Back
Cody Frederick

Cody Frederick

@spcoder

Nachobase.com Get Gabbin' SCORE-O-MATIC
Load previous page…

I think this is an excellent question because there are a ton of different ways to skin the CSS cat.

If you're using plain old CSS, then what you'll need is a reliable technique to organize it because it tends to get large over time, especially if other developers are working with you. The method that I like is described at rscss.io/.

If you're using a single-page-app (SPA) framework like React, Vue, etc. I would suggest giving tailwindcss.com/ a spin. Tailwind CSS is considered utility-first, which means that it provides you with class names very similar to CSS keywords. You use these class names to apply the styles. Standard JS components are then used to enforce a consistent look-and-feel. Tailwind is an acquired taste, but I've enjoyed it so far on a couple projects. Here's a helpful link about its utility-first design. tailwindcss.com/docs/utility-….

Personally, I wouldn't go with a CSS framework unless you need pre-canned designs. CSS has evolved to a point where creating a responsive grid system is very easy. CSS grids (css-tricks.com/snippets/css/c…) and Flexbox (css-tricks.com/snippets/css/a…) are very well supported now and make this a breeze.

this is gold! thanks a lot cody!

ya I've checked out the grid and flexbox as well, that's what got me thinking the frameworks I was using was not as good as I wanted them to be.

seems like almost everyone else is pointing to Tailwind so I'll try it out.

really appreciate it Cody!

The site that I use the most is libhunt.com. They have a decent interface and do a lot of analysis to determine the state of the library or tool.

You might be able to compete with libhunt.com if you simplify the interface, and provide both library/tool and SaaS grades.

I'm mostly concerned about three things when picking a library/tool or SaaS product. Is it popular in the community? Are there plenty of resources and documentation? Is it still alive and kicking?

Maybe you could create some type of DevTools (r) number that is essentially a rotten-tomatoes-like score that grades these products on those three key areas?

I love the idea of a Rotten Tomato approach. Originally, I was just going to have the tools that I've used personally and grade them personally. But I realized how difficult that would be to scale. It could help me to expand my territory in development by adding more skills (example: diving into Ruby, Java or Go and writing articles about tools in in those languages).

Libhunt is massive!! They have so many topics.

Maybe just ask for the email with a checkbox to also subscribe to the newsletter?

It's not too much to ask for an email to gain access to a slack community. Especially in the age where anonymous email forwarding tools exist, such as anonaddy.com.

I wouldn't worry too much either way. If your concept is so popular that people are piling onto your Slack, then that's a good problem to have. You're much more likely to be overwhelmed by apathy.

In my day job, I work for a company that provides technical coaching services. This question is right up my alley.

It's one thing to learn how to program, but it's an entirely different thing to learn how to do it well. I think that could be said of many things.

Sites like roadmap.sh do an excellent job of showing the path towards learning how to code, but they stop at how to run a software project with a team.

The primary goal should be to organize your project in such a manner that by adding more developers, it doesn't interrupt the flow of building your product. Start day one pretending like it's not just you writing the code. Think about future people who will try and pick up what you're doing.

You should write automated tests that validate your product. Every commit to Git should cause these tests to run and let you know if your product is broken. This is called continuous integration (CI). CI is a more significant subject we could talk more about at a different time.

Production deployments should be automated. This is typically called continuous deployment (CD). This will involve learning about Git branching strategies, writing automated tests, database migrations, etc. Once you have this setup, then your team can get into a flow in which they're just focusing on your product.

In my opinion, the best team size is either 2 OR 4 developers. It's a whole other subject we can talk about later if you'd like.

Make sure you're getting developers who've worked in a team before, and somehow test if they're good product owners. Maybe ask them how they'd improve on your product or even just another product they know a lot about.

The technology doesn't matter as long as it's popular in your geographical area (if you're not hiring remote workers). Otherwise, just let the team pick the tools they enjoy. Happy developers can crank out products like nobody's business.

I believe this could be talked about quite a bit more. If you'd like to talk more about this, then please feel free to reach out to me directly on Telegram.

Hey Cody,

I am not sure how to write you on Telegram. Just signed up for wip and never used Telegram before (installed it already though).

What is the coaching company you are referring to?

