AI Coding Assistant Lingma:FAQ

To resolve the issue, follow these steps:

1. Check the plugin version and runtime environment:

   • Make sure you have upgraded the Lingma plugin to the latest version. Confirm that your operating system and system architecture support Lingma.

   • If you use a JetBrains IDE, make sure that the IDE's heap memory size is sufficient. Here are the steps:

     1. Click File -> Invalidate Caches.

     2. Clear the cache.

     3. Restart the IDE.

2. Check network connectivity and proxy configuration.

   • Make sure you can access the following addresses. If you can't, contact your company's IT department to add them to the whitelist:

     • Run this command:

       curl https://lingma-api.tongyi.aliyun.com/algo/api/v1/ping

       The return value should be pong.

     • Run this command:

       curl -I https://devops.aliyun.com

       The return status code should be 302.

   • If your company's firewall has restrictions and you cannot add the two URLs above to the whitelist, you will need to manually configure the proxy address (contact your company's IT for the specific address):

     Proxy format: http(s)://username:password@proxy_server_address:port

     • If the domain name is not saved correctly, you can change the configuration file manually:

       Path: C:\Users\username\AppData\Local\.lingma\config.json

       After modifying the http_proxy field, terminate the Lingma process and restart.

   • Clear DNS cache:

     • Windows: Execute the command ipconfig /flushdns.

     • macOS: Execute the command sudo killall -HUP mDNSResponder.

3. Clear the local cache and restart the IDE. Start the IDE with the admin role, open a non-empty project folder, and try the following steps:

   1. Terminate the Lingma process.

   2. Delete the .lingma directory:

      1. Windows: Path is C:\Users\[username]\.lingma

      2. macOS: Path is ~/.lingma

   3. After completing the above operations, restart the IDE.

4. Manually start the Lingma service if the issue is still not resolved:

   1. Go to the directory: .lingma/bin/x.x.x/CPU_architecture_64_system/

   2. Run this command:

      Lingma.exe start

      After the startup is successful, click the login button again.

5. (Optional) If you see an "incompatible program" message on Windows, it might be because the program wasn't fully extracted. Here's how to fix it:

   1. Click the Lingma icon in the bottom right corner, and select Advanced Settings.

   2. Change the extraction path to a non-C drive path, and make sure the path ends with an empty folder.

   3. Restart the IDE and try again.

      Note

      Some company intranets block binary programs.

6. Modify firewall or security software restrictions. If the Lingma process won't start normally or if you get no response when logging in, it might be because the firewall or the company's security software is blocking it. Follow the steps below:

   1. Add Lingma to firewall whitelist:

      Open Control Panel -> System and Security -> Windows Defender Firewall -> Allowed apps, add Lingma to the whitelist.

   2. Add process file to whitelist in company security software:

      Path example: C:\Users\xxx\.lingma\bin\2.1.x\x86_64_windows\lingma.exe

      Contact your company's IT support personnel for assistance.

   3. If the JetBrains plugin version is 2.1.4 or higher, switch Transport Type to stdio as shown in the figure below:

• Invalid logon parameters can cause a logon failure. Try clicking the logon button again later.

• If your network environment requires a proxy for internet access, upgrade Lingma to the latest version and ensure the following hosts are accessible. Add them to the whitelist if necessary. See network proxy configuration.

  curl https://lingma-api.tongyi.aliyun.com/algo/api/v1/ping
  curl -I https://devops.aliyun.com

• After the configuration is complete, terminate the Lingma process in the process manager and log on again in the IDE.

For remote development in VS Code and JetBrains IDEs, use the AccessKey pair of your Alibaba Cloud account to log on to Lingma. For more details, see installation and logon guide.

You can log on to Lingma using the AccessKey pair of your Alibaba Cloud account. For more details, see installation and logon guide.

Lingma supports HTTP, HTTPS, and Socks5 proxy configurations. Configure them on the plugin settings page. For more details, see network proxy configuration.

Run the following command in the Terminal:

curl https://lingma-api.tongyi.aliyun.com/algo/api/v1/ping
curl -I https://devops.aliyun.com

If pong is returned, your network is most likely to connect to the Lingma server.

For enterprise customers, Enterprise Dedicated Edition is available. See Billing for details.

Log on to Lingma using your Alibaba Cloud account. In the Lingma window of the IDE client, click the logon button to access the Alibaba Cloud logon page. For more details, see register an Alibaba Cloud account.

Enterprises interested in the Enterprise Edition can refer to Quick start for Enterprise Dedicated Edition.

JetBrains IDEs are compatible with versions 2020.3 (203.x) and later, including IntelliJ IDEA, PyCharm, GoLand, WebStorm, and others.

Visual Studio Code is compatible with versions 1.68.0 and later.

Visual Studio 2022 is compatible with version 17.3.0 and later.

• JetBrains IDEs: Windows 7 and later, macOS, and Linux.

• Visual Studio Code: Windows 7 and later, macOS, and Linux.

• Visual Studio 2022: Windows 10 and later.

Lingma supports mainstream programming languages, with enhanced performance in Java, Python, Go, JavaScript, TypeScript, C/C++, and C#.

Lingma does not store your code data:

• During code completion, your code context is needed but will not be stored or used for any other purposes. You retain full ownership and control of your data.

• For AI chat, chat records (excluding code) are processed only after you click thumbs up/down. This data is anonymized for algorithm improvements.

See Alibaba Cloud International Website Product Terms of Service for details.

The training data consists of high-quality open source code, technical articles, Alibaba Cloud service documents, and Alibaba Cloud SDK and API documents.

No. The system does not share your code snippets with other users. When using the large language model for code completion, we need to get your code context information to complete the completion, but the context information will not be stored or used for any other purpose.

The code generated by Lingma is only suggested code. Its usability cannot be guaranteed, and developers should review the code themselves and decide whether to adopt it.

• Case 1: Your code might consume too much CPU or memory resources during code indexing because of the large files in the project. To fix this, create a .tongyiignore file in the root directory of your project. Remove directories or files that don't need to be indexed. This will lower CPU or memory usage. The file works like the .gitignore file. After you change the .tongyiignore file, manually terminate the Lingma process to make sure the changes work.

• Case 2: Local Machine Model is enabled. This model uses about 500 MB of memory. If you do not need it, you can disable Local Machine Model on the Settings page when you use the inline completion feature.

The following error message is a known issue in VS Code 1.56:

"Error loading webview: Error: Could not register service worker: InvalidStateError: Failed to register a ServiceWorker: The document is in an invalid state."

It occurs when some Windows users run VS Code as an administrator. You can try the following solutions:

• Method 1: Start VS Code with the --no-sandbox command line flag:

  $ code --no-sandbox

• Method 2: Close VS Code, then try running killall code in the terminal.

Query no more than 10,000 data entries per page. Limiting the number of entries helps the system run smoothly and makes the user experience better. You can change this number based on how the system works.

API data is updated every 15 minutes, starting early in the morning each day. Several verification checks are used to ensure that the data is up-to-date, reliable, and flexible. At 02:00:00 in the morning, you can get the latest and accurate data set from the previous day.

To ensure a stable and efficient service, the limit for API requests is 100 queries per second (QPS).

Click your profile in the upper-right corner of the Lingma console, click Management Console, and then copy the Organization ID from the Basic Information page.