Post by sturmghost on Initialization of visualization variables and cyclic code execution ^{CODESYS Forge talk (Post)}
I'm looking for a smart and short way to implement initialization of visualization variables depending on the visualization input. For an easy example consider a rectangle which rests at XPos := 0 when the input state is false and at XPos := 50 when the input state is true. My visualization variables look like this: VAR_IN_OUT State : BOOL; END_VAR VAR XPos : INT; END_VAR I put this rectangle via a visualization frame element into another visualization and link a frame reference variable with the state to it. If the variable is true, the rectangle should rest at XPos := 50 and false at XPos := 0 at visualization init but how should I assign the 50 or 0 to the internal visualization variable XPos? I would need some init-methode for the visualization but I dont want a global init-method for such tasks. I want to do it inside of the visualization element but I can't see any solution for this? It would be good to be able to define ST-code within the visualization element which runs cyclic at each VISU_TASK task-cycle then I could just check the input state and change the XPos accordingly. Does someone have a solution?
Last updated: 2023-10-01

Post by captaincookie on increase default string length in queue ^{CODESYS Forge talk (Post)}
Hello, I'm using Codesys V3.5 SP18 Patch 4. In the ElementCollectionExample Project from Codesys, I test the SimpleQueueExample in a Control Win V3 x64 environment. I try to add a string of 95 characters length to a queue. The default length of strings is defined as 80 characters. In the initialization of a string variable, it is possible to increase the length by the definition of e.g. STRING(1000). But when I write the string defined like this to the queue, only 80 characters are written to it and the rest is missing. I think the default length is still set in the queue definition, so it is necessary to change this, isn't it? Is there any option to increase the default length of strings in the queue? Attached you can find the used project. Thanks in advance. If any information are missing or my description unclear please let me know.
Last updated: 2023-10-05

Post by mondinmr on Jitter problems on imx8 ^{CODESYS Forge talk (Post)}
We are experiencing serious jitter issues on an ARM64 imx8. Until now, on rpi and beaglebone derivatives, and on Intel, we never encountered jitter issues unless they were already present on the device. In this case, as can be seen from the attached graph, we have excellent maximum latency. The cyclictest on the isolated core dedicated to IEC CODESYS tasks shows a latency of 37µs, with an average of 14µs. Usually, on all other devices, with an RT kernel, with the appropriate parameters to avoid frequency throttling, with properly managed IRQs, and with the disabling of large page sizes, the jitter measured by CODESYS is very close to the cyclictest latency. However, on this imx8, we are noticing a fluctuating value of +/- 300µs that seems to be added to this latency, as if something occasionally miscalculates the sleep times by exactly +/- 300µs. Two screenshoot attached. - Result of cyclictest - Jitter on CODESYS Tasks are assigned to isolated core!
Last updated: 2023-10-18

Post by martinlithlith on Mux I2C ^{CODESYS Forge talk (Post)}
Hi! I´ll try to get this thread going again as i´m having simliar problems. Right now i get a red triangle on my connected devices (i´m testing with a BME280 sensor and a PCA9685 resvo driver). The TCA9548 have the green symbol beside it. On the "red" devices i get that the bus is not running. I have tried to either run it using the parent bus cycle setting or creating a new freewheeling task prio 1 but nothing helps. I have compared my settings to the settings from the example from the MUX package and i have read (i think) every thread in forge that could be of interest. EDIT: The ic2detect only shows the mux (70-70) and the servo drive (70-76) so i´ll look in to if i managed to burn the 280-sensor while soldering). Another question regarding the 280 sensor. I have a sensor GY-BM E/P 280 that should record humidity and temperatur. Could this be a problem with the library? Does anyone have an idea of what i might be doing wrong? I have some experience from Codesys but not much so this could be an easy one. All suggestions are welcome. I tried to upload the project archive but it did not work. Should i upload the export or how could i show you my horrible project? Best, Martin
Last updated: 2023-10-22

