Year end
North Coast's first fiscal year ends with this month, and I have a bunch of small news items and progress updates.
First up, the Leapfrog VCF. This is the big flagship module that I think people are waiting for, and I was supposed to have the next prototype in hand by now, but stupid things just keep happening. I wrote earlier about ordering the circuit boards, right? So, I did that, and then a few days later I got a package from China... containing a batch of circuit boards completely different from mine and an invoice addressed to one "Kaelin Wilson" of Alameda, California! It appears that the PCB contractors messed up and packed someone else's order in the box that had been addressed to me. It also appears that Canadian Customs didn't even glance at the declaration form. I went after the PCB company's support people by email and (when that didn't work) Twitter, and they now assure me that although they don't know where my boards are, they'll re-do the order and send me a fresh set. So, fine, I just face a delay. But I do hope that they get their act together before the end of the month, because apparently Chinese holidays start on October 1 or something and there won't be any more boards manufactured for a while after that. Meanwhile, I decided to do a good turn for the other customer and I stuck the boards I'd gotten, in the mail to Alameda.
It's just one thing after another. I am reasonably hopeful that the next prototype will be the last. You can see the earlier prototype, with the green boards, in the online store, and even listen to audio demos there; it seems to work quite well, so the only things for which I need another prototype, as far as I know, are making sure that the new panel layout works, and that there have been no errors introduced in reconfiguring the boards for that. I have all the parts to build a few of them as soon as I get the circuit boards, so if all goes well it won't be too much longer before I actually have Leapfrogs for sale. If all doesn't go well and I need to do more prototyping and development work, well.
Be aware that the bottleneck for the Leapfrog VCF on the manufacturing side, is the panels. I have plenty of the other parts, but I only have four usable panels right now, and I need at least one for the prototype, and since panels are the most expensive and hard-to-source components and I can't afford to spend a lot more on development costs before I sell the first few units, I'll be ordering new panels a few at a time as orders for modules and kits come in. Modules and kits take priority for the panels I have. I may actually take "PCB and panel" off the list of options when the Leapfrog goes live; if I do keep it, I will at least be keeping the available quantities low, and I strongly encourage you to buy kits and modules instead. The margin on "PCB and panel" is terrible (I end up taking a hit on the PCBs in order to be able to bundle them with panels); "PCB and panel" sets are lightweight but still cost just as much as kits and modules to ship because shipping doesn't go by weight, with the result that customers get sticker shock on seeing the shipping fees; because the nominal price tag on "PCB and panel" is lower, I have to sell a larger number of them to stay in business, further compounded by the poor margins; and almost everybody else selling "PCB and panel' sets is really selling "PCB and pretend fake panel that is really another PCB," which understandably costs less and so puts even more pressure on my prices for the genuine article. Who needs this?
EDIT: Comments on shipping charges at the time this was originally posted removed, North Coast now pays the shipping on all orders.
In Eurorack news, Mutable Instruments has announced that Braids is discontinued. It sounds like the "menu-diving, bleah!" whiners finally piled up too deep, and it's the designer's right to stop making the module, of course, but it feels like a tragedy. Braids has been one of my own wishlist modules for a long time and I always thought I could afford to wait until I had the money to spare. Beyond my own system personally, though, I was also using Braids as my go-to suggestion for newbies building a Eurorack starter system; my thought was that if you want to buy just one module, and have it be both something that will be fun by itself and also continue being useful as your addiction modular synthesizer grows, Braids would be a good choice. Both for the eventual reworked version of that article, and for my current skunk works project, I need to have something to recommend for that role - a complex, preferably digital, multi-mode oscillator that can produce interesting noises all by itself. Some people have suggested the Klavis Twin Waves, and that may be the right answer.
I wish my modules were so popular that I could discontinue one just because too many whiners were buying it.
Preparing for the financial year-end is a bit of an adventure. I am doing my bookkeeping with Ledger-CLI, partly because it integrates well with other systems I use (Subversion, random Perl one-liners, and so on). My accountant is being a good sport about it, but he's clearly never seen anything quite like what I'm doing, and there's a bit of a learning curve on both sides. Frequently, something he expects will be very difficult for me turns out to be easy - or, fortunately less often, vice versa. There are two big advantages of this system over others, though: one, being that it's open source and based on text files, I know how it works, and that makes me confident that I actually understand what's going on in my business. How much is that peace of mind worth? And two, also a consequence of being open and text-based, it integrates well with my other systems. I can type "make" and get a nice TeX-formatted PDF file explaining exactly where my parts inventory stands, how many Leapfrogs I can build with the parts I have on hand, which are the first parts to run out, and so on. I think this would require several thousand dollars worth of ERP software licenses, followed by as much work configuring them as I spent writing my own report system from scratch, to accomplish with commercial GUI software.
As I said recently on Twitter, if you ever want to know what it is like to fight a war against a malevolent AI, just get a Google AdWords account and try to use it. I don't even like the concept of Internet advertising, and I run aggressive blocking plug-ins on all my browsers. I wish that people would just visit my store because their friends told them it was cool and we could put the whole concept of Internet advertising where it belongs. However, third-party referrals haven't been happening in anything like the amount that would be needed for it to work, so, off I go to AdWords. And I end up facing nonsense like this:
I think - and I'm not even certain this interpretation is correct because the message is unclear and the help is useless - that what this is saying is that their Really A Human Employee And Not A Machine Learning System, Scout's Honour! reviewer thinks my ad extensions infringe a trademark and thus can only be permitted to be shown in jurisdictions where the law permits trademark infringement. (Surprising how many such jurisdictions it turns out there are!) And the "trademark" they think I infringed is the words "only you." Can you see, in the image, where they think I referred in a trademark-infringing way to the phrase "only you"? I had to read it several times over to find the place. In fact, it seems to be where I wrote "PCBs only - you supply the rest."
I've also been advertising on Project Wonderful, a system which I used about a decade ago to advertise my Web comic. (Disclosure: that's an affiliate link, and I get a tiny percentage if you sign up for it after following that link.) It seems to be still the same as ever - and unfortunately, it still isn't a truthful auction. Instead of paying as much as the highest losing bid, the winner on each ad box pays one bid increment more than the highest losing bid, except only their own bid in the case where there was a tie and the tie-breaking went in their favour. "One bid increment more" could be as much as 100% more when it hits the $0.01/$0.02 or $0.10/$0.20 boundaries. This sounds like a minor difference from the truthful-auction rule, but the consequences are absolutely disastrous. You can often pay much less in an auction you were already winning, by reducing your bid, and you still win; therefore it is NOT best play to bid the most you'd possibly be willing to pay and just let it ride. Instead, advertisers end up playing cat-and-mouse games with their bids, reducing each bid just far enough to stay on the winning side, those who aren't willing to spend effort on that drop out entirely, and it drives prices down - which has the knock-on effect of driving high-quality publishers out of the system because they can't get high enough prices, and prices being so low forces the "house" to take a large cut in order to keep the servers running, which makes the returns for publishers that much worse, and it spirals further. The result is that I can get clicks from Project Wonderful at prices so low as to be negligible (i.e. why not do it?) when I'm also dealing with Google, but it's not clear I can expect any of those clicks to really turn into sales. Also, and a new development since I last looked at Project Wonderful in the mid-2000s: it now seems to be All Furries And Bronies All The Time, which may be a good thing for my modular ads because I think there's some overlap with the modular synth crowd. I'm not sure just what kind of special offers or new module designs I need to appeal to the furries and bronies, but that might be a good direction to look.
Muffwiggler has been down for a couple days now, and the rumour on Twitter is it'll be "a few days" more. There is apparently a Facebook group where Muff, the guy who runs it, has been in communication with users, but I was unable to find that - Facebook's search only turns up the former Muffwiggler Store in Portland (which is now called something else and run by other people). So I don't know what's up with that. I want to blame Cloudflare on general principles, even though I have no evidence for suspecting that the current outage is Cloudflare's fault in particular. Second choice would be that the social justice idiots finally got so upset about the name of the Web site that they DDoSed it into a smoking crater. What's interesting for me is seeing how big an impact it has had on traffic numbers for the North Coast Synthesis storefront: basically, 50%. Half my traffic, until Muffwiggler went down, was coming from there. I had some clues about that from statistics reports before, but seeing the actual drop in the total when MW went down, really made it stand out. And since it's been weeks since I posted anything specifically about North Coast there, and the threads specifically about North Coast have worked their way onto the back pages that nobody ever reads, it seems almost all of that 50%, a few tens of visitors per day, was just from people clicking on the links in my signature. So, I guess those links are more valuable than I thought.
◀ PREV Fun with fractal chord progressions || Notes on notes on the plane NEXT ▶
