Original article has been posted 26 of December 2013 on habrahabr.ru
Remembering the questions of habrahabr.ru users to one of my previous posts, I'll remind you in a nutshell what this is all about.
We are the virt2real team, enthusiast geeks with a fondness for creating all sorts of things with remote control and video surveillance over the internet. A controllable webcam, racing over the internet, and then developing our own microcontroller, taking pre-orders, taking orders, releasing the first batch, and then Bond's ride.
Normally, we only publish the results of our work, and don't say much about what happens behind the scenes. Therefore, in times of silence on air, questions start coming to us – 'Are you still alive at all?' In this article, I would like to expose the inner workings of the team. There will be no intelligent conclusions or references to fundamental works on project management – this article came out as a documentary, in essence. A description of our journey, our mistakes and shortcomings, which hurt.
The first thing I want to say is that the project proved to be incredibly much more difficult than we had expected. There were hardware problems, a gigantic showstopper with software, and production issues. There are so many emotions that I want to express. As to not dilute the whole article with them too much, I will try to concentrate them in a small metaphorical description.
You know, at the start of the project we felt like this chipper, swift, experienced hunting dog coming out into an open field to chase another rabbit. We came out into the field, saw a rabbit in the distance – hooray, hooray! Drive, adrenaline, excitement! And we set off. Only we were approaching the bunny somewhat slowly. Havig had quite of a run and starting to get tired, we realized that the bunny is quite far away and does not look quite like a rabbit... When we finally got close, it turned out to be quite a huge buffalo! Well, we snapped onto its nose at full speed, and it easily pulled us off the ground. After a long battle we are dented, with broken and dislocated legs, and tattered skin. But the buffalo also got tired and fell to its knees. So here we are, preparing for the final blow.
I'll mention in my story the end of 2012. By that time, we had managed to get the first batch of 10 virturilkas to working condition, which allowed us to test it in different usage scenarios. The result was the first article on habrahabr.ru.
What happened then was incredible! We were buried in emails, questions and proposals. The reviews were very different, ranging from “It's cool, where do I buy one” to “We conceived something like this a long time ago as well, but never got to doing it” and “What the hell is this thing good for?”. But the vast majority of reviews were positive, which gave our morale a great boost. On a separate note, I would also like to mention a wave of emails from people who have their own projects and tasks, asking us how they could use the virturilka. From smart intercoms and an in-car DVR to diesel generator controls and agrorobots. There were offers of help – in development, manufacturing and software. Several retail and online stores asked about collaboration. The first proposals for funding came along as well – both from funds, and from people saying, “I know this uncle Bob with money, let's talk.” The near future seemed bright and cloudless.
Newcomer to the team and lean-startup
In the first wave of emails we received a message from Anatoly, who later joined our team. I must say that we are all hooked on the technical side of the issue, and Anatoly brought the lean-startup method. From it, we learned a few interesting tricks and tools. For example, being asked the question “What is your solution all about?” we would start a long story about the the capabilities of the board, drowning in technical details and possible usage cases. And every time this story came out different, both in content and in duration. Having tried the lean canvas tool in practice, and having brainstormed with the whole team, we came to the definition “Low-cost platform for quick construction of systems with remote control and video surveillance over WiFi, cellular networks, and the internet”. It seems to be one sentence, but to shrink all that we wanted to say into one sentence was impossibly difficult. We also gave a good thought to who might need this solution (we came up with DIY geeks and corporate customers), what can earn us some money, market niche, etc., etc. Overall, it was really helpful to understand and define for ourselves what we have made and what can be done with it.
Two questions dominated the incessant torrent – how much will it cost and how to buy one. Here, we could have taken the classical path – find money, run a small print and try to sell it. Or we could first try to understand who is ready to buy the virturilka and not spend on production. Basically, to count people willing to buy.
No sooner said than done: we got a quote and estimated the cost of a kit in a print of 1,000 units, created a landing page, and started taking pre-orders. Here is the page.
