240 karma · joined September 22, 2017
Like the other commenter said, treat this as an incredible compliment. They see the value of your product and want a slice of the pie. Please approach your other potential customers (ie their competitors), and start removing their leverage on you.
Best of luck! Good problems to have!
But like many things, there's no replacement for getting your hands on a real life system and gaining real life intuition.
The best way to learn this? Get/make a test bench with safe-ish motor and encoder system. Play with values and measure at the performance in the system. Increasing the P value gains, makes the system "stiffer" but at the expense of stability. Put your hand on the flywheel and "fight it" safely. You can feel the I (Integral) value winding up against your steady state errors. It's quite magical.
As you start to chase the highest performance for your system, you start to think about alternative control strategies. Eventually you will dream up the concept of feedforward control loop. Try to automate the the picking of PID values, then you start to learn about Ziegler–Nichols tuning method. Gain Scheduling. Non Linear Control theory. It never stops.
(If you do any of this, please make sure the motor/flywheel won't kill you if it goes unstable, and please wire a E-stop that's easy to access when it's unstable)
I just purchased a 150W c02 laser from China. When you import a laser from China. The FDA starts to care. They require that the import lasers that have an associated FDA Accession number for the product. Basically saying the laser is safe and has proper safety etc.
Having a customs broker for your first transaction is worth it's weight gold.
In terms of importing these lasers from China and selling it in the US market? I think you're a couple years late on the market for that one ;) There quite a few players that do that now.
Particulates and VOC (Volatile Organic Compounds). Think of particulates as specs of dust with varying size and VOC's as random chemical compounds such as spray paint, smoke, oders.
The general go-to clean room/industrial filter stack is Prefilter(F5 Bag)->HEPA->Activated Carbon. The pre-filter catches all the big particulates to prevents clogging the HEPA filter. The HEPA filter catches all of the really small particulates. The activated carbon captures most of the VOC's/oder/smoke.
The official way to monitor when a particulate filter is bad, is to measure the pressure drop across the filter, and compare to manufacture specs. For the carbon filters, you would install a gas sensor measuring whatever chemical your trying to scrub. When there is a spike in that particular gas compound from your sensor. You know the carbon filter is saturated, and should be changed.
It looks like the the filter mentioned in the article video,combines all the filter products into a single product. Nice! If your air handler can manage the increased pressure drops, please buy the most legit filters you can buy. It's literally the cheapest insurance you can buy.
A laptop can connect to 27 inch (4k monitors), your prefered keyboard/mouse setup. It also comes with a slick integrated uninterruptible power supply (ie the built in battery).
Need to go on a business trip? Undock it. Need to take some work home? Undock it. Want to show a colleague in another building what your working on? Undock it.
If you had a dual computer setup (Desktop and Laptop), you're now in charge of making sure state/configs is mirrored across the devices. I've wasted hours of productivity from this issue. Drove to coffee shop 30 minutes away. Forgot to git push the repo on my desktop. Drove back, git push, Drove back. 1.5 hours wasted...
For 95% of use cases, laptops are pretty hard to beat. The outliers are AI/ML workloads, serious video editing, CAE software etc. Basically anything, where long-term steady state thermal management would become an issue.
I kinda wish the Macbook or Thinkpad-lines would support dockable compute, in a slick way. I know E-GPUs are a thing. But still find the packaging clunky. On the road (4 core i7, 16 GB RAM). Docked? Now it becomes Dual-xenon-like with 64 gb ram, 2x gpu. All state would live on the laptop. A pipedream? Yes.
But for using an IDE to write Javascript, Python, C++. A post-2012 laptop typically has the horsepower. I'm currently rocking a T430. I just wish it had a 4k touchscreen. Then it would be perfect.
Here's my thoughts/background:
Background: Doing small scale training/fine tuning on datasets. Small time commercial applications. I find renting top shelf VM/GPU combos on the cloud to be psychologically draining. Did I forget to shut off my $5 dollar an hour VM during my weekend camping trip? I hate it when I ask myself questions like that.