I definitely have it on my to-do list to set up unit tests, though its a bit tricky with my product as its computer vision based and uses real-time input. But I could at least start setting it up for some parts of the software.

I would very much suggest using Grammarly or Hemingway.

I plugged in a sample from your home page and it found quite a few corrections. Here's an example of a fixed entry.

Remote Leaf hand-picks thousands of remote jobs to find the right one for you.

We aggregate remote jobs in one organized place from tons of remote job boards, company career pages, LinkedIn, Reddit, Twitter, Facebook groups and Hacker News hiring

Are you sure that web scraping couldn't solve this problem? I think the problem is a bit vague to provide the best suggestions.

Here are some incentive ideas.

What if you gamified the process of collecting data?

How about an achievement system with badges, rankings, etc.?

What if you created a type of digital currency that users could build up and then use to redeem later to enable parts of your system? For that matter, I think Dogecoins are super cheap, and people like to collect them.

What if you did pay users, but consider it part of your marketing budget?

These ideas might be garbage, depending on your project. I'm open to thinking this through more thoroughly with you if you'd like. Just hit me up on Telegram.

Thank you Cody.

I love those ideas. Your initial statement matched well with what Marc recommended me to do during a chat I had with him so I'm going to just do the leg work myself initially. I do think that the achievement system maybe something I consider in the future.

I think it depends on the nature of your project.

If you're looking to offload manual data entry, then you might want to look at Amazon Mechanical Turk. It's not precisely "crowdsourcing" because people who perform Turks get compensation. I've used it for personal projects, and the process works pretty well. You'll probably have to sanitize data afterward, but it's much quicker than the alternatives.

Again, this all depends on the nature of your issue. If you would like to talk about this further, then feel free to ping me directly on Telegram.

Looks good!

Here are my notes:

  • How quick and easy is it to create a new alias? That would be important to me. It's there a Chrome extension or something that will simplify the process?

  • Building trust with your customers is imperative. So, maybe highlight that my emails will never be lost, that you don't store them, guaranteed delivery, how my login credentials are saved, etc. I know some of this is covered in the FAQ, but maybe promote those higher in the page.

  • Move the roadmap to its own page? I think it might disrupt the potential purchaser flow.

  • "You can decide and customize what information you want to share with a website." instead of "You can decide and custom what information you want to share to a website."

  • On the pricing page, instead of "Number Alias" maybe use "Maximum Aliases"?

  • I don't think that custom domains are important to me. I'm not concerned about what my email looks like for login forms. Maybe that feature falls down the priority list in the roadmap? But to be honest, I haven't thought this one through. Maybe it is a useful feature.

  • I think there's a market for your site because of sites like burnermail.io. Also, I think you're pricing seems fair. But, there might be a disconnect between the login feature vs. the email alias feature. It's almost like it's two different things. Maybe two different plans, products, or even websites?

  • Personally, I'm only interested in the email alias feature. I'd find it useful to see which sites are leaking my information and causing spam. Maybe that's a selling point. I think you could sell just the email alias part of the product for your current asking price.

Thanks a lot for the review Cody 🙏! Yes a browser extension is a must-have, I've been thinking about it for a while!

The custom domain is requested by some users who want to have a bit of "independence" and I understand their points. However I'm hesitating between the priority of this feature vs the browser extension.

The login feature is indeed confusing to users. I'll maybe hide this section for now to avoid users getting lost.

Have you tried the product? It's 100% functional (I use it almost everyday since 2 months 🙂).

Pretty interesting idea. I see this idea best implemented as a type of game. Maybe people could guess phrases that the community made to earn points for a leaderboard.

Thanks for the feedback Cody, hadn't thought of that. Could also be used to generate better tags for GIFs.

I like the idea, and the minimalist feel. But, I think you need a primary logo or brand.

Also, I see a couple spelling and grammar mistakes. Plug your content into Grammarly or Hemingway for proofing.

You might want to elaborate on how "the magic" works a bit more. I specifically want to know how I can control the order in which blog entries appear.

I don't know how to say the name. Maybe include some type of pronouncing aid.

Thanks Cody. Great feedback. I'm working right now on finding designer who can collaborate with me, so that one should be addressed. Spelling I'll check with Grammarly. I have a feature in roadmap - that will proof read Montaigne content automatically. Need that badly myself