Detailed descriptions of four major functions of clearcase (build management)

Using clearcase, the process of constructing software can be compatible with traditional methods. You can use a self-made script or local make program for the data controlled by clearcase. However, the clearmake and omake tools for the upward compatible setup of clearcase provide important features for the construction: automatically complete tasks, ensure reliability of reconstruction, storage time, and support the establishment of parallel distributed structures.
Supports the establishment of makefiles for UNIX and windows
Clearcase includes two independent setup programs: clearmake and omake. These two programs combine the main creation features of clearcase, including configuration lookup, binary file sharing, and configuration records. The clearmake program is mainly applicable to users who use Unix makefile (gnumake. Omake is mainly applicable to users who need to be compatible with Windows programs (including Borland make, Microsoft namake, intersolv configuration builder, and opusmake.
Automatically detects the associated original file, including the associated header file
Clearmake and omake use of the configuration of the current original file (to some checked and checked files) to flexibly create the entire or part of the software system in the view. Clearmake and omake do not need to describe the associated header file (or any associated original file) in makefile ).
In the clearcase development environment, the original object plays a decisive role. The source object is a file object or directory object created by clearmake and omake. Typical source objects should include the object modules generated by the document system, executable programs, library documents, specification documents, and content tables. Source object components include: the object name created as the target; independent source object ID; Data container pointer (the file that stores the data generated by the creation script ); configuration record pointer (information list); and reference calculation (indicating the view number currently displayed on the source object ).
Automatic Tracking and creation to generate a permanent data list
During the execution of the build script, clearmake and omake run a build trace in the Multi-version File System of clearcase. This mvfs records the version of each read or executed file in the connected vob; it can also comment on which files are created (or overwritten ). After the creation script is executed, clearmake and omake write the tracing data to the configuration record and store it in the vob database. The vob database pointer allocates configuration records to the source object of each creation process.
A configuration record is a list of source object information, including its content and useful information at the time of creation:
■ Versions of file elements stored in vob and used during reconstruction-including source files and tools (such as compilers) under the clearcase control ).
■ Every private view file used during creation.
■ Non-clearcase files used during makefile.
■ Create script text and all extensible macros.
■ Operating system version and CPU type.
■ The user who executes the creation process; The host that executes the creation script; the view set by clearmake and the date and time when the creation process starts.
The clearcase configuration file allows source objects to be compared-instead of relying on Object Data, it relies on their configuration information. The diffcr (compare config REC) command can output differences between different configuration files, including:
■ Source file difference, not the timestamp of source file modification.
■ Differences in script execution during creation, including different macro values in makefile.
■ Unnecessary differences that do not affect the creation of clearmake or omake, including creation time/date, host name, and view name.
Developers share binary code, storage of time and storage space
Based on makefile, this is an important aspect to avoid unnecessary setup. Clearmake and omake have excellent creation policies. They are specially designed for parallel development solutions. Clearmake and omake can detect the on-site conditions through the configuration file to detect which source objects can be shared among multiple views. This tool also supports disk storage and time storage. Clearmake and omake provide three alternative creation methods:
■ Reuse existing targets in a view-clearmake and omake use a technology that is more proficient than timestamp. The configuration information can be used for source version comparison, script comparison, and option comparison.
■ Run the transfer creation script-make, clearmake, and omake run the target creation in the same way. However, clearmake and omake can track the creation process and distribute the tracing information to each file related to the reconstruction process. File and Its configuration information group become the source object.
■ From a wink-in source object in a specific view, clearmake and omake can learn about multiple instances of the same target created in other views earlier. After verification, the correct source version, options, and creation scripts are used to create other instances. clearmake and omake perform a wink-in operation on The View. A source object can now be shared by other views.
Parallel distributed creation across different models
Clearmake supports distributed creation (using execution scripts on other hosts) and parallel creation (consistent execution of creation scripts ). For example, clearmake can be established by three Parties, and all processes are executed on a multi-processor computer server. In a LAN, it can be distributed across all workstations.
Clearmake also supports cross-development environments.
Automatic balanced loading and distribution across multiple hosts (UNIX)
Clearmake has a cutting-edge load balancing technology that can optimize distributed execution. The user-specified function is equivalent to the distributed server creation host, and sets variables, including: time, machine loading, and control the user ID created by each machine. Clearmake can automatically balance loads across these hosts for distribution creation.