Background of the invention Field of the invention: The present invention relates to automatic irrigation with humidity sensors, and more specifically to a system and method for more efficient automatic irrigation based on a large number of cheap humidity sensors and cheap automatic faucets, which can optimize the use of water so that less water is wasted and each plant or group of plants can get the optimal amount of water that it needs. Various solutions shown in the patent can be used for example for gardens, agriculture, and flowerpots (for example at homes or in plant nurseries). Background The most efficient water irrigation systems today for gardens and/or fields typically use dripping systems that release drops of water at certain distance intervals for example for about 30-60 minutes per day (for example every 30-100 cm of the pipe there is dropper that releases typically 2 titters of water per hour) and are typically controlled by timers that start or stop the water in the main pipes. However, although this is in general more efficient than systems that do not use droppers, this can still be far from optimal since it does not take into account different needs for each area, depending for example on the individual needs of each plant, heterogeneity of soil type, different amount of Sun or shade in each part of the garden or field, different number of plants in each area, etc. In other words, irrigation systems based on pipes with droppers, typically controlled only with a timer, which are the most common form of irrigation used today, suffer from one very basic weakness, which is that they have no feedback, so they are in essence working blindly. Another problem, which is related to the above lack of feedback, is that there is no efficient way of self-monitoring, so typically, since the system is not aware of its own condition, it also cannot report problems, such as for example breach of main pipes that can cause flooding, or, in the other direction, various pipes or side-channels becoming blocked. Therefore, these systems typically can still waste a lot of water on the one hand and neglect many plants on the other hand, so that some plants get too much water and others get too little water. For example, a raspberry plant or a weeping willow tree typically needs much more water than other plants. Similarly, to the best of our knowledge, there is no simple solution for efficient cheap automatic irrigation of plants in multiple flowerpots that can be used easily with ordinary flowerpots, for example in homes and in plant nurseries that sell plants, except for inserting a pipe in each flowerpot and opening and closing large groups of them by time control, which suffers from all the drawbacks described above. British patent 2281182 describes a closed container of water covered with a capillary mat on its top on which flowerpots are placed. However, this can reduce the efficiency of water uptake compared to placing the flowerpots directly in a water-filled bottom dish, and also the container is filled manually. Many patents describe the use of a water container coupled to a flowerpot that automatically lets the flowerpot draw water when needed, but only few of them, such as US patents 5918415, 4083147, 4546571, and 4557071 describe a truly automatic refilling of the container. However, even those typically use a complex configuration that can't be used with normal flowerpots or requires a complex control valve. Anyway, in practice in homes and even in many plant nurseries the plants are still typically watered manually. Therefore, many plants either get too much water, or are neglected and dry out. Saving water is very important, since according to the World Watch 2000 report we are depleting the planet's water resources at the rate of 109 billion gallons of water per day. Many areas in the world already suffer shortages of water, and others will suffer from it in the coming years. Israel, for example, is now in a critical stage of water shortage, with the Kineret sea's water level already at a critically low level. Therefore, in addition to more desalination of water, more efficient irrigation systems are essential for our survival on this planet. In order to improve the efficiency of the automatic irrigation systems, humidity sensors are needed, however, although many types of humidity sensors exist, they are typically quite expensive (typically between $150 to even thousands of dollars), and automatic faucets are also typically relatively expensive (costing typically at least a few dozens of dollars each, since they typically contain an electric motor, good insulation between the water and the electrical parts, etc.), so they are not used for controlling more optimally the amount of water for each individual plant or for each small group of plants or small area; AlsOj many of the known methods for humidity sensing suffer from various limitations, such as for example limited range of response, sensitivity to changes in the salinity of the ground, sensitivity to changes in temperature of the ground, etc. So clearly cheaper good sensors and cheaper automatic valves are needed. Such a cheap solution would also be very attractive to customers and encourage them to use it, since a cheap enough system that saves a lot of money on watering per month while also improving plant growth, can preferably pay itself back in a few months or even less and start actually saving money for the customer. Summary of the invention The present invention tries to solve the above problems by providing much cheaper humidity sensors that are still quite reliable and also much cheaper automatic faucets, so that preferably each plant or (preferably small) group of plants can be automatically watered by an individual set of moisture sensor and automatic faucet. The attainment of cheaper but reliable humidity sensors is preferably done by using durable cheap sensors that do not degrade quickly and are preferably immune to or able to cheaply compensate for changes in temperature and in salinity. The attainment of cheap automatic faucets is preferably done by using at the end nodes of the system low water pressure, so that much less force is needed to open and close the local waterway, and then either using much simpler electrical valves that do not require engines, or circumventing the need for electrical valves altogether, by using mechanical sensors that control a mechanical valve or directly exert pressure on a flexible pipe, as explained below. Another possible variation, instead of mechanical sensors and valves, is to use for example some chemical control that takes advantage of the behavioral tendency of the water itself, so that for example the water is supplied by a device similar to a plant's roots, except that it works in reverse, so that water is supplied at low pressure to the artificial "root" from above, and the "root" adds water to the earth instead of absorbing it, and stops supplying the water to the earth when the earth has reached a certain humidity level, which automatically creates an equilibrium for example in osmotic pressure between the artificial root and the earth. A similar variation of this is adding a preferably synthetic material that tends to behave like a normal root preferably at the edge of each side channel (and/or in other places), so that the "root" counter-balances the water supply and reaches equilibrium with it when the soil becomes wet enough, based preferably on asymmetric capillary material or materials, as shown for example in Fig. 6. Another possible variation is to use for many plants or at least for each sub-group of them a common water tank like a Niagara, but air-tight, and one or more pipe leads from the common tank to the plants where preferably each side branch for example goes preferably more or less vertically into the soil in a flowerpot or (if it is in a garden or field) into the soil near one or more plants, so that each such side-branch has a humidity control based on air passage, as explained in Fig. 7. The solution for flowerpots is similar, except that the sensing can be done even more efficiently and even more cheaply, and also the control of the watering itself can be done more efficiently and more cheaply, by taking advantage of certain features of flowerpots, as explained below. Therefore, the solution or flowerpots can be regarded also as a smart-home gadget, since it uses smart and cheap automation to both save work and time and to save water. In gardens and agricultural fields preferably one or more main pipes are used with sufficient water pressure of for example 1 or more atmospheres, and each pipe preferably extends into smaller channels that go for example sideways, each preferably with a much lower pressure. This way, the valve that is needed to control each of these small channels needs much less force and therefore can be much cheaper than an ordinary electronic faucet. The reduced pressure can be created for example by using long twisted small conduits at or before each side-channel that easily lower the water flow (such as for example in the pipes by Queen-Gil), which is very cheap and efficient. Another possible variation is using for example a set of prefer-ably small water collectors that work like a toilet's Niagara (preferably one for each side channel), or using mechamcal pressure reducers (however these last 2 options are less efficient). This general configuration is shown in Figs. 1 & la-b. The sensing can be for example mechanical, so that for example a sponge or wood or hair (or other material that changes it shape when it becomes wetter or drier) closes or opens a valve or for example applies pressure to a flexible pipe for example directly by its own mechanical change of shape or indirectly through activating an electrical element (Preferable solutions for this are shown for example in Figs. 2a-i). One of the most interesting of these variations is mechanical sensors based on a bi-material of two or more materials which expand differently when they become wet, thus converting the difference of the expansion into convenient movement. Or the sensing and control can be done electrically, but preferably in very cheap and efficient ways, as described below (Preferable solutions for this are shown for example Fig. 3a-d & 4a-c), or the sensing and control can be based on physical and/or chemical tendencies of the water itself, preferably by using asymmetric and/or irregular and/or strong capillary material or materials, as explained above and in Fig. 6. In flowerpots (plant pots), the solution can be even more efficient, because of the very fact that the plant and its earth are isolated and typically placed over a bottom dish that prevents excess water from running away. This opens up a few interesting possibilities that are harder to accomplish in gardens and fields: The sensor can be placed preferably on the bottom of the bottom dish, so that it merely has to sense if it is in water or in the air, which is much easier than sensing the level of humidity in the earth, since it does not have to face all the problems described above. This can be done for example by a simple electrical circuit that is closed when it is in water, or for example by a simple preferably "small element with a floating part, that preferably moves up when there is water in the dish and down when there is no water or less water and opens or closes a valve mechanically or electrically. This on/off method is free of all the problems described above, and also is optimal in the sense that the earth in the flowerpot is always kept at more or less maximum humidity, and yet it is very efficiently since the reserve water is always kept at the bottom dish, instead of going down deeper into the ground, as it would do in a garden or in a field. The actual watering of the flowerpot is preferably done by letting the sensor control a valve on a pipe that enters or comes near to the flowerpot soil from above. This ensures that the water will go through the soil from the top down before it reaches the dish. Another possible variation is that the water pipe drops water for example directly into the bottom dish, which has the advantage of making the device even simpler, and due to capillary action, the water is absorbed in the soil anyway even if it comes only from below. However in this case, preferably there are more holes and/or larger holes at the bottom of the flowerpot. Another possible variation is the use these features in combination with using each dish in sharing with more than one flowerpot, for example by creating a round or square large area (for example like a large bath) in which the flowerpots are together side by side, or for example using an elongated dish that supports many flowerpot next to each other in a line. In these variations preferably the dish is balanced horizontally, so that the water is more or less evenly spread around it. This way, preferably one sensor is enough for the entire dish, and if the variation of watering the