We filmed the video at Sergei Gol's, but that's a different funny story. We decided not to overcomplicate the page with lots of fields and options, we didn't even set up a captcha – the goal was to provide a way for people to quickly and easily make pre-orders. Of course, there were some jokers who sent orders on behalf of Putin for 100,500 units, or with addresses like firstname.lastname@example.org. Later, we had to go through the whole list manually because many people mistyped their email addresses (mal.ru, gmal.com, etc.). During the first day we got about 1,000 pre-orders (the final total was about 2,000).
I must say it was the most exciting day. Jitters, anxiety, waiting for a miracle... It's difficult to convey in words. I periodically informed the team of the current number of pre-orders. Tried to put up an order counter on the site – couldn't get it to work immediately, but made a crutch and the counter was visible to team members logged in to the site with necessary rights (thanks to the Webform module for Drupal 6). After our article got into the top 10, the stream became very active – sometimes several orders per second. In essence, these pre-orders have become for us a measure of our idea's viability – if there would have been a hundred pre-orders, then we would have to give it a good thought and understand what we had missed. But when the counter first passed the 100 mark, then 300, and finally crossed the mark of 500 pre-orders – our hearts were much easier.
On the day we were taking pre-orders, we started receiving calls from various publications, and emails started pouring in. The next day articles appeared on quite serious portals – here are some links that we managed to keep track of. We got a chance to meet many publications because the topic was interesting. Some portals told us that the clickability of our story had been higher than that of their other news that day. Of course, it all added positive feelings and boosted the team's morale. There was a feeling that this was happiness – pre-orders were coming in, we were in the news, and it all would start to grow tremendously.
Of all that had happened, we concluded that the product has demand, is quite interesting and has quite a real future. The only thing left to do was to find money for production and to release it. In December 2012 we created a legal entity, Virtoreal Ltd., and went to the next stage. And then we had a lot of great discoveries.
As I wrote above, we were approached by many who wished to make a financial contribution to the project. I would divide them into several groups.
Fraudsters and 'smart-asses' of all kinds.
The first group was not numerous but outstanding. Some said, for example, “This is private and confidential information – but we have this project, the state has allocated 10 billion roubles for technical means for the disabled, we cannot reveal any details, but there sure is a lot of money – let's work together.” The second call already changed the amounts and tasks, but there was constant references to unbelievable sums of money. Instinct warned us that something was not right, and such requests were quickly cut off by saying “please send your technical requirements – we'll have a look and discuss it.” I guess these kind of people just wanted to get some money for “preparing documentation in a tender for a hundred thousand million” or the likes – a well-known scheme in IT-procurement. There were a few calls from people who offered to help getting loans from banks, with no collateral but in exchange for 15% of the loan amount. There were unexpected requests like “We have this project here, a lot of money was allocated, but there is no solution – can you make your PCB round, in a different colour, remove your name from it – and we'll say that it's our product.” No kidding – they actually asked to make the virturilka round! All in all, only keeping a cool head and thinking rationally saved us. I must say, however, that many of the callers were well versed in conducting such negotiations – the feeling was that this activity is their speciality.
The second group were, perhaps, the people that surprised me the most. I am not generalising at all – just sharing impressions from my personal experience of talking to them. Maybe I just met the wrong kind. Or maybe the format and presentation of our articles attracted people of a certain kind – I cannot judge here. Preparing for meetings, I expected to see lively, active entrepreneurs with real money, sharp questions about the project and ready to discuss terms. In reality, it turned out that I was meeting intermediaries who informed me that they could try to find and bring money from here and there. From all these conversations I had the impression that I was trying to sell a used car through Craigslist, with some touch of suspicion and discomfort.