I would rather spend the $2k upfront and ride the depreciation curve, than have the "constant" VM stress. Keep in mind, this is for a single instance, personal/commercial use rig.
I feel that DL compute decisions aren't black/white and should be approached in stages.
Stage 0: If you do full time computer work at a constant location, you should try to own a fast computing rig. DL or not. Having a brutally quick computer makes doing work much less fatiguing.Plus it opens up the window to experimenting with CAD/CAE/VFX/Photogrammetry/video editing. (4.5ghz i7 8700k +32gb ram +SSD)
Stage 1: Get a single 11/12 gb GPU. 1080TI or TitanX (Some models straight up won't fit on smaller cards). Now you can go on Github and play with random models and not feel guilty about spending money on a VM for it.
Stage 2: Get a 2nd GPU. Makes for writing/debugging multi-gpu code much easier/smoother.
Stage 3: If you need more than 2 GPU's for compute, write/debug the code locally on your 2 GPU rig. Then beam it up to the cloud for 2+ gpu training. Use preemptible instances if possible for cost reasons.
Stage 4: You notice your cloud bill is getting pretty high($1k+ month) and you never need more than 8x GPUs for anything that your doing. Start the build for your DL runbox #2. SSH/Container workloads only. No GUI, no local dev. Basically server grade hardware with 8x GPUS.
Stage 5: I'm not sure, don't listen to me :)
He wrote a sub-40 page book of his experiences over the years. It's incredibly interesting to read some of his viewpoints and perspectives during his time.
There are nuggets of golden information hidden within this book. Such as talking to customers,product development,conversion rates, and having actionable data at your fingertips.
Some of his advice and experience truly stood the test of time. At the same time some of his predictions were inaccurate (E-commerce in 2000 is a fad).
I would go as far as saying that this is the pre-internet version of the Rie's Lean Startup.
Hope you guys enjoy it as much as I did.
https://docs.tryretool.com/docs/tagging-images-for-machine-l...
Don't worry about QR codes, it kills the vibe of an "organic" meeting. Just do it via GPS.
I wouldn't worry about liability, look at the physical contact Tinder is promoting :)
I think the rating should be on the group overall, and not on an individual level.
Negative individual ratings should only be done with consensus, if everyone says Bob was a bully. Then Bob's a bully. But if only Debbie says Bob was a bully, then maybe Debbie just overreacts.
If Debbie constantly negatively rates against the consensus, then Debbie probably isn't a good fit for group situations.
No profiles, no media uploads, no digital showboating. The best relationships I have with other people fall in this category. Real life,common interests,no digital uploads.
Just add attributes of your geolocation and hobbies [biking, yoga,tensorflow,etc] or let it crawl your digital profile(s) to autobuild your attributes identity. The system will group people with similar personalities/attributes. Every once in a while, "the system" it sends you a text "Free at 9am? Mountain biking at X park for an hour with beer after with 3 new people? Text yes to opt-in"
Get revenue by having businesses bid on being the beer spot. Refine the algorithm by having an anonymous rating system after the fact.
I like Meetup but it must be taxing to always manage the organization/logistics of a meeting.
Yes. I just realized I blended the show Betas,an episode of Black Mirror, and the 2016 movie Nerve together.
Edit: Grammar
Front-end: React SPA (Via Create React App),Redux, React-Bootstrap Why: I had no prior framework experience, single command setup, and plenty questions on Stack Overflow and Medium articles to learn with. Going the plain Javascript route with no frame work,seemed like I would be on a slippery slope for spaghetti code down the line. React's one-way data flow and component architecture made things incredibly easy to mentally digest. I found React's documentation VERY well organized and explained, and Vue's documentation was intimidating when compared to React.
Version Control: BitBucket Git Why: Free private tier for solo use and enjoyed previous experience with Atlassian products
CI/CD: BitBucket Pipelines Why: Already using Bitbucket, this just worked seamlessly with it. Only $10 for 1000 more build minutes.