Post by vassilis91 on EtherCat-Rexroth Drive lost connection or no? ^{CODESYS Forge talk (Post)}
Hi all, i am encounter a mystery problem with a build with rexroth drives at my work. At my company we are make a machine with two Rexroth Indradrives . The two drives are goes by ethercat protocol . To be more specific i am going from my plc to the Beckoff El1100 coupler and after to the first drive and from the first drive to the second. The second drive works with no problem but the first(motor) sometimes without any pattern start to make a mysterious noise like stop and suddenly goes on again. If i couldn't hear the noise from the motor i couldn't Imagine that this happen. After a lot of research i see that sometime i have some strange zeros from the drive to the plc? Really i am in a deadlock and I don't know how to continue. From the other side the values on the other drive(second)are stable . The problem maybe: https://youtube.com/shorts/g7PCFLUaDUI?si=eiq0F2IzhQZax2e2
Last updated: 2023-10-25

Post by nizrahel on OpC UA server on Raspberry Pi 64 SL runtime ^{CODESYS Forge talk (Post)}
Hi all! I'm working on an home project using a Raspberry Pi 3 device with Codesys for Raspberry Pi 64 SL runtime. I would like to create the HMI interface with Siemens WinCC Unified (or Advanced) configuring the Raspberry as OPC UA server. I tryed everything, but I didn't find the way to let the OPC UA connection works using UaExpert. I tryed to use anonymous connection, authenticated connection and the ip address instead of the server name in the path connection. There is no way to connect to the OPC UA server and I was asking myself if this feature is really supported for the runtime version used. Another strange thing I noted, is that is not possible to quickly create an OPC UA certificate in the Security Screen, as show in any tutorials. I have correctly created the Symbol Configuration enabling the OPC UA feature. I attach to this topic some printscreen of the actual configuration (sorry for the italian language interface, but I didn't find the way to switch to english). Thank you very much in advance for your support and advices!
Last updated: 2023-12-03

Post by ofey on Testing of Codesys program ^{CODESYS Forge talk (Post)}
Hi! I thought to hear what you guys think is the best way to test the program in my scenario. I have some PLC programs that have been made and deployed on different PLC's. The program has added the specific devices and IO that is applicable. I have made a test environment for everything outside of the PLC, ie. the same SCADA system that is talking to the PLC, and a program running locally is simulating the plant (level regulation of different tanks). Pr. now I have had to manually change the variables that are reading/writing values from the IO so they read/write via OPC instead. I also have to change the device tree from the specific PLC rack, to the soft PLC. So I kind of end up with two different programs that I have to maintain. I was just wondering if you believe there is an easier way to do this. So I can have the exactly same program and kind of switch between the soft mode, and the "real" mode.
Last updated: 2023-12-07

Post by abinvest579 on TCP Client ^{CODESYS Forge talk (Post)}
Hi All, I am using codesys based plc to read and control magna power DC supply over LXI protocol using SCPI command. To connect device required socket programming for that i am using Net base service library function block TCP_Client to establish connection over TCP/IP , TCP_Write to write SCPI commands and TCP_Read to read data from Magna power DC supply. I have attached screenshot for logic developed. I am able to connect and write commands to device but i not receiving any information from device. For writing Commands i also tried various end of line character like \n,<nl>, A, D, 10, 13, 0x0A, 0x0D but not received any status from device. For testing purpose i am using simple command '*IDN?' to read data. Also if tried to check on hercules software what data actually plc read, first it read correct data after that is read wrong data as see in screen shot. 'Hello World' was correct read and after that Codesys in replaced with 'Codesysorld' that is not correct way. Same issue also appeared for send data. Any one have face same issue how i can read data directly from device correctly and send data please suggest solution on this.</nl>
Last updated: 2024-07-01

Post by wbj0t on What the right way to update TCP/COM slave device holding registers ^{CODESYS Forge talk (Post)}
Hi all :) I have an plc. This plc has option to be as slave device. There are TCP -> Serial Slave Device and COM -> Serial Slave Device. There are 4096 ModBus holding registers with mark "internal modify" for the both devices above. So, COM Slave has 4096 registers and TCP Slave has 4096 registers. And, finally, I have an ARRAY[0..4095] OF WORD. This array glued with COM Slave and TCP slave. Yes, I need plc's opportunity to be communication both TCP and COM ways for the same registers. For example, the next task structure: * TASK_0: 1. pou_READ_Registers 2. pou_0 3. pou_1 4. ... 5. pou_N 6. pou_WRITE_Registers At the first pou I read registers for the other pous, after job we have sort of state of the plc. At the last pou (6) we write modified registers. It is a good scenario: some one wrote registers for the job, after job we save these (modified) registers. But, what if registers wrote between 1 and 6 pou by operator, for example, at the 3 pou -> the last pou 'pou_WRITE' will rewrite this request from operator. What the right way to work with this logic? Thanks you.
Last updated: 2024-07-01

