Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Learnings from One Year of Building an Open Source Elixir Application

achariam.com

Post reply on HN
Print view View original

11–20 of 24 posts

Re: Learnings from One Year of Building an Open Source Elixir Application

#11

Post by ploggingdev »

Interesting post, more so because I am working on a project that has many characteristics of a link aggregator.

While implementing the ranking algorithm, which is very similar to the one mentioned in the article, I decided to run a periodic job every 60 seconds that updates the rank for each submission and stores it in the database so querying the ranked data is more efficient than recalculating the rank on every page request. Are you doing something similar or did you take a different approach?

Ranking all stories works well if the total number of submissions is a small number, but I imagine the approach is a little different for large sites like HN. Ranking all submissions periodically seems like a waste since people rarely view submissions beyond 10 pages. One approach is just to rank submissions from the past n days, where n depends on the average daily submission volume.

For the part that displays time since a submission was made, I implemented the HN model, where it displays only minutes, hours and days. Python code here : https://dpaste.de/5d1w

> Elyxel was designed and built with performance in mind. Styles and any additional flourishes were kept to a minimum. My choice of Elixir & Phoenix was driven by this consideration as well. Most of the pages are well under 100 kilobytes and load in less than 100 milliseconds5. I find it's always helpful to keep performance in the back of your mind when building something.

Once you start to scale, the bottleneck is rarely the application layer. For the typical crud web app it's likely to be the database.

Re: Learnings from One Year of Building an Open Source Elixir Application

#12

Post by wolco »

achariam wrote: post #3
darkmarmot wrote: post #2

Having a sign up button or perhaps context or text on the site might've been helpful...

You're completely right, it seems a little counterintuitive but one thing I learned early on while building a different community site (Designer News) was the way to build a wholesome community was to be invite only in the beginning. That and real first and last names. Which means it's a poor user experience. Sorry for the trouble, I wonder if any folks have better ideas here?

An invite only for designers because the barrier to entry is so low it makes sense.

Not sure that would work or is as helpful in this case. The community seems more friendly than most and the general goals of many are to learn something new, powerful and fun. If you build a community around those themes I think it will attract the right crowd.

Re: Learnings from One Year of Building an Open Source Elixir Application

#13

Post by tim333 »

ploggingdev wrote: post #11

Interesting post, more so because I am working on a project that has many characteristics of a link aggregator. While implementing the ranking algorithm, which is very similar to the one mentioned in the article, I decided to run a periodic job every 60 seconds that updates the rank for each submission and stores it in the database so querying the ranked data is more efficient than recalculating the rank on every pag…

Not sure if it's relevant but you can avoid having elapsed time in the formula and instead inflate the scores of the newer stories (https://news.ycombinator.com/item?id=9889750)

Then you only have to recalculate when there is a submission or vote.

Re: Learnings from One Year of Building an Open Source Elixir Application

#14

Post by WillPostForFood »

rpazyaquian wrote: post #7
deedubaya wrote: post #6

One of the interesting things about the Elixir/Phoenix community is that building your own authentication system seems to be encouraged. Even if you're using one of the plugs available (guardian?) you still have to do a ton of manual lifting.

This seems to be the case in many places outside of Rails. I know that Clojure, at least, seems to encourage implementing things yourself by way of leveraging smaller libraries and modules. Auth is genuinely hard, and turnkey solutions often aren't enough.

I agree Auth is hard, but that makes me want a good turnkey solution even more.

Re: Learnings from One Year of Building an Open Source Elixir Application

#15

Post by agentgt »

> Elyxel was designed and built with performance in mind. Styles and any additional flourishes were kept to a minimum. My choice of Elixir & Phoenix was driven by this consideration as well. Most of the pages are well under 100 kilobytes and load in less than 100 milliseconds5. I find it's always helpful to keep performance in the back of your mind when building something.

