Experience of Using Multimonitor Workstations Under XFree86 4.x in VEPP-5 Control Room

TL;DR

This paper discusses the implementation and experience of using multi-monitor workstations with XFree86 4.x in the VEPP-5 control room, highlighting technical solutions and human factors involved.

Contribution

It presents practical deployment of multiheaded XFree86 4.x workstations for control systems, including modifications and human interface considerations.

Findings

Multiheaded workstations improve screen space for control applications.

Xinerama mode offers a better control interface than multiple separate screens.

Human-related issues arise in multi-monitor configurations, requiring adjustments.

Abstract

Modern PC workstations often provide more CPU power than required for most control applications. On the other hand, the screen space is always in short supply. One possible solution is to use more PCs, but in fact we need only more screens, not more keyboards, mice etc. PC architecture allows using more than one videocard, and X Window protocol is also aware that there can be more than one screen. But until release of XFree86 version 4 there was no freely available server capable of driving multiple ``heads''. We have been using multiheaded workstations under XFree86 in the VEPP-5 control room since early 2000 (currently 4 4-headed PCs plus several dual-headed). The ``Xinerama'' mode (one-large-screen) is better suited for accelerator control system than ``several separate screens''. When moving to this configuration we've encountered a number of, mainly human-related, problems, some of which required modifications to X server Additionaly, the ``style'' of performing control has slightly changed.