Firebase: Hosting/Cloud Functions/Firestore/Auth Why: Everything is automagical HTTPS static website hosting for the create-react-app (This is what first got me started) Then came Storage,Firestore,Cloud Functions, and Auth. Wish it had an automagical SQL product. Low Entry Cost
Cloud Provider: Google Cloud Why: Had no experience with any cloud provider. I found the pricing/product offerings really easy to digest and interpret when compared to AWS.
Email Sender: Sendgrid Why: Good starting free tier. Decent Docs. Easy to setup.
Backend Compute: Google Compute Engine +Docker+Python-based API app Why: I can do local dev easily with Docker, push the image to Google Container Registry, then pull the image to the VM. It's easier to do than learning to setup a CI/CD piped GKE cluster Eventually I want to pay the "tax" go to a GKE CI/CD Kelsy Hightower setup,so I can git push/PR/merge a change, and it's in production. I don't do enough changes to the backend right now to justify a CI/CD piped GKE setup.
Other: Email Client: Gsuite Why: Custom Domain Name I'm already familiar with Gmail
Project Management: Trello Why: Integrates with Bitbucket Multi Device Support (Phone,Tablet,Desktop) Easy to mind map and organize features
What I value: Good Docs Free tier for solo devs/low use, and progressive options for paid plans (Gives me breathing room on cash while I figure things out) Lots of questions/answers on Stack Overflow or in-depth Medium articles.
The PDF is an excerpt from Shingley's mechanical engineering design textbook. It's is considered to be one of the Mech E bibles. If you like buying random interesting books, buy it.
Food for thought: Most welding applications that "normal" people would do would experience forces that range from 1lb to 2000lbs. (This a bad assumption, but whatever) If you actually do the math, it's even a small weld bead would have the strength of several hundred pounds force.
Another general rule of thumb is that the base material will fail before the actual weld, if done properly.
The easiest way to mitigate the risk, is to weld the entire seam, but you risking warping your part.
The other thing is making sure you have full weld penetration when doing the weld. Penetration is affected by a multitude of parameters(amps, speed, technique), easy way to check is to look at the backside of the weld. Harder way is too cut a cross section at the weld. Or if you're going to ball hard. X-ray it.
Have her program a solution that allows her to solve her own problem. For example, have her write a server script when a SMS image message is sent via Twilo, it will take that image and display it on your Raspberry Pi digital photo frame on your living room wall.
Little projects like that make in enormously easy to have momentum when learning for the first time.
Unless you're strapped for cash, I would avoid learning "for the market" at first. Since if you put React/Angular in front of my face when I first started programming.
I would say "What's the point?". I didn't understand the problem they were solving,until I had a problem that they solved.
The process that this mirror goes through, just redefined high stakes for me.
Think about it, if somebody or something makes an error with any significance. Bolt wasn't secured, sensor calibration was off, etc. The entire project is derailed by years. Just getting new material, will take at least a year based on the video.
In aerospace, usually it's alot of money to scrap a part. But then you call up your supplier, and they can make another one "quickly".
With this, you have no wiggle. At all.
I'm curious how they manage risk/failure modes on their team/facility/machines with a project so critical.
Especially since the entire project has a QTY output of less than 10 over a few years.
The telescope lab is directly under a stadium, their process for a single mirror is several months/years long. I wonder how they manage their process. Do they just pause and continue once the people leave? I also wonder how damaging the starts and stops (if they do it) are to the final quality of the product.
My overall impressions are good! The program completely redefines my thoughts for capabilities of DOM based applications and WebGL.
