It’s been coming a long way, but we’re confident it was worth it: the new research system is finished. Actually you can try it out right now, since we’ve created a new public test server for it. All the information you need to access the PTS are at the end of this blog, but first let’s see what this is all about.

Tech trees

The main concepts of the new techtree-based kernel research system have been covered in the previous posts, so I’m not going to repeat them all. Here are the most important “good-to-know” points, and some decisions that we have made for its final incarnation:

  • Researching a kernel happens the same way as before: right-click, research. 1 kernel contains 1 RP (Research Point), but robots drop multiple kernels now.
  • Every item in the knowledgebase requires up to 2 types of RP to unlock. Usually this means factional RP plus common RP, but for T4 items the common RP is replaced by hi-tech RP. Common technology uses only common RP, but double the amount (as if it were common+common), and hi-tech comes in for T4 items here as well (this is of course not doubled anymore, since it replaces one common part). Colony buildings use pure common RP for T1, common+hi-tech for T2, and pure hi-tech for T3.
  • Robots have a 5x RP multiplier.
  • After much consideration, we have decided for a 3x RP multiplier for the corporation tech trees.
  • You cannot research kernels directly into the corporation RP pool. First you need to research kernels for yourself, then donate the points into the corporation pool. Donation is strictly one-way, there is no option to get back what you have donated.
A tech tree in the new knowledgebase
  • If you research an item in both your personal and corporation tech tree, that item will receive a production bonus. This bonus halves the material requirements when producing a prototype item (so a 5x material multiplier instead of the normal 10x), and adds a fixed 50 points material efficiency bonus when producing the item in a factory. Items receiving the bonus are marked with yellow text in the facilities.
  • We have introduced 6 new complexity 1 research extensions, for each type of technology (Common, Nuimqol, Pelistal, Thelodica, Nian Industrial Trust, and Modular Private Colonies). The tech trees in the knowledgebase are laid onto a 10-column grid, and each level of a research extension will allow you to unlock items in that column for the relevant rechnology (e.g. Research: Nuimqol technology level 4 enables unlocking items in column 4 of the Nuimqol tech tree).
  • Researching in the corporation knowledgebase requires the same extensions, there are no separate corporation or CEO extensions for this purpose.
  • We have a new Research access setting for corporation members, with 3 settings: none, limited, and full access. None means obviously no access at all to the corporation knowledgebase, it remains invisible for that member. Limited means the member can see and use the benefits of the corporation knowledgebase, but not spend RPs to unlock items. Full access means that the member can see and use the known tech, and unlock items. CEOs and Deputy CEOs have forced full access.
  • It is possible to research kernels and unlock items while on the terrain. (Kernel research on terrain will not work, sorry.)
  • Kernels on the market can be now found under "Industrial implements".

Knowledgebase conversion

We have determined the new Research Point cost of every item in the new system. During the patch we will take every character’s current knowledgebase with all its known tech, calculate the RP cost of all known items (including partially researched items’ cost as a percentage), and return the RP sum to that character’s available RP pool. This effectively means that you could rebuild the same knowledgebase you had before (provided you have the required extensions), if you wish to do so.

We are aware that the new extension requirements could be an obstacle for some in achieving this, however we also feel that the ability to specialize and pick exactly what items you want to unlock compensates for this quite well.

Kernel conversion

The current 99 different kernels types will be replaced by only 6: Pelistal, Thelodica, Nuimqol, Industrial, Common and Hi-tech. The first 4 factional kernels will be dropped by the appropriate NPC factions. Common kernels will be dropped by rank 1-3 NPCs in all factions, and Hi-tech kernels will be dropped by rank 4-5 NPCs in all factions, as well as observer and officer/starred type NPCs. (As there will be no special observer kernels anymore, their special research properties will be turned into much more dropped kernels instead.)

The drastic reduction in kernel types not only helps us in easier balancing, but should also revitalize the trading of kernels on the market, which has been hopelessly fragmented so far.

One kernel will simply contain 1 RP of that type. So for example a 3rd star Devastator, which is a rank 3 Arbalest NPC, will drop 160 Common and 80 Nuimqol kernels on average, and you’ll get exactly that amount of RP when you research them. (We use a +/- 20% interval in the case of NPC kernel drop randomisation, so that could mean something between 128 and 192 Common kernels in this example.)

