done to improve the implementation of the ideas behind the current kits that, for whatever reason, didn't make it into the finished products. I totally agree that the greatest strength of the LEGO system is the fact that you don't have to build it exactly the way they tell you to and leave it at that, but that's really beyond the point at hand. For the purposes of this thread, the real question is, what is the out-of-the-box experience? Do the kits, as is, engage your enthusiasm? You say these are about what you expected, I find them a little disappointing - that's fine I like coffee ice cream but know many people prefer chocolate and I will even accept chocolate when coffee is not available. My MOD to Gandalf's cart was a trivial fix, I'm sure I'm not the only one who's made such changes, but my point is that _I_ shouldn't have had to make it and a six year old who wants two people to be able to ride in it or open the tailgate might not have the parts or skill needed to "fix" it once they realize the cart was too small. It was an obvious defect that would have been easy to address for relatively little additional cost, but they didn't - that's where I see a problem. Maybe it's not that I expect more from the LOTR or The Hobbit kits, it's that I expect more from TLG, I've been collecting this stuff for 40+ years and I know they could do (and have done) better. I'd buy that for a dollar - or ten, as the case may be.

In the case of the "Mines of Moria" set, I'd start by giving it a different name like "The Tomb of Balin" to downsize expectations as, after all, it's not the MInes, it's one room from the mines (this also opens up the possibility for a "modular" Mines of Moria sub-theme where various sets connect to form a giant play set). Next, instead of a couple disconnected wall segments with finished tops that make them seem more like exterior ramparts, I'd give the it the "hinged doll house treatment" that we see in sets like the Haunted Mansion, Gringotts, and the buildings of Medieval Market Villiage such that you're building "one big box" not several small builds that were packaged in a big box. A common baseplate wouldn't suffice, since it's supposed to be an underground chamber, you really need something that implies a ceiling (even if it's just a few large plates angled inwards). The could unfold to give a similar backdrop to what you have now, but when closed up, it would form a sealed structure that can hold minifigures and props within it. It doesn't need to be a 5000 piece set, but should be large enough to hold a cave troll when closed. The kit already sells for 80 USD, at that price point why not make it 99 USD, add another 200 pieces and give it a more "complete" kit that "feels" like an interior space? A bigger, more to scale, version would be great, but unrealistic with respect to TLG's price points. The cost of building to Minifigure scale in commercial sets is that their environment will always be unrealistically small or built to scale but trivial. I didn't say I liked Weathertop, I said I was "pleasantly surprised" by it. I entirely agree that, as a representation of the "real" (progenitor?) Weathertop is is barely passable, mostly due to the issues of scale, but it is an interesting and coherent build and comes with a couple Nazgul so I'll give credit where credit is due. I recognize that I'm not the target audience for the kits TLG has released to date for LOTR or is planning to release for The Hobbit and that demographic may not care so much about scale as price and may actually want flick fire missiles. The Bag End kit falls into this same vein for me. It's too small to be a "good" Bag End, but it looks like, as a building and play experience, it might be a "good enough" Bag End for its price point. Personally, I like the idea of mixing battle packs with scene builders, like they did with the Helm's Deep expansion set (extra wall segment, siege weapon, figures). I agree there's lots of opportunity for this in the source material. For example, Warg riders could be packaged with small brick built trees to build up forrest as well as the army that tears through it; soldiers and horses could come with some terrain elements on a plate like rock outcroppings or shrubs to build up the battle field as well as the corps. Even interiors lend themselves to this sort of marketing. Consider a Mines of Moria battle pack that comes with a handful of goblins and a brick built column with base that can be connected via technic pins to other such columns; in addition to building up a goblin army, you can slowly be creating that giant hall of columns where the Fellowship got swarmed just before the Balrog showed up. I have no idea about the wording of the license agreement, but if it's like the Star Wars one, TLG has to tread carefully about minifigure to parts ratios (with Star Wars, they can't sell "action figures" but can sell "construction toys" that include character representations. If this is the case with The Hobbit as well then, bundling figures with scene builders keeps things (arguably) in the realm of construction toys while offering a wallet friendly way to build large armies.