Pros: Use anywhere on any device(Linux)! (DOM Based) Free for personal use. Scripting based design automation seems like first class citizens with the part monitoring and profiling features Sheet metal capabilities are easy and intuitive. Something that I can't say for Solidworks. You can straight up delete things in Onshape, Fusion 360 makes that a bit hard. I like the fact that parts and assemblies are two explicitly different things.(Solidworks does this) It makes it easier for my brain to cut the "pie". Fusion does it a more NX'y way where "a part is a part/assembly". I don't have much experience with surfacing, so I can't comment about that aspect. Git like branch/merge functionality with parts.
Cons:
It ain't the whole widget.. At the end of the day it's "just" a CAD program. Which is fine, but I feel that most consumers are getting "finer" tastes with the advent of "the whole widget" design. Thanks Apple. (I don't mean that in a negative way either.) The thing that blew my mind about Fusion 360, is that CAM is integrated and just works, CLOUD Based FEA/Rendering just works, FEA Analysis it's there, Topological optimization there too. Rendering it's there also.
I see that on the apps developers on Onshape's store, are taking notice. Now you can download CAM add-on, render add-on, etc. I'm
scared that this is going the way of Solidworks/Jira/Confluence. Where you buy the base program. ($100 per month, then add CAM
/rendering/FEA separately). That $100 per month now looks like $200 per month. Now you are tied to vendor specific systems. Now you're
dependent on vendors for fixing their stuff when OnShape pushes an update. If the 3rd party vendor folds. Now all your legacy data is
effectively worthless. Or if there are problems, then 3rd party vendor blames Onshape, or Onshape blames vendor.
If Onshape makes the whole widget, then I have high confidence when an update is pushed, then CAM still works, render still works, and the UI is very
similar feeling. It makes me have trust that my "mechanical engineering workflow" just works.
When I use my mechanical engineering workflow. I want to be able to go from rough idea to CAD to FEA to render to 2d part drawings to CAM to "Part is
literally in my hand". With as little friction as possible. Adobe CC and Fusion 360 do this very well. I feel that with these industries where the data
model is incredibly complex. Own the whole pie is a requirement. Does that mean you need to lock-in your users? No. Give them escape hatches so they can use
familiar programs at whatever stage. Export as X, works well for this situation. But the main focus is making the UX/UI seamless between CAD/CAM/CAE. Fusion
360 in my opinion demolishes onShape in this regard.
Random side note. In the software engineering field, Companies that control the entire widget is "frowned upon". With good reason. The "data models" are
usually only text based (.C, .py, .js, .yaml,etc), and have open source standards associated with it. Meaning if you write Python in Visual Studio, the
same Python code will work in PyCharm. This is good. That means past the IDE stage, The files coming out are "clean and non-propriety", which allows you to
use any git-client you want. Git is open source and has a non-proprietary data model. So now you have the choice of using Github,Bitbucket,Gitlabs, etc. If
you don't like one, then switching is "painless".
Same goes for CI (Travis,Pipelines, etc,) To cloud providers (Kubernetes on Google Cloud, AWS, Azure etc). You're never locked in, since the data model at each stage is straightforward (text-based, that is interpreted to commonly agreed upon standards).
The CAD/CAM/CAE industry is nothing like this....which is why owning the whole widget, until/if common data models become prevalent.
Fusion 360 is free under ($100k of revenue per year), $300 a year for basic, and $1500 a year for Ultimate. (Basic will work for like 95% of users).
OnShape is free for personal/non-commercial use and everything needs to be publicly accessible. Then to go from personal to commerical/private. It's $1500 a year...just for CAD. Want the other features that Fusion 360 has? Open your wallet...
Fusion 360 in many respects, functions as a great drug dealer. They give you a little for free. They supply more value for you, as you need it. But the pricing is never so high , where you need to make hard call, and stop using their drugs. You grow with your drug dealer. Onshape is like "Here's enough for one hit", but if you want any more than that. Here's a brick, and I won't give it in little bags so you can give it to your friends at parties. You need to do it yourself....It's also a felony if you screw up when you try to buy in big.
