Fighting the Startup Battle

Startup Bugs

Let’s face it. A startup never begins without its bugs, hiccups, or caveats. Utilities.Games is no different. Our tool has come a long way, even through bugs, server troubles, and “life-first” events. The members of this team are helping create this tool as a hobby, and as such, often times must shift priorities a bit. Does this mean we are not yet successful, though? Our concept of success is simply that we keep working towards improving as we can. We ask that our small audience thus far is patient with us as we wage war on the Startup Routine.  

Startup Bugs.

We have already squashed many bugs. We’ve been fortunate to self-discover many of them as we test new or existing features. A particularly tricky one happened recently. Our hosting company sent a planned update to the underlying infrastructure our server uses. This caused a seemingly phantom process to absorb 80+ percent of our available CPU power. This was a drastic change, as the average 95 percentile load over the course of the development path thus far had been barely over 3.5% CPU utilization. Now, seemingly out of nowhere, it was pegged at a solid and consistent 80% or greater. 

As is usually the case, user-error was to blame. Due to a small degree of unpreparedness which lead to the server not being gracefully shut down before their planned upgrade (Yep. Our bad.), the bug surfaced. The core of our app somehow managed to spawn a twin of itself outside of the utility that typically controlled it. Long story short, this was three whole days spent racking our brains with it. We were assuming the worst, that either Malware, Rootkits, data corruption (thankfully this was not the case), or some other negative instance that could happen was to blame. Instead, simply re-synchronizing the server with our local data was all it took to get the site back up and running again. Thankfully, this, so far, has been our worst battle and we have begun taking measures to mitigate future concerns of this nature. 

Life Comes Before the Startup

To be very clear, this is a hobby for us, at least for the time being. We all have primary jobs and other areas of focus in which we invest the majority of our time and effort. That said, there may be lulls in which there seems to be little-to-no progress in regard to this tool. This should not be seen as an indicator of abandonment or hopelessness, but simply a side-affect of needing to continue to actually support ourselves out in the real world. This then allows us to continue our pursuit of our hobbies, in whatever shape they take. 

Everyone involved with this tool has become very passionate about it.  We continue to show incredible amounts of interest, even when the ability or effort is not able to be expressed explicitly. As the tool continues to evolve and become whatever it is destined to become, we again ask for your patience. We are all in this to create something we love and plan to continue doing so.

So until next time, keep the games running. Don’t forget to remember what matters most, whatever that may be.

Lighthouse Performance Snapshot and Site Comparison

Lighthouse Report

[EDIT 10/31/2018] An additional performance test was run on Utilities.Games after performance updates, still before ads. Previous results are crossed out.

We’re close to implementing Ads to the site. Again, sorry, but we’ve gotta pay the bills. Anyways, throughout initial development I’ve regularly used Google Chrome’s Auditing tools (powered by Lighthouse). This tool has been tremendous help in preparing our site for PWA. Additionally it’s been used to compare our site to our “competitors”. I use quotes because I respect these sites and look at them as healthy competition, a means to encourage the best quality and fastest website.

Test Conditions

For this comparison, I navigated to each site (except ours) directly from the first search page from Google using Google Chrome (
v70.0.3538.77). To clarify, each page was supposed to be the info page of the Pokemon Crobat. Then, from the Developer Console (F12), I ran an Audit with the following conditions:

  • Device: Mobile
  • Audits:
    • Performance
    • Progress Web App
    • Best Practices
    • Accessibility
    • SEO
  • Throttling: Simulated Fast 3G, 4x CPU Slowdown
  • Clear Storage (Enabled)

After running the audit, I saved the results of the report for future reference and took a snapshot of both the results and the mobile rendering simulation.

The Results

Here are the results with an additional “Score” which simply averages each score for the given site. The table is in order of the highest “Score”.

Site Performance PWA Accessibility Best Practices SEO Score
Utilities.Games 55 83 92 100 93 87 100 85.4 92.6
PokemonDB.net 94 58 67 87 100 81.2
Smogon.com 72 54 52 87 78 68.6
Serebii.net 80 46 42 80 67 63
Bulbapedia 25 31 74 53 91 54.8
GamePress.gg 31 27 58 67 82 53
Pokemon.com 3 31 25 73 100 46.4

I was pleased to see that our site scored the highest, however, it is also important to note that this is pre-Ads. All of the other sites had Ads enabled and they are of course “fuller” websites for the moment. I wanted to make sure I ran these tests before we implement Ads to see a difference, if any, it will have on performance.

