Thursday, 28 April 2016

Phantasmal Dev Log (Part 7)

So cool the game is finally released.

It was pretty awesome, it even lasted a little while in the Popular New Release list, and also one of the feature panels for a while (probably a few hours only).

But hey I can't complain - it was my first crappy game and it got shown on Steam for a little bit :)

Release day was probably the most hectic - tons of key requests came through from the press.  I had to respond to a lot of forum activity.  These are the times  I wish I had a community manager - but unfortunately no one from the existing team had stepped forward, so I ended up having to do it.  I guess I knew the game/product the best anyway.

It was a pretty exhausting time for the next 2-3 days, but finally traffic peaked and started to die away.

Fortunately for me I had planned out the entire lead up to release day really well: every single task was in a Gantt chart and we had ticked everyone of them off like clockwork.  If nothing else I was pretty proud of the planning and execution.  I can take that plan away with me when I do this again next time.

I think truthfully this project was only ever a test.  We got what we wanted out of it: namely the education.  I learned so much via a trial by fire over the past couple of years.  I had been flailing around a lot, but now I feel like I could do this again - faster, better and more confidently.

I'm not the smartest, most technical, or most creative person around, but I am glad my work experience forced me to learn rapidly.

I will need to sit down at some point and write up a full post mortem.  This will be critical as I will want to encapsulate all of our mistakes so that we don't make them again.

I had written up a preliminary one in TIG Source:

So there we were - we had finished a game project all on our own.  Who would have thought? :)

Sunday, 3 April 2016

Phantasmal Dev Log (Part 6)

So luckily we found what the cause of the severe memory issues was last Friday.  It turns out that it was to do with the Astar Pathfinding package that we were using.  Honestly speaking it was my own fault:  we had hypothesised that increase the area of the grid graph, which is used to control the pathing for some of the creatures, wouldn't impact the performance of the game.

We naively theorized that next to the graphics, the information stored would be negligible without actually looking deeper into how it worked.

The dev lead decided to fix a major section of the sewer section, and that threw us.  The game started crashing to desktop horribly, and we thought it had something to do with the new sewer changes.  We rolled back the change but this didn't do anything.

We looked through the other things that had changed and realized that Astar was one of them.

I incrementally took everything out, and found that by decreasing the Astar area, that the game crashed less. Eventually I decreased the size of the grid graph so that it was back to the original size.  At that point the game NEVER crashed again.

We must have been operating on the brink of failure the whole time and had no idea whatsoever.

By the time I had figured it out, it had been a panic stricken 2 days where I was at the verge of giving up on the whole thing.

Fortunately that got sorted out and I could get back to marketing. A slow trickle of responses started coming back - people actually wanted to play our game!

That gave me a huge sense of relief after several days of silence.  Granted, most of them were low tier youtubers, but it was a start.  There were a couple of mid tier guys at 100k subs, and one guy who was 2M that responded, but we really needed 1 or 2 super high profile ones to give us that momentum for day one.

It was looking good, but we needed more. Fortunately the technical side was stabilized, and looking really good actually!  I had taken a 'break' this weekend to tweak and fix things which is what I enjoy most.  By the end of the weekend, I felt the game was actually pretty fun!  The polish was there too - I'd even addressed a lot of the little niggling remaining issues all by myself!

So from here on in, I will need to focus hardcore on marketing and PR.  It will still be a hard stretch to the finish line, but at least half the battle (i.e. the technical) is now resolved!