From concept to launch in 24 hrs and why there's no reason you can't do this too | Hacker News Reader

From concept to launch in 24 hrs and why there's no reason you can't do this too

kylewritescode.com

68 pointskylebragger30 comments

It’s a project I’ve been wanting to build for a while, [...]

I think that makes it more than 24 hrs from concept to launch. I suppose you already designed a good deal of the website in your head, if not layout-wise then certainly functionality- / interaction-wise, right?

I knew that I wanted to do some kind of reverse Q&A, but fleshed the entire thing out within the 24hr period.

But we all have something like that in our head. The point is, that's not the part that keeps us from deploying - its execution, and the thesis is (paraphrasing) that should not be a problem any more.

One reason: design.

I have virtually no design ability. Believe me I've tried and failed miserably. Over the years I've read a number of books and tutorials, I've tried to emulate styles and designs from sites I like, I've spent countless hours in photoshop and I've asked for input and feedback on public forums. Nothing seems to help. It takes me a disproportionate amount of time to complete a design task compared to anyone with even a basic proficiency.

When I see apps like this one and others for example, from rails rumble put together over the course of a few days I'm truly envious of people's design ability. I've lost count of the number of projects I've abandoned because of frustrations with the design.

I've finally just accepted that I was never cut out for design work, that I'm much happier and far more productive writing code. For the next project I start, I'll save enough money to hire a designer first.

Don't let that stop you!

Not only do I have zero GUI design ability and no color sense, my software background is as far from web applications as it gets and I've still been working on an app for the past month. Use very basic design elements, understand the CSS box model and learn to look at everything around you -- designed objects are everywhere.

We're surrounded by design and you don't need to be able to create something from scratch, just see the elements you like around you and analyze them to understand why they're appealing. Just like programming, design can be broken down into component parts and each one understood on its own. Once you see how they work, you can put those parts back together in different ways.

...at least it seems to be working for me so far and my screens look a little less butt-ugly every day :-)

Another thing that I've been using to punch out mvp type projects fast is this for the UI:

http://www.designinfluences.com/fluid960gs/

A grid system is awesome to begin with to help with laying stuff out well. But this also includes some basic css boilerplate for typography, forms, tables, etc.

It's purposefully in black and white (see the grey box method http://v3.jasonsantamaria.com/archive//2004/05/24/grey_box_m...) which doesn't look half bad until you decide the mvp is worthy enough of some better style.

Yeah, using a grid-based css framework is a huge time saver. I find things tend to get a bit less readable when you've got "grid_5 alpha push_2 foo bar" classes everywhere, but hey, it does wonders for getting an MVP done.

Also, that JSM post never gets old.

Compass ( http://compass-style.org/ ) lets you use CSS frameworks while bypassing the problem of non-semantic class names by using SASS mixins. It's yet another layer of abstraction to learn, but worth a look if you don't know about it already.

And while we are on the topic :) these are some sweet buttons you can create if you are using compass:

https://github.com/imathis/fancy-buttons

I would add CSS compilers like CleverCSS, SASS et al and CSS3 in the same timesaver category as grids. They make prototyping and iterating much faster with small and large projects.

Btw. Here's a nice CSS reset & typography to go with 960.gs and other grid systems http://borderleft.com/labs/

Oh God, thank you for this. I'm just starting a new project and I think this will save a lot of time.

A key part of this is that you should start with the MVP[1]. Get the (minimal) features together that make your app work and release it.

Then follow a user-centred design (UCD) process by testing it on real people! Iterate early. I'd be more than happy to discuss what I mean here if anyone requires clarification. Most people end up following a UCD process without realising it, and it tends to be the way that the most successful products (physical and digital) tend to be created.

  [1] MVP: Minimum Viable Product http://en.wikipedia.org/wiki/Minimum_viable_product

Could not agree more. The litmus test I've been trying to use is "does it feel like I forgot something" — if yes, it's probably a good stab at an MVP. Once it gets into the hands of users, the things I felt that I forgot or otherwise omitted from the initial iteration will either make their way back in, or get discarded entirely, depending on user feedback.

The second "key" is a developer's proficiency with design. I am a developer and I know little to nothing about design. This is a major distraction since I spend a disproportionate amount of time getting a barely passable design. Would anyone recommend a good design resource for developers? Books/links/blogs all work. I try to read stuff like Smashing Magazine and Core77 but I mostly feel out of depth there.

A good starting resource here: http://boagworld.com/design/when-developers-design/

I wonder if http://threewords.me/ influenced this one? That was my first thought when I saw Kyle's tweet about Facto (nice name/url btw).

Reading Mark's HN post the other day definitely helped get me pumped to build something.

Do you possibly still have the link?

http://news.ycombinator.com/item?id=2051288

If you're serious about building a product, 24h isn't gonna cut it. Don't drag it out, sure… but there's more to a product than just a proof of concept.

That was my feeling back when 'we made this site in 3 days using Ruby on Rails!' was all the rage.

Frequently it was really 90% of an app. I like for things to be polished, have no error messages, handle all different types of input gracefully, and so on. The sites made in 3 days never did that. It's like 'the first 90% took 3 days, the last 10% took 3 weeks'. Details are difficult.

But then, that's where cutting out features or releasing an MVP comes in. If it's a really slim product, it's possible to make something usable, at least, in 24 hours. But why not just take a week or two and make it really good? There's no rush.

Oh, I'm definitely not advocating building everything in 24 hours. Moreso, I'm saying that I think at this point it's possible for nearly anyone to round out their toolkit and prototype their ideas into a working MVP. Doesn't have to be polished or detail-heavy; focus on the minutia later, just get something tangible in front of users now.

Sure, it's similar to thdconcepts propounded in 37 Signals' 'Get Real' manifesto, right?

The mistake is thinking that '90% of an app' is actually 90% of an app.

Very cool!

I just slapped together http://www.mileagebrain.com/ in two days to do a proof-of-concept and convince myself that it was feasible, and to gauge interest from the frequent flyer community. They love it so I'm moving forward with making it a real product.

". I’m currently on the $15 shared database, which accounts for probably 99% of my costs so far (the rest being marginal costs for S3 [likely a few cents])."

Why not on the FREE plan?

I'm already over the 5MB free limit.

I am using mongolab [1] for my small projects, that give me some more room to grow (free plan = 256 mb)

[1] https://mongolab.com

that sounds great i have some ideas, have little knowledge in both design and developing fields, but need to strengthen my programing background to get them alive. i have a question about hosting the my site, with a very low budget, what is your advice?

Facto is a great app... Very original idea!

Reply on news.ycombinator.com