I've been a big fan of Tolkien's work since I first read the books some 40+ years ago and a collector of Middle Earth memorabilia for decades, so I expect I'll eventually be picking up at least one of each. That said, I can't really say that I'm "looking forward" to any of these sets in particular (in the way I couldn't wait to get my hands on kits like the Emerald Night, The Robie House, The Super Star Destroyer, The Taj Mahal, etc.). These kits sort of quietly saunter over to me and say, "You know you're going to buy me eventually, let's just get it over with." As opposed to some kits that wake you in the night and say "The LEGO Store is only ten miles away and it will be open in three hours!!! Don't forget your credit card!!! They may only have 10,000 of me in stock so you'd better hurry up!!!" I do like the look of these sets and think TLG did a better job with the initial release of Hobbit offerings than they did with their LOTR kits - then again maybe my expectations were different because I think of the Hobbit as more fitting fare for pre-teen boys so I'm more forgiving of some of the design choices. Still, I don't find any of them particularly compelling beyond the basic appeal of the theme. I appreciate (and agree with) the points made and seconded by several people regarding the lack of "completeness" for a number of the Middle Earth sets announced and shipped to date. I like the look of the Barrel Escape but at the same time it seems to have a lot in common with the Mines of Moria set (my personal choice for "most disappointing set" in the LotR line - could have been great, but wasn't) It _feels_ piecemeal, it's not a big build; it looks like a bunch of small builds packaged together in a common box (I think the Goblin King shares this weakness), a bit like a LEGO Advent Calendar on steroids. I understand marketing and appreciate why poly bags and, what I consider "stocking stuffer," sets exist - hey, I _WAS_ one of these kids who couldn't afford most of the kits I lusted after as a child and considered myself lucky when I got one of those low end, few-thrills sets. I know I've been critical of things like Gandalf Arrives (mostly because the cart is too small for Frodo and Gandalf to sit in side by side) but if someone had given me that kit 40 years ago I would have been overjoyed. I'm less sympathetic to supposedly "big build" high end kits that pander to short attention spans. I was pleasantly surprised by Weathertop, still a bit pricey and I have no idea why there are gratuitous flick fire missiles, but a good build for the part count. I expect Bag End will be similar (though I've seen better (more organic) looking MOCs of the same subject). Several of other kits, however, seem to up the piece count without increasing the size of complexity of the item(s) being built. Like the Mines of Moria, it's "build this stand-alone item, now build this bit of scenery, now build this other prop, now link them all together with a common table top..." They just _feel_ like backdrops for a play, not a coherent model like the Monster Fighters Haunted House, PotC's Queen Anne's Revenge or something from the Modular Buildings line. As for battle packs, I'll let that slide for this initial line-up, but as we near the point in the narrative for the Battle of Five Armies, I want to be able to stage the BATTLE of five ARMIES not the "Mild Disagreement of Five Guys, a Warg and a Horse". Perhaps, as part of the battle packs, they could each include a few gold pieces (chalices, swords, coins, etc.) so as you build up your armies, you also can build up the size of the treasure hoard they're fighting over.

Welcome, I too lurked for some time here before getting an account but I'm glad I did and I think you will be happy here too. Like you, I've been a fan of LEGO since my first kits (actually manufactured in the States by Samsonite) at a very early age, though, despite the intervening decades, I freely admit to playing with them ;-> I'm in no position to question a cat name Purr, I once had one named Mouse, and speaking of cats and being accident prone, I once tripped over my cat and dropped a 5000+ piece original LEGO sculpture that had taken me a month to design and build. My cat refuses to answer to any of the names I called him that night... As a (near) total aside, my favorite incarnations of Doctor Who are David Tennant, Jon Pertwee, and Tom Baker (in that order) and when Matt Smith's time is up, I can't say that I'll miss him. Welcome again and I look forward to seeing some of those animations or yours.

Hmmm, they've gobbled up Pixar, Marvel Studios, Jim Henson Productions, ABC, Touchstone, ESPN, Hollywood Records, Hyperion Books, Hulu, Lucas Films and now Hasbro? Who would have guessed that the world's biggest evil empire would have started with a mouse? I think SheepEater could be right, if Disney gets Hasbro, several of TLG licensed themes could be in danger of not getting renewed (SW, POTC, Cars, PoP, Marvel SuperHeroes, etc.) Disney has a pretty long reach.

I've long favored Redbeard, but Brickbeard's sneer won me over. There's something about his expression that just says "Arrgg" to me while everyone else is just standing around smiling.

When you say: Custom decals, stickers, cut outs and official TLG stickers are allowed Does that include custom (potentially printed) fabric for unique/not yet available capes, cloaks, sashes, etc.?