During the patch we will convert all kernels in personal storages, corporation storages, robot cargos, and field containers to the appropriate number of new kernels. We will also have to cancel all sell and buy orders on the market for all kernels, since we can’t convert them while on the market (and their prices will likely drastically change too).

Structure kernels will also cease to exist and won’t be available in the Syndicate Supplies store anymore. The new system uses Common and Hi-tech kernels for colony building research, and we’ll convert existing structure kernels to them as well.

Prototype robots

We’re introducing prototype variants for all standard robots, so it will no longer be possible to reverse engineer a standard robot. In order to produce a standard robot CT, you will have to unlock it in your knowledgebase, produce a prototype variant of it in the prototype facility, and then reverse engineer it. All prototype robots roll out with pure white paint.

Prototype robots receive a few bonuses over their standard counterparts:

  • +1% to CPU and reactor performance
  • +10% to sensor strength
  • 10 seconds base locking time instead of 12.5 seconds
Prototype robots

Test server

As mentioned, we do have a public test server for all the new stuff, so you can try it out and then we hopefully don’t have to make 3 server restarts and 2 rollbacks on patchday.

Now, our test server is only one lousy machine as opposed to the 3 beefy ones that run our live world, so initially there will be only one island active, New Virginia. If everything is smooth we'll activate a few more islands later, but testing the NPC->kernel->research->prototype->produce process should be doable on a single island too.

We have copied over a recent snapshot of your accounts and characters, so that’s what you can log in with. However, since we’ll only have New Virginia, all characters have been transferred to TM Alpha terminal. That means you won’t be able to access your stuff that isn’t already there (or one of the NV outposts), but since every one of you received a gazillion of NIC, it shouldn’t really be a problem. Every account on the test server has also received an extra 300 000 EP for testing purposes.

The installer for the test client is available for download here:

The install process is just the same as for the normal client, and it even won’t overwrite your normal Perpetuum shortcuts anymore like in the past. Unless you install it to your normal Perpetuum directory, cause it will obviously overwrite the whole thing then :)

After it’s installed though, feel free to copy over your Local.gbf file from your normal installation so you don’t have to configure your settings and arrange your windows again.

There is a dedicated forum topic for this testing session, where you can post your feedback and any bugs you find, and where we’ll update you on the status of the testing:

Testing will commence throughout the weekend and will end some time in the middle of next week. The plan is to deploy the patch to the live server towards the end of next week.

That is all, see you on the test server, and thanks in advance for the help!

I’m happy to tell you that the development of our new research system is slowly approaching the finish line. Last time I have shown you the tech tree layouts, and as promised, now you’ll get a glimpse of how it will actually look in the game. Not only that, but we’ll also have a look at how our entry to Steam is coming along, which is also being developed in the background.

For science!

Your new playground for item research will have a separate place of its own in the main menu, so we’ll remove the current knowledgebase from the already overcrowded Agent profile window.

Now this is only a design mockup, since the guys are busy with its integration even as I’m writing this, but you’ll get the idea:

Research GUI mockup

As you can see, the window includes both tabs for personal and corporation research, and just like in the case of the colony management interface, the latter will be only visible if you have the proper “read rights” in your corporation.

On the top you’ll always be able to see your or your corporation’s overall research progress for great e-peen justice, and of course the Research Points available to spend per type. The "Donate" button, available only on the corporation knowledgebase tab, will allow you to transfer RP from your own pool to the pool of your corporation. (Doing it the opposite way won't be possible.)

The grid is obviously where the magic happens, all the items available for research will be laid out there, like you’ve seen in the previous post. Locked items will be dimmed and will show the required RP amounts and types. What may not be obvious at first are the icon frames: green is the color of the personal knowledgebase, and yellow is the color of the corporation knowledgebase. Unlocked items will have a solid frame with the appropriate color depending on which knowledgebase you’re looking at, and a dashed frame will mean that the item is unlocked in the other knowledgebase. Items with a double green-yellow frame mean that they have been unlocked in both your personal and in the corporation knowledgebase as well, and receive a production bonus.

