I just pushed a lot of changes for the lag. Reminder that these things can take days, because it requires looking at where the pain points are, and then testing a fix. We can't just click a button or rewrite a line of code and expect it to resolve the problems.

There is a lot of misunderstanding going on in this post, and I'm going to address a few of these points.

1. Taking the site down is never the answer. I'm not sure why this one keeps getting repeated on every one of these threads, but "taking the site down for a few months" isn't going to make rewriting portions of the site any easier. Subeta runs on a shoestring budget (reminder for those of you threatening not to purchase GAs or saying that you enjoy the site free for 10 years but also want it to run perfectly!) and we can't take the site down for three months. We can't take the site down for two weeks, unless you'd all like to give us $50,000 upfront to make that happen. A rewrite is happening in the background now, with more and more features being moved over to our rest API. This isn't easy and takes a lot of time. Taking the site down isn't going to help that.

2. I understand being frustrated about communication. It's a double edge sword, and we try to walk the line as best as we can. Someone like or can only say "we're aware of the problem". We didn't discover it was a memory leak until early this morning. We can be better at making sure that those updates are in places that all users can see, but it's not particularly useful to do communication at that point while we're still trying to spend our time investigating.

I think the thing that needs to be communicated more is that Subeta is not sustainable. We're a business that has to pay all of the staff you see making the site happen, in addition to a 10,000 a month infrastructure budget. Displaying thousands of avatars, running a wardrobe system where users can have more than 40 thousand individual items - these are all incredibly difficult technical problems at scale. You'll notice that Neopets, Gaia, etc have all sold multiple times and for less money every time because this little corner of the internet is dying. We could resolve all of the problems tomorrow and that is still the hard reality here.

I appreciate all of you offering to help on this thread. Maybe we can break pieces of Subeta out to be open source that can be worked on by community members, but I don't feel comfortable letting users jump into our code base that has information for the hundreds of thousands of users who have played Subeta.

What do you think the cash shop goes towards?

Running something that brings us personal joy, and puts something good and useful into the world.

That's totally fair, and I completely understand! I'm not trying to shame or blame anyone into purchasing a GA.

We're here to run Subeta for as long as we can, we're really proud of the site and the community. I want Subeta to be around for another ten years.

Popping back in to say that I never meant to say that non-paying users don't get to complain or raise concerns.

I want everyone to feel like they've got the space to do that (otherwise we wouldn't have these forums, would do more to get rid of off site forums that are used primarily to target the site, etc) but I also want to just remind everyone that it costs money to run Subeta. Loading all of the avatars on this page costs CPU to generate, bandwidth to get to you, artists to draw the items, etc. Just because you don't pay anything doesn't mean that the site doesn't cost anything.

You're making the assumption that we don't know. Every action you take on the site is logged, and we know what users are spending the most of their time doing. What they hop on for two minutes to do, what they spend thirty minutes doing, and what the last actions they take are before they stop coming on the site. We know what people dedicate themselves to doing, what achievements they work toward, how many pets they are training at one time, what quests they do the most of, what item prices they are willing to pay in a wizard quest that'll have a payout of x. Data is a lot more useful to us that polls.

We have a pretty exciting roadmap for the year, more group features that involve some cool ways to earn prizes. I do think that we could slow down on frequent item releases and instead focus on these features.

You're fitting a lot of assumptions of bad faith into that last sentence. It's been requested a number of times and judging by the number of users who've entered (who aren't vocal on the forums) it's pretty popular. We have made contests for people who can do other things, and continue to make that a priority during events. However, wearables are the primary way that most people interact with this site.

As for polls - we have forums for feedback, channels on social media, the shoutbox, etc. There are plenty of places where we see and collect feedback that aren't just hard data as well. the assumption that "we don't care to ask you" is hard to make when and make threads for the work that they're doing and invite and collect feedback on them.

Taking a quick look at the update threads from and that are filled with taking user suggestions and quality of life updates speaks differently. Amber posted on a news post that we were looking into the memory leak that was causing the lag, and while we've agreed here that better communication is warranted for things like that - we were looking at it.

I'd also say those threads (and both of those staff members posts on threads in the forums) point to great staff interaction. I'm done responding to claims that we don't interact with users, or take suggestions, when there are so many indicators that point to the opposite. Just because we don't do what you exactly want doesn't mean that we aren't listening. In fact, the reason that this line of feedback is possible is because we do interact and take it seriously.

Just wanted to let you know that we think you raise great points and we are discussing your suggestion. <3

Now that TravisCI is working, I've pushed out some infrastructure updates that should reduce overall lag ?