Post by psychoengineer on Cannot Save CSV File on Windows Computer Using CODESYS and Control Win ^{CODESYS Forge talk (Post)}
I am currently working on a project using CODESYS and Control Win on a Windows computer. I've encountered an issue where I am unable to save a CSV file. I have followed the standard procedures and used the appropriate file handling functions, but the file is not being created or saved as expected. IF xStartTest THEN sFileName := '\logs.txt'; sFileName := CONCAT(gvl.g_sMainDirectory, sFileName); sFileNameCopy := '\CAAFileCopy.txt'; sFileNameCopy := CONCAT(gvl.g_sMainDirectory, sFileNameCopy); //eWriteState := FILE_OPEN_TRIGGER; xTestDone := FALSE; xError := FALSE; xStartTest := FALSE; END_IF VAR_GLOBAL g_sMainDirectory : STRING := 'C:\LogFiles'; // file path for CSV file END_VAR Despite this, the file does not appear in the specified directory, and no error messages are being generated. I've checked the directory permissions and ensured that the path is correct. Has anyone else experienced a similar issue or have any suggestions on how to resolve this? Any help would be greatly appreciated. Thank you!
Last updated: 2024-07-04

Post by dekelec on Codesys 2.3 & Peak PCAN ^{CODESYS Forge talk (Post)}
I use Peak USB adapter daily to download from CoDeSys 2.3 and 3.5 to IFM, EPEC and other controllers. The process: - First you need to install the appropriate driver. Link: https://www.peak-system.com/Drivers.523.0.html?&L=1 - Restart the computer - Start the CoDeSys application - Change the name of the adapter in Communication parameters, as mentioned in previous comment. Write the name exactly as written "Peak_USB", as it could also be case sensitive (defines which .dll file to use). If this doesn't work I would contact the vendor of the controller to find out the procedure. In the attachment I've added a FAQ document from IFM regarding using PCAN USB. P.S. I've noticed in your picture of communication parameters a channel via TCP/IP is mentioned. In this case a USB to CAN connection is not being used. First you should change the channel/gateway to connect via CANbus or connect using an Ethernet cable.
Last updated: 2024-07-09

Post by simone on Cannot connect to webvisu on localhost ^{CODESYS Forge talk (Post)}
As I wrote before on the title, I have difficulties to connect to see webvisu from localhost. The story: while waiting for the plc+hmi to arrive, I'm preparing the software, mainly the visual parts trough the control Win3 x64 (same version as the ide 3.5.20.10). The problem is that I can't see the webvisu with the installed webbrowser. I don't know if this is a problem, but I changed the port to 9001 (randomly "free") because 80, 8000, 8080, 8088, 3000 are all occupied by other projects (all on the Apache server). Is there something wrong with the config file? This is the url: localhost:9001/webvisu_app.htm and here the config file part "CmpWebServer". [CmpWebServer] ;The handling of startup/shutdown of the webserver: ;0->automatically start, shutdown only on shutdown of the runtime;1->automatically start, could be shutdown on demand ;2(default)->start/shutdown on demand;3->start on demand, shutdown only on shutdown of the runtime StartupType=0 ;The TCP port the webserver listens on WebServerPortNr=9001 LocalAddress=localhost ;LocalAdapterName=LAN-Connection LocalAdapterName=EthernetPCPLC ;LocalAdapterNameUnicode=L\00A\00N\00-\00C\00o\00n\00n\00e\00c\00t\00i\00o\00n\00 LocalAdapterNameUnicode=E\00t\00h\00e\00r\00n\00e\00t\00P\00C\00P\00L\00C\00 Thanks for all the help!
Last updated: 2024-07-11

Post by trusty-squire on CNC - How to manipulate SMC_GeoInfo objects ^{CODESYS Forge talk (Post)}
I have an application using CNC GCode interpolation, but I need to modify the GCode provided to the PLC based on certain parameters. I am currently attempting to modify the SMC_GeoInfo objects in the SMC_OutQueue using the code below. Note that all the other code is pretty standard and works fine, but when I add the below it errors. PROGRAM TEST VAR fbReadCncFile : SMC_ReadNCFile2; fbCncInterpreter : SMC_NCInterpreter; arrCncInterpreter : ARRAY[1..99] OF SMC_GeoInfo; pGeoInfo: POINTER TO SMC_GeoInfo; giGeoInfo: SMC_GeoInfo; // ... END_VAR // ... Some code here in order to read CNC file using SMC_ReadNCFile2 and provide to SMC_NCInterpreter pGeoInfo := SMC_GetObj(poq:=ADR(fbCncInterpreter.poqDataOut), n:=1); IF pGeoInfo <> 0 THEN giGeoInfo := pGeoInfo^; // Do some manipulation here, then update the queue at the same position MC_SetObj(poq:=ADR(fbCncInterpreter.poqDataOut) , n:=0 , pgi:=ADR(giGeoInfo) ); END_IF It throws an error when I get to the line giGeoInfo := pGeoInfo^; Error: EXCEPTION [AccessViolation] occured: App=[Sim.Device.Application], Task=[PathTask] How do I use SMC_GetObj and access the data? It creates a pointer with the value as shown in the photo, but all the dereferenced values say dereference of invalid pointer.
Last updated: 2024-07-26

Post by grant32 on RAMP_REAL FB is retaining my output value when i try to re use the FB ^{CODESYS Forge talk (Post)}
Hello, I'm relatively new to PLC programming. I am trying to creat a ramp down function, I have an example here of what I am using. I can get the function to work, the issue is that the OUT (variable3) value is being held by the FB so after I use it once it no longer works. In this example i set variable3 to 100 and the rest is as seen, over the course of a minute it ramped down to 10 (as desired). But when I deenergize TEST_START, set variable3 back to 100, and then re-energize TEST_START it immediately jumps back to the previously held value 10. The reset on this block just pauses the function and holds the last OUT. In my case I will need the function to start OUT at a set value, ramp down to the IN (variable2) value, and then accept a new OUT value to ramp down next time the block is energized. Hopefully this made sense, if anyone has any advice that would be awesome!
Last updated: 2024-07-29

Post by denkihitsuji on Content update through Data URI on Web Brower on Visualization Toolbox ^{CODESYS Forge talk (Post)}
Objective: I want to dynamically update the content with strings and JPEG images received from a TCP/IP socket. For this purpose, I am exploring the possibility of utilizing the Web Browser feature of the Visualization toolbox. Example: In a standard web browser, entering data:text/plain;charset=utf-8;base64,SGVsbG8gc3RyYW5nZXI= in the URL window will display "Hello stranger" on the webpage. Issue: When attempting to use this method in the Web Browser of the Visualization toolbox, the content does not display as expected. Instead, it shows the error:'Navigation to the webpage was canceled.'. Request: I seek guidance on how to resolve this issue so that the Web Browser in the Visualization toolbox correctly interprets and displays the data URI content. Alternative Solution Welcomed: Or, if it is possible to directly use Codesys Structured Text (ST) variables within an HTML page, this approach would also meet my requirements. Thank you for your attention.
Last updated: 2024-07-30

Post by mubeta on parker servo and position ^{CODESYS Forge talk (Post)}
In general, drives have parameter sets to adjust behaviour in such cases: position recovery, adaptation with resistant torque, out-of-position window failure, etc. (Much depends on the type of configuration set for the axis: positioner, speed controller, torque controller, etc.). I have used that type of drive very little, in only two cases and a long time ago. To tell you what is normal is impossible. I know by reputation they are very simple objects, without too many pretensions and adjustments possible. Try running traces by monitoring actual position, motor current, torque and following error. These are the most sensible curves to understand what is going on. Maybe by fine tuning the motor tuning, the current controller or the speed controller you will get something closer to what you want. For example, from DS, the motor has a standstill torque of 3 Nm. Observe with traces whether this is effective: motor current trend. And afterwards, when you have your own accurate picture of the situation, you can also hear from the product's local support to see if you can steal a few more things from it performance-wise.
Last updated: 2024-08-05

Post by paulorb on STRUCT AT %MW1000 ^{CODESYS Forge talk (Post)}
Hi I am trying to use STRUCT to parse commands from Modbus/TCP, Modbus client will fill the bytes related to a specific command (type, parameters), then PLC (Codesys) will access the command parameters. A STRUCT will simplify a lot the process of parsing the data, so I don't need to parse byte per byte, address will be mapped to a struct and I can use it for parsing the data. From what I understood, the compiler does not allow me to specify the location of the struct for a %MW register. VAR_COMMAND AT %MW1002: structGenericModbusCommand; This will throw: Bad declaration, use '%ML' for 'structGenericModbusCommand' variable But If I change it to %ML, when you go online you see that it is actually allocating the struct in another memory address. When we go online, on GVL we can see: VAR_COMMAND. structGenericModbusCommand %ML1002 commandType COMMANDTYPE %MB8016 ... %MB8020 I am not specifying a pointer or a reference it is a struct type, why is it mapping to a %MW location (random one). How can I control where it is mapping to? I need a consistent way to map STRUCT to specific modbus address. Is it possible? Thanks, Paulo Note I am using a Codesys OEM: Schneider Machine Expert Logic Builder (M241)
Last updated: 2024-08-13

Post by abauza on Image Pool using in other libraries ^{CODESYS Forge talk (Post)}
Hi im Albert, im developing libraries for my PLCs and i have some issues on visualization part. First of all I have done a library that have a image pool with a list of 4 gifs "ImagePoolErrors" (i've tested in other projects and it works) but the problem comes when i create another library that in the visualization part contains images from my image pool library, when i simulate a program that includes a instance of my second library object the visualization images dont show, any of them, i've checked all the references and tried some options of the image properties but without any difference. Here is the screenshots of the first library (that contains Image Pool) and the second library (that includes images of the image pool first library) and the visualization simulation (with the missing image gif issue). Any idea of what i forgotten in the process? If you need any detail of the library or of the project or you need the library file, ask me, thank on advance
Last updated: 2024-08-19

Post by ruobian on Analog Input Delay Timer ^{CODESYS Forge talk (Post)}
Hello there, I am new here and in programming. I need help with the basics. I am trying to do what I mentioned in the title. I have an analog input. So I have a real or integer data type value. I want to delay it. TON and TOF only work with bool. I think there is a function block that has two inputs, 1 for real or int and 1 for bool. And if bool is true, it will give the output as real or int. I don't know but I need something like this. Actually, it is not exactly like that. In other words, it will not show the real value at the input at the output after a certain period of time. I want it to show the real value from 2 seconds ago continuously. The purpose of doing this is to compare the real value I received with the value from 2 seconds ago and find out whether it went up or down. I am using only FBD. Please help me with this. Thanks in advance.
Last updated: 2024-08-20

Post by rossanoparis on Clarifications regarding the appropriate runtime system to use ^{CODESYS Forge talk (Post)}
Hallo all I would appreciate some clarifications regarding the appropriate runtime system to use. I've been using 'KUNBUS CPUs' and 'CODESYS Control for Raspberry Pi MC SL' for the last 4 years. Recently, there have been some changes in the CODESYS licensing policy. [ https://store.codesys.com/en/howto_applicationbasedlicenses ] My questions are mainly two: 1) If I need to start a new project using 'KUNBUS CPUs', which is the correct runtime system to use. 2) Old projects built using 'CODESYS Control for Raspberry Pi MC SL' should be updated to another RT system? Current RT systems ... a)CODESYS Control for Raspberry Pi MC SL b)CODESYS Control for Linux ARM SL c)CODESYS Virtual Control SL May I count on receiving an explanation to understand the differences among the runtime systems listed above? A detailed and well-organized answer would be greatly appreciated. Kind regards Rossano
Last updated: 2024-09-23

Post by lbartik on J1939 TX/RX PDU1 ^{CODESYS Forge talk (Post)}
Is it possible to transmit a PDU1 PGN using the J1939 manager and local device? What if I need to transmit a PGN RQST via 0xEA00 (PGN 59904)? How would I do this P2P (Destination < 0xFF) or broadcast (Destination 0xFF)? I don't think this is possible for any destination type. I don't think it's possible to receive broadcast PDU1 messages either. For example, an address claim PGN 60928 (0x18EEFF80) is addressed to all nodes (0xFF) and not any specific local device so it will be filtered out. These are major oversights in the IoDrvJ1939 library to not support every PDU1 RX/TX scenario. IoDrvJ1939 supported scenario: 1. Receive (RX) PDU1 (P2P) destination-specific (PDU-specific < 255) to local device with matching address IoDrvJ1939 unsupported scenarios: 1. Receive (RX) PDU1 (P2P) global (PDU-specific = 255) 2. Transmit (TX) PDU1 (P2P) global (PDU-specific = 255) 3. Transmit (TX) PDU1 (P2P) destination-specific (PDU-specific < 255)
Last updated: 2024-09-23

Post by bjarne-pagaard on Communication between applications on same device/controller/runtime (Win RTE 3.5.20.20) ^{CODESYS Forge talk (Post)}
Hi, I would like to divide a project into multiple applications - as a minimum: one handling visualization and Alarm Manager, one handling I/O and plant control logic. But how to exchange variables between the applications? What have you done to get such a solution? In versions 3.5.19 and earlier, you can have Child applications, where the children can access a GVL in the Parent application. Children apps is no longer possible in 3.5.20 - but you can have 'sibling' apps - Several apps directly under PLC Logic, that is. But how do they best communicate? The Communication Manager / Data Sources Manager is sort of possible via OPC UA, but it seems like overkill and with some limitations - for example no ARRAY OF STRUCT possible this way. If you have separate devices in your project, you can exchange data via 'CODESYS ApplicationV3', but not other applications in the same device (see attachment Datasources.png). You can do it via 'Select the project type'->'Other Project' and select the same project file, but this leads to crashing the Development system when working with the variables afterwards. It would be great to hear your thoughts / experiences on this topic - Bjarne
Last updated: 2024-09-27

Post by bjarne-pagaard on Communication between applications on same device/controller/runtime (Win RTE 3.5.20.20) ^{CODESYS Forge talk (Post)}
Thanks for this - I have now been looking in to just having an extra (non-RTE) runtime on the same machine, as you suggest. I will probably proceed this way. Even though it still seems very odd to me, that variable exchange to a different runtime is easier than between applications on the same controller. There will be some fun with the licensing, it seems. Does anyone know if it is possible to specify which license container/specific license the runtime is going to use at runtime. Let's say I have two available licenses with different resources available, for example one license allows multicore and multiple EtherCAT channels, but not many visu tags (for the RTE). Another with less resources but a lot of visu tags (for the Visu controller). Let's say the program in one of the controllers could be fully licensed by either of the two available licenses, but picks the 'largest' first. A bit later, the larger application starts, but is left with the smaller license.. How can I make sure that the 2 runtimes select the right license at startup?
Last updated: 2024-10-01

Post by egau on Hard shutdown: no code on device after power on ^{CODESYS Forge talk (Post)}
Hi, We have a machine running Codesys on a Windows IPC (CODESYS Control Win v3 - x64). When we hard shutdown the machine, the code on the PLC sometimes becomes "corrupted" after a power on (When trying to login to the PLC, we get the message "The application 'Application' does not exist on device."). I've noticed these errors in the logs, but I'm not sure what to make of them. We are not using any RETAIN variables in our code, although I'm not sure about referenced librairies. (we are using the persistence manager). I'm pretty sure that not doing a graceful shutdown is the root cause of this. This being said, what is the proper way of doing a graceful shutdown? Is putting the Codesys application in "stop" sufficient? I know how to implement this, so if it works then it would be a quick fix. However, I think we need to add a UPS, which would detect power loss and inform the PLC that it needs to initiate its shutdown, and then the PLC would shutdown gracefully. Any help with this will be greatly appreciated :)
Last updated: 2024-10-03

Post by egau on Hard shutdown: no code on device after power on ^{CODESYS Forge talk (Post)}
This explanation aligns with the issues we’re experiencing with our machines. The scenario I described is happening with one of them (let’s call it Machine A). This machine is identical to another one (Machine B), except for some custom code that facilitates communication with an external Beckhoff PLC for MES integration in the production line. Machine B has been powered on and off daily for at least three months and has never had this problem. Given that, I highly doubt the MES custom code is the cause of the code corruption. This being said, your explanation does seem plausible. However, if faulty EtherCAT cables were the issue, why wouldn't the error occur during normal operation? Why would bad cables only cause problems during a hard shutdown? One last question: When you encountered this behavior, did you see a similar error in your logs? i.e, an "AccessViolation" exception in the EtherCAT Task?
Last updated: 2024-10-03

Sort by relevance or date