epicsdev 3.1.5

Helper module for creating EPICS PVAccess servers using p4p

epicsdev

Helper module for building EPICS PVAccess servers using p4p.

epicsdev is designed for:

• Rapid PVAccess server development
• High-rate data simulation and stress testing
• GUI-based monitoring and control
• Rapid instrument integration
• AI-assisted automatic device support generation

It integrates following EPICS IOC services:

• Autosave: automatically saves the values of EPICS process variables (PVs) to files on a server host, and restores those values when the server restarts.
• IocStats: provides support for PVs that show the health and status of the server, plus a few control PVs.
• caPutLog: logging of PVAccess put operations.

---

Installation

python -m pip install epicsdev

Quick Demo

Start the demo PVAccess server:

python -m epicsdev.epicsdev

Control & Visualization

Install optional GUI and plotting tools:

python -m pip install pypeto pvplot

Launch the control interface:

python -m pypeto -c config -f epicsdev

This provides:

• Device control panel
• Live waveform plots
• Real-time parameter monitoring

The screenshots can be seen here: control page, plots.

Phoebus Display

An example Phoebus display is provided: config/epicsdev.bob. Screenshot.

Multi-Channel Waveform Generator

epicsdev.multiadc generates high-throughput synthetic data for stress-testing EPICS systems.

For example, the following command :

python -m epicsdev.multiadc -s 0.1 -c 10000 -n 100

Will start a server, which generates:

• 10,000 noisy waveforms per second
• 100 points per waveform
• 40,000 scalar parameters per second

Monitoring GUI

python -m pypeto -c config -f multiadc

Text Put Logger

epicsdev.putlog hosts a writable PV named dump and appends any written text to a file.

Start the logger server (required argument: output file path):

python -m epicsdev.putlog /tmp/putlog.txt

Default PV prefix is putlog0:, so write text to:

pvput putlog0:dump "hello from client"

---

AI-Assisted Device Support Development

epicsdev is structured to enable automated server generation using AI tools such as GitHub Copilot.

Workflow Example

1. Create a new GitHub repository.

2. Provide an AI prompt such as:

   Build device support for Tektronix MSO oscilloscopes 
   using epicsdev_rigol_scope as a template and the 
   programming manual available at <PDF link>.
   

3. Within ~20–40 minutes, the AI can generate a pull request.

4. Review, test, make minor corrections if needed, then merge.

Real-World Example

Using this method, a server implementation for Tektronix MSO oscilloscopes was:

• ~99% correct on first generation
• Required only minor adjustments

---

Requirements

• Python 3.8+
• p4p 4.2.2+

Optional:

• pypeto
• pvplot
• Phoebus (for .bob display files)

---