4 Comments

Launched landing page for a programming language that we are building, looking for feedback :)!

by Martinsos

We have just launched our landing page - https://wasp-lang.dev - pls give us feedback!

Target are developers in general, especially web developers.

I will say nothing more so as to not influence you!

Martinsos

posted to

Landing Page Feedback

on February 10, 2020

Say something nice to Martinsos…

Post Comment

3
A few other ideas off the top of my head:
1. Develop your own cloud platform to conveniently release wasp apps to. Providing everything (deployments, monitoring, analytics etc.). Make a CLI for that. Make the money this way.
2. Make some styling framework a first class citizen. Like tailwind or whatever. So that its classes are natively supported and maybe even type checked in the source code.
3. Auto generate docs for the JS source code.
4. Auto generate Swagger specs for auto generating API clients in other languages.
5. Get rid of React smell (like where does this.props... come from with "Main"?) or stick very strongly to it. Since you want declarative I'd say get rid of it, it's unnecessary except for maybe JSX.
6. Create a CLI tool that "just works" and handles the full life cycle of the web app. Bundling, deployments, health checks, fetch analytics etc.
7. Generate Typescript. Shouldn't be too hard since it's basically declarative (and thus immutable) anyway.
8. Compile backend code to something really performant if you can. Like Rust or something. Or even Haskell which I assume wasp is written in anyway. Show benchmarks and then brag, proudly.
9. Profile wasp for some specific industry. Like the media industry. Build features to support adoption there (ads etc.).
10. Make wasp somehow modular. A huge factor in React's success is component based modularity. Full blown frameworks are less often capable of allowing such modularity, which may hinder the development of the ecosystem.
11. Explain it very clearly why it is necessary to learn a completely new language for web development for people who already know JS.
12. Strive to make wasp "the new, better WordPress". May sound lame but WordPress has factually been a huge success. It's aging and needs a modern equivalent for SPAs and for more interactive websites. How to do it? Make a web app generator app. A web app which generates a ready made (wasp) web app. API for custom themes.
hopia

·
6 years ago
·
Reply
1. 1
  Thanks again for amazing feedback and all the effort you put in it, I really appreciate this :)!!
  
  This is one of our ideas for monetization, so I certainly agree :)! We also plan to offer easy deployment to other platforms, but we are thinking, if we make our platform optimized for Wasp built apps and full-featured as you mentioned, it could provide enough value to be monetized.
  
  Yes, that woud be great! We thought about that, but have not put much investigation into it yet. I will take a look at tailwind!
  
  Auto generated docs -> yes, certainly! Also, clean code with comments when needed.
  
  Swagger specs for auto generating API -> hmmm interesting, we thought about generating API directly but this would be very interesting, I will take a look into it. We can certainly do many different interesting things here.
  
  this.props -> Ah you noticed that :D! That is our next thing on the TODO list, we just did not have time to remove it yet. Plan is to get rid of strong dependencies like that, and if we can't get rid of them yet, then make them very explicit (which in this case we did not do, so it is ugly / smelly, for sure - "where did that come from?").
  
  CLI tool -> Yes, certainly! That is one of next steps also, and really important for this seamless experience that we are aiming for.
  
  TS -> it should be even easier long term for us, since we can be more certain generated code works (due to types), but I have to admit we are both more experienced with JS so that is why we are pushing with it for now. If people show enough interest in Wasp, we will certainly at some point switch to or add support for generating TS code (and also other tech stacks).
  
  Yes, Wasp is written in Haskell :). That is interesting idea, although for us right now it is most important to generate backend in technology that users use -> if we generate it in Node, they can extend it with javascript. Also, if they eject from Wasp, they can continue with Node. And right now I don't think Rust or Haskell are popular enough to be first choice for us. Nor is performance so important on the backend for most web apps that it should be the first thing we optimized. So for the start I think we should go with whatever is most popular (node, python) and then we can add other languages + optimizations. And brag :D.
  