And the main icing on the cake was the fact that none of them understood the essence of the project – all the people were non-specialists, having little idea about electronics in general. They could hardly understand that we created a DIY platform for constructing ready-made solutions. All the conversations went to something like “Let's develop this end user product and produce it in huge quantities.” They went into creative musings about incredible new smart devices that will conquer the world (is that really induced by our posts?). Switching the conversation to “Great idea! So here is our platform – develop your end user device on it, reproduce it and make money from it, since we have already done the hardest electronic part” did not work. After several such meetings, I began filtering out people through preliminary email correspondence and telephone calls, which allowed to avoid useless meetings. There was only one adequate representative of this group of investors who said, after an extensive conversation, “this is really very interesting, but not my cup of tea.”
The third group was the most numerous. By the nature of my main activity I had not come across investment dealings before. There were many discoveries about dilution of shares, rounds of investments, a lot of new terms from the investment business. There are loads of articles on this subject online, so I won't repeat those. One thing was clear – getting investors all excited with only pretty words is unrealistic. One needs a standard set – at least a presentation and a business plan. Our new player on the team, Anatoly, actively joined in on this one. We made a presentation, having extensively studied the history of similar electronic projects (Arduino, BeagleBone, Raspberry Pi). As for the business plan, it had caused very ambiguous thoughts since the beginning. We are a working team that has created a product and is actively seeking a business model. We tested the first hypothesis – there is a DIY market for us. It's all according to lean-startup. In other words, we don't even know for sure whether the markets that we see will work in the right direction, or new ones will open. But without numbers conversation is useless, so we conceived our ideal model and constructed a business plan based on it. It was a beautiful three-dimensional multi-parametric model that we could play with and study possible developments in different scenarios. From it we made a business estimate and calculated the necessary funds. And negotiations began.
We were able to talk to most Russian funds. I must say that cooperation did not happen with any of them. The most interesting observations were as follows.
First of all, the focus of our project didn't fit common patterns. We are hardware people. There's no way to play with us a game of “put five programmers to work and a year later the project is worth a billion”. With hardware everything is slower, the cost of failure is thousands of times higher, there is a huge amount of risk at all stages. And, furthermore, substantial amounts of money are required from the very beginning.
Secondly, there was a problem with expertise. We were told exactly that – there is simply no one to turn to evaluate your solution. Nevertheless, two funds found third party experts who beheld in the entire solution only a TI DaVinci processor and gave detailed reports on the performance of its ARM core. For some reason, they did not consider the product as a platform. One of the funds thought for a long time after talks and then disappeared from the radar. And after some time, four other funds approached us themselves, saying “you're interesting, let 's talk.” It turned out that the first fund, not understanding what we are, had quietly passed us to others so that they would try to figure us out and give their assessment.
Thirdly, again, we did not fit the pattern of typical project development. This is the conclusion we came to, having analysed a dozen meetings and suspecting that we have some kind of system error, which brings conversations to a standstill.
In essence it's like this: the project has developed the actual product (idea) itself, and developed all structural add-ons – a team, a legal entity, resources, management model, etc. It has the related structure and internal activity. But it turns out that in terms of organisational development we are a team of enthusiasts from different cities, in which almost everyone works on other regular jobs. We don't even have an office. But in terms of the product, instead of being at the stage of “we have an idea – we should put it on paper”, we have already passed the stages of idea, technical design, implementation of prototype in hardware, debugging of prototype, releasing first small series, and preparing of serial product. And the funds we need are already for manufacturing a completed product. It's as if a 7 year old son would come to his father saying, “I need some money – I already dug a hole near the forest, poured a foundation into it and built the ground floor of a 10 storey building of my own project – I need to finish the other 9 storeys. Oh, and I already have perspective buyers for most apartments.” Something like that. We got the most precise formulation of this situation in the Skolkovo innovation centre – we had a good sensible conversation there. We had already passed those stages of development during which we could have gotten a grant precisely for developing our idea. We did all that with our own money. And we are now in that stage where funds are already granted for production – with a portion of the financing coming from the fund, and part from third-party investors. But some companies are deliberately trying to pass as small projects in order to get money for production under the guise of funds to develop their product. Hopefully, I haven't disclosed any secrets here.