The past few days have been very taxing so I picked up the new SW B-Wing UCS set (10227) and the Haunted Mansion (10228) to take my mind off of things. I finished the house but may save the B Wing for next weekend.

Even Modular Western could be, indirectly, running up against licensing issues. I think the idea is great and I'd definitely buy them if produced, but there may be a snag with the upcoming Lone Ranger line. The same way that Castle was put on hold for LOTR and Pirates has been (temporarily) shelved in favor of Pirates of the Caribbean, MB's Western Town modulars may be stuck in limbo due to contractual obligations with Disney. TLG may not have the option of producing "competing" western themed sets and/or they may not want to cannibalize their own revenue stream on a theme licensed from a third party. This is purely a guess on my part, but a guess based on past precedents. If Modular Western Town does fail for such reasons, I think it will be a real black mark on CuuSoo's record. It already seems like people's enthusiasm for CuuSoo is waning (maybe this is just a false impression since they cleaned up the activity feed streams to cut down on the spam-feeds). Good ideas are languishing in the Hell of Forgotten Projects, fewer new projects are showing up, new part proposals seem dead on arrival and licensed themes are getting vetoed as being too out of date or inappropriate for little kids. If a viable, original idea gets shot down because of something like the Lone Ranger contract, it will really take the "wish" aspect out of CuuSoo. Instead of asking the crowd "What would _you_ like TLG to produce?" the implied question becomes "What can you come up with that doesn't conflict (or could be construed as conflicting) with anything we currently sell or plan to sell, that isn't licensed off an IP with more than a PG rating or is currently held by someone else, that doesn't require us to make new molds, is recent enough to be popular but not so much of a fad such that it will still be popular a year and a half from now when the product finally hits the shelves and priced realistically with respect to current sets with similar piece and mini-figure counts?" It kinda makes you wonder if CuuSoo is really about proposing fresh ideas that will be taken seriously or about TLG creating a new channel to get fans to promote the brand on Facebook, Twitter and in personal blogs.

I don't know where you've been shopping but I've got about 500 or so Tolkien related pieces in my collection and only a couple dozen are from the post-Peter Jackson film era. I've attended entire convention auctions dedicated to JUST LOTR art and merchandise back in the 1970's (Peter Jackson would have been in high school back then) so it is a mischaracterization to suggest that without the movies there's no market or that there has never been vendors actively feeding that market. When the number of items available for purchase exceed both your budget and your storage capacity, I consider it "substantial." Before United Cutlery ever made a single prop replica there was the Franklin Mint, Museum Replicas, LTD and Del Tin Antiche. Before Games-Workshop introduced their take on things, RAFM, Ral Patha, Grenadier and Mithril all had lines of 25mm and 28mm scale LOTR figures for fantasy wargaming. Roughly twenty different Tolkein Art calendars in about a dozen different languages have been published each year since 1969 - ony in the past decade have images from the movies appeared in any of them and non-film related ones are still being released (and selling well). You are absolutely right that the mass marketing campaign to _kids_ both surged and is waning, but there's an entire Middle Earth sub culture that's been around at least since the 60's that has little to do with the recent films and spends millions on Tolkein related items every year. If anything, the recent film _hurt_ the marketing of a lot of the stuff because the film-related merchandizing didn't want competition from unlicensed merchandise or alternate visions from licenses granted decades ago. Jackson's film didn't create the market; they capitalized on it; diversified it; and yes, alienated part of it (as not all LOTR fans share/endorse _that_ vision). I agree that the current kits are tightly coupled to the films and without them in recent memory, pre-teen boys won't be putting LEGO LOTR at the top of their Christmas wish list in years to come, but there's money to be made selling to adult LOTR fans if TLG can figure out (both legally and marketing-wise) how to tap into that (film independent) demographic.

If you define "toys" as something that would appeal to your average 6-8 year old boy, then I'd have to agree, but if you're talking about "collectables" that will appeal to LotR fans in an older age bracket, there I must differ. I too have been a fan of the books for most of my life and even going back 40+ years ago there was no shortage of collectables (calendars, art books, resin sculptures of iconic buildings, pewter and/or bronze figurines of major characters, etc.) Granted, other than lead figures designed for table-top war gaming and jigsaw puzzles, few of these items were designed to be "played" with, but simply as display pieces there was enough consumer demand to get these things shelf space in retail stores ( Internet direct marketing was still decades away ). The movies have certainly created a new explosion of genuine "toy" elements to the marketplace (and, for better or worse, consolidated the "look" of Middle Earth and its peoples) but, in the long term, I doubt that demand for Tolkien collectables will be any less than it has been since the 1960's once Peter Jackson stops making films. The more germane question is whether or not TLG will expand the line to include "display" sets that appeal to adult LotR (but not necessarily LEGO) fans, as they eventually did with Star Wars. It's a different demographic with different taste in sets and a different target price range, but it's also a market they haven't really tapped yet.