dish directly is used, than also preferably only one water supply and one valve is needed for the dish. This way automatically each plant that needs more water absorbs more water from the common pool into itself and therefore into its soil, and as long as there is sufficient water in the common dish and yet the water does not overflow, no plant is underwatered and no plant is overwatered, even for different types and sizes of plants, different soil types, etc. Also, this can lead to much more optimal conditions for plant growth, so plant nurseries can make more money because the plants grow bigger and faster, so by the time they sell them they can get a better price for them. The elongated dish variation has the advantage that it's more practical and more easy to balance, and also allows easy access to every plant, whereas a dish extended in two dimensions would make it hard to access the inner flowerpots without stepping into the dish. This can be used very easily for example in balconies in homes, or in plant nurseries. An array of such large multi-plant dish rows in a plant nursery is shown for example in Fig. 5f. Another possible variation is to use for example dishes that are closed on the top and have for example holes in the top part for inserting the flowerpots. This has the further advantage that less water is lost due to evaporation directly from the dish. This is a major advance over the current state-of-the art of methods of irrigating plants in flowerpots. Another possible variation is to connect a number of such preferably elongated bottom dishes for example with side pipes, so that one set of sensor and water supply can take care of more than one dish. Another possible variation is to combine the above for example with time control, so that for example the dishes are kept with water in them only for example for a few hours each day. This gives more flexibility in the moisture content in the soils, so that lower moisture levels can also be used. Another possible variation is to use these multi-flowerpot dishes in any of the above configuration with manual filling of water in the dish, which is of course less efficient then automatic control, but still, if for example a plant nursery is divided into a number of rows, each with an elongated dish that serves for example dozens of plants, this is already much more efficient than the current state of the art, since all the workers have to do is water each of the elongated dishes, which is much more efficient and easy than having to water each individual plant, and yet each individual plants gets more optimal conditions than by the normal method of watering each plant individually. Of course, various combinations of the above and other variations can also be used. Preferable variations of these solutions are shown for example in Figs. 5a-g. Also, the same methods or principles described for gardens and fields can be also used with flowerpots, however that could be less efficient, except in the case of asymmetric capillary materials, which might be the best method also for flowerpots. Another possible variation is to use similar principles like those of the solutions for the flowerpots - also for gardens and/or fields, for example by inserting (preferably more or less horizontally) a water-blocking material (such as for example a preferably strong plastic or nylon) below the plants, for example by removing 1-2 meters of earth, adding the material, and adding back the earth on top, preferably before planting the plants. The blocking material is preferably also hard enough so as not to be distorted in shape too much by the pressure from above and by the contours of land and rocks below, and preferably has also for example vertical walls around itself, so as to create one or more large pool isolating the earth with the plants above that pool from the rest of the earth below and around. This way, although the humidity sensors have to work more like in the solutions described above for gardens and fields than the special solutions that can work with water dishes with flowerpots, still the usage of water can be much more efficient since the earth in the area of the plants can be kept at higher humidity levels with less water than in a normal garden or field where excess water can always escape further below into the ground. This can work even if the blocking material does not seal the area hermetically but only significantly reduces the rate in which water can escape away downwards. Like with the flowerpot dishes, the water blocking material can also be for example based on an array of elongated structures that look like bottom-halves of large pipes, that are inserted into the ground and covered with a layer of earth upon which for example vegetables or other agricultural products can be grown more efficiently. Another possible variation, which can be applied in combination with any of the other variations, is to supply the plants with the same water supply system, also with other nutrients in addition to the water, such as for example liquid fertilizers and/or minerals, and/or for example air or CO2 or oxygen (for example by using Soda water with various degrees of CO2 melted in the water) in order to further help stimulate the plant's growth. Such additional materials can be added for example all the time in the desired quantities as a certain percent of the water, and/or part of the time with the aid of an automatic time schedule, and/or together with additional sensing (for example when the naturally occurring electrical potential in the earth indicates too low salinity, or when there is indication of too little air in the ground, and/or for example depending on the level of humidity, etc.). The addition of air or CO2 or other gases is especially important, since, apart from speeding up plant growth, it can also protect its roots from rotting, since he main cause for rotting in roots is the lack of air when they are immersed too much in water. This addition of gases such as for example air or Oxygen or CO2 can be used also in combination with hydrophonic or hydrostatic irrigation methods, since the main problem that limits the use of such methods to only a limited variety of plants is that in many plants the roots rot under such conditions due to lack of air. However, adding for example air or Oxygen instead of CO2 is more preferable, since the absorption of CO2 in water makes them acidic. Since (unlike leafs) the roots need Oxygen, adding Oxygen to the water supply can help the plant thrive even at levels of 100% humidity. Another possible variation, which can be applied in combination with any of the other variations, is adding a feedback system for automatically reporting problems for example to a central control unit, such as for example flooding or blockages. One way of accomplishing this is by allowing for example each valve or sensor at the side channels to report back the approximate amount of water passed by it and/or the percent of time it remained open and/or for example to report significant changes in conditions, such as for example suddenly finding much more humidity in a certain area, or finding that the area remains dry despite the attempts of the sensor to open the valve. (However, increase in humidity can also be caused by rain for example so this is preferably reported to the user by the central control only if it deviates significantly from other sub-areas). However this can make the system a little more expensive. Another possible variation is to use a hierarchy of more than 2 levels, so that mere are not only main pipes and side channels but also one or more intermediary levels, and preferably intermediary junctures are responsible for indication and/or reporting of such problems, which is cheaper to implement, since in this case only these junctures have to be smarter. For these junctures the more preferred variation is that they simply have a cheap water-meter and report back to the center and/or to the main supply of each main pipe the approximate amount of water consumed per time period, and/or each main pipe for example has its own water-meter and reports this, and then either the human operator or a preferably cheap processor at the center can easily notice if there are significant deviations. Brief description of the drawings Fig. 1 is a top-view illustration of a preferable general configuration of a main pipe with sufficient water pressure which extends into smaller channels with a preferably much lower pressure that are preferably each controlled by its own cheap humidity sensor and cheap valve. Figs, la-b show a few preferable variations of methods for lowering the pressure at the side channels. Figs. 2a-i show a few preferable examples of mechanical sensing based on materials that change their shape when they become wetter or drier and thus efficiently close or open or gradually move a cheap and efficient valve electrically or mechanically and/or exert pressure for example on a flexible pipe. Figs. 3a-d show a preferable example of a cheap and efficient electrical moisture sensor that is both reliable and durable and preferably is not misguided by changes in temperature and/or salinity of the soil. Figs. 4a-c show a few examples of cheap electrical valves that preferably work with low water pressure. Figs. 5a-h show a few preferable examples of cheap and efficient sensors and water supplies that take advantage of the bottom dishes of flowerpots. Fig. 6 & 6b show a preferable example of using reversed capillary pressure at the end of the side channel, so that when the earth has reached a certain humidity level it automatically creates an equilibrium for example in osmotic and/or capillary pressure that stops the water flow until the level of humidity of the earth has sufficiently decrease again. Fig. 7 shows a preferable variation where each such side-branch has a humidity control based on air passage, with shared or individual air-tight containers. Important Clarification and Glossary: All these drawings are just exemplary drawings. They should not be interpreted as literal positioning, shapes, angles, or sizes of the various elements. Throughout the patent whenever variations or various solutions are mentioned, it is also possible to use various combinations of these variations or of elements in them, and when combinations are used, it is also possible to use at least some elements in them separately or in other combinations. These variations are preferably in different embodiments. In other words: certain features of the invention, which are described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention, which are described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination. The word "flowerpot" as used throughout the text, including the claims, can mean any type of pot or container for growing plants. The words "automatic faucet" or "automatic valve" as used throughout the text, including the claims, can mean generally any type of automatic control, including one with no moving parts, such as for example when using asymmetric capillary materials. Detailed description of the preferred embodiments All of descriptions in this and other sections are intended to be illustrative examples and not limiting. Referring to Fig. 1 & la-b, we show a top-view illustration of a preferable general configuration of a main pipe (10) with sufficient water pressure (such as for example 1 or a few atmospheres) which extends into smaller channels that go for example sideways (11) with a preferably much lower pressure that are preferably each controlled by its own cheap humidity sensor (13) and cheap valve (12). Each such side-channel can go for example to an individual plant, or to a preferably small area surrounding a number of plants, as desired by the user. This way, the valve (12) that is needed to control each of these small channels (11) needs much less force and therefore can be much cheaper than an ordinary electronic faucet (solenoid), which typically contains a motor and is: designed to deal with much higher pressures. Preferably the sensors are not too close to the end of the side channel in order to sense the real humidity in the near earth and. not to be influenced too much by immediate feedback of humidity at the end of the side channel. Of course other shapes and angles can also be used (and me channels can for example go only to one side instead of the two sides), and in each garden or field preferably more than one main pipe is used. Another possible variation is to use for example a hierarchy of more than 2 levels, so that between main pipes and side channels there can be also 1 or more intermediary level pipes, preferably with intermediate water pressure. The reduced pressure can be created for example by using long twisted small conduits at or before each side-channel that easily lower the water flow (such as for example in the Queen-Gil pipes, except that the side branches are preferably at larger distances from each other than the 10 cm interval in the Queen-