While we’re at it, I should mention that this bonus will affect production in the factory and in the prototype facility as well. In the factory the item will receive a yet-to-be-determined material and time efficiency point bonus, and the prototype facility will work with halved material multipliers (5x instead of the normal 10x). This should make it well worth to aim for this combo research bonus.

Since it has caused some uproar on the forum, you have probably heard by now that we’ll also introduce new research extensions, which will enable the unlocking of items on a grid column-per-extension level basis. The level of these extensions and consequently the availability of research will be indicated by the dark blue and lighter blue grid background, as you can see on the above image. I should also tell you that these extensions will have a complexity of 1, so they shouldn’t be so heavy on your EP-wallets.

The current “Research” extension will be removed from the game and we’ll of course reimburse the EP for it. We have considered to keep it and let it provide a point bonus when consuming kernels, but it felt like a mandatory extension for everyone interested in research, so we dropped the idea.

We are also considering a buyback option for already researched items, where you could regain spent RP with a certain loss of points. This may not be available from the get-go, but we’re developing the system with this in mind.

Now the only major thing left to do is to determine the point cost of all the items and how many kernels the NPCs should drop. (As mentioned earlier we’ll replace the current annoyingly large number of kernel types with just a few, 7 to be precise.)

Once that is complete, we’ll try to put together a test server for anyone interested in helping squishing some bugs. We’ll keep you posted when the time comes.

Building up steam

Besides pushing the new major features, the Steam integration is also coming along nicely in the background. Although it wasn’t mentioned in the patch notes, the recent updates did some fixing with the Steam overlay, which works properly now. We already have a working (but yet hidden) Steam account linking option on our website, which already makes it possible to log into the game through Steam, without having to enter an email and a password in the game.

How exactly account linking will work and all the details will be revealed in a later post, once we have it all set up and ready.

I’d also like to stress that using Steam will be entirely optional: if you don’t have it or don’t want to use it, you will be able to continue playing with the normal game client just like now.

Doesn't look like much yet, but it works :)

Lastly, what would a Steam integration be without achievements? I’m happy to tell you that we’re working on that too, and we’ll have some interesting achievements to aim for when we launch on Steam.

When? If all goes well, we’ll be there by the end of summer.

In the last blog I have explained about the basic concepts and the grid layout of the new tech trees, which will be introduced with the revamp of the kernel research system.

By now we have completed the layout of all of the trees, have a look:

Nuimqol tech tree
Pelistal tech tree
Thelodica tech tree
Industrial tech tree
Common I tech tree
Common II tech tree
MPC tech tree

As you can see we'll have 7 separate trees, since we had to separate common and industrial tech, and even had to split the common one in two. This doesn't really have any influence on the research mechanic itself, it's just for usability reasons (less complexity, less scrolling).

As I have explained before, the items are laid out onto a grid, and the grid columns determine the research point cost of the items. The news here is that we'll also introduce new enabler extensions for each faction which will be used to unlock the research of the columns of that specific faction one by one. As there are 10 columns and 10 extension levels, this is really convenient (thanks Simmy for the hint!). Eventually when new items get introduced into the game we'll also create "advanced" extensions for this, but for now this will be enough.

I have also mentioned that we intend to make prototype variants for robots to give a reason to research them, and you have pointed out that ammunition and colony buildings don't have prototypes either. While this is true, we don't intend to make prototype versions for those due to balancing reasons. I also forgot to point out again that despite that, actually there will be a reason to research everything you want to mass produce, since you'll get a bonus for that if you have an item researched in both your personal and the corporation knowledge base. I believe this will come in handy for both ammo and buildings, since the former is in large demand, and the latter needs a lot of commodities to build so there is a lot to gain there.

In the next blog I can hopefully show you how the tech trees will actually look in the game. Until then feel free to provide some feedback on the tech tree layouts, we'll of course consider any changes if they sound logical.

I know, I know, I'm late with this. I could tell you about the reasons for this delay, our mildly excessive optimism regarding the timeframe, and the bugs we had to fix instead of developing, but I think you would rather hear about how's the new stuff going.