Thats much appreciated but I was kind of hoping you (the author) would go into more details about request time. Most people can (and should) do the above. However I would be more interested in what Elixer + CRUD is like particularly for TTFB. Like does the author do streaming (I don't necessarily mean websockets or comet)?

After all if the TTFB is really slow the CSS optimizations and what not matter little.

In traditional request per thread (or whatever is analagous) web framework paradigms the request is a single thread and often waits for the database to finish before moving on to display the page. I would imagine Elixir has a better answer at least for read only pages.

Re: Learnings from One Year of Building an Open Source Elixir Application

#16

Post by fny »

rpazyaquian wrote: post #7
deedubaya wrote: post #6

One of the interesting things about the Elixir/Phoenix community is that building your own authentication system seems to be encouraged. Even if you're using one of the plugs available (guardian?) you still have to do a ton of manual lifting.

This seems to be the case in many places outside of Rails. I know that Clojure, at least, seems to encourage implementing things yourself by way of leveraging smaller libraries and modules. Auth is genuinely hard, and turnkey solutions often aren't enough.

Auth is genuinely hard, and a hand-rolled solution's definitely not enough...

Given the massive attack surface for a web application, it's absurd to think someone could (or should*) develop an entire auth framework from scratch for all their projects. Turnkey solutions, like Ruby's Devise, are a godsend. Even in situations were a custom flow was needed, it's saved both me and my clients hundreds of hours.

In addition, I benefit from the community around a turnkey solution. Think of all the years the software has been tested by in umpteen production environments. Think of all the people that have an eye on the code and report security flaws while you sleep. Your custom session management implementation will never have that benefit.

Also, why the hell are you building auth when you could be building app? Is that really such a critical experiential part of your app that you can't possibly rely on something turnkey and then move on to the features that actually matter?

Re: Learnings from One Year of Building an Open Source Elixir Application

#18

Post by abrookewood »

agentgt wrote: post #15

> Elyxel was designed and built with performance in mind. Styles and any additional flourishes were kept to a minimum. My choice of Elixir & Phoenix was driven by this consideration as well. Most of the pages are well under 100 kilobytes and load in less than 100 milliseconds5. I find it's always helpful to keep performance in the back of your mind when building something. Thats much appreciated but I was kind of hop…

Time To First Byte?

Re: Learnings from One Year of Building an Open Source Elixir Application

#19

Post by jacquesm »

fny wrote: post #16
rpazyaquian wrote: post #7

Earlier quoted context omitted.

This seems to be the case in many places outside of Rails. I know that Clojure, at least, seems to encourage implementing things yourself by way of leveraging smaller libraries and modules. Auth is genuinely hard, and turnkey solutions often aren't enough.

Auth is genuinely hard, and a hand-rolled solution's definitely not enough... Given the massive attack surface for a web application, it's absurd to think someone could (or should *) develop an entire auth framework from scratch for all their projects. Turnkey solutions, like Ruby's Devise, are a godsend. Even in situations were a custom flow was needed, it's saved both me and my clients hundreds of hours. In additio…

> Even in situations were a custom flow was needed, it's saved both me and my clients hundreds of hours.

And it possibly saved you from embarrassment as well.

Re: Learnings from One Year of Building an Open Source Elixir Application

#20

Post by pdimitar »

agentgt wrote: post #15

> Elyxel was designed and built with performance in mind. Styles and any additional flourishes were kept to a minimum. My choice of Elixir & Phoenix was driven by this consideration as well. Most of the pages are well under 100 kilobytes and load in less than 100 milliseconds5. I find it's always helpful to keep performance in the back of your mind when building something. Thats much appreciated but I was kind of hop…

On my virtual Debian, the development version of a relatively complex website -- which is probably anywhere between 5 to 30 times slower than production -- has 450ms for the first ever request after the server starts. Anything from then on goes between 0.2 to 12ms.

Not really telling you much, I know, but these are pretty standard numbers for Elixir + Phoenix apps.

Post reply on HN