Historia Cossacks [eng]
Z pAmIęTnIkA nIeGrZeCzNeGo AnIołkA
http://cossacksworld.ucoz.co.uk/index/0-16Beginning:
The history of project "Cossacks” began in 1997, when Microsoft presented its strategic game age of Of empires. In it, there were many original motions, which simply cast a spell on players, this was the first revolution in the genre, the new words in strategies, which, left tracks in our hearts.
At that time we dealt with the creation of online multimedia encyclopedias, whose sales began slowly to go to down, and the development of games became a question of further survival of the company. One way or another, we decided to try ourselves in the new sphere. Proceeding Age Of empires far and wide, we studied all its special features, we tried to understand how it gathered an audience. In one-and-a-half year of active game the necessary ideas appeared and an understanding of the fact that we were capable of making a product even better than Ensemble Studios. We were assured in our forces, and started project "cossacks”.
The first question, which did arise on the very first times of development, in which time frame or period to place the game? In Age of Empires the epochs from the Stone Age to the antiquity are dealt with. We decided "to beat” in - the 18th Century, because Microsoft would for sure select the middle ages for Age of Of empires II, and cossacks would come out by its logical continuation. Actually, we thus made "by third” OF AOE, which does not overlap the first and second parts. This was our marketing calculation.
Time showed that we were not mistaken. The setting and time of action of game were right which led to the commercial success of the project. As a result we succeeded in not only making a good game and obtaining pleasure from the process, but also earning rather well.
In the first approximation, "cossacks” appeared as follows: opposition thought between Russia and Ukraine, and in all nations there were four - Russians, Ukrainians, Europeans and Turks. be sold game had to on the domestic market.
In 1998, when the project was started, in the development team it was only four people. On the present measures, the programmer, designer, historical consultant the leader of project. In two years, when "the cossacks” already left to the finishing straight line, a quantity of developers grew rapidly to 12 people. Although, in addition, this a little. However, it should be noted that the project at the different stages were connected the creators of maps and testers, so that the size of development team varied.
The first steps
The first major step toward the commercial project became the demo version, which we in 1999 brought to the exhibition MILIA in [Kannakh]. This was the present "appearance” - we not only showed our project of the interested audience, but also acquired by useful acquaintances with the publishers and other representatives of play industry. There we for the first time heard council to make "Zazakov” with the sight for the entire European market. Strictly, for us this was shocking, since about the potential of our "toy” with confidence they spoke authority, in professionalism of which there were no doubts. The solution to increase a quantity of nations to 16 and to publish the game throughout the world was accepted.
Strictly, the success followed us from the very beginning project. As it proved to be, we bit into the apple with the name of the project, since Cossacks - in view of historical reasons - they know both in England and in France. In Europe this word is unambiguously considered the synonym of Russian coloring as "Kalashnikov” or "the vodka”. Success was complete - the learned, but necessary brand brought to us in the West enormous sales.
Continuing the list of successful moments, one cannot fail to speak also about the date of realising it helped the fact that in the year to the output Zazakov the market for strategies was overfilled by projects, moreover not very high quality. Genre ceased to draw developers, and they reduced revolutions. We left exactly when players "they were famished” on qualitative RTS. Furthermore, in the game was begun to operate the original temporary space: from the middle- ages to the 18th century. Important role played very colorful screen shots, on which we "beat” into one of our main enormous crowds of units At that time this was new, and our hordes arrived on the soul and the public, and the press, which was not stingy to the appreciations on the sites and in the periodicals.
But to take user we intended not only by being crowded on the screen [yunitami]. Differences from another RTS were sufficient: in particular, in "The Cossacks” adapted the system of formations with the possibility of the retention of position (stand of ground). As a result forces obtained by bonus to the parameters of protection and attack. Furthermore, we realized the possibility to seize the peasants of enemy and the unprotected buildings and to in parallel develop the nation of the former enemy. As the basis of this, economic part of the game was assumed, the output of service lives as in the majority of strategies of that time. In "the cossacks” their number was equal to 6. Basic resource was the food, which was used for building of all units and was spent on their content and improvement. Tree, stone and food were obtained by peasants directly on the map, and the remaing 3 resources (gold, iron, and coal) - in the mines. Coal, for example, was spent on each shot.
During balancing of game we made all resources on the map infinite. Plus of this diagram - could be played by one base as conveniently by developing its empire. Naturally, they were and inevitable minuses - is lost the part of the game, when you have the troops and there is no possibility of building units, so that sum depends on that how you will conduct last battle. This strongly shortened the time of play session.
