Vagabond goes for a Walk

#amwriting

2017-03-06T15:50:00.001+00:00

A snippet for Fowl Play

For a while now, on and off, I am writing a book. Obviously, as a non-native speaker, writing a book in English is something of a challenge, but - slowly - I progress. Right now I am very close to 70k words, and planning to write another 50k, or more. To put this in perspective, an average page is somewhere between 350 and 400 words in a normal fantasy book, so we are talking about 170-200 pages so far. That's a lot.

So, what is it about?
The genre is litRPG. Not many people can really place this in the western part of the world, although more and more books are written in this style. Instead of trying to define this genre, let's just take a look at the snippet above: it's basically a fantasy story, but written as someone playing a game, with all those game related details added. Although this might look inferior to a "normal" fantasy book, I'd dare to say that I enjoy these stories more than those.

The story itself is mainly centred around Yang, a successful streamer, who is lately experiencing a certain lack of motivation when it comes to gaming. When he is challenged by another streamer to prove his excellence in games, he is happy to accept it. However, it soon appears that there is a bit of foul play, and the challenge is not exactly what it seems to be at the first glance.
The title of the book - somewhat related to the plot explained above - is Fowl Play.

As opposed to some hardcore litRPG books out there, this one is not about the life of the player, or the end of the world as we know it. It is just focused on the reputation of our main character - which, at best, can affect his livelihood, but not much more than that. Maybe this will make the story a bit weaker, but maybe not. The part that really matters - in my opinion - is the adventure ingame.

I am planning to self publish this book on Amazon, once it's done an properly edited. There is still a bit to understand about what that will require me to do, so I am not in a hurry to publicise the book anywhere just yet. Nevertheless, I am happy to show the first chapter for those who are interested in either the genre, or the book itself. Please ping me on gplus about it.

Salem Sandbox

2017-02-27T15:20:00.000+00:00

Boston, the seventeenth century of our Lord. The early colonists have built a nice little town, but everyone is now looking towards the west, where unconquered lands offer unimaginable riches - or at least a good life for those who are not afraid of hard work to survive.

I am one of them. As I get off that horrible ship, I quickly talk to a wilderness scout. We take a long trip, he helps me to build a leanto, and disappears. Hey, where did you go?!

That jerk, just left me here. Oh well. Let's get to it.

Salem is a unique sandbox game. The graphics are ~~ugly~~ charmingly simple, yet the content is way beyond what other contemporary games offer. The main reason for this, I think, is the difference between survival and sandbox games; the emphasis here is on building and growing. Growing , however, makes us more desirable targets for other players, who are not above raiding and destroying our livelihood.

So, let's look into a few areas of the game, to see how the mechanics work, and to get an understanding of the complexities of the game.

In Salem, there is a difference between crafting and building things. When crafting simple things, such as a bark cup, the materials must be present in our inventory. Building, on the other hand, places a marker on the ground, which acts as a stockpile. Once the character collected all the required materials, and added them to the stockpile, it can be built.
Craftable things stay in the inventory, and if dropped on the ground, they disappear after a while. One example of this is the bark cup, but there are many, many things the character can craft.
Built things, on the other hand, are never shown in the inventory. Some of them will stay exactly where they were built, such as the leanto on the picture above. Others, like the basket, can be carried and placed somewhere else.

Nota bene: You can keep up to two containers in the leanto, which can only be accessed - other than the character itself - by high level players, who learned criminal activities. However, these activities usually leave a scent, which then can be used for requesting punishment. So, this storage is a fairly safe option for a new player.

There are certain material categories of craftables and buildables, starting from stick & hay, then progressing to wood, to stone, to metal. Not everything can be built of everything, there are no metal houses, for example.
Obviously, multi step processes are also part of the picture. You might start a process with cutting wood, then creating planks, then treating and drying said planks, and at the end you use them for building a special piece of furniture that gives a bonus to the attribute improvement activities. Branching out and specialising in something is not just possible, but desirable as well. A single character can learn anything, but it's better to rely on friends who can do something for you better than yourself. There are a few bigger roles which are useful: Miner/Metalworker, Farmer/Cook, Animal Handler, Hunter/Gatherer. All of these can give the character a good hour work every day, or even more if they are doing that work for a whole town.
One interesting aspect of this game, especially at the beginning, is the scarcity of coin. The NPC merchants in the city only buy a few things, most of them what you can't craft for a long time yet. The precious few that you can, will need you to spend a considerable amount of time out in the wild, scavenging. But, to be fair, this is something that you will do a lot at the beginning. Because...

.. MAD SKILLZ!!! The skill system is faily complex at the first sight, but it is well designed, and it will quickly make sense. There are a long list of different categories (on the left), such as `Perennial Philosophy'. These represent the character's understanding of the different aspects of the surrounding world. Each skill require a certain level in some of these aspects. To have the required skill level, you need to Study 'Inspirational Items'. (example and further explanation below). When you study an item, it will give you a score to a few of these aspects, raising the "current" value of it. The current value, of course, can not go over the maximum value. Once you have all the required current values on all the required aspects, you can learn the skill, which then offers you new actions, craftables, buildables, and so on.
However, as you go along, you will notice that some skills require a higher current value than your maximum. This is where you need to learn a second mechanism: how to raise your maximum.
And this is not too complicated either. If you fill up the current value to the maximum, you can improve it, and it will be raised by a 100 points, or so. At the beginning, it is very easy to increase the max, however, you will need to be aware of the waste. More often than not, you will fill up several bars to their maximum, and as soon as you increase one of those, you lose all current values from every single one. (this is also true when you learn a skill) Planning your development is important.

As I explained above, Inspirationals are very important for the characters skill development. But what are these exactly?

There are 3 ways to get inspirationals: gathering, crafting and receiving them as a side effect of some other crafting. The one on this image is one that I found while walking around collecting stuff. It gives a current score between 150 and 450 to six different aspects. This is obviously not a lot, if we consider that your starting maximum is 1000 points. On the other hand, it's a nice bonus, if you need just a bit of knowledge, and also, it's a food item (eating will be explained later).

Crafting inspirationals usually takes a bit longer, since you need to collect all the materials to make them, but it is a more reliable way to get targeted points.

The random items are really just a nice bonus, but they are often coveted by more advanced players for their very high score values.

There are two limitations which you will need to consider when Studying something. One, there are diminishing returns from reusing the same items several times between learning skills or increasing the skill maximum. Using the same item ten time will take the acquired value down to about half of the maximum value. Obviously, it makes sense to introduce as much variation as you can.

The second limitation is the amount of Study points you have. For each character, there is a maximum study points they can spend, and then they need to wait until they replenish. On the picture above, there is an "Inspiration required" line, which shows how much points you need to spend to Study this item. For an advanced character, even to raise the maximum once, you need to spend a lot of points, and that slows down the speed considerably.

Naturally, this second point puts further emphasis on the variations of Inspirationals you use.

Another dimension of the characters development is the humour system (from the original latin expression meaning bodily fluids). There are four of these, Blue, Red, Yellow and Black, and they are representing the characters ability as well as the resource attached to that ability.
For example, let's say the blue bile represent the strength of the character. The damage caused by an attack is calculated from the current value of the blue bile of the character. Strength related abilities will also use up the blue bile, as well, so if the maximum value is 20, and we lift something, we will notice the value to go down slowly for as long as we do that activity. To replenish this, you need to eat food that provides that kind of humour. For example, take the chestnut we looked at earlier as an inspiration item. There are also different values there, with different colours, four next to 'Heals', and the other four after 'When gobbled'.
Healing is, as you would imagine, the direct effect coming from eating the item. This is fairly straightforward.
There is another mechanic here, though, which is similar to that of raising the knowledge maximum, although it's much more complicated than that.
Some tasks will require you to have a minimum value in one of the humours to work effectively. For example, you can build stick fences, wood fences, stone fences or brick fences. To be able to destroy a piece of these fences, you will need a value of, say, 30, 50, 80, 120 in blue bile, respectively. This is obviously much more than the original maximum of 10. (My example character have already improved these humours considerably)
This is what "gobbling" is about. Once your character has a big and carefully selected pile of food in their inventory, they can start a gobbling session by clicking the double fork button just below the humour. This is an interesting minigame, where the character needs to fill up one bar by eating specific food, however any piece of food will also add "resistances" for different food types, so effectiveness of the gobbling will decrease. Once you finish, the improvements will be added to the maximum. While at first this is an easy task, there are a few details which are good to know before you get to the higher levels. First, your highest humour value will determine the length of every bar, so if you improve, say, blue bile to 30, and leave red at 10, then improving red will be exactly as hard as improving blue again. Second, there are different types of food (nuts, meat, fish) and it is important to control the effectiveness of all of these while in a gobbling session. Third, the values provided by raw, foraged food are not really sufficient after a while, so you will need to pick up cooking, and learn complex recipes, if you want to develop your stats beyond a certain point. Obviously, cooking has a lot of levels as well beyond the foraged foods, which slowly open up as the character progresses.

So, what else is out there? On the picture above, there is an example of a small farm, that shows what kind of activities are out there for you to learn and do. There are fields and pots. There is a kiln, a haystack, a compost bin, a butterfly house, a hive, drying frames, a mine, a hut, walls with gates, trees and bushes planted by the player. Next to the river, there is a fishing net. What are these good for?
Farms can be used for mass production of food items & cotton. There is a very complex system for farming, where you need to consider how you rotate your produce, and how you fertilise, based on the kind of outcome you want (which can be speed, quality, quantity, and other things).
Pots are used in gardening, where you can grow a single plant with a small amount of foodstuff or similar. This is the path you need to take to quickly improve the quality of your foodstuffs grown in pots and on farms. Also the only way to grow certain foodstuffs.
Without going further in the details, let's just say that even a small farm like this one requires the player to spend at least two months in the game before it can be created. But, under this farm there is another huge storage space (in the mine) where all the other workplaces used for metalsmithing and leather curing are placed. And there are other things out there which I did not mention, like animal husbandry, and expeditions, which are also important activities for advanced players. And yet there are still more stuff to mention, such as crime, witchcraft and illnesses and player stalls and player towns. I will not explain these now, since most of them are advanced topics.

Yet, one important aspect I did not mention previously - although it is a significant part of the game - is combat. This is a must have in any MMO description, so let's see what we can know about it.
There is a really small selection of wildlife in the game, but most of what is there is important for the character as a source of food as well as a source of materials and inspiration items. Now, combat is a bit weird. It is much harder than you would imagine, and most of the good stuff comes only quite late in the game. (e.g. the owner of this farm above is probably able to kill grasshoppers, rabbits and spiders, but a deer or a bear would be a challenge for them, if not downright impossible) It takes a while to learn to do it well, and most of the animals have different abilities that you need to learn to handle. The good thing is that animals do not kill you permanently, so you can practice, if you want. Still, dropping everything in your inventory is a painful experience. The combat is pretty much the only activity that affects your red bile, but it also uses other humours for determining your damage and measure your stamina.
(An earlier, related blogpost about combat & salem: Snails of Love )
Besides that, there is open PvP, which comes with the possibility of perma death. However, this is not a very frequent thing, since you can't always tell whether the other player you attack is stronger than you, and there is always a risk of being perma killed.

In short, Salem is a game that, if you can look beyond the graphics, offers you a lot of joyful hours of gameplay, with new skills and activities to discover even after you have been playing for months.

Secret World Flavour

2017-02-08T22:28:00.000+00:00

Almost a year ago, StarShadow wrote a blogpost where she compares ESO to GW2. Somehow I got around to read it today, and offered to make a similar list about TSW. So, here it is.

As I wrote here earlier, I did not like the Secret World at the beginning, then later it became my favourite MMO. It's one of those games where you need to immerse yourself in the lore to really appreciate it. I hope to explain the reason why this is so.

Graphics:
Maybe, compared to the latest AAA games, TSW is not beautiful, but it has a great atmosphere, and many, many places which make you gasp in surprise. In my opinion, the visuals are more than satisfactory. The world looks realistic, and, because of the partially horror related theme, it uses fog and mist very well at some places. The cutscenes are extending the world, and they feel natural. If I need to compare, I would say that it has better visuals than GW2 or ESO.

Characters:
They look realistic as well, especially considering that there are no bulky armours here to hide the human shape. To balance that, there is a great selection of cosmetic items, some of which are rewards for completing quests and achievements. Here as well I would state that the characters are better looking than in either of those game.
However..

Races And Classes:
.. there is only one race, which is human. There is a good selection of customisation options, but at the end, you will still look human. There are no classes, as such, only about 500 skills that you can learn. To help those who don't want to spend hours with build optimisation, the game offers decks. These define a set of skills to learn and use, as well as give the character a uniform when they lean all the required skills, however, there is no obligation to use the deck exactly the way it was created.
Obviously, TSW can not compete against ESO or GW2 in terms of races, and the flexibility of the classless system might also be against TSW in some cases.

A thin slice of the skills you can learn

Animations:
I think these are ok, not necessarily good, but acceptable. Can't really compare them against ESO or GW2, as my memories of those games are somewhat faded. The combat animations of the characters - as well as the enemies - are unique enough to be helpful.

UI:
Functional, not really good looking, but maybe not too bad either. I personally don't have any issues with it, but I had a friend who hated it with all his heart. Far less overloaded than many of the traditional MMOs, but less sleek than some of the new ones. I prefer GW2's ui to this one.

It's... colourful?

Gameplay:
While there are some quests which are very simple (collect 9 essences of whatever), the average quest is often a chain, spiced up with lore and random encounters. The quest system is fairly complex, and exploration is often rewarded with new sidequests. Many times you come across signs of past events which - even though they have nothing to do with your current quest - give you a deeper understanding of the world and the happenings.
The game limits the number of quests you can have at the same time, which at first seems to be annoying, but later you realise (hopefully) that it helps you to focus on them. This system is very well balanced by the abundance of chain quests that I mentioned earlier.
The game also introduces interesting new techniques to make questing more interesting. In the first area, there is a quest where you need to follow a flock of ravens around, and this literally means watching the direction where they fly, and move that way until you locate them again. It is more fun than it sounds :)
Open world events are usually tied to seasonal content or the introduction of new DLCs. I think the ideal balance would be somewhere between that of GW2 (which is very frequent) and this one.

Combat:
TSW did not break the tradition of "Press 1 for normal attack" formula of the old MMOs, which is a shame, because it would be a good candidate for a more action oriented type of combat. Yet, there is an interesting depth of the combat system that makes it less repetitive and boring than some of those old ones.
There is a limit of active and passive skills that the character can equip, and the game puts a huge emphasis on the synergies between your skills for a better effect. You will also need to be aware of your positioning in the combat, as many skills work around or in front of the character. There is also a system of resources, e.g. hammer resource, which you get by using the "standard" hammer attack. This is then used in many other hammer related skills, where the damage, additional effect, etc. depend on the number of resources you spend. Of the 9 basic weapon types (3 melee, 3 ranged, 3 magic) the character can wield two at the same time.
I personally like this system better than the combat system of ESO or GW2, but this really depends on the player. To be very effective, you need to understand exactly what you do, and this requires a lot of dedication.

Controls:
Nothing special, really. Can be used as is. In theory, there is a partially action oriented usage, which I tried a few times, but somehow never liked.

Immersion:
It's easy for a game that plays in an alternate reality of ours to be immersive, as what you see is very lifelike, for example the doubledecker bus in the (extremely small) London area. The visuals are masterfully used for emphasising your feelings, like the ghost of a kid in an abandoned facility, that you often only see for a second, or the cultist graffiti in Egypt. It is easy to lose yourself in the game.

Target Dummies are demons.

Crafting:
Not really a fan, but one of the better ones. Somewhat similar to the "shaped" crafting of minecraft. The crafting system is sometimes used in quests, though, which I think is a great idea.

Fun:
The game is fun, but it is rarely funny. There is epic, sombre, depressed, heroic feeling, but I can't recall a time when I just laughed (although I think there were a few of these occasions, but not many). Worth playing, but not when you feel a bit down, and need a pick-me-up game. It's not.

The anatomy of a bot

2016-05-09T13:59:00.000+01:00

I can't really explain why, but the last @codingame competition did not motivate me much. I knew that I should collect all the data on the board, calculate scores for the next turns, and the like. Just didn't really want to do it. However, at the end, in the last two days, I created a bot, that got into the Gold league, and at the end, finished somewhere around the 800th place (of 2400+)

Take a look at this match:
Incredible Victory

I am the player called LateTide, and, watching how my bot plays, you can easily spot a few patterns. It likes to build L shaped pieces, somehow it intentionally misses some opportunities to have a better one a step later, and it likes to take some risks. Obviously, I have spent some time to work out how to do this stuff.

Which is not true. This bot is a great example of what we call emergent behaviour. Actually, the bot has only a few rules:
- Every exposed block has a value calculated for them, that shows how long of a chain there is at the moment, and also, how deep they are.
- Chains of 3 worth more points than chains of 4+ (the chains are removed from the board if they are 4 blocks or longer), which in turn worth more points than chains of 2. A chain of zero doesn't give you any points.
- Masking a block that has a good score for another color gives negative points.
- The score that you get on the side of a block is three times as much as the score of the block underside.
- Blocks that are lower down worth a bit more than the ones further up
- A block of any colour is slightly more likely to be dropped on top of a column that has blocks of the same colour in it, if all else is the same.
- The blocks will be dropped to the place that offers the best (highest) score for the two combined blocks we drop.

What the bot does not do is any calculation of future state. Although it could go and check the blocks coming in the next 7 turns, or it could calculate the score of the enemy, it doesn't do that. It only uses it's own current state to calculate the best place and rotation to drop with. It doesn't even track how much score a move will make - considering the additional chains built from falling blocks after a move, this is a pretty significant factor.

So, let's see why this very simple emergent behaviour is successful:
- The bot is more likely to build than to get scores. This works well, because it gives the ability to have secondary chain eliminations once it decides to start it.
- Chains are extended to the length of 3, so it only takes a single same coloured block to start the reaction.
- Same coloured solo blocks will more likely go above the similarly coloured blocks, so they might start secondary reactions later.
- Places with high potentials are left open unless there is a comparably good situation.

So, as a matter of fact, for a bot of this complexity (less than 300 lines of code), being in the top 30% is actually a huge victory.

AI contest ideas

2016-05-09T13:12:00.001+01:00

Today is the last day of the latest Coding Game Contest. After a few of these, and a few of previous Google AI Challenges, I have to say that it is great to write these one vs. one battles, but it would be interesting now to do something different.
So, here are a few ideas which could work well in an AI bot challenge:

Game: Paintbomb
Simplified rules: 2-10 players play against each other in an arena. The arena has static walls (but different setup between arenas) and dynamic hotspots (generated at the beginning of each turn). In each turn, the players place 1 bomb and their avatar inside the arena. The range of the bombs is given. The goal is to collect the most points under a 100/1000 turns.
Match scoring: Blowing up a player earns positive points to the owner of the bomb, and negative points to the player who was blown up. In case if the player is blown up by more than one bomb at the same time, the positive score is equally shared between the bomb owners, and the blown-up player receives the full negative score. Hotspots work in a similar way: if the player stands on a hotspot and survives the turn, they receive the value of the hotspot (random value between 1-5). If they stand on a hotspot, and they are blown up in the same turn, the value of the hotspot is taken from their score. (the owner of the bomb do not receive any positive score in this case)
Blown up players are not removed from the game, they can continue until the end.
Leaderboards: At the beginning of the match, an average Elo score is calculated from the score of the participants. At the end, those above the average point are considered as winners, and those below considered as losers of the match. The amount of received Elo score depends on the final score of the player.
Programming challenge: To be able to compete well, the bots have to be able to predict actions based on patterns from previous ones.

Game: The Chase
Simplified rules: This is an asymmetric game, where one player controls a fast, single bot (the Prey) and the other player has three slower ones (the Chasers). The matches are played in an arena consisted of rooms which have random openings to other rooms around them. A single chaser alone can block a doorway. The goal of the chasers is to catch the Prey, and the goal of the Prey is to stay free for as long as possible. The chasers catch the prey if it gets into an adjacent sector.
Match scoring: Assuming X turns as the time limit and T as the number of turns it took the Chasers to catch the Prey, the Chaser player receives X - T points, and the Prey receives T.
Leaderboards: There are two leaderboards, one for Preys and one for Chasers. A player can play on both sides, but their bots never play against each other. The received score depends on the match scoring.
Programming challenge: Spatial segmentation and movement predictions

Game: Cooperative Reversi
Simplified Rules: Two to 4 players start from the corners of a square. The square is made up of sectors which can be lit and unlit. All of the sectors of the square start as unlit, and every time when a player moves to some location, the sector and the ones around it revert to the opposite state. The players can jump exactly two steps in one move. The players can also vote for staying at a given position. The goal is to turn 80 percent of the sectors lit in the least amount of turns. The player moves are sequential.
Match scoring: None. The players only get the same score at the end.
Leaderboards: The characters have an average step number from their previous matches, but the scoring is actually calculated from their ability to cooperate. To earn a positive score, the players must have a lower number of turns at the end than the highest personal average in the team.
Programming challenge: Cooperative minmaxing, prediction of the behaviour patterns.

Indie Game Emails

2014-07-18T12:00:00.004+01:00

Short, educational conversation with a journalist about how to notify them about your new game:

The Secret World - A little side project

2014-05-30T10:58:00.000+01:00

I started playing The Secret World again with my friends lately. Soon we realised that creating a good build is really hard without seeing the DPS and similar statistics from combat. However, I couldn't find a log analyser that provided just the right information.

But this is not a big problem for programmers. With just a little effort, and quite a bit of testing, we wrote our own. There are a few rough edges here and there, but it is already quite useful and we love it.

Here is an example of the generated result:

StormTorn: Imaginary Let's Play post

2014-04-10T11:34:00.000+01:00

Hi everyone, this is Rob from LateTide, and today I'll be playing StormTorn.

While it's loading, a bit of description: StormTorn is a nautical game, where you go around on this huge map of hexagons, and do stuff like trading, fighting other ships and exploring ruins. I have started my current game about a week ago, this is my third character on this map. I've hidden a chest of treasure with the previous one, and now with this one I am trying to find it.

Okay, this is the main menu. I could press "continue" now to pick it up from where I left it yesterday, but I want to show how this works: This button takes me to this screen where I can select the world I want to play in - it is similar to Minecraft, except that if your character dies, it won't respawn, but you can create a new character in the same world. This one at the top is my latest, so let's select it, and press Play.

The world is loading... here we go.

So, this is my ship. If I click on it, I can see the cargo I have, I can manage the crew, see the captain's stats, etcetera. This hex around me is where I can do stuff, at the moment it's almost empty, except for this school of fish. If I click on it, it shows what I can do with it - the only action I have now is "Make a Note", but it depends on the skills of the captain and the crew. My previous character was a trader, I was able to fish these with him, and sell the raw fish at a settlement. This one is an explorer, if I make a note of this, I will be able to sell this information somewhere.

There is a settlement on that island - I see it because I visited that sector before. The darker shade of the hex on the screen and on the minimap shows that. If I open up the large map... this is where I have been before with this character. The map is huge, by the way, it takes a lot of time to cross it, even if you don't stop to do stuff. Anyway, let's go and visit that settlement now.

You can use both the keyboard and the mouse to control the ship, today I will use the mouse, so that you can see where I am going. If you click on one of these little arrows at the edge of the hex, the ship will automatically move to that hex. After the camera catches up, you can start clicking around in the new sector. If I click on this house - which means that this is a normal settlement - I can see the what I can do here. It's the same as the fish. With this action I can sell the note about the fish I just made, for 10 gold coins. I have over 2k now, so that 10 is not really important, but sometimes it goes for much more. This is the trade window, I can sell what is in my ship, and buy what they have, but I will not do that now.

Ok, so our goal for today is to go around and find the island where the treasure was hidden, if we can. The main problem with this is that the characters start at different places on the map, and they do not see what the previous character have seen. Actually, at first you don't even have a minimap, and the big map turns around as your ship turn, so you don't even now which way is north. After a bit of playing you can buy a compass and a telescope. The compass obviously sets the map to fixed - north upwards - and the telescope turns on the minimap display. I read somewhere that you can buy sextants too, that shows the coordinates on the map, but so far I didn't find a place where they were sold. Also, they are not cheap, I think.

Anyway, let's go now. I remember seeing the Vishva tribe around when I hid the treasure - which, fyi, is about five thousand gold coins - and there was a cluster of three islands forming a triangle. It was in an icy environment, so I guess we need to go south. This is totally new area for me with this character, so I don't know what to expect here, but I guess that is the point of the game.

New sector. Not much to see, keep going. I will not comment on the empty sectors from now on, as they are not really interesting anyway. South. South. What is that?

Ok, so this is a floating shipwreck. This is a new thing again, shipwreck can usually be found on the shore. Actions... I can send a crew to explore and loot it. You can set the size of the party on this screen, the more people there are, the more they can find, but also the more people you may loose if something happens. I have a bad feeling about this, so let's keep the size small.
Oops. That's five guys minus, and the morale of the remaining crew just plunged a bit. Morale is important, because if it goes down, there might be a mutiny, and that means creating a new character. There is this other stat, called discipline, I don't exactly know what it does, but it can also help with mutiny. Or not having one.

On we go, further south. There is an island there in that sector, which is good, because we need to replenish our food store. Let's go there. South, nothing here, southwest, and this is it. No enemy ships so far, which is a surprise, because they are usually a bit more frequent. There is no settlement on the island, but that is not a problem, we can gather food. The way you do this: You click on the ship, press the anchor icon, which takes it to the island, and then there is this whole new bunch of actions that you can do here. There is a bit of a tear and wear on the ship, but I don't need to repair it yet, and I don't want to mount an exploration party just now, so let's use this one, which is 'Gather food'. From my experience it also gives a bit of the morale back to the crew, I guess because of going to the land and eating fresh food does that. It seems they have collected two crates of food, which is lucky, and the morale went up by only one percent, which is a bit less than the normal.

Let's go south. I see the first signs of ice, which is good. South, I can see an island on the minimap, let's go that way.
In the meantime, there is a pack of mermaids swimming there, under the ship. I don't think we can interact with them, at least I can not click them. There are some similar things, whales, jellyfish, but I think they are only visual elements, like these rocks down there. The weirdest object I have seen so far was a huge skeleton - much bigger than the ship - of something that looked like a lobster.
Wow, we might have just found the place, there is a triangle cluster here. The treasure was hidden on the island that is on the north corner, so let's go there.

Good news, this seems to be the place. Bad news, there is a hostile ship in the sector. Normally, I can try to flee at this point, but because I want to stay in the sector, I will stay and fight. There is only one ship, so I have a chance. If I wait for a few turns, it will attack me automatically, but I will be in a better position if I attack first. So, let's click the ship and push the attack button.

This is the combat screen. It is basically the sector you are in, just bigger. There are basically three things you can do here: move around, do manoeuvres and fire your cannons. Every turn you can do one movement or one manoeuvre, and you can fire the cannons every 3-5 turns, based on the crew and the ship. It is actually quite complex, so I will not go into this in detail now. The battle will take a while, so let me just speed this up.

Ouch. Saying that this didn't go well is quite an understatement. I have lost, which means that my character is dead, and I will need to create a new one. Let's wrap this episode up with this, thanks for watching, and please subscribe below.

The IDE that actually helps

2014-04-02T22:34:00.001+01:00

While I was working as a Java developer, my main IDE was Eclipse. Earlier, I used Netbeans, but Eclipse did more, and was free as well - at that time, IntelliJ Idea was not. Eclipse, as they say, is a stone hatchet with a diamond edge - crude, but it is perfect for what I need it for, and it really helps me to be productive. There are small things, for example if I mention a class in the code that doesn't exist, and I want to create it, I can do it easily with pressing two keys. Or, the fact that I can use aER in a type search, and Eclipse will automatically search for not just aER, but also variables/types that fit this abbreviation, for example alertExtensionRequest. There are a lot of similar examples which make development quick and efficient, without compromising good programming practices, such as proper naming of the types.

This is my biggest issue with the current development I do. Because I need to use Visual Studio Express / MonoDevelop / SharpDevelop, which are fundamentally .net oriented IDEs, these features are missing altogether.

Just to illustrate the difference, here are three screenshots: the new class creation dialogue in VSE, SharpDevelop and Eclipse respectively.

Apparently, the Eclipse one is much more complex, offers a lot. For this, it sacrifices some of the simplicity, but to balance this, the default values (such as package, modifiers, etc) are good enough in 90% of the cases. However, what it offers is priceless: it creates a lot of the boilerplate code that you would manually need to do otherwise, such as abstract method creation (including the ones coming from the interfaces).

I really, really miss this.

Unity Boo problem - Error in valid code

2014-03-19T23:05:00.001+00:00

I have faced this issue several times now in Unity since I started using Boo, and I guess a lot of other people can just run into the same, so here is a tip how to fix it forever.

First of all, the issue itself:

"BCE0005: Unknown identifier: '"
or
"BCE0043: Unexpected token: <INDENT>"
or
"BCE0044: expecting "DEDENT", found '<INDENT>'"
or
"BCE0044: expecting "EOF", found '<DEDENT>'"

and the code seems to be perfectly valid.

The root cause of the issue:

Boo uses indentation for determining the beginning and the end of a block. The boo compiler that Unity uses accepts both spaces and tabs for indentation. You can even mix these up, so the triple indent can be a tab, tab, 4 spaces, or 12 spaces, or anything along the same lines, with one single caveat: The elements in the same indented block MUST have the same combination. "Same indented block" in this case means subsequent code lines with exactly the same indentation level, ie.

else:
            sectorDesc = sectorPool.CurrentSectors[key]
            if sectorDesc.DrawLevel != visibilityLevel:
                sectorDesc.UpdateSectorVisibility(visibilityLevel, absoluteLoc)
            sectorDesc.DrawLevel = visibilityLevel

Here, the lines starting with "sectorDesc = " and "if sectorDesc" are in the same indented block, but two lines below "sectorDesc.DrawLevel" is in a different block, although it is at the same indentation level.

If there is a mismatch between the spaces and tabs, the errors will be reported AFTER the issue, not on the same line.

The solution:

Three things:

Search and replace all \t characters in your code with 4 spaces. Not the "\t" literal!
Set up your preferred code editor (in my case it is SharpDevelop, but can be done with MonoDevelop as well) to replace tabs with 4 spaces.
Locate the following file: "...\unity\Editor\Data\Resources\ScriptTemplates\82-Boo Script-NewBehaviourScript.boo.txt" and replace the \t characters with 4 spaces in it.

Now you are good to go.

It didn't work?

Ok, two possibilities:

There is still a sneaky character somewhere in the code, which looks like a space, but is not. Turning on the visible spaces and tabs helps a lot here. It might also be the number of spaces too, 3 or 5 instead of 4 somewhere.
Maybe your error message is a valid one? Leaving a "pass" in a function after valid code can cause a similar issue, for example.

I hope this helps at least some of you.

Sustainable F2P modell

2014-03-13T12:26:00.001+00:00

Let's just directly jump in: I don't believe that selling cosmetic items and xp potions is a sustainable business model for the long term. The best F2P model I have seen so far - and I know this will annoy a lot of people - was that of LoTRO.

Lotro basically sells content. There are some cosmetic items there too, and quality of life improvements (say, you can buy riding at lvl 5, or you can earn it at lvl 20), but the best deal is to buy areas, classes or packages, which combine a few of these. They also sell extension packs, but it's very similar to a really big package.

Now, this might sound bad. I, for one, quite liked it, for I knew that I am expected to buy some stuff now and then, but I was in control of what and when. Also, I hate snowy areas, and I simply had the option to skip those and go for other ones I liked better.

There was just one big issue with that game: Their economic model had a fault that simply inflated the points, so after a while the price of new stuff started rising to accommodate for the fact that a lot of people had a lot of accumulated score on their account.

This fault was caused by one single decision: They gave X points to each subscribers every month.

If you are a subscriber, you basically have it all. You can access all areas, get riding for free, can use all classes, etc. Now, besides, you get 500 points a month. That is 6000 points a year, which basically means, that if you don't spend it (and there is not much you can spend it on now) then you can pretty much buy anything. Lots of subscribers mean lots of accumulated points, that basically means that Turbine needs to raise the prices in the shop.

In my opinion, the good solution for this would have been a discount system instead of adding permanent points. Every month you can spend 500 free points in the store, but you can't carry it over to the next month.

And now the funny fact: As much as I see, The Secret World is on the right track with this. They found a balance that doesn't inflate the Funcom points, and they sell you content (issues) for it. Win-win.

Let's talk about Roguelikes - again

2014-03-12T12:09:00.001+00:00

After writing the post yesterday, I pinged it over to a few people on twitter (Including Tanya and Lars) and posted it on Google Plus. Got some really good feedback there and here, so I decided to write a follow-up post to list these issues and concerns, and to give an answer for them.

The Quest for the New Term

`I think we need neither "roguelike" nor a term of similar general purpose.'

In the gamedev circles some people are struggling when they need to define their games. There are similarities with roguelikes, but the games are not strictly roguelikes (based on everyone's definition), and they don't want to fight for the right of using it for their own games. Sure, they could use Roguelikelikes, Roguelites or PDLs instead, but personally I find these terms even more ambiguous and a bit forced.

Inclusivity and the Knights of the Round Table

`Inclusivity is great when applied to people, but not useful as a principle in marketing and communication..'`Definitions need to define things - to make the boundaries clear, and to create useful distinctions between things.'

There is truth in this, however I think there is an error in this kind of thinking. Genres are not definitions, they are categories. Categories must be both inclusive - so that everything can be categorized - and (preferably) mutually exclusive. But this is not easy to achieve when you need to categorise a piece of art or creative work. In the case of games, a genre is not an exact definition for what the game must and must not contain - rather, it should give you a feeling about the gameplay. If I say that my game is a Platformer/RTS crossover, you will form a mental picture of it, even if you will not be able to say whether it is 2D or 3D, what kind of units you can build and how do you fight, and who is the enemy. This is because both of these categories include a lot of different types of games.
Roguelike Elitism doesn't help, though. Keeping the five games (or one, or two, or x, depending on who you ask) on the pedestal, and declaring that your game must be exactly like them, is hurting creative vision.

Breaking Bad with 'Roguelike'

`I'm not sure what "Epic Adventure" does to distinguish, say, Nethack from FTL from Starflight - and yet I consider them distinctly different genres of game, and only one of them is certainly Roguelike and another can make a good case. '`It's even harder when you try to tell an existing community to pick a new name for themselves...'

Right, this is my mistake. To be perfectly honest, the title of the previous article was somewhat misleading. I am not aiming to actually rename roguelikes, as they are. What I propose is to define a genre that is bigger than roguelike, and more relaxed. This genre would include straight roguelikes, rogulikelikes, roguelites, PDLs, and other games, which has a lot of common with these, ie. Skyrim or the other, non procedurally generated rpg adventure games. For games like StormTorn, or Proven Lands, you could just use Epic Adventure (or similar term) without offending half of the internet. :)

Going on an Epic Adventure

`An epic is long, involved, complex, storyfilled, and deep, stuff I wouldn't really attribute to Roguelikes.'`I am also unconvinced that roguelikes often have ANY real story, never mind long or interesting or compelling arcs.'

I think 'long' would describe most roguelikes well, but the things about story and deep stuff are correctly pointed out. Maybe 'Epic Adventure' is indeed too generic, and it means something else. What we need is an expression that will tell you that you are facing something impossible, you need to go a long way to achieve it, and you will need to sacrifice a lot. Maybe heroic adventure, or high adventure? Since I am not a native speaker, I am sure you people can come up with a better expression.

Sincere thanks to Tanya, Lars, James and Ocho for pointing out the above.

Let's rename Roguelike games

2014-03-11T17:03:00.002+00:00

Roguelike became quite an ambiguous term lately, which almost always generates a row about whether it can or cannot be used for this game, or that. If you take a look at the Berlin interpretation, it is a nice guideline, but in a lot of cases, people disagree with even that.

You can spot the same, even if you are only reading some blogs. Tanya X warns you to "Never Say Roguelike". Dustin Anglin thinks that "You should say Roguelike more often". Lars Doucet offers a better term: Procedural Death Labyrinth.

Well, I do agree with all of them. Yes, I know they are telling exactly the opposite things, but there is a grain of truth in all of the above. I actually prefer what Lars tries to achieve, however, I think that his term is still very limiting and exclusive.

What I propose is to broaden the genre a bit with a different, more generic term. For to be able to do that, I will approach this problem from a different angle.

For most roguelikes I played, there are certain factors present: The games are fairly long; the story have a great story arc and high stakes; failing has dire consequences; there is a constant and measurable improvement of the avatar; and usually there is only a single avatar.

Now, let's jump to a different topic for a minute. In ancient cultures, such as Greeks, Chinese, Babylonians or Scandinavians, there are tales of heroes. However you call these, Eddas, Eposes (Eposi?), Chronicles, they are usually about a single hero who will face mortal perils for to be able to achieve something that seems almost impossible to achieve. They need to use their wits, strength, magic, helpful companions, even luck for to be able to survive and win at the end. Oftentimes, countless other heroes fail before they go and do it.

Sounds a bit similar? I think it does. Thus, I believe that using the term "Epic Adventure" for these kind of games would be a good idea. It is inclusive, but it does suggest a certain kind of gameplay.

There are certain elements which are not included in the above, but following normal game design practices, will fall in place: To be able to create a really long game, one obvious solution is to procedurally generate content. That also server well for replayability, which will be required because of permadeath, which is one logical solution for making the consequences dire in case of failure. High stakes and the long, interesting story are the motivation factors for the player to go on and try it again, and again. The constant and measurable improvement is also a factor for keeping the players involved and also to give feedback about progress, as well as control the pace the player can progress with.

There are other factors, which are absolutely not mandatory, but might also serve the game well: in case of turn based games the player is not under time pressure, so they can make better decisions. Non modal , hack and slash gameplay helps the Flow, grid based levels are much easier to handle with procedurally generated environments than continuous spaces.

Now, I imagine a lot of people at this point saying "How about Game X?! That is not a roguelike, and your description fits it!" Well, as I said, I am trying to be inclusive here. I'd love to include games like Spelunky in this, if they want to fit in. Saying that a game is an Epic Adventure Platformer is totally ok.

So, let me know what you think. I am genuinely curious.

The reason why you should always go for first person..

2014-03-03T14:56:00.002+00:00

Greenlight woes

2014-02-26T10:43:00.001+00:00

Free tools and engines make game development easier and easier, which is, in a sense, a good thing. However, as a side effect, more and more excellent games are getting lost in the noise that this generates.

The solution from Steam for this issue - which they are now planning to replace with something else - was Greenlight, a system where the gaming community can go and vote on games they like. There are obvious issues with this approach - for example, that most people will never go there, so the result only shows what a certain circle of people likes - but it is better than nothing.

One of my friend recently put up a game, and now I am asking him for updated stats every day. It is somewhat disheartening: After a bit of surge on the first day, the votes are now slowing down. And this is actually an excellent game with nice visuals and a lot of polish..

I am already afraid that in one year, when I am at this point, I will run into the same trap as a lot of indie gamedevs: I will not know how to advertise my game so that those people who would like it learn about it.

To make the long story short: The Architect is now on Greenlight, and could use some help. Please go there and vote if you like puzzle games which require you to actually think.

Someone should make a game of this..

2014-02-20T12:27:00.001+00:00

Sophistication and Hubris

2014-02-13T13:37:00.000+00:00

I have seen a girl today on the tube. She was reading Dostoevsky. Was she doing this because she just loves savouring the words of the Brothers Karamazov? Possibly.

This made me wonder: How many other people on the same car were reading Dostoevsky in the last year? Probably 0. Same train? Maybe 1.

Now, let's take a look at the bestselling books list on Amazon. In the top 20, you find fiction books, lifestyle books, children books, romance books, but no book that could be considered as a classical piece. Similarly, if I would look into the list of bestselling movies, I am pretty sure I would not find Casablanca there. Or, to take another example, most painters do not make a living with painting a masterwork in two years, and then selling it to the national gallery for millions. Rather, they paint a lot of average pictures, and sell them by the dozen to average people like me who wants to have something nice on their walls that doesn't cost a fortune.

Apparently, when people need entertainment, they are not craving for deep thoughts and deeper meaning every time. I would dare to say that most people never or very rarely does. What they need is something that doesn't require them to think, they can relax and enjoy it without investing too much energy in it. And don't take me wrong, I am not judging these people; our jobs and other daily activities alone can be quite taxing on our energy. It's nice to just sit back and relax in our free time.

The astute reader can already guess what my reason was to write this post. My point is: people seem to be puzzled by the success of Flappy Bird. A lot of gamedevs, especially in the Indie Bubble, say that it's too simple or stupid to be such a success. It's "just Flappy Bird". I think this is the hubris speaking. Instead of hating the developer, let's congratulate him, and then decide: do we want to create a game that - with a lot of luck involved - can be similarly successful and played by millions across the globe? Or, do we want to create a game that is a bit more complex, uses deeper emotional triggers than the happiness felt when we finally pass the 10th pipe, and basically, one that has a meaning? If the latter, then we should realize that we are not making the game for everyone - we are targeting a slice of the gamer audience, who can spend more time and thought on playing a game than what the average commuter has on the train in the morning.

Once we make this decision, let's act accordingly.

Simple game? Let's learn from flappy bird. God forbid, let's copy it, and try to sell it. The goal is not quality, but quantity.

Or rather, the complex game? Accept that it will never be an "instant hit", but make it an excellent game. There is still a chance that it will sell exceptionally well, but just don't expect to earn a year worth of your wages in one day. Instead of that, enjoy the fact that you have created something valuable.

Of course, the world is not just black and white. But this decision helps to keep the right goals on our horizons.

Update: Added a link to the Indie Bubble for reference.

A Game of Games - A Metagame Idea

2014-02-11T10:00:00.000+00:00

You are one of the parents of a 3 year old child (for simplicity let's assume that the child is a girl). The child just got over the hard period of finding her own voice, and now is on the way to build up a personality and the required self defences before she is introduced to the wider society: pre-school education. Your task is to help the child with this by helping her to play the best games.

There are several things you can do to make this happen. First of all, you can observe as the child plays. You will see which toys she likes more, what kind of games are her favourites. You can also join in, and help her to improve an aspect of her play style: Dexterity, Imagination or Creativity.

You also have a budget for each month that you can use to buy props - toys and games. All the props have different properties and themes - property is about what it can improve, and theme is about the type of game - superheroes, dolls, pets, etc - that determines the relation between the child and the toy.

The game is broken down to turns. One turn is a week; in this time period you can decide what to do with the child. At the end of the turn you will see the observations and improvements of the child.

Your goal is to build up all the different abilities of the child up to a minimum level. A side goal is to build up a talent: something she will be exceptionally good at.

The child's affinities and abilities are randomly generated, but somewhat controlled by the gender, so there is a point in playing the game more than just once.

I think I have mentioned earlier this book: Imaginary Games from Chris Bateman. It is an excellent book that digs really deep down into games and imagination, and a lot of other, related, things. It is actually so dense, that every time I pick it up, I read only 20-30 pages in one go, and then put it down for another 2-3 months. :)

The reason why I mention it again is because this game idea is based on that book. I don't think this could be a good commercial game, but maybe it is a good candidate for a jam game. As I don't have time now to work on it, I thought I share it with you.

IndieDB

2014-01-18T23:52:00.001+00:00

Following the advice I got from a gamedev on twitter I registered the game on IndieDB. Then, seeing what it has to offer, decided to use that as the main channel to communicate about the game. So, instead of the blog mentioned in the previous post, please find the game here:

Moving StormTorn

2014-01-16T21:39:00.001+00:00

It seems that these days I only blog about StormTorn, which, in a sense, is correct. But that basically means that this blog, that was originally meant to be about MMOs and games are now solely dedicated to game development. This is not what I want, so I decided to move the StormTorn related blogposts to its own dedicated blog.

Meet you there.

StormTorn: To get the feeling

2014-01-15T09:13:00.002+00:00

After laying down quite a bit of groundwork, finally I am able to show the description on the currently selected item. The screenshot is far from impressive, but still, this was a big jump forward.

Also, a short video can be found here.

StormTorn: A simple explanation about the visibility levels

2014-01-08T14:42:00.000+00:00

As I mentioned earlier, the game will have discreet (hex shaped) sectors. The contents of the sectors vary - some will have islands, some will be totally empty, or you can find natural resources here, or a native tribe there. But, to make the game interesting, and exploration rewarding, most of this information is hidden until you enter the given sector.

The basic concept is that the content of the current sector (the one where the ship is) is always visible. You have also a good visibility in the surrounding sectors (first circle) and some visibility in the second circle (two steps away from the current sector.) Also, the visibility in the first and second circle must be different if the player have visited a sector already, because they will know that there is a native settlement there.

Technically, this is solved by using visibility levels and the visited flag. The levels are explained on this diagram to the right.

Level 1 is the sector itself, or the sea, in other words.

Level 2 is an island or similar, static object in the middle of the sector.

Level 3 object(s) are the ones which are attached to the static object. In case of the island, it can be a settlement, or some natural resource, like sugar canes, or something like these.

Level 4 objects are the dynamic ones, pirate ships, mist, and decorator objects, such as the compass.

Now, how does this work? In the current sector, the visibility level is (almost) always 4, so we show everything on and below Level 4. This practically means that everything is displayed, including the animations and particle effects. In the first and second circle, the default visibility level is 2, but for visited sectors, it is 3. The difference between the first and the second circle is more at the level of detail. Animations and some particle effects (say, the smoke coming from a volcano) will be only visible in the first circle.

I've yet to implement this system. The refactoring I did yesterday was the foundation for this, making it possible for me to start coding this. I hope I will be able to show this either with screenshots or with a video next week.

[GameDev] Refactoring Love

2014-01-08T00:06:00.003+00:00

I actually like refactoring. Usually. Not so much this time, when I was trying to refactor a piece of code in StormTorn which handles the level rendering. The reason being: I have no unit tests whatsoever, mainly for the lack of proper support.

Nonetheless, I just spent about one and a half hours breaking one part of the code, and rebuilding it so that it provides the same result when I am running it. The main difference is that it is now much more suitable for my planned future implementations, which is always a bonus. It also helped me to understand what would be the proper way of implementing a feature I was planning, but didn't really understand in depth so far.

It might take several weeks until the visuals change again, but this is important work too.

StormTorn: The state of the game

2014-01-06T22:31:00.000+00:00

Luckily, I spend much more time with developing StormTorn than writing new posts for this blog. Still a very limited amount though, but at least there is a visible improvement over the time. This post is really just an update on how the visuals look at the moment:

I hope you like it.

The Berlin Factor

2013-12-10T23:09:00.001+00:00

I believe I mentioned before that StormTorn is a (will be a) nautical roguelike, with a strong emphasis on exploration. In this article, I'd like to expand on this a bit, using the Berlin Interpretation for the definition factors.

But, before I go into that, here is a bit of explanation: There is an ongoing "conversation" on the internet about Roguelikes. Some say you should never use this expression, for you should not define a genre by the first game. Some others argue with this. I think it is an expression that means much more today than just a reference to the original game, especially that most of us have not played with Rogue whatsoever. So, when I use the term "Roguelike", I use it in accordance with this definition:

"Roguelike" refers to a genre, not merely "like-Rogue". The genre is represented by its canon.
The canon for Roguelikes is ADOM, Angband, Crawl, Nethack, and Rogue.

Having said that, let's go through the list: (for the definitions please see the link I mentioned above: Berlin Interpretation)

Random Environment Generation: Yes. The worlds are randomly generated, using some rules to make them interesting and playable. I have a huge list of different places and events to make it interesting.
Permadeath: Yes. If the character dies, there will be no resurrection/respawn. But, there is a twist here: Worlds can be reused for playing with a new character after the previous one dies or if the player decides to abandon it.
Turn-Based: Yes. Actions will take place only when the character acts. There will be "waiting" commands to support this.
Grid based: Yes and no. The world is represented as a hex-grid of locations, but a single location is more similar to a room or dungeon level in a classical roguelike than to a single point that the player or the monster can occupy.
Non-modal: No. The game will be modal. Combat, trade and other similar interactions will be separated from the main window. This is the direct consequence of the previous point.
Complexity: No. Not in the classical sense, but the way character advancement and the ship improvement works, they will allow the user to use different strategies for overcome challenges.
Resource management: Yes, although with certain differences. The player will need to collect resources from the start to improve the ship, but trading will have a strong role in the game. In a nautical game this approach is better that finding ship parts and crew members randomly on the map.
Hack & Slash: Sorry, no. There will be occasional battles, but the emphasis is on the exploration.
Exploration and Discovery: Hell yeah!
Single player character: Yes
Monsters are similar to players: Mainly yes. Naval battles will be against very similar enemies who will be affected by storms and other environmental factors the same way as the player. Battles against land dwellers will have a very different flavour, these can be considered as events instead of actual fighting.
Tactical challenge: Yes. Moving further away from the civilization will increase the challenge factor, and will also introduce new types of challenges.
Ascii Display: No.
Dungeons: No. Instead, there will be locations in a huge, explorable overworld. At these locations you will be able to use different actions, will face enemies and natural events.
Numbers: Yes.

So, can I say it is a roguelike? I believe it is not a lie, especially if I also say that it is nautical. I guess no one will expect me to move an ascii ship around in a maze built over water, bombarding orc warships at every step and using enchanted sails I picked up from the water, while drinking my Potion of Greater Hull Repair. Also, the exploration focus means that I need to somehow relax the steep challenge curve and allow the player able to simply enjoy the voyage sometimes, instead of fighting all the time.

Let me know your thoughts about this, I am really curious how people think about this.