  Interesting, well that would make sense from the "niche to win" standpoint. Although we are not yet there, I mean ready for high lvl features like that, but when we get there, this would be a good idea, I agree. Let's see what will be the best direction.
  
  Modularity -> I see what you mean. React -> yes, modularity, and I believe simplicity. I also love using languages / tools that have only very few basic building bricks but you can do a lot by combining them. However, since Wasp is DSL, whole ideology is somewhat opposite - big choice of premade bricks, which somewhat stack together, but you can't do anything you want. And that is where the power comes from. I would love to also have this composability in Wasp, on the other hand. Right now we have this option to mix Wasp with external code (js/react, css, ...) but that is really extensibility, not composability. Certainly something to keep our eye on as Wasp evolves, I would love to recognize some parts that we can make composable like that.
  
  Explaining why learning a new language is necessary - hmmm, I was hoping that promise of less lines of code and a lot of stuff done for you would be enough :D? Also, since Wasp is declarative DSL, learning it should be relatively simple, since concepts will be mostly horizontal and language will be very close to the web app domain. Do you think explaining this would help? Maybe you could help me here and elaborate a little bit on what you had in mind?
  
  Hah, funny that you mention this! Sometimes we do explain Wasp to people as "Wordpress for web apps", but we did not use that too much exactly because of that - it did not sound very exciting. But yes, conceptually Wasp is "Wordpress for web apps". Ok, although it is language, and not visual editor, but then we can build visual editor on top of Wasp (also possibly a way to monetize).
  
  I am still amazed with amount of super high quality feedback you gave us here, and I feel your thoughts are actually very aligned with our ideas! That is very encouraging :). This is great advice and we will take it into account for further development for sure. I can easily say this is most detailed and thought through feedback we got so far, I feel like I am talking with somebody who has been working on Wasp with us for at least some time :).
  
  One extra thing: we are now trying to get email subscribers, github stars and build community, to prove that there is interest for what we are building and to get things rolling. I don't know if you left a star or subscribed to email list, but if you didn't (which is completely fine, you don't have to), it would be great to learn why not, and if there is something we could have done to get you do that but we missed the opportunity.
  Martinsos
  
  ·
  6 years ago
  ·
  Reply
3

I think you need to refine some clear advantage, like THE reason to use .wasp over other choice available.

So brag with something in comparison with other web development platforms. It could be performance, smaller bundles, type safety, or less lines of code to achieve the same result. Show some visual proof to enhance the message. Like code comparison, benchmark, bundle size graph etc.

The obvious competitor you'd need to pit this against is React, sounds like that's your primary audience anyway. So you need to come up with an angle with marketing that appeals to that crowd.

I know it'd be massive effort, but another choice would be to take the abstraction level even higher. Basically approaching a low code or no code solution to build a web app. That would make web development accessible to a much larger crowd - a major achievement by itself.

hopia

·
6 years ago
·
Reply
1. 1
  
  Hey, thanks for great feedback :)!
  
  Defining advantage - certainly, that is something we can emphasize more. Right now we have comparison of lines of code (wasp vs generated code), but we could certainly have some stronger indicators! Type safety and less lines of code are on the spot. From less lines of code also follow easier maintenance and faster iterations.
  
  While React certainly is competitor in some way, it is also an important part of Wasp (since you define custom components in React and plug them into Wasp), so angle we are thinking about is not presenting Wasp as replacement for React, but for all the "boring" parts of web development.
  
  Taking abstraction level higher -> maybe that was not obvious, but that is the plan. We are still very early so that is why we have only very basic stuff, but plan is to introduce higher lvl concepts like login, deployment, navigation and what not, so that it is easier and easier to build a web app. Maybe we should make this vision clearer on the landing page, I don't think that was very clear right?
  
  Martinsos
  
  ·
  6 years ago
  ·
  Reply