Disputes about the vision of project began at the very beginning developments. We spent week on the explanation of that, to select for us 2d or 3d (isometry- prospect), and only by the strong-willed decision of Sergey [Grigorovich] they decided to stop at the two-dimensional cursor. Now this solution can seem strange, but it is not worth forgetting, that the breakthrough into 3d could manage too dearly for that time. Furthermore, in favor two-dimensional character told the fact that with the development of technologies with its aid of those days it was possible to make more qualitative and detail units. Finally, our "crowds of units”, important and original feature of game, could be realized only into 2d.
There was one additional important moment, because of which we did not begin to break into the third dimension. At that time we did not want to compete with the important developers, who packed enormous sums into the development of the new generation of 3d- strategies. Market 2d RTS rapidly was reduced, and we proved to be almost by sole representatives on the Ger. after being dismantled with the measurements, we began to experiment with the angle of the slope of camera. Versions there were several - from 90 to 45 degrees, but we stopped at 30. It turned out that in this case our sprite troops appeared most attractively and did not shun units confronting after them, which simplified management of important armies. As we already spoke above, initially game was planned to the output only in the territory of the CIS. Since the basic chip of game - simultaneous mapping of 8000 units for us it was necessary to develop their technologies of mapping drawing and error of the logic of game taking into account the fact that the future players will have computers only below average level. Generally, the selection of main features cost the large credit to our chief programmer - the rates of writing the code were about 300-400 lines during the day! The volume of the clean code "Zazakov” at the moment of realising (1.5 mb.) could be compared by the volume [s] "Anna [Kareninoy]”. The chief programmer with the outstanding memory, who always remembered, "which where lies”, without resorting to the commentaries of the code.
Some of the basic problems in the development of cursor were the conclusion of the large number of units and three-dimensional landscape, the rapid search for way and the artificial intellect. Several it was necessary to bustle with the multi-user game and the game through the Internet. During the work on Cossacks we constantly held in the mind main rule - our code it must not impede. We consider that this succeeded for us. The basic concept, on which was done "rapid” code - function were not bulky, but all algorithms linearly depended on a quantity of units.
Search for the way
In post mortem"Zazakov” one cannot fail to mention one of the major problems, which encountered developed. This is the search for way - fundamental task of play developments. In our case it was necessary to contend with the fact that the usual search algorithms, with which the map is divided off into the passable and impervious zones, and from the point of outcome to the point of destination are plotted a curve, envelope "of impassability”, it does not approach. All were abutted against so many times mentioned in this post mortem 8000 units, which godlessly impeded on the classical algorithms.
The two-level search for way, which consists of two parts, helped to leave the situation. The first of them - "global”, way in it searches for on the topological zones, which had interconnections and certain conditional weight the second part of the system - "precise” search, which was used for displacing from within the framework 1 topological zone (general search for way on the map, broken into the squares into 16x16 pixels).
The System worked as follows: in order to find way, units searched for the shortest way, and then, when they entered into final region, checking the straight lines, which connected the centres of topological zones, was selected way in that indicated by player point, using a precise search. The presence of topological zones healthily helped with the creation by theAi. As has already been spoken, each zone had its weight (this index it influenced the presence and the number of enemy troops). By the Ai, was built way on the topological zones, going around dangerous places by side. On the map there were as a rule, about 700 topological zones, which considerably facilitated the search for way, in comparison with 16384[kh] 16384 of possible points.
However, this successful in every respect search for way skidded with the presence of the walls, by which they partitioned off topological zone. In order to go around this problem, we realized the algorithm of a dynamic change of the topological zones in the real time, which proved to be very complex task.
Other problems
There were some problems, also, in the graphic department. The enormous volumes of two-dimensional drawing, which it was necessary to operate, led us to the creation of its own size: each sequence was packed separately, using an algorithm LZ with the general dictionary. A good compression ratio was reached due to this, but access to each sequence did not require unpacking previous.
Problem with the error of the fog of war arose. Since to dynamically calculate a radius of survey for each of 8000 units, without straining in this case processor sufficiently complicatedly, we devised the following solution. Around units it was arranged from 1 to 8 points - map of the fog of war. And further by the method of solving Laplace's equation was reached the smoothing between these points. As a result the fog of war appeared reliably, smoothly and did not impede. One additional pleasant solution was connected with mapping of water, with which we were straightened with the aid of the rapid and promising algorithm, which, by the way we was utilized, until now. Small waves on the water were integrally and rapidly solved by the physical real equation of wave processes. And [rendering] of one "aqueous” texture with the size of 256[kh] 256 of points on the computer 200 MHz occupied fractions of millisecond.
