Hightlight the value proposition of the starter-kit

[japf]

I think the current documentation currently lacks explaining the value proposition of the starter-kit compared to going straight with the anguar-cli. In the README, we currently have:

Web project starter kit including modern tools and workflow based on angular-cli, best practices from the community, a scalable base template and a good learning base. Benefits

• Quickstart a project in seconds and focus on features, not on frameworks or tools
• Industrial-grade tools, ready for usage in a continuous integration environment and DevOps
• Scalable architecture with base app template including example components, services and tests

I think it would be a nice addition to clearly explain the advantages of bootstraping a new projet from the starter-kit instead of the angular-cli. Back to the Angular 1.x world, because there was no cli so everyone was looking for a starter-kit to speed up their development. Now that the angular-cli is out, I think we should explain why using a starter-kit still makes a lot of sense.

Ideas?

[sinedied]

That a good idea, but finding the good words for it is a difficult task ^^

The ideas behind the 3 points I wrote initially (not the cleared way to say so, I know...):

Quickstart a project in seconds and focus on features, not on frameworks or tools

=> Provides a ready-to-use environnement with all frameworks/tools needed to start development (= no need to add other libs or tools to begin), including UI framework, i18n and utility libs. Documentation with best practices and coding guides are also provided, along with an interactive documentation system.

Industrial-grade tools, ready for usage in a continuous integration environment and DevOps

=> Linting is improved (+ HTML + SCSS) and stricter than angular-cli, tasks are ready for CI with unit tests and coverage reporting and a PhantomJS runner (vs Chrome and no preconfigured UT reporters for angular-cli). Additional tool for translation workflow is also included.

Scalable architecture with base app template including example components, services and tests.

=> Comprehensive examples for common use-cases are provided, along with best practices, so beginner (and experienced) teams have a base learning ground to properly develop their app.

All of this could be used I think to build a better value proposal, but I agree that the current text is not very clear/attractive vs bare angular-cli. I'm open to suggestions here :smile:

Maybe a straight bare angular-cli vs SK comparison table?

[dancancro]

Hi. Great project.

Is your end goal more to increase the marketshare of this project specifically or to foster convergence of the many starter projects out there? I would be happy to help you with the latter.

[sinedied]

@dancancro As for convergence/extensions plan you should take a look at https://github.com/angular-starter-kit/generator-ngx-app/issues/33

The primary goal is to make a base web/mobile angular stack based on Bootstrap or Ionic 2 (maybe others later), then allow to customize/complete this base with addon generators, as no starters can fit all the needs out there 😀 I can't maintain a starter with all the existing tools/frameworks on my own, but if we split the job we can definitely rationalize the many starters out there, especially when the addon generator feature is done.

[sinedied]

Ok since I'm getting tired to explain on reddit the differences with plain angular-cli, the readme should definitely be updated with some additional info ^^

Here's a copy of what I quickly replied on reddit as a reminder, the form needs to be improved a bit:

• documentation, coding guides for TS/SCSS/HTML, Angular onboarding guide
• improved tooling: SCSS & HTML lint, stricter tslint rules, markdown-based wiki local server for docs, i18n out-of-the box with dynamic language change
• template with example app structure tailored for scalability, with examples and boilerplate code for everything: unit tests, directive, components, routing, authentication, http service extensions...
• internet proxy support (almost always needed in enterprise environment!)
• API proxy support to easy development using a remote server
• add on support to plug in your own generator additions: your enterprise theme, SSO authentication, backend, whatever your need

[sinedied]

I made a quick readme update on the readme to go along with the new major release, what do you think @japf ?

@dancancro I know it's been a while, but if you're still up to it now that the foundations are ready (multiple targets, UI frameworks and add-on support) I would be glad to work more towards (Angular) starter kits convergence 😉 Do you have any specific thoughts on this?

[japf]

Like it a lot ! I think it greatly showcase the advantages over the bare Angular Cli, great job !

[dancancro]

Hi @sinedied

I like the new docs. Sure, I have some thoughts. And some solutions.

---

1. Tech Tradeoffs Database

• This is a project I started about 4 (or 14 depending on how you look at it) years ago. It is a collection of technologies, their selling points and dependencies. It contains over 1400 standardized technology selling points and 500 or so technologies all with citations. From this data the system will compute a given technology's full complement of selling points introduced by itself and by inheritance from its dependencies. You can then view reports to compare things like this side by side, or like this side by side by side.

Solution No. 1: If you would like to clarify the advantages/tradeoffs of ngx-rocket wrt angular-cli

Add information about generator-ngx-rocket to this spreadsheet and then produce a complete comparison between it and others. You can begin by filling in [this]() little form and I'll give you comment permission to the spreadsheet.

---

2. Great Big Example Application

• This is a full-stack Angular/Java Spring Boot example application generated by and updated with JHipster. I've been working on it full time for about a year by incorporating into it the best practices of all the best starter projects I could find. It has 150,000 lines of code, 200 unit tests, 16 main routes, all of these selling points, demonstrates best practices found in several small, authoritative demos and harmonizes all the disagreements in approaches. This was no small feat, but mandatory for anyone who wants to make a commercial-grade app from lessons learned from tutorials. I spent very little time on new functionality beyond that of the source demos. The rest was spent just on negotiation the different opinions of them. These are tutorials written by experts with the same framework, Angular, and it still takes forever to reconcile them. That's how hard differences in opinion make life and why reducing fragmentation is so important. To my knowledge this is the biggest, most orthodox open source Angular example application out there. I honestly did not want to have to make this project but I just could not get bloggers, educators or the Angular team to work together. I don't understand why Google hasn't produced and open sourced something like this, but they haven't. They test Angular releases against hundreds of sophisticated internal applications but zero are visible to the public. So I had no choice.

Solution No. 2: If you would like to accelerate starting point convergence

You're not gonna like this one, but...

Work with me. Instead of continuing to develop ngx-rocket separately, help improve my project (made from genius demos built by other people) and JHipster by working on its task queue or other important enhancements, features, documentation or innovations delivered by ngx-rocket. The more consistent and objectively superior a single option is, as measured by its benefits, the more people will invest in it and contribute to it instead of branching off in more and more and more and more and more different directions.

[bursauxa]

Here is my take on this issue (it represents my personal opinion and should not be considered as the project's official stance):

• the original issue was a very real issue, and it was properly addressed by @sinedied's documentation effort in 9a8ba6f ;

• the first point raised by @dancancro, about his tech tradeoff database, seems out of the scope of ngx-rocket. I do not mean that it should not be present in the database, or even that people from the team should not be the ones doing it. What I mean is, the absence of ngx-rocket in that database is an issue with the database, not an issue with ngx-rocket ;

• the second point about stopping development of ngx-rocket and instead integrate into the Great Big Example Application does not align with what I am looking for in a starter kit. I am very little faith in full-stack examples, as there are too many technology combinations out there. Focus in Angular feels like the better thing to do for us.

Any and all comments are welcome :open_hands: