AntiERP

I focused on coding the past week. It seems I finally love results: no intentions to rewrite everything, and I'm very keen to open source the code. Maybe soon.

Remember a friend who may join development in his spare time? We had a few conversations introducing him to the current project state, and he had a chance to contribute to the codebase to get familiar with it. No help from him yet, and he spent a little of the time (which is understandable).

Also, I asked a designer I know to design some UI for the project. He did a few sketches. I'm not happy with the results. Useless crap, which is not what I'm dreaming.

So, I'm continuing my solo journey.

Since I'm the only developer on my project, and I never did front-end in this Millenium (besides some JavaScript back in 1999), only touched Vue.js briefly. I decided to visit Vue.js and Frontend Developer Love conferences in Amsterdam to understand better what's going in this space.

So I did it, and I think I will never do it again. Do not get me wrong -- Vue community and ecosystem is great, and Vue.js is an excellent piece of software I'm happy to use.

However, I'm much disappointed with the quality of the talks at these conferences. Sometimes misleading, incompetent, telling false facts or bold statements on controversial things... But in general, they provide zero information on subjects. So I skipped the last day, deciding it's better to work and/or rest.

One good outcome of the conference is I learned about the existence of AWS Amplify. In some parts Amplify compete with a framework I'm working on. I shall take a look at it, but I do not expect any interesting findings there, according to what I hear in the talk.

Another nice thing was to see Vue.js creator Even You live on stage. Other than that, Amsterdam conferences are crap and a waste of time.

Cheers.

My rant a week ago was mostly about the premature perfection of things (code design, UI design, UX, etc.). It is not how people build products these days (at least not in the mainstream). They go with MVP and sketch out things to ensure they work and serve users' needs.

However, in the field of ERP, CRM, and other types of business applications, I would not succeed building yet another CRM system MVP and final product. I'm going to build great technology underneath, excellent user experience, and impressive design; and this is the first and most important product I have to create in my journey.

So my MVP is not a business application(s), but MVP of technology, UX and design. That's what I'm focusing on, building MVP of underlying technologies and concepts. And I'm already the first and very demanding user of this MVP.

Once I "sell" MVP to myself, and I will be convinced that this will evolve into a great foundation for business applications, I will go forward with technology and applications on top of it.

So, I'm still working alone, did not work on any UX this week, but I have good progress with the Core Framework design. I did throw out 90% of the things I made since the project start (actually, I'm throwing out 90% of code every week :)... but, finally, I feel I'm happy with the result.

It's the second month of product development and I feel I'm going forward into well-known pitfalls.

First, it's perfection. I would like to build a solid foundation for my code and I'm rewriting core again and again. I know every version I'm throwing away is designed well enough to work for years. Also, the concept does not change, and I'm walking around details, and I know many developers would accept each version well.

Similar to premature optimization (which is considered as evil among most of the engineers), premature design perfection looks like evil too. But I can't resist it. Partially, this is because I have no deadlines. On the other side, I also know I can meet a reasonable deadline I set, but I would not be satisfied deep in my mind. You have to love your product from every perspective. You can easily lose motivation by doing sloppy work.

I believe collaboration with other engineers would solve this problem. They would argue that the current design is good enough and it's hard to make better. Or they could propose changes for better, also people who use your code would stop me from rewriting everything as well as naturally set deadlines because I should not block them.

So I need to find an engineer to work with me as soon as possible. It should be an experienced one as well, a newbie would accept everything I deliver, and I do not expect any good improvements from a junior.

An engineer I mentioned on Jan 25 so far not found any time and contributed nothing. My goal is to solve this problem as soon as possible and find a good engineer for my team of one.

The same perfection thing happens on the UI/UX side. I know people do not care much about design for product prototypes. I mean they are OK if this is not world-class top-notch design quality. By world-class, I mean better UI design than most of the well-established competitors have at the moment. I should not care about this either, but deep inside I'm not satisfied. I want to see how best UI and UX emerges together with the code. That's a problem as well.

While I was in my previous business it took me three years to find a great designer, and when I lose him, it takes three years again to find a great one. Any I need a designer now :). Perhaps I will try to contact good ones on dribbble.com upcoming week.

Also, there is an option to hire a leading design agency from Silicon Valley, but their prices start from 100K USD per project, the upper amount is unlimited. I do not have experience working with such agencies and I'm expecting to work closely with a designer at least for a month and getting back to design issues constantly with small iterations. Full time until look-and-feel settled down, and smaller constant iterations while product evolves.

I found a couple of agencies which work increased my heartbeat, https://clay.global, and milkinside.com. Maybe I should contact them as well. But if I would love to work with one of them, I should secure funds also, which may be not a fast process either.

So, things above distract me a little. Of course, if I had a novel idea, I would try to build a prototype fast and validate my idea with real people. But I'm going to play in an established market. In the beginning, people will see my product as yet another CRM (or ERP). Similar to hundreds of ones existing on the market. And this would not change if I bring really next-generation stuff on their table. Masses are blind to innovation until they try and understand it. Until an innovative thing goes into masses.