Game through the Internet
The technology of net game was realized as follows: Let us assume that 1, 2 and 3- play time. Let us assume that player -2 presses an [icon] to build a [unit] at a certain moment, and packet with the information about this order is sent away to the computer of player -1. In this case the computer of player -2 carries out this order on 3 times, when packet will theoretically be in player -1. Thus, occurs the synchronization of the computers of player -1 and player -2. If packet does not reach (it is lost) and player -1 did not obtain information about the order, then he stops at 3 stages, but player -2 goes further, to following time, since he sent packet.
Player -2 does not obtain any packets from the player -1 even he thinks that the packet is, probably lost and sends packet again. Player -1 obtains packet, sends away his packets, and itself goes to 4 time. After player -2 obtained the packet of orders from the player -1, game was synchronized, and they further continue to be exchanged packets. And although situation with the loss of packet appears extremely rarely, nevertheless it was necessary to solve this problem. This method is good fact that after the sending of packet the player -2 must not wait packet with the confirmation of the fact that the order is obtained. Also in the development stage a problem arose, with the real time, which goes differently in the different machines. With the game of more than hour the difference appeared somewhere in 1-2 seconds. This problem especially was manifested with [pinging] more than 1000. for eliminating the problem of machine the [pinging] and it was calculated real time. The packets of data were sent for this, was fixed the time of sending and [ping]. Further time was equalized.
Multi-user game the Internet
The problem of the de-synchronization of game on two computers appeared in most different situations, and to [otsledit] it was possible only miraculous. The searches for the solution occupied weeks. Sometimes with the design schedule of sines and cosines, which was used everywhere (for example, for the search for ways or flight trajectory of projectiles), on the different computers were obtained different values.
This problem arose because of so called of floating of point of inconsistency (nonconformity of value with the use of the floating point). For the successful search for precise place in the code, which became the reason for de-synchronization, we made a generator of random number (which it influenced almost everyone: the arrangement of units the won back personnel of animation, the returned orders… on the whole, everything which it had to be synchronized), which was put into the doubtful sections of the code.
The values between the machines were compared after the end of play cycle and if numbers were different, which means, it occurred de-synchronization. In this case the programmer obtained the name of file and the number of the line, where occurred "the failure of random number”. Further by the method of assumption and near-scientific "spear "was located the specific place in the code, which the causal de-synchronization could become. Obtained bug mercilessly was destroyed. In spite of titanic efforts on the finishing of Internet- game, it should be noted that interest in the net battles in "The Cossacks” was not very high. By reason to that - the step by step publication of game in the different countries, because of which the community of Internet- players in no way could collect critical mass, First "Cossacks” left in Germany, then England and France, then in Spain, Japan, Italy and so on. In all our game left in 17 countries of the world, but interest in the Internet- game grew too slowly. The very large number of player`s online was about 500, and on average - 150-200 in the daytime even 10-50 at night.
The quantity of players necessary for the appearance of valuable community capable of involving new users, we at that time evaluated into 2000. Important companies (Blizzard, Microsoft, EA) publish their games simultaneously throughout the world, attaining the huge amount of players on the servers and creating stir around the Internet- possibilities of its production.
Finishing straight line
At the beginning of the work on the project all were proud of its work and project. But in the end colleagues began "to burn down”. Doubts about the success appeared, which developed the syndrome "infinite repair”.
Someone began to indicate that the drawing is insufficiently detailed, someone - that the game there will not be popular and it is even now uninteresting. They concluded project, having literally gritted teeth. When we left to the stage of beta testing "Zazakov” and sensibly were estimated the volumes of works, it became understandable that another minimum of 6 months will be required to the project. In order not to allow disruptions, we send to the extreme measures and they introduced the system "of military time”: working day began at 6 in the morning and lasted to 11 in the evening with the dinner interruption one hours of purpose were clear as, and from us it was required not to be breathed out on the finishing straight line, in this most complex section of any round.
The most severe order of day was somewhat smoothed out by the addition of dual wages and by free dinners. All came without the delays and worked without putting away the hands military time it lasted for the development team of long 3 months, during which of beta forged himself to "gold master” of the project. After the final version "Zazakov” was sent to publisher, and thehalf-dead development team was in full strength sent for leave.