Here's a good use case: I'm an at home inventor that wants to do a Kickstarter/Sharktank product. With OnShape, since it's commercial I need to front $1500 a year to solve only one piece of my product pie. Then more cash after that. With Fusion 360, I pay $0. If it takes off and I make over $100k. Paying $300 bucks a year is literally not a problem.
With this use case: OnShape isn't even on the radar for a CAD product to use.
But I always root for the underdog. But they have some work to do before they can have the same impact as Solidworks when came in out in the 90's. The DOM based app is really cool, and has alot of potential. But I don't plan on doing CAD on my Pixel2. The CAD/CAM/CAE workflow matters an order of magnitude more. Here's the cool thing. All of that can be computed in the DOM, so it can be developed down the line.
For year of 2015+ products, Fusion has the crown now.
Nice thing about those designs, is that you don't need to accelerate the motors/wiring back and forth. Giving better acceleration rate per unit of motor.
The downside with those systems, is that they tend not to be stiff (due to large amounts of belting), or react to external forces well.
That (H-bot/Core-XY) design would be poor for a convention CNC router, that has to react the forces associated with conventional machining.
But for this application, it doesn't matter. The forces acting on an extruder are basically non-existent, with the exception of (the mass of the extruder * its designed acceleration rate).
For going that fast real time, it might have some trick servos, or a hybrid motor (Search for Technic ClearPath Motors).
If you like that video, then you might like this one.
https://www.youtube.com/watch?v=NgGiu_0x7tg (Note: Not afflilated with that company, just like the video)
I like how they have a free tier for public projects. It lowers the cost of entry dramatically for people just want to "feel" out CAD. I remember when I was little, that I had to visit some sketchy websites to "try" out CAD. It's a great time to be alive now and play with things.
Just checked out their website. Linux support! I'm already 50% sold. My gripe with Fusion 360 is that their is no Ubuntu support. That and Adobe CC is the only reason why I haven't dissolved my windows machine.
In my experience, I found Fusion 360 to be a better experience. The ability to do parametric modeling, surfacing, CNC toolpaths,PDM data storage, "cloud" FEA solving. Everything just works. I don't need to manage licensees from multiple vendors. Set up runboxes for simulations. Manage my own PDM server (PITA).Or re-learn 3rd party vendor specific UI/terminology.
On boarding new team members to projects is as easy entering their email address.
Pricing is transparent,straight forward, and reasonable. All my data is auto backed up to "the cloud".
Free for personal use, and commercial use under 100k of revenue. After that the pricing is very reasonable and straightforward. $300 year for basic, $1500/year for ultimate.
The Ultimate version includes shape optimization, 5 axis machining toolpaths, and a few other things. If you had to piece that together from vendors. Your looking at $40k a year per seat. The value is incredibility hard to beat.
Solidworks does have the lead in having a huge 3rd party plugin system,weldements, wire routing. I did get thrown off from the idea of a "mate to the idea of a "joint". It forces you not to use hacky practices when building assemblies.
I feel that Fusion 360 has better product direction compared to Solidworks, and with their SAAS based model can update users more frequently, when compared to Solidwork's yearly basis. But honestly both products are VERY good in general.
Note:Not affiliated with Autodesk in anyway shape or form, other than as an end user. I just really like Fusion 360.
What if you fail publicly?
Who cares. As long as you do not have a big ego, or talk more than you walk (I'm looking at you Theranos). Then nobody is going to judge you. It's 2017, apparently starting your own business and "pivoting" (ie failing) is what all the cool kids are doing.
If someone puts you down and blindly says that you are going to fail. Then stop talking to them, and get them out of your life. They are probably too small minded to have a legitimate conversation with anyways. (At least in my experience)
What if you go broke?
That's fine. How much money does it take for you to be truly happy, with no external influences? For me, I need a workshop, a laptop,the gym, the local library, and my significant other. I can theoretically sustain that lifestyle working a minimum wage job full time. Not being spartan, just minimally comfortable.
