CodeR: Real-time Collaborative Code Editor - Key Points

Abstract

• CodeR is a real-time, web-based collaborative code editor and execution environment.

• Enables writing, running, displaying terminal results, and real-time collaboration with other users.

• Supports C, C++, and Java; includes chat and collaboration features within a project workspace.

1. Introduction

• Software industry trend: IDEs moving from desktop to web to reduce configuration time and enable ubiquitous access.

• Real-time collaboration (RTC) combines communication and collaborative editing in a single environment.

• RTC building blocks: Unified Communications, Presence Information, Contextualization, and E-Collaboration Portfolio.

• RTC enables two- or multi-person editing, sharing, and synchronization of code changes in real time.

1.2 State of the Art

• Goal: an online code editor with real-time collaboration, syntax checking, run/build through a terminal, and chat.

• Features include workspace creation, editing, running/building, auto-complete (C, C++, Java), and import/export of projects.

• Aims to improve programmer collaboration and project development efficiency.

2. Related Research

2.1 Real-time Collaborative Programming

• Real-time text editing allows multiple programmers to work on the same document/file and see updates immediately.

• Web-based editors/components for embedding in IDEs: Ace and CodeMirror; examples of collaborative systems include EtherPad and Google Docs.

• Collaboration involves joining/leaving sessions; a join protocol includes JOIN, START, READY, and FINISH messages to synchronize new participants.

2.2 Integrating Collaborative Program Development and Debugging within a Virtual Environment

• IDEs with real-time collaboration for coding, compiling, running, and debugging (e.g., Idaho Collaborative IDE, ICI).

• Applications enable teachers/students to collaborate remotely and share development tasks in real time.

2.3 Real-Time Collaborative Coding in a Web IDE

• Collabode: a web-based IDE for Java with real-time collaboration; uses EtherPad for editing and integrates with Eclipse features.

• Models of collaboration: test-driven programming, micro-outsourcing, and mobile instructor.

3. Design & Method

3.1 CodeR Features

• Login via the Facebook plugin for easy access; supports C, C++, and Java.

• Run/build, download source code with folders, real-time collaboration, and chat.

• Email for bug reports; after login, users see a workspace with projects/folders/files; real-time updates with auto-save (inactive for 1 minute).

• Collaboration is project-based: invite coworkers, set access restrictions, and communicate via chat.

3.2 Operational Transform Algorithm

• Real-time collaboration relies on Operational Transformation (OT) to maintain data consistency across users.

• Components: integration algorithm (receives/transmits/executes operations) and transformation functions (combine concurrent operations).

• Transformation example: when op2 arrives after op1, op2 must be transformed to op2' so that applying op1 then op2' equals applying op2 then op1.

• Case distinctions: Update vs. Insert/Delete.

  • Insert/Delete changes the linear address space; Update does not.

  • If an update is transformed against an insert, its position may shift (e.g., the update position increases if it occurs after the insert).

• OT in CodeR supports concurrent editing, ensuring consistency across all collaborators.

4. Result

4.1 Application Architecture

• User authentication leads to loading of workspace data; the Facebook friends list is used to add collaborators.

• Chat and collaboration are integrated into the workflow; collaboration data is exchanged via socket.io.

• The project/file tree is displayed; actions include create, delete, rename, move, open, download, and upload.

4.2 Collaboration Features

• Real-time collaboration: invite friends from a friend list; anonymous users can still connect with other anonymous users.

• Chat supports the display of user names and dates and search functionality for inviting specific friends.

5. Discussion

• Evaluation used Apache Benchmark to measure HTTP server performance under load.

• Test setup: a URL tested with -n 1000 requests and -c 1000 concurrency (i.e., $n= ext{1000}, c= ext{1000}$).

• Hardware test specifications:

  • SPEC 1: Pentium Dual Core E2180 with $4$ GB RAM.

  • SPEC 2: Core i7-2600 with $8$ GB RAM.

• Results (SPEC 1):

  • Total data transferred:$3575000$ bytes; Time:$28.842$ s; Requests per second:$34.67$Time per request (for 1000 concurrent): $28841.650$ ms; Transfer rate:$121.05$KB/s.

• Results (SPEC 2):

  • Total data transferred:$3501000$ bytes; Time:$0.442$ s; Requests per second:$2264.06$; Time per request:$0.441684$ ms; Transfer rate:$7740.70$KB/s.

• Overall, the average load times and throughput improved with higher-spec hardware; mean/max values reported for each test indicate response-time behavior under load.

6. Conclusion

• CodeR provides a web-based, real-time collaborative coding environment with a terminal for execution results, chat, and project management.

• Supports languages:$C, C++, Java$Main features include workspace creation, real-time collaboration, chat, and terminal execution.

• Future work suggested: automatic UML design generation from code, multi-file editing across projects, and repository-like features akin to GitHub.