The conclusion from talking to funds was this: we will go through one more stage ourselves, smooth out the rough edges in development, and then come back to this topic. But for now we're good on our own.
A bit about production...
Meanwhile, we decided to test the entire production chain on a small print – 100 pieces. We decided to order exactly in such a place and such a way so as the production of the first thousand would go on a beaten track. We produced the first ten boards (we call them green among ourselves) completely in Russia. To make a final decision on the production, we decided to order the print for evaluation to several companies in Russia who came to us themselves, and to test the option with China.
Briefly, the production cost estimates we got in Russia were significantly higher than what we had expected and than the announced retail prices. Plain exorbitant. We later found out that companies that said they would produce even a 10-layer board in Russia, still would order it in China and then bring to Russia for assembly. We were somewhat upset with the approach of some producers: “tell us the amount you're willing to spend and we will go from there.” The aim is clear, all the technical documentation is there, the volume of production is known – evaluate, give us your quote. It's like having a conversation in the bakery: “I would like a loaf of bread, please. – How much are you willing to pay?” If they at least had given us a choice – doing so is more expensive, and doing so is cheaper, and if you don't test you'll save some money, but there are risks. A loss of time and nerves.
We decided on production in China. I'll digress on this for a moment. We are used to doing everything ourselves. We develop our own hardware within our limited resources, we make our own websites, we do our own graphic design (yes, we know that everything is bad, but at least it's there), we code our own Linux and create our own examples. We only hired an external graphic designer to design the pre-orders page and help with the logo. That's why we wanted, for example, to organise production in China from scratch – with trips there, choosing factories, buying parts, etc. But we were lucky on this one. Through habrahabr.ru we met Dmitry (dzhe), who helped us organize the entire production process. It can be considered advertising, but we got everything we wanted. Every step was clear to us, the whole chain of procurement, production, delivery. In each case we had a choice – for example, to buy NAND flash from a reliable supplier, but more expensive, or less expensive from another one, but it could be culled. We communicated directly with suppliers in China and discussed the obtained information with Dmitry. We found out which RAM is better considering what is manufactured and sold in China as compared to Korea (where there's quality and where there are culled chips from brand manufacturers). Which operations to do on machines, and which ones manually. Our partners' Russian friend living in Hong Kong, who can resolve all sorts of things there, was a true godsend. By the way, it's one of the developers of the Radio-86RK... We discussed peculiarities of factories – one place is more expensive, but they buy components themselves and take full responsibility, another is cheaper, but we must buy components ourselves, which saves us some money. One place will make the boards in a week, but more expensive, while another will do it in three weeks, but cheaper.
Current problems – like replacing a purchased lot of electronic parts, in which the Chinese had mixed up the prefixes 'milli' and 'mega' in the nominal value of resistors, – were also solved quickly and painlessly. In the end, even such a thing as notifying the FSB was done for us officially.
In this batch of 100 PCBs surfaced the first operational problems, and not just operational. It's worth noting that the design of the virturilka had changed significantly – this required a major rework of the architecture and changes to the board's construction. Surprises could emerge from anywhere. And so they did. One hardware error has crept up from the green version – so we had to patch the first hundred by cutting one track and putting in a hair. The calculated values of some resistors turned out to be wrong – this we found during GPIO tests, for which the first version of a normal driver had been done. We changed the camera module – we found a module with higher resolution for the same price. Hence, we had to write a new driver. And during all this time we were actively working on the creation of a normal Linux on kernel 3.9 – it turned out that nobody had done this before us, and all that existed online was done on kernels many years old. Overall, it was a lot of fun.
...and about Bond's ride
All this time, Sergey was also doing Bond's ride. On 11th of March the topic was born.