So if I go broke, I know I have the skills and work ethic to make at least what a full time min wage worker makes per month. So in my mind, money doesn't matter.
(This changes a lot if you have kids. On the flip side, if you go broke your kids will have such an enormous respect for money, that they will probably be financially responsible when they are adults. In a really twisted way,you might be doing your kids a favor by going broke.
Opportunity cost?
This is the real killer. Let's say your 25 years old. You start your company and put 5 years into it.
Now your 30. Your business isn't a unicorn, but you didn't explicitly fail. So lets say your making $150k profit a year (non-passive), working 40+/hrs a week. Plus the weight of the business on your shoulders. You look at your buddies working at Google/FB/MS/AMZ, they are pulling in $250k+, working less hours with less stress, and they spend more time with friends and family. Your other buddy wanted to bring you into his early stage business when you were 25, you declined and still struck it out on your own. Your buddy just made a 7+ figure exit.
Did you make you make the right choice 5 years ago? This hypothetical question is what keeps me up at night.
Here's what scares the hell out of me. Once you lose time, you will NEVER get it back. You are that much closer to being dead.
How do I deal with it? Ask yourself this question. Am I happy(but stressed) right now? If you say yes, then nothing else matters.
When I was working for the Fortune 500's, I would look out the window and wish I was working on my own project with my own terms. That's what I'm doing now. After starting my business,I never for a second, wished I was working for a Fortune 500. I only wanted the Fortune 500 paycheck, not the daily work/routine/office politics.
Perspective is an incredibility underrated way to manage stress.
There is one thing that I would like to add.
4. Task making/Goal setting- The domain of this MUST encompass your entire life. (Personal/Professional/Health/Hobbies/Relationships).
Try not to put an emphasis on professional/business things. Generally those have the highest priorities, and don't need as much active attention in this context.
Goals should start at 5 years out and work backwards to next week.
Every morning/ start of each week/start of each month- review your goals and allocate "hard time" to them.
Want to learn guitar? Good. Tomorrow at 3 o'clock go to the store and buy one. Don't research the best guitar manufacturer for weeks and eventually buy it, just ask the guy behind the counter his recommendation and buy it on the spot. Next week, Tuesday 6pm- Sign up for music lessons for every Wednesday at 6pm for the next 6 months. No exceptions. Do not let anyone or anything take that time slot away from you.
But you may say, I'm the only one that can do "X" task and I'm needed at that time slot every week. Then build fault-tolerance in that task. Either with multiple people knowing how to do what you do, through documentation or automation. Engineer a solution to your time problem.
Or if you're the founder of your startup, and you can't justify taking a 3 hour break once a week.
You're lying to yourself.
I highly doubt you are completing work at 100% duty cycle if you are working 60-80 hours a week. Getting a few hours a week of "you" time each week is a small time investment, for huge returns in more effective decision making for the rest of the week.
This may vaguely fall under meditation, but knowing that you are tangibly and actively working towards things that you value, will prevent you from getting "time depression".
Yes, I just made up that term. "Time Depression" as I call it, is doing the "same" thing day-in and day-out. As a result, you get depressed since you "feel" like you didn't accomplish anything.On paper, you closed 90 issue tickets last month and committed more code than anyone else on your team.If you did all that, and don't have a feeling of accomplishment. Then you need "you" time.
Now go buy that guitar.
Tie in data from home via sense.com sensor, car travel data via google location services and a user selected vehicle.
Then give tips to user on better habits. Such as putting home in "eco" mode when you leave via Nest, or converting house to LED, etc.
Like if you change your Ford F-250 to a Civic, you will save $2000 a year based on your driving habits and distance.etc.
Or after driving your car. It says you spent about $10 in gas and 30 mins by driving just now.
The thing that inspired me was was the real time MPG gauge on my Honda Civic. It single handedly had the greatest MPG increase for my car. Since I felt the "pain" ($$) of smashing the throttle every time.