So as I wrote in the last blog, we're working on the new research system. This will use a fixed tech-tree system instead of the random lottery thing we have now. Consuming kernels won't yield direct research results, but research points, which will accumulate and can be spent to unlock items in the tech tree. Pretty standard stuff I guess.

What we have done so far is working out a concept how the trees should be built up, and did an internal tool for this which works directly from the database. So once we have all the tree layouts, it's just a matter of some proper UI skinning to get them into the game.

Here is a picture of the tool featuring the complete Nuimqol tech tree (still work in progress):

Work in progress layout of the Nuimqol tech tree

And these are the main concepts that we have laid down so far, additionally to what I already wrote about in the previous post:

  • We'll have separate trees for the 3 main factions (Nuimqol, Pelistal, Thelodica), one for the modular colony system, one for the industrial items/robots, and one for the common stuff. The latter two might be even merged, this is yet to be decided.
  • Each tree will have multiple subtrees with their own root points from where you can start out your research. In the case of the main factions, this looks like one tree for offensive items, one for defensive items, and one for EW/faction specific items, with robots intermixed.
  • The trees are laid out on a grid: you work your way from left to right, from the standard items towards the high-tier items. The grid columns also determine the point cost of the items in that column. (Research points are discussed in detail in the previous blog.) This is the reason why the T1 ECM root is placed further to the right as you can see in the picture.
  • We'll possibly have some multiplier for the points required for robots. (i.e. they will cost more than the modules in their column)
  • We'll introduce prototype versions for every standard robot in order to give another reason to research them. This also means that you'll only be able to reverse engineer and create CTs from prototype robots.

Once we have worked out all the trees, we can start integrating the system into the game client. Obviously it won't look like this in the game, the tree nodes will be square and use the proper icons and tier markings.

Extended combat logs

I'd like to quickly touch this topic too. Some of us who are waiting for the others to finish their research system work had nothing better to do and started working on an extended combat log system. "Combat log" stands here not for the realtime log that you see in a window while shooting each other, but the history that we save to the database for further use and datamining.

The interesting thing for most of you here would be a whole new level of trophies/losses logs (or as you better know it: killmails), which would be tapping into these logs and include anything you want starting from repaired armor points to transferred energy amount and yes, even EW usage. Not to mention that this also paves the way for an achievement system.

Our plan for the first step is to set up the detailed logging and make the current format killmails work from that, and then expand from there. We also intend to send out all of the raw logs via our API, so the nice killboard-hosting guys will be able to create some battle statistics like damage dealt over time or ECM accuracy or whatever you can think of. Of course we also intend to do some fancy graphs and stats of our own in the game once we're done with the more important things.

This is it for now, I promise you won't have to wait another 3 months for the next update.

The end of the year is near and we’d like to offer you some insight on what’s going on around Perpetuum, tell you about what’s currently in development, and get you a glimpse of what the near future will bring. Be warned, it’s a long read, but I didn’t want to miss anything - and I’m sure you’ll have many questions even so.

Behind the scenes

We know we’ve been a bit quiet lately, and this has its reasons. In the past months much of our development power has been put into things that are not directly related to the game itself.

For one, we’re of course getting ready to release Perpetuum on Steam. This involves a lot of paperwork, preparing marketing materials, and not to forget about getting the game itself fit for releasing it to such a broad audience.

Steam Greenlight - a highly public event - happened during the autumn, however we've been spending most of our time since the Gamma Expansion working on getting Perpetuum the extra funding it needs so we can take a quality leap forward in terms of world content. We realize that to create the type of content that would allow Perpetuum to grow to the next level requires a team much larger than ours, and all of us are working very hard on making that possible. This process involves all of the devs to some extent, and is mostly why we've been so quiet during these last months. These things however take time, and the new year will see us continuing the effort.

Meanwhile we’re working on the game itself as well of course, so in the rest of this blog I’m going to talk about what’s coming up. The time for small changes is over, we’re getting the big guns now.

As we stated after the gamma expansion, we’d like to focus on the PvE part of the game for a while, since that part is in desperate need of some fresh blood.

