19 years of legacy bullshit.

March 15, 2018 | Filed under: business

We lost our cat jadzia this week to kidney problems :( we had her 16 years. I mention this because something I notice about my website is that there is still a picture of her in the root folder of www.positech.co.uk at this link. This is the image of her from when we got her 16 years ago, when I already had that web address. Through about 5 different ISPs, several different web-hosts and plenty of website redesigns, that link is still live, mostly because I hate having 404s on my site. There is no technical reason to ever kill off a web link, we have redirection tech, its just laziness for when people redesign their sites.

In the same week our cat was put to sleep, my website domain expired with zero warning from my domain name registrar. Suddenly positech was unreachable, and was theoretically up for grabs. Cue very angry cliff on the phone, and in web-chat demanding to renew the domain immediately, no matter what that involves, for the longest time imaginable, no matter what that costs. You will now find positech.co.uk is mine until 2028. Bwahahahaha.

Anyway, the reason I mention these two things is that they are two symptoms of the ‘problem’ of having a long lived indie games business. First… what kind of muppet uses the root of their website to store an image of their cat? Don’t we have at least a flipping ‘images’ folder? What a dork, and yet it, and a bunch of other files are forever stuck there (yeah I could redirect but meh…) Secondly, and more relevantly, what the hell am I doing with a custom-coded (HTML) website rather than word-press anyway, and why for the love of all that is holy, is it sat on a dedicated server that is madly over-specced for it:

Of course the reason is…legacy. Virtual servers were a ‘bit dodgy’ back in the days when I first got a dedicated one. The inability to easily reboot the server if things went wrong was an issue. And I needed bandwidth, and a lot of it, because serving up thousands of copies of game demos that might be 100MB soon adds up, and because I was getting a lot of traffic to my site for my games, my forums, my blog, my GSB online stuff, and metrics & sales reporting, showmethegames.com, and so-on. Plus I was selling my own games, plus Big Pharma, and redshirt, and it looked like that may expand. I couldn’t trust all that to run on AWS in the early years of AWS.

And back then… youtube was not a big deal for game devs 9which is my my youtube channel is my name, not my company, and used to have silly videos of me shooting arrows at melons on it), and facebook for business wasn’t a thing either, and thus a much larger proportion of gamedev-related stuff about *my games* came to *my site*. Back then you HAD tto have a site, and it had to be FAST, because it was a vital part of your business. Thus…fuck wordpress, lets do hand-coded html, and lets use a CDN account with Amazon S3 etc etc..

These days would I bother? TBH, if I was starting out as a gamedev now… I would probably have a simple wordpress blog for my game, with a buy link to sell direct (always sell direct etc), but would probably make youtube my #1 route for talking about my games and communicating with customers. Maintaining your own internet forums is a CHORE, yet I have forums with 74,000 posts on, so scrapping them now would be nuts. This means I have to split my time interacting with the community between reddit/facebook/steam/youtube/my forums/my blog/twitter. I get *some* value from each of those, but TBH, its way too many. I would like to prune that at some point, but am not sure how. Either that, or get a proper, trained-up community manager and have them deal with it all (but people much prefer talking direct to the dev anyway).

This isn’t just a problem with websites and web-interaction though, it also happens with biz stuff, and  publishing. I have some games on itunes. TBH if you put a gun to my head and asked me which ones… I’m not 100% sure. I have some on macgamestore too, I even sell the odd game through outdated fastspring links. I’m now so busy I only settle up with my old mac-games partner at redmarblegames pretty much annually. I have contracts with almost everyone, deals with almost everyone, and some of them have been worth it, and many have not. Over the years… it thins out, and I dont release new games on platforms that bring in too few sales. I just wish I’d been more fussy from the start. I should cancel my YMLP account and my S3 account for starters.

My advice on avoiding all this mess? make decisions about new channels, new websites, new forms of interacting with people, very seriously. Don’t just try and muddle through like me. After 19 years it gets messy as hell.

 

4 Responses to “19 years of legacy bullshit.”

  1. Steven W says:

    Nice cat, sorry hear she’s gone :(

    Can relate to the legacy website issue, had my website since 1999 and before the days of having places to share content with people I used to put photos, videos, etc up on my site to quickly provide a link to someone, then never bothered to remove the file. Strange things to be found in the root directory….

  2. CdrJameson says:

    If you’ve had a perfectly good website for 19 years I’d leave it well alone.
    How much maintenance effort does it cost you?
    How many security patches do you have to install to a hard-coded HTML site?
    How many nested links to third-party JS repositories hosted goodness only knows where?