When people use their electricity at home, they don't feel " real time pain" when they turn on their 5kw AC unit or 3kw water heater. It's a crazy amount of power. They only feel once a month, when they get their energy bill for $200+ a month.
In conclusion, game-ify the hell out of dropping an end user's individual energy consumption number. By reducing energy demand at the source, then it's much easier to swallow eco friendly options.
For example, with poor energy consumption habits you may need a 10kw solar array @ a cost of 20k). But if you drop the user's average energy consumption. You can have the same quality of life at 5kw solar installtion @ $10k. Spending $10k is easier to swallow than $20k. (I made up all the $ numbers, I'm not sure where pricing is at now)
Well these are my random thoughts, I hope they inspire you.
Why I like it:
-CLI makes it easy to use.
-Deploying to the internet is as easy as "firebase deploy"
-Comes with HTTPS for free and with zero config
-Automatmagically uses Firebase's CDN
-Can serve with your custom domain name
-Low cost (for my needs/traffic levels)
-(Mostly)Straightforward documentation/references
Pain point(s):
-Slightly fragmented documentation. This was only one instance, so I might be blowing it out of proportion. I wanted to have Clean URLs in my website, there was no mention of that capability in the Firebase docs -> "Customize hosting behavior".
But the only instance I found this feature was from a random Firebase blog post from 2 years ago.
https://firebase.googleblog.com/2015/12/a-host-of improvements_61.html
-Deploys can be slow(30 secs -2 min), but honestly I should setup a CD pipeline and stop staring at it while it's deploying.
I originally was going to use Google Cloud Storage, but they don't support HTTPS out of the box.
To learn the actual "art of compute". Go from bare metal and then up. In your free time casually learn about logic gates, SR latches and clock cycles. Then to Registers, ALU, Instruction Sets and pushing/popping things on stacks. Then to memory mapping and data buses.Work your way up to the C level of abstraction.
It is a long and brutal road, but it gives a great CS/EE foundation.
I am recommending this route since when they talk about pointers, references,heaps in the C world. It will give you a tangible foundation to built these abstract ideas on.
I vaguely learned the "art of compute" from college via Digital Logic, Microprocessors, and other related classes. Since your in college, I would just sit-in the EE/Comp Sci classes. It is much easier to learn from lectures when you don't care about tests/notes or labs. Then use the internet/office hours to supplement your curiosity/questions.
In conclusion,I recommend casually learning from bare metal and then go up to C then Python. But in terms of actual programming, do Python and work your way down to the assembly level as needed.
I like Python since their are "defacto" ways of doing things. Matrix Math -> Numpy. Lightweight HTTP framework -> Flask. Everything-include web framework -> Django. Web scraper->Scrapy. Need packages -> pip. I know there are other contenders for the things I mentioned, but based on my perception, this is what the Python community is rolling with.
I wanted to recommend doing Javascript/HTML/CSS. It's really cool to work on a project, then publish it on the internet for the world to use/see. But things change so fast, it's hard for a beginner to "sort" through what's needed/useful. React or Vue? Grunt or Gulp? Packages npm,bower,yarn? Async/Promises or callbacks? Don't forget about webpack. Ember or Angular? But wait. There is Typescript,CoffeeScript,JSX which needs to be transpiled. ES6? Here comes Babel. Want to make dev process quicker. Start hot reloading.
The point is,it is a little tough to "sift learn" when picking up Javascript and friends.. The community is moving fast, but it incredibily exciting to watch/participate. But I don't reccomend starting with it.
If you really want to learn HTML/CSS/JS. Start at the Mozilla Developer Network and use vanilla JavaScript. Ignore all the trick frameworks and tools until you understand the problems they are solving. Use them only if they solve problems that you are having.
To go from Python to C. Write some Python app. Profile it. Then write optimized C functions to replace the slow parts of your Python program. Make a Python wrapper with SWIG. Best of both worlds in my opinion.