We’ve chosen three large systems currently in the game, the ones where we feel that reworking them would substantially make the game more fun to play: the kernel research system and the knowledgebase, the assignment system, and the new player experience with emphasis on the tutorials.

So let’s see them one by one in detail. The ideas presented here are still work in progress and may change as we work out the smaller details, and of course we will consider your feedback as well.

Research revamp

What’s wrong with the current system?

  • Random research results are not fun. This is especially true when a technology line in your knowledgebase is close to complete, at which point kernel research becomes vastly irritating. Due to the randomness the system doesn’t allow for any serious planning or strategy.
  • There are too many types of kernels. Kernels have been partly intended as a second income besides plasma, but their sheer number fragments the kernel market and has a negative effect on their trading potential.
  • Corporations are not a part of the system at all. They are forced to feed one or two of their agents with kernels to stay competitive. This is very risky, as losing those agents leaves the corporation with close to no prototyping capability.
  • Research has much more potential. Why waste it simply on prototyping, a feature used by not many players?

The new system:

  • Point-based. Instead of getting technology directly from kernels, you consume kernels to gather research points (RP). You then spend these points in the knowledgebase for specific item technologies.
  • Choose your research goals. No more random results - you choose what you want to spend your RP on, using a fixed technology tree. It’s completely plannable, research only what you really need. I’m sure you’ve seen such a tree in many games already - just imagine a branch with many expanding nodes, starting out at the base from the lowest tier technology, and unlocking your way up to the most coveted items.
  • Reduce the number of kernels to 6 main types. They all hold a different type of RP, which accumulate into separate pools. To help keeping the system calculable, kernels will always contain a fixed amount of RPs. To keep scalability, NPCs will drop not just one of these, but many, depending on their ranks.
  • The RP technology types are: nuimqol, pelistal, thelodica, common (used for industrial too), hi-tech and “new-tech”. Unlocking item technologies in the knowledgebase will use a varying mix of these RP types, depending on the item’s faction, tier, or even novelty. (Nasty details: the latter is what “new-tech” will be used for, and we’ll always use it when we introduce new robots or modules into the game. When the time comes where “new” items become “old”, and we’re ready to introduce a new batch of technology, accumulated “new-tech” points and kernels will always be converted down to hi-tech points and kernels - this is to prevent unlocking new technology on the first day by prior RP accumulation.)
  • Separate knowledgebase for individual agents and agent-independent knowledgebases for corporations. Agents always accumulate RP to their own knowledgebase, but they will have the option to donate any amount of unspent RP to their corporation at any time. The corporation-level knowledgebase can then be researched using the donated points by anyone with the necessary access.
  • Having an item researched in an agent’s own knowledgebase will unlock prototyping option for it (like it is currently), and having an item researched in a corporation’s knowledgebase will allow every member with proper access to prototype that item.
  • Here’s the twist: having the same item researched in your own knowledgebase AND in the knowledgebase of your corporation as well will provide a factory production bonus for you. We think this provides an incentive to develop both knowledgebases, your own and your corporation’s as well.

So what will happen to your current knowledgebase when this gets into the game? Well it will be wiped of course and you’ll have to start over, ha-ha. Nah, I’m just kidding, once we figure the exact RP value of every item in the new technology tree, we can give you the matching amount of RP for your actual knowledge, so you can basically rebuild your knowledgebase from scratch. Now, we can’t guarantee that you’ll have the same amount of “raw research value” after that, but I’m sure that being able to select specific technologies to research will mostly provide you with a better result overall. We still have to figure out what happens to unresearched kernels, so we’ll get back to that.

Some of you have been also wondering what happened to the promised new tier of items using colixum. Well, this happened :) We didn’t want to build upon a broken system and create an even bigger chaos. This new system will allow us to introduce new item technology in a much cleaner and nicer way, and that’s what we’ll do once it’s ready.

A new assignment system

I’ll be honest with you: the current assignment system is something we put into the game almost 3 years ago just to have an assignment system. There wasn’t much concept behind it, and it shows - it’s bland, grindy, and not really fun.