The news exploded like a bomb and knocked us out in the process. We had never expected that besides the national press and TV the article would be translated into English and draw so much attention in the West. Most of the information in the 'Press about us' section is exactly about Bond's ride. In addition to being mentioned on such monsters as Wired, Gizmodo and National Post, we were very happy about the article on TopGear and, of course, on the Croatian Playboy website. TV crews came from Russian TV, RussiaToday and foreign channels (Discovery Channel). Enthusiasts from automotive projects started approaching us. After this, we did a site redesign (actually, we deployed a new one on Drupal 7) and included support for Russian and English languages, although we don't yet have the time to translate everything. The old site remains at old.virt2real.com.
Bond's ride added notoriety to the project, but did not bring any money directly. However, it helped us to get a second custom project that brought us funds for the production run.
...and a bit about the Great Depression
In the spring, another risk came into play – all of us have permanent jobs. Because of this, the debugging process stretched for a very long time. While taking pre-orders, we promised to start shipping in the spring, but we were very much mistaken. Sometimes the work stopped for a few weeks due to extreme workloads that one or other team member had at their main job. The amount of tasks to accomplish grew exponentially – but resources were not adding up, so the team's speed of work decreased. In emails and pm's on habrahabr.ru we started receiving questions like “Well, where are you?”, and comments like “Well, just as usual, started on a merry note...”, “It was a cool idea, pity”. I would say it was the most depressing part of the year, when all that could go wrong was going the worst way possible. The problem with the Linux video subsystem that took us out for a month, the negative outcome of negotiations with all the investment funds, the null result on all funding sources, the stoppage of work on the hardware and the first print run. And amid all this, we were receiving proposals of cooperation from different companies – we were simply physically incapable of processing it all adequately. Wound up, feeling trapped in a vicious circle between financing and print, a huge and rapidly growing field of work – yes, we were seriously inflammable within the team. And we had no way of switching to the project full-time – we all have families, kids, loans. And the project was clearly getting stuck. So spring and summer passed in a mode of titanic tension and internal devastation. But the mixture of anger and helplessness gave new force. One of the consequences – we only managed to start working on the promised second storey daughterboards after quite a long delay.
I still consider it a miracle that our team has been doing this project for almost two years, was able to put it in hardware and continues moving it forward.
So, on 13 of June we received the first samples of the new series, which was a joy.
And back to money – projects, crowdfunding
We needed money for production. The first 100 pre-series boards (the virturilka 1.1, “the black one”) was paid for from personal funds. I can say that during the production of the first hundred a lot of work has been done to prepare for the main print – various one-off costs. Each virturilka of this 100 print cost us $180.
At the end of 2012 we were approached by some guys with a custom project. By the terms of the NDA I cannot tell you any details except that it weighed 60 kg and was controlled from a tablet. This project brought us the means to support our pants, but could not solve the production issue. Work on the project went on for several months, along with all the events of production and finding finances.
And then we decided to try crowdfunding. Our first thought was, of course, Kickstarter. We even started preparing a video and other content. The only thing that bothered us was that we were not known in the West, we hadn't really made any noise yet (this was before Bond's ride). But this tale was over faster than we had expected – Kickstarter blacklisted and banned our project since they classified the virturilka as surveillance equipment. That was it.
The next thought was – we wait for the first hundred PCBs from the print, make sure that everything is in order, and try to run a crowdfunding project on habrahabr.ru. After all, we did already have 2,000 pre-orders – that was something.
To raise funds, we revived the online shop we had – mikrogonki.ru, which had been put into conservation due to its main product – microcars – being pulled from production. But the engine was working, so we decided to use it. As it turned out, due to its age, advances in technology and a change in the API of our payments aggregator, the shop had made it quite hard for those who tried to make a payment. Traditionally, all bugs surfaced right at the busiest time – when an active stream started pouring in of people willing to help the project and make a pre-payment for their virturilka. The crutches introduced on the go solved some problems. Basically, we decided to break the vicious circle and try to use the last chance to raise funds for the print run – in one night the shop had been redesigned, redone, and filled with new content. And on 19th of June, 6 days after receiving the first five new generation boards and testing them, we published an article about taking orders.