Your end user is the 6th grade version of myself. (Curious and interested in computers but no mentors/guidance to get started/lay breadcrumbs). My family/friends came from blue collar background, so their programming skills are effectively zero. Asking for help doesn't really exist since I didn't know anyone to talk to or know where to go on the internet.
When I was much younger (10 years old), a big part that delayed me from getting involved in programming, was not understanding "the point" of programming and figuring what to search for when trying to learn.
I (10 year old version) seriously thought that all programming was: cout << "What is your name?"; cin >> name; cout << "Hi << name; return 0;
Based on the C++ book I had access to, I didn't understand what I could tangibly build from knowing the idea of int,strings,arrays etc. I also thought that when people made games, they hardcoded the raw pixel values for every possible state/character/action. (Dumb, I know.)
I also thought you programmed "cool" games, by throwing a stupid amount of time in hardcoding each pixel value.
So I paused on programming for 10 years and picked up paintball/cars/Mechanical Engineering instead.
Make and install custom Ubuntu image with all of the needed packages/software to do introductory tangible dev work on. Make it work for 3 large dev camps(Python, Web Dev(Basic HTML,CSS,Javascript),Embedded electronics (Arduino). Sure the 10 year old kid can install everything via command line scripts themselves.
But knowing that you need a './" in front of 'script.sh' isn't entirely obvious at that age. It is also incredibly demotivating when your getting your feet wet, and can't figure the './' out.
Have links on the Desktop to the Mozilla Developer Network, Python Docs/ Game Tutorials, the github explore tab, Intro to Python/Web Dev online courses and StackOverflow. As well as a challenge for the 10 year old: Like make a pong-like video game.
The mission statement for this entire idea, "We gave you all the tools and they work well. Now the only thing stopping you from being a hacker/programmer is yourself."
Distributing it via old hardware would be nice, because most kids/parents are worried about "ruining" their existing computer by migrating to Linux and losing Windows.(At least I was)
Having a separate dedicated physical box, will give the kids a good environment to try things and fail in. Without worry about accidentally deleting everything on the file system. Like family vacation photos.
Do FB marketing and target parents that have kids in middle school and want them to go college and be doctor/lawyer/engineer types.
Sell it at (cost + $20 +shipping). Plug and play.
I probably don't fit as your target customer, so please take my feedback with a grain of salt.
After digging around your site for a bit, it seems like your "special sauce" is to bring incredibly powerful search features for multiple data types/sources all from a single authoritative place (Shelf.io)
Based on your YouTube videos, I really like the ability to parse rich data on the internet and input it for my team to see. Seems like a great product marketing/engineering folks during the high level brainstorming phase.
Use case: 1.Software Engineer: Reads a cool article on Medium about some software technique and wants their team to consider it for their their new beta project.
2.Marketing person: Gets inspired by a video/image for a new ad campaign that they are designing and wants to share it with their team.
Also as an end user, I really like being able to see your actual product and how it works without needing to signup or give out my information. If the product seems like it can solve my pain point within 45 seconds of an intro video and 5 mins of playing with a demo site, then I'll commit to a email signup. Basecamp's "How it works" tab does a great job of this.
On the flip side, I feel that Salesforce doesn't need to do this, since they already have market share and a great reputation for their products.
I would take your "Intro to Shelf" YouTube video, cut the first 25 seconds out and embed it somewhere convenient on your website.
Highlight an example use case through a short video. I think Grammarly does a great job with this in their "new social media manager video" FWD TO :15 seconds https://www.youtube.com/watch?v=Ak-Y56SfkS0
Also for your voiceovers in the Youtube videos, I would invest in a good audio recording mic. The audio seems a little distant,and hollow. I use a Zoom H5 recorder, it is a decent product for what I need/require.As a lay audio person, This is my reference point for good voice audio. https://www.youtube.com/watch?v=hUhisi2FBuw&t=15s.
I know you guys are in start up mode, and need to ship. But as random end user, this is what I'm perceiving. Hope this helps! Good Luck!