At first the main problem is that the player is presented with a huge list of available assignments. This can be quite overwhelming and you just don’t know what to do. Later on you figure out which assignments are the easiest to do or give you the most rewards, and then you’ll grind those over and over, effectively destroying your own game experience. And then we introduce the 6/10 rule - insult to injury.

So we’ve been planning a complete revamp of the system for a while now, and unlike in the case of research, randomness could be our friend here. This is a huge undertaking, and luckily we can separate it down into two stages: the first stage will use the current assignments, but will change the way they are provided and accepted, and the second stage will change the assignments themselves.

When you open the assignments window in the new system, you won’t see a list of specific assignments, only some buttons for types of assignments - combat, industrial, transport, and so on. These will be arranged into a grid, where rows mean assignment levels and columns mean assignment types.

Once you click a button in this grid, you’ll be assigned a random assignment of the selected type and level. Then you can either complete that given assignment, or you can choose to abort it, in which case you’ll be given a penalty. However, that penalty is not a deduction of relation points like now, rather a penalty multiplier for your next assignment’s reward. This means that the more assignments you abort in a row, the less rewards you will get for the assignment which you finally complete. The penalty will be cleared after a certain time or when you complete an assignment.

So this random assignment provider would be the first stage. The second stage is where we also randomize the objectives within the assignments. This system is still heavily under development, but it would use assignment templates, where we only set the number and types of objectives that you have to do. The exact NPCs you have to destroy, or the minerals you have to mine, or the items you have to transport would be randomly selected from a pool of objectives around a certain radius of the location where you take the assignment. This would maximize objective variety and minimize boring walking time. In the new system we also plan to scatter small assignment terminals around the islands, so you will be able to take them there too and not just in the big bases.

Assignments in squads

I'm sure most of you are familiar with the current mechanic for doing assignments cooperatively in squads, which is far from being optimal or useful.

As it stands today, the members of the squad and the assignment they are doing have a very loose connection. Everyone is doing their own assignment, it's not shared among members. One member's objectives are not completed if another member does them. The only actual mechanic is that you can share the money and relation reward among your squad members if you chose to do so when accepting the assignment.

We had a little poll on the forum where I asked you which of the two new squad mechanics you would prefer. While the result turned out not really decisive, it turns out that in a system where assignments are given randomly, only the first option makes sense, because there is very little chance that the squad members can get the same assignment.

So we’ll stick with option number 1, where every member of the squad is working on a common goal and contributes to one shared assignment, the squad leader’s one.


I left this for last because frankly this is the topic where I can share the least details - it’s very much work in progress. However I wanted to tell you as early as possible that instances are coming.

Before you pick up those torches and pitchforks and yell that instances have no place in an open sandbox world, I’d like to explain why they are needed, and what for.

What we definitely don’t want is to drain population from the open world into an infinite number of hidden pockets - the islands feel empty even so, we don’t want to make it even worse.

However there are a number of features and mechanics where we just simply can’t go around instances without having a negative impact on gameplay experience. In an open world you can’t set up a preset game event without pitfalls, however carefully you plan and script it. There are so many unknown variables that a single speck of dust can break the machine.

This is especially critical for a new player - when he can’t finish a tutorial objective because someone mined out all the titan ore before him, he will rarely ask for help, and probably leaves for something else. That’s why first and foremost we’ll use instances for our tutorials, where we can be (almost) sure what happens to the player, and where we'll have much more options than in the open world. Think for example temporarily giving a heavy mech to a new player just to give him a taste what the future holds for him, or to let him try various factions before he makes his actual choice.

Of course it would be a waste of technology to use instances only for a tutorial. It holds many possibilities, like small dungeons that you can scan down and enter from the open world to clear out, or controlled PvP arenas with proper limits, teams, timers and everything.


That’s pretty much all I have for you right now. These are the features that we’ll be working on in the next 2-3 months, to give you a rough timeframe. Of course once we’re ready with one system we’ll patch it out, you won’t have to wait for all that time.

And don’t worry if all of this is a bit hard to grasp now, I’ll post a follow-up to this blog a bit later when we already have some GUI-mockups and stuff.

Post your opinions and questions below, or feel free to create separate feature discussions on the forums, and last but not least, we wish a Happy New Year to all of you!