First we gave the opportunity to make a payment to those who had made pre-orders. Technically, this was done quite simply – we sent invitations to the site to all the email addresses on pre-orders, and pre-created accounts with rights to buy and make payments. After a while, when the stream dried up, we removed the restriction – anyone could order and pay. In the first wave, a total of about 500 virturilkas were paid for. Those who paid for their order within a week from the start were offered to choose a bonus daughterboard from a list of planned ones.
It turned out that the money wasn't enough for a print run. If we were to make less than a 1,000, the price would rise substantially. Besides, during this time the dollar had managed to go from 27 to 32 roubles, which also added to our delight. And just at that moment another custom project came our way. Without breaking the NDA, I can only say that this thing weighed in at a little less than four tons. And the idea of approaching us came up after Bond's ride. This project, logically, demanded a serious shift of attention to it from our main development.
And what do you think? Even with this outside project the money still wasn't enough. We had gathered slightly more than half. We were already seriously considering the option of declaring to the community “sorry, we've got this problem here”, and refunding all payments. It's quite an unpleasant feeling, I tell you – we promised, we took it upon ourselves, we prepared the production, and even collected money – and then bang! In emergency, we began thinking of all possible ways to raise funds. The happy ending happened by taking personal loans that did not kill us, but made us stronger. Then, print began. It came to us on 28th of August. Here's the photo report.
Early adopters – user skills levels and entry threshold
Remember the struggle with the Linux video subsystem? On the first five boards it broke out with renewed vigour. We were waiting for the print and had to make adequate software before its arrival. We assumed that the second storey daughterboards would come immediately together with the main print (or a bit later), and our happy customers would be able to start playing with the virturilka full on straight away. This struggle is still going on...
In the article about the virturilka we said that we see several levels of user skills – from beginners to experts. And here we had a reason to be upset. Among those who had ordered the virturilka there were many people without a serious level of programming skills, but inspired by our articles. They got the virturilka with no daughterboards and with software that made even experienced Linux users strain their brains. This is clearly seen in the first discussion threads of our forum. Basically, it turned out that literally only a couple dozen of people out of the 500 that had bought the virturilka were able to work normally with it using the software that we had when the first print came to Russia. A very high entry threshold. We were only finishing the lower, basic, level of the software, upon which there will be a user-friendly set of tools for working with the board.
Guys, we simply didn't have enough time. Not yet. Essentially, only three people in the team are working on hardware and software, and it is very hard. But we are trying very hard to cope with it. And we will!
On some issues, we turned to the community on our forum – as a result, there is now a native app for iOS to work with the virturilka (open source), for Android, and there's also a web steering app. The first “Virturilka for Dummies” article has been written, reviewing the most basic concepts. On our forum, the “projects based on virturilka” section, where colleagues share their developments, is getting new additions. And we are very grateful to all forum users that help newcomers on the board!
I must say that both the forum and the wiki added another front of work – community support and creating content for our information resources. And we feel real help from the community – it's just great!
One cannot foresee all
As I mentioned, an error in hardware development is manifold as expensive as one in software. When developing software, you can quickly release a patch to solve the problem. With hardware it's a totally different story. And we had one such surprise. In the production of the first 100 we included, in particular, WiFi modules. The module was very simple in construction (compared to the 10-layer virturilka with 400 components). A random check of the modules showed that all was well. When the batch arrived and we started testing, everything was good as well. So we did the print run. To be completely sure, we tested each set fully before shipment (I'll write about this below). And then we got hit below the belt.
It turned out that about 25% of the WiFi modules did not work. The devil was that some of them would start up normally, but would die after a few minutes of work. Through comprehensive testing we eliminated most problematic modules, but some had leaked to end users. Of course, we replace them, no problem, but each replacement means mailing. And that means time, loss of money, and a slight disappointment for the buyers. And we only had to take the time for additional testing of the modules at the factory, which we skipped to save time. It's our fault.
