Ask HN: How can I ship a web project if I'm not a web dev?

15 pointsvarlock5y ago14 comments

I'm not a web developer - more of a back-end one - but I had an idea for a side project that requires a dynamic web UI which I though was worth pursuing, so I've embarking on a journey leveraging ReactJS.

I managed to slowly and painfully create a MVP (very minimal in fact), and many features are lacking - such as, login support, automatic deploys, etc.

It's all fun and great source of learning but I have to admit that, at this pace, it will take me ages to complete it - and even then, most likely never at the standard of a web dev.

At this point, what are my options (beside ditching the project altogether, of course!)?

14 comments

quickthrower25y ago

Probably find an open source project that has the stuff you need as a basis. There are even templates for sale to get you started too. Choose something familiar language wise. I’d try to avoid JS frameworks if you are not used to web dev. Back end, render html, use a tiny bit of JS maybe for form validation. Use bootstrap - I am recommending because it will have the most documentation and stack overflow coverage which is what you’ll need as a newcomer who wants to get something done quickly. Consider hiring for odd jobs where you are stuck.

mrkwse5y ago

I'd second this and also add that there will be plenty of good tutorials going through a lot of this stuff. They'll be very opinionated and produce particular solutions, but if this is your first project you don't want to worry about which framework etc. to use, you just want to ship something.

Netlify produce a lot of very thorough blogs that will cover a lot of the things you're yet to sort out. You'll end up in their platform as a result, but it shouldn't be too difficult to figure out how to migrate away once you've got something up and running (or you might find that Netlify fits your use quite neatly).

varlockOP5y ago

I started off a template, which took away the need to think about web design. For React, I did follow quite a few tutorials - from how to route through pages, to hooks to unit tests - but the process is quite slow (though I do enjoy it).

And then there's setting up gitlab pipelines, docker images, deployments to AWS... it's a lot for a one-man band with only a few hours to dedicate.

I'll have a look at Netlify, maybe it could take some of the above away. Thanks for the suggestion!

cercatrova5y ago

Read up on Pieter Levels. He has a shitty PHP app, by his own admission, that makes around 1 million a year, NomadList. It's about whether you can solve people's problems, not what the tech stack looks like.

There are also no code tools like WebFlow, Bubble, etc, and you can also learn more frontend by doing more sample projects.

p0d5y ago

Your best option is to be more confident.

I am 50 and do not call myself a webdev or programmer however I have made a six figure sum with my side projects down through the years.

Most webdevs will never make anything they can call their own. They will make for others. Make something and put it out there.

varlockOP5y ago

Thanks for the confidence boost! It's always too easy to self-criticize.

dchest5y ago

The best cure against doubts is shipping as soon as possible, and then slowly improving what you have. Lots of successful software began as an incomplete and ugly version of what they became. So, solve your initial issues as quickly as possible to not lose motivation (e.g. plug an auth library that you don't have to spend too much to write code for, use Heroku/Digital Ocean App Platform/Vercel for deploys) and ship something. After that point, you officially become a web developer who adds features to their product :)

martinopp5y ago

I think you basically have two options if you think there is still a lot to be don before launching a good MVP:

Option A) Using a no-code tool as buuble.io, Webflow, or similar. The choice depends on your requirements.

Option B) Hire an agency or freelancer. Probably nearshore or offshore to maximize your investment.

Both are great ways of validating an idea before actually investing tons of money and time.

varlockOP5y ago

I wasn't aware of neither bubble.io nor webflow. Good to know they're there. Given what I've got in mind seems more to be an app rather than a website, I'd probably give a shot to the former.

thedevindevops5y ago

Do you feel there is no value in converting this project into a high quality documentation API and letting the creative minds that consume it do the UI?

varlockOP5y ago

Maybe not, as I'm actually leveraging existing APIs, so effectively it'll be a wrapper around those.

gostsamo5y ago

Use for the backend something that gives you most functionality as a plugin. Ruby on rails or django. Don't reinvent the wheel. Add js on the frontend as needed.

varlockOP5y ago

In the beginning I actually thought about Django or Flask (my background is mainly in Python), which I wrongly dismissed because I thought the main part would have been JS - so I chose ReactJS given the wide availability of tutorial and support. Flask would have solved some issues such as login etc. As they say, hindsight is 20/20 :)

fiftyacorn5y ago

Use next.js it's simpler than vanilla react

j / k navigate · click thread line to collapse

14 comments

quickthrower25y ago

mrkwse5y ago

varlockOP5y ago

And then there's setting up gitlab pipelines, docker images, deployments to AWS... it's a lot for a one-man band with only a few hours to dedicate.

I'll have a look at Netlify, maybe it could take some of the above away. Thanks for the suggestion!

cercatrova5y ago

There are also no code tools like WebFlow, Bubble, etc, and you can also learn more frontend by doing more sample projects.

p0d5y ago

Your best option is to be more confident.

I am 50 and do not call myself a webdev or programmer however I have made a six figure sum with my side projects down through the years.

Most webdevs will never make anything they can call their own. They will make for others. Make something and put it out there.

varlockOP5y ago

Thanks for the confidence boost! It's always too easy to self-criticize.

dchest5y ago

martinopp5y ago

I think you basically have two options if you think there is still a lot to be don before launching a good MVP:

Option A) Using a no-code tool as buuble.io, Webflow, or similar. The choice depends on your requirements.

Option B) Hire an agency or freelancer. Probably nearshore or offshore to maximize your investment.

Both are great ways of validating an idea before actually investing tons of money and time.

varlockOP5y ago

I wasn't aware of neither bubble.io nor webflow. Good to know they're there. Given what I've got in mind seems more to be an app rather than a website, I'd probably give a shot to the former.

thedevindevops5y ago

Do you feel there is no value in converting this project into a high quality documentation API and letting the creative minds that consume it do the UI?

varlockOP5y ago

Maybe not, as I'm actually leveraging existing APIs, so effectively it'll be a wrapper around those.

gostsamo5y ago

Use for the backend something that gives you most functionality as a plugin. Ruby on rails or django. Don't reinvent the wheel. Add js on the frontend as needed.

varlockOP5y ago

fiftyacorn5y ago

Use next.js it's simpler than vanilla react

j / k navigate · click thread line to collapse