That's why I should come with a polished and exceptionally looking product in the beginning. I'm going to build Ferrari and McLaren of CRM/ERP, and you know these cars look great and different from mass-market cars not by accident. Their exterior reflects great technology inside. I shall do the same thing too. That's a challenge.

So I need an engineer and designer to work with me, and I need them asap :)

I back home from http://fosdem.org, disappointed. Technical talks were boring and useless. I did hope to learn about new developments in tooling, languages, frameworks, or find some emerging tech, which may be helpful to stay on the bleeding age. Big things often come silently, and almost nobody knows about them in the beginning. I discovered eclipse in 2002, years before the first EclipseCon, a platform unknown to the masses those days, and built good business on top of it. The only thing I enjoyed at FOSDEM was meeting my old friend Andreas Rumpf - creator of Nim programming language http://nim-lang.org.

From a business development standpoint, I expected a lot from 'Why the GPL is great for business' talk by Frank Karlitschek. I want my ERP product to be open source and thinking about the model. I do not know yet if it will be Open Core or fully open-sourced under EPL (a love its commercial-friendly nature). But open-source products facing some threats from cloud providers. A cloud provider can take your product for free and distribute it to customers. So all potential revenue streams may bypass you and go to a provider who sells and support your product.

That's why many successful open source products such as Redis or MongoDB changing their open-source licenses, to protect them from AWS, Google, etc. not allowing them to sell their product as a part of their cloud offerings. As for my EPR, of course, I do not expect Amazon or Google would be interested in selling my product. However, some reputable business infrastructure providers can do it. Open Core model as GitLab will solve this issue. But the Open Core model limit benefits for the end-user I want to achieve with being open source. I want end-user to be able to customize any parts of the product (despite it's doable without full access to sources) and make them more confidently rely on the product in mission-critical tasks.

No doubt, FOSDEM speaker has a kind of "GPL brain damage" syndrome. He has the only message during the talk: "GPL is good, everything else is bad." As for Redis and MongoDB license changes, he said it's terrible and makes their licenses "proprietary." What? Redis Commons Clause only restricts cloud providers from gaining commercial benefit from software that was not developed by them. Indeed, from OSI perspective, new Redis and MongoDB licenses are not open source licenses, since an open source license cannot include limitations. I'm not a lawyer, but do not open source licenses define how people can use source code, including restrictions and obligations?

Anyway, if OSI (currently) does not consider such licenses as open-source, I think it's too strong to call them "proprietary." You can name these licenses as "sources available," etc., but not "proprietary." Frank Karlitschek argues that the Commons Clause is terrible because there is an article on the Internet, which tells that Commons Clause is good. And the author of that article is... prominent investor. A guy who makes millions from investing in software writes that Commons Clause is good, and he is an evil capitalist who acts in his interests! You should not believe what capitalists say, because they exchange our freedom for money.

So this talk did not add anything to my thoughts about protection from cloud vendors, besides being advised that I should not protect because it's not open-source, and I should not exchange the Freedom for money. OK. Also, I wanted to hear something about license choice. As I mentioned above, I tend to license under EPL, but GPL is also on my radar. Frank brings some points on why GPL is good, but those points are also pretty valid for other open-source licenses. Finally, he said that GPL is excellent because "developers love open-source, and I will have benefits in hiring ones." I hired at least 300 engineers during my life, and working on the open-source project is far from the top developer's interests for sure.

The final question from the audience will summarize my rant. Someone asked what Frank will do if Amazon takes his project and sell via Amazon Cloud. Frank's answer was like, "Um... If we become big and Amazon will be interested... Yes, they can do it... But... Hmm... We have a trademark, so they can't use our trademark... Hmmm... Of course, they can sell it under Amazon Supercloud name or whatever... Then... Good question... I do not know, but we are not big enough to think about it..."

Think Big.

Happy day! I talked to a great engineer I did work with in the past, and he is joining me to work on the project in his spare time.

It's hard to work solo. You need to share ideas, you need another point of view. And two extra hands are very helpful.

Hope we'll go forward to success together.

Past two months I learn technologies, which I plan to use for my project. It is nice to program in TypeScript and I'm excited with Vue, HTML, and CSS. Last time I used HTML and JavaScript was 1999 😀. Web technologies made a giant step forward within 20 years. I also played a bit with MongoDB. Looks nice!

I think this is a great technology stack to build my project.

You know it's impossible to think about the future when depressed. But I did recover, and it's time to start something new.

When I was young in 1995, I made an ERP system, which I did not push to success (I got a job). But I never stop thinking about to build an ERP product.

These days, an ERP system is not something that engineers happy to build. It sounds boring. But not for me. It's my dream, and with a new decade just started, I'm starting my journey...