About underestimating small details
Oh, China! How wonderful you are for your low price! This is about testing and packaging. The virturilka itself, the Wi-Fi and camera modules came to us in anti-static zip-bags fastened with a snapping white plastic strip. The bag is small, the lock sturdy – after a dozen of bags your fingers literally start hurting.
To test one kit:
1. Open lock and pull virturilka out of bag
2. Open lock and pull Wi-Fi module out of bag
3. Open lock and pull camera module out of bag
4. Peel orange protective film from virturilka's connectors (2 connectors)
5. Peeling protective film from Wi-Fi connectors (2 connectors)
6. Plug wire into camera and into virturilka
7. Screw the antenna on
8. Insert wire into MicroUSB plug and connect USB stick (for testing USB)
9. Insert MicroSD card with test firmware
10. Connect mini-TV to AV output
11. Connect power and wait for firmware to load and tests to run, displaying results on TV
12. In case of failed tests, replace faulty component and repeat test
Test passed? Okay, pack it up!
13. Unscrew antenna
14. Unplug camera wire from virturilka and camera board
15. Disconnect USB stick
16. Take out MicroSD
17. Pack everything in bags
18. Put foam and contents in box
19. Close box
20. Mark order as complete
For all these operations we elaborated a conveyor system – during loading and testing we switched to mass unpacking and peeling protective film, or to assembly.
Why am I mentioning all this? This is valuable experience. For the future, our main conclusion is that absolutely all mass simple operations must clearly be done in China. It will cost a little more, take more time, shipping will be more expensive (when everything is packed into boxes the volumetric weight of the shipment increases dramatically), but it will definitely pay off. Not foreseeing such small details, one can drown in routine. On some days I finished testing and packing by five in the morning, as a consequence – mistakes in compiling orders: forgot to put in a cable or mixed up order numbers and put two virturilkas instead of one.
The cardboard boxes for the virturilkas were ordered from a company we found on Yandex in 10 minutes. After we published pictures of the boxes a lot of people asked, “Where did you order those? They are super!” The boxes turned out to be more interesting than the virturilkas.
Ordering the filler foam, we made a mistake with the size – it was 1 cm longer than necessary. Such a small thing made the cardboard bite the extra foam when closing the box. Big waste of time correcting the biting problem.
When sending mail orders by Russian Post, the post-office workers cursed while packaging our mini-box with tape on all sides. The solution was an elegant one – I was given a branded tape, and we packed all the boxes at home before going to the post office.
Housekeeping note: shipping 100 parcels at once by Russian Post will be much tastier if you do it at the main post office, which works around the clock, late at night. And with proper diplomacy you can leave the parcels to the nice lady there and come back for receipts in the morning.
We have assembled and are testing prototypes of second storey daughterboards. We already have the PCBs and the parts as well. We will assemble in Russia, for it's a small batch. We expect them to be ready in January. We already published technical descriptions in our wiki – a motorshield, a relay module and a debugging board for prototyping.
Results in one paragraph, briefly and self-critically
So, in one year we managed to:
+ Develop a serial model of the PCB and extra modules
+ Find funds for production
+ Organize production in China
+ Run and bring the print to Russia
+ Send the virturilka to everyone who wanted one – “let it out into the world”
+ Develop a programming environment for work
+ Do several client projects
+ Keep and even expand our team
We didn't manage to:
– do everything on time
– attract third-party funding
– avoid mistakes when receiving hardware from manufacturer (WiFi modules)
– make the software accessible to users of all skill levels
– start working with corporate clients
I should add that the current volume of project tasks is becoming incompatible with my main job. Currently I am passing on my responsibilities and completely leaving my job to work on our project full-time. Other members of the team are also actively considering this now.
I would like to thank all those who believe in us and help the project. And apologize to those whom, in the heat of the battle, we didn't reply to in time or with whom we were “slow” in our correspondence.
Translated by Paulo Fino