You'll get no argument from me on the question of the breadth of appeal, but to be fair, that's really something for the review stage to suss out. I think part of why non-AFOL projects leap past many, very good MOCs is that too many of us want to _play_ reviewer rather than just be a supporter (or not). We get overly critical of this, cite some reason why TLG would never do that , claim that we could build a better model than the one provided and before you know it, the project has slipped off the new and active radar screen to languish in the Hell of Forgotten Projects. I, for one, wouldn't buy Purdue Pete (except possibly for parts if the price were right) but I could see a clever marketer coming up with a limited poly-bag DTC line where the 'consumer' in this case is the university bookstore or an alumni association looking to buy 10,000 units to give out at reunions and fund raisers. In _that_ capacity, the idea is brilliant, especially for universites with engineering schools (we engineers are big on construction toys) Purdue could have their Pete, MIT could have a Beaver, Tufts could have Jumbo the Elephant, Stanford could have Tree (maybe the Realistic Trees project should give them a call), etc. I suspect putting these on a shelf at ToyRUs would redefine failure, but as "contract sale" in bulk to an institution willing to assume the risk of resale (or to write off as development promotional items) I could see TLG turning a modest profit for minimal risk.

I think the "condition" of having one's LEGO ambition exceed the feasible limits of one's collection is a very common problem regardless of how extensive (or not) that collection actually is. In my youth, I had a very modest collection and even my simpliest MOC designs seemed impossible to realize. I'm very fortunate these days to have been able to expand my collection and I _still_ have a notebook full of ideas I've put on hold while I try to amass the parts needed to build them. Maybe I'm an addict and "thinking big" is just my excuse to buy more LEGO lots, but I'm pretty sure I'm not the only one...

I think it's very true that _most_ of the people jumping on the Purdue Pete bandwagon probably won't go out of their way to buy the set (if produced), but I have noticed that since sandsanm changed her tags to be more generic and CuuSoo friendly, I've definitely seen a bump in several of _my_ projects and (tracing the activity feed) most of the new support is coming from accounts nominally created to support 'Pete. Granted, I'm talking about 1 in 400 Purdue Pete supporters sticking around to support at least one of my projects, but I'm only tracking my stuff and some of the new users I've checked out are already supporting over a dozen projects. Purdue does have a good engineering program, and engineers do love their toys after all, so maybe -just maybe- this project might have the side effect of bringing a small number of people _back_ to LEGO or at least carrying a few other CuuSoo projects along with it (even if only a few dozen scattered votes here and there).

Decades ago I wrote a spy novel where a public page for Shayd DeGrai was a run signal to warn a deep cover operative that his active persona had been compromised. It was an acronym for "Someone Has All Your Data, Drop Everything, Go (to) Roost, Await Instructions". I rather liked the name, so years later as a GM for a fantasy role-playing game, I used the name Shayd for any agent of a information brokering organization. Shayd agents lurked in shadows and lingered in taverns eavesdropping on conversations and collecting rumors that they could resell later on. When e-bulletin boards, the Internet and chat rooms came along I started using the id Shayd as, in my mind, it was similar to "lurker" but not as common. When I started joining LEGO fan-sites I discovered that Shayd was not uncommon enough and that despite being "Shayd" on scores of other sites, I'd need to mix things up. That was when I looked at my LEGO collection and realized that more than three quarters of it was either black, white or some shade of gray - so It made perfect sense to me append the Shayd's original last name to my sign on. And in answer to questions asked at various web sites: A) My used of "ShaydDeGrai" dates back to the mid 90's and has nothing to do with the book "Fifty Shades of Gray" B) Yes, both the name and the book title _could_ apply to the color of my hair, but it was _supposed_ to refer the dominant colors of my LEGO collection

Sounds like you got a bargain, congrats. The Emerald Night is a personal favorite of mine (it reminds me of the Flying Scotsman, a real train where my grandfather worked as an engineer) and Lego Trains are a great way for an AFOL to get back into the greater brick hobby. Welcome Aboard!