Performance

Utilities.Games PokemonDB Smogon Serebii Bulbapedia GamePress Pokemon
55 83 94 72 80 25 31 3

The performance metric basically measures the speed and responsiveness of the site. “How quickly can the user do stuff on the page?”. In terms of performance, we have some room for improvement in comparison. Surprisingly we’re at least ahead of the repeatedly highest ranking site on Google. For us, we have a few things in the immediate future that we will be implementing such as bundling. Also, it can be difficult to truly compare each of these sites since they each utilize different web services, hosted in potentially different areas around the world, and process pages on the server differently. Overall though, this metric is supposed to highlight the impact on the end user. So, it could help determine whether our current web technology stack is still viable or if we should consider moving in another direction.

Progressive Web Apps (PWA)

Utilities.Games PokemonDB Smogon Serebii Bulbapedia GamePress Pokemon
92 100 58 54 46 31 27 31

The PWA metric essentially checks the boxes for whether the web application is PWA-compatible. I am personally excited for this technology and am very pleased that we scored so high. For those that have no idea what this is, basically you can install our web site on your phone like almost any other app. We have a download link in the navigation menu if your device is capable of installing PWAs. This works on newer Android and iPhones as well as Windows 10 devices (desktops, tablets, & phones). I’m not surprised that the rest of the sites didn’t score very high since this technology is really just now emerging into the market with the help of sites like Twitter and Pinterest exposing its capabilities.

Accessibility

Utilities.Games PokemonDB Smogon Serebii Bulbapedia GamePress Pokemon
93 67 52 42 74 58 25

The accessibility score highlights how well a site can accommodate users that are hard of seeing, hearing, etc. This is certainly a topic I’m not too familiar with, but developing this site gives me a great opportunity to learn how to best provide for this audience. I was certainly inspired by Microsoft’s development of the Adaptive Controller for the Xbox One. I personally think that video game culture is a great place to cultivate great culture and acceptance not to mention the importance of understanding and inclusion. Alright, I’ll get off my soap box, but ultimately even though accessibility scores are low across the web, a lot of browsers are picking up the slack. At a certain point there won’t be as much requirement on developers like myself to fully understand this realm in order to accommodate these users.

Best Practices

Utilities.Games PokemonDB Smogon Serebii Bulbapedia GamePress Pokemon
87 87 87 80 53 67 73

The best practices metric is kind of a moving target, but ultimately it boils down to staying on top of security and efficiency practices. As long as you adhere to the standards this value should improve. For the most part I treat this metric as a checklist. When I’m in a lull developing features, I try to visit this metric to make at least minor improvements.

Search Engine Optimization (SEO)

Utilities.Games PokemonDB Smogon Serebii Bulbapedia GamePress Pokemon
100 100 78 67 91 82 100

The SEO metric is what everyone wants to be 100%! This highlights how easy it is for people to find your website. Which, if your means of paying the electricity bill is based on Ads, you need more traffic to your site in the hopes that the Ad placement will payoff. While revenue is certainly the most important influencer for improving SEO, I personally have a bit of pride whenever I see my site show up in a Google search. Searchability is something I design in my day job, so it’s especially helpful for me to practice those skills on this site. However, my day jobs websites aren’t open to the public, so it’s not directly applicable all the time. Regardless, the hope with this is site is to provide different approaches to playing video games and we’d like to share those (hopefully) cool ideas with as many people as we can because that’s what gets us excited to develop more features.

Summary

Utilities.Games PokemonDB Smogon Serebii Bulbapedia GamePress Pokemon
85.4 92.6 81.2 68.6 63 54.8 53 46.4

Our site’s overall score is great in comparison to “the competition”, right now… Like I said, this could all change as we implement Ads and other site-wide changes, but we’ll keep our fingers crossed and tackle each metric one at a time.

Let us know in the comments your thoughts on the comparison and/or what metrics mean the most to you. As with our features, we’re always ears-open to what is most important to our users. If you need more Accessibility options, or have poor internet connection and need a lighter website, these are important things to know so we can try our best to deliver.

One Last Thing…

Here’s a snapshot of the mobile renderings and scores from when I ran the audits: