Our first closed beta test ended yesterday. Thank you all who participated and helped to make it possible!
In general, the test went a lot better than I originally anticipated. I was expecting a lot of problems with the launcher, inability to enter the game, server crashes… But seems like it went pretty smooth in general. Everyone was able to join (not without problems, but still), the main game systems were working very well, and looks like the game in general was surprisingly stable and playable.
Of course, there were still a lot of problems, which we will try to fix during the next stages of the beta test, but the main goal for this test was to make sure that all the systems of the project, all pieces of the infrastructure, work well together and there are no big problems that can prevent people from playing the game. And this goal has been successfully achieved.
And the greatest part is that the people were sharing their experience, asking questions, reporting bugs, making suggestions, and actually playing the game together. It allowed me to collect so much valuable feedback I had never hoped. It made extremely clear what people are expecting, what is lacking from the game and what to focus on next. Thank you everyone!
So, what did we find out during the test?
The server
The server performed extremely well during the test. I totally didn’t expect that. There were only two major problems with it:
- Personal zone instances were not destroying at all after being used. Which meant that the memory and the CPU consumption on one of the servers was constantly increasing each time someone triggered a personal instance (first two zones of the main story quest for example).
- A memory leak was discovered in the monster despawning system. Which also meant that the memory consumption of the servers was constantly growing.
The first problem was not a big deal as the personal instance server was isolated from the other servers and it was very easy to restart it without any interruptions for the players. The second problem though forced me to restart the whole world on the second day because of the lag. I’ve made a temporary fix for that and will solve this problem properly before the CBT2.
Other than that, it was great. Yes, a lot of stuff was not yet implemented (like geodata for the NPCs, because of that they were constantly glitching through the terrain and surrounding objects), there were some small bugs in game logic (like few broken skills, incorrect pricing calculation, etc.), but nothing critical.
The client
The client on the other hand, wasn’t so stable. A lot of people had reported long load times, performance drops, crashes, and a ton of small bugs in general. Also, a lot of people were struggling with false-positive alerts from Windows Defender. At least (or as far as I know) everyone was able to play eventually, no one was stuck with a totally broken client.
We’ll try our best to fix most of these problems (or at least, the most critical ones) before the start of the CBT2.
Game content
This was probably the main problem with the game during the CBT1. People ran through the initial quests instantly and then it was just grinding. Classes and skills were unbalanced, unfinished.
Because the skill transfer and parties were not working, a lot of classes were left extremely weak and almost useless while some other classes were totally overpowered. Total lack of the guidance on how to progress or do some basic stuff like change the class or weapon, left many players confused.
This is the main problem we’re going to work on for the next beta test.
Some statistics
The test lasted for about 4.5 days
We started as planned, on August 7th 8:00am UTC and the server was successfully running (with the exception of a few hotfix restarts) since then until the final restart at August 11th 6:00pm UTC. After that, we did a very small event just for fun. The server went down at about 8:30pm UTC.
282 issues reported
Everyone did an awesome job with the testing and reporting bugs, and making great suggestions. It was hard to keep up with the number of posts being created on the forums during the test. Some of those are duplicates of course, but it’s totally fine. Now I need to sort them out and check every one of those. Yay! 
But I’m actually really glad we have everything documented like that, and in most cases with screenshots and even with video! It’s extremely helpful.
In addition to that, I have collected tens of problems I’ve noticed myself or heard from the players but never seen those on the forums. We have a lot to work on now!
Total amount of people invited: 137
Among them who had actually entered the game: 81.
We invited about 70 people for the initial start of the server. About half of those either had participated in the development of the project or were long time supporters; The other half was taken from the CBT application form on the forums.
After the server had stabilized and we fixed the most important bugs, we started inviting more people every day (the second batch of invitations was sent about 5 hours after the initial start). At the end, everyone who had applied for the CBT was invited, up until the announcement of the CBT closing.
Peak players online: 32
As expected, the players were mostly active during the weekend and the amount of people online dropped as soon as the work week started. Averaging at about 23 during the weekend and during the EU prime time on the weekdays and about 10 during the night time and working hours in EU. On total, the average players online were about 16.
Quests
Only 26 persons had discovered that there is a small sidequest on the boat.
Total of 633 boxes were delivered from the Hoddmimis Post Office to the harbor.
And more than two thousand Kui-Kuis and Pui-Puis gave their lives to decorate the Hoddmimis Inn. R.I.P.
Skill books
On total, Cattleya had sold 866 skill books to the players for a total amount of 38 Zeny 5799 Rupees.
Top 30 characters
Show the results
| # | Name | Total Base Exp | Base Level | Class |
|---|---|---|---|---|
| 1 | Darkt | 91500 | 15 | Assassin |
| 1 | Лайф | 91500 | 15 | Priest |
| 3 | fyfe | 55095 | 13 | Knight |
| 4 | Neris | 36207 | 12 | Adventurer |
| 5 | NewGun | 33292 | 11 | Enchanter |
| 6 | iLarch | 32886 | 11 | Berserker |
| 7 | RooftopNinja | 30259 | 11 | Knight |
| 8 | ThatOneGuy | 28439 | 11 | Berserker |
| 9 | meta | 28328 | 11 | Enchanter |
| 10 | Echoladone | 28227 | 11 | Berserker |
| 11 | Edzik | 26883 | 11 | Berserker |
| 12 | Oilvert | 24678 | 11 | Assassin |
| 13 | ExKudesnik | 24330 | 10 | Enchanter |
| 14 | TheSoaringSokol | 21132 | 10 | Enchanter |
| 15 | Abbey | 20127 | 10 | Berserker |
| 16 | Arqueduck | 19131 | 10 | Recruit |
| 17 | Атиони | 17110 | 10 | Berserker |
| 18 | Rincew1nd | 16789 | 10 | Enchanter |
| 19 | Morang | 16768 | 10 | Assassin |
| 20 | Aquarii | 16088 | 9 | Adventurer |
| 21 | Omoto | 15886 | 9 | Recruit |
| 22 | RaviOlly | 14939 | 9 | Recruit |
| 23 | ezCaos | 13737 | 9 | Assassin |
| 24 | Skroyan | 13321 | 9 | Adventurer |
| 25 | Inoue | 13278 | 9 | Recruit |
| 26 | DRAMA | 12425 | 9 | Recruit |
| 27 | Shrike | 11242 | 9 | Knight |
| 28 | Гестия | 10369 | 8 | Priest |
| 29 | Schenog | 10347 | 8 | Enchanter |
| 30 | Shidon | 8956 | 8 | Assassin |
And no, I didn’t use any GM commands to gain advantage. 
Screenshots
Here are some screenshots taken from the media sharing channels in our Discord:
So, what’s next?
Now we have to solve a lot of problems that came up during the test (or at least the most important ones) and finish/implement everything planned for CBT2.
For now, I can’t name any dates, stay tuned for the news. I’ll try to post regular updates on our Discord server.
And again, thank you everyone! It was great!