Differentiating Hard and Soft Dependencies in Software Transport
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software development and deployment systems fail to effectively differentiate between hard and soft dependencies, leading to inconsistencies and errors when software objects are transported from a development system to a destination system, as consistency checks only verify syntax and not the actual dependencies within the destination environment.
Innovation Solution
The system identifies and classifies dependencies as hard or soft by recreating the destination system environment at the source, using an object versioning database and a binary predictive model to determine if a dependency would result in inconsistencies, thereby ensuring consistency checks consider the destination system's environment, and utilizing a machine learning model to train on parameters for accurate discrimination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If consistency checks only verify syntax without considering destination environment dependencies, then the checking process is simple and fast, but inconsistencies and errors occur when software objects are transported to the destination system
Solution Approach 1:
The patent creates a copy of the destination system environment at the source system by accessing an object versioning database that stores data characterizing versions of objects at the destination system. This copied environment enables the source system to perform comprehensive consistency checks including hard dependency verification without requiring direct access to the destination system, thus improving reliability while managing complexity through virtualization.
Solution Approach 2:
The patent performs dependency classification and inconsistency detection before the transport request is released from the source system. By classifying dependencies as hard or soft types and identifying potential inconsistencies in advance using the binary predictive model, the system prevents erroneous transports before they occur, improving reliability through proactive validation.
2Reliability
If all dependencies are treated equally without differentiation, then the dependency checking process is simple, but errors and unintended software operations occur at the destination system
Solution Approach 1:
The patent changes the parameter of dependency classification by introducing a binary predictive model that categorizes dependencies into hard dependency type (indicative of inconsistency) and soft dependency type (indicative of consistency). This parameter transformation enables differentiated handling of dependencies, improving verification accuracy while managing detection difficulty through automated classification.
Solution Approach 2:
The patent replaces manual or simple syntax-based dependency checking with a machine learning-based binary predictive model. This substitution automates the complex task of dependency classification and inconsistency detection, improving accuracy while reducing the manual effort and expertise required for dependency analysis.
3Reliability
If comprehensive consistency checks including hard dependencies are performed, then software object consistency is ensured, but the transport request processing time increases
Solution Approach 1:
The patent performs comprehensive consistency checks including hard dependency verification before the transport request is released from the source system. By completing all necessary validation operations in advance, the system ensures consistency reliability while enabling faster processing at the destination system since the validation work has already been performed.
Solution Approach 2:
The patent creates a virtual copy of the destination environment at the source system, enabling comprehensive consistency checks to be performed locally without requiring repeated communication cycles with the destination system. This copying approach maintains high validation reliability while improving processing speed by eliminating network latency and repeated verification steps.
Data Source
AI summary
Data is received at a source system and characterizing a modified first software object for transport to a destination system via a request and to update a first software object deployed on the destination system. A first dependency of the modified first software object on a second software object is determined by the source system. An inconsistency between the modified first software object and the second software object is identified by the source system and using the first dependency. Data indicative of the inconsistency is provided. Related apparatus, systems, techniques and articles are also described.