P030502a: ODMref1.0 Mitigation Experiment #1

A running diary and job jar of work items and notes on ODMref1.0 Mitigation Experiment #1

Status	Date	Description



done	2005-03-27	Stitch P050302 to the material we have so far.
done	2005-03-27	Rename DeRun01 as a Test Harness, not Test Program.
	2005-03-27	The ODMA DMS entries must support aggregation. It is not clear whether they are confirmed to operate properly when aggregation is not requested. We don't want to offer aggregation in DeRun01 so we can keep it simple. It will be interesting to see what happens.
	2005-03-27	The package needs bug-report and other communication and engagement functions. This means that the ODMref engagement must be setup before this is announced more broadly.
	2005-03-27	The command-line parser should accept "-" always and also whatever character is prescribed in the environment variable that is used for that purpose. This should be documented as a platform dependency.
done	2005-03-27	Correct the DeRun01 command-line examples to use option "--dll", not "-dll".
	2005-03-27	Need to make sure we pay attention to the creation of batch files and their customization in the P050302e or related context. Maybe it is a separate page on shadowing/compilation, etc. Either way, it should work off of environment variables set up in other batch files so the customization is all in one place. I don't know what to do about the MS-DOS editor, though. -- Moved from my Journal.
	2005-03-27	`DeRisk01` will use its load location (if it can find it) as a way to decide where to put its log. We need to look at how it can use configuration information to do this in later versions.
	2005-03-27	Although we get things working using DeRun01 against the sample database and the dummy `DeRisk01` DMS, it really matters when we manage to link to `DeRisk01` through the ODMA32 Connection Manager. So we need that setup and piece of the experiment too. Looks like piece P050302g
done	2005-03-27	Add 0.01 to the anchor blocks and histories of P050302a (here), P050302c, and P050302 [already done]. Perpetuate that with the others being added. [dh:2005-03-27 This moves into move pages via cloning from existing ones.]
	2005-03-27	There need to be tutorial pieces to link to from the download materials and the SourceForge file descriptions. These can be developed after mitigation and as part of engagement.
	2005-03-27	We need a "what resources you will need" piece in viewable text that can be consulted before any downloads occur. This is to make sure that people know what they need, and can know how to get it. There can be more information inside the package, but there must be some known even for using the package. This includes for viewing .MHT files. This might be P050302f, and there needs to be more over in the ODMA /dev section where this material is developed and housed.
	2005-03-27	There needs to be a piece of the project that accounts for the organization of materials, the gathering of same, and then setting up the binding to resources from the directory that has the mitigation material in it. Let's say P050302e. This is not entirely critical to the mitigation experiment except that it is part of the "burden" of development for ODMref 1.0. It takes over my journal backlog item "Create a batch file for setting parameters to be used in compiling, editing, etc."
	2005-03-27	There needs to be a piece of this project about having identified where all of the needed Win32 definitions are found and how to include them in a minimalist way. Let's say P050302d.
done 2005-03-27	2005-03-26	Make page P050302c.htm for creation of DeRisk01.c
done 2005-03-27	2005-03-26	Add 0.01 to the anchor block for P050302b.
done	2005-03-26	Figure out the proper command-line syntax for specifying a DMS DLL to be used. It should not be confused with looking up a DMS in the registry, so we need to be clear about the format. Check the Visual C++ Toolkit 2003 on that topic. [dh:2005-03-26 I can use the GnuPG model for now, with - on simple options, -- on the "verbose" ones. --dll can be used for the DLL case.]
done	2005-03-26	Make page P050302b.htm for creation of DeRun01.c
	2005-03-26	OK, here is how I want to do ODMref Mitigation Experiment #1 [dh:2005-03-27 and reflect in P050302.htm]: 1. Make a derisk01.c program that is for providing the entry point and compiling into a DLL that only handles the basics -- the "nobody home" case. This is just a DLL that gets loaded, an entry point that gets entered, and a result code that gets returned. Then a log file gets written about what happened. [dh:2005-03-26 P050302b.htm is the starter for this.] 2. Make a derun01.cpp program that is a little command-line tool that loads an ODMA DMS DLL and enters its basic entry point. It does the absolute minimum, logs what it can, and then releases everything it gets, also unloading the file when its loader determines that the DLL is not being used by any accessible procedures [dh:2005-03-27 P050302c.htm is the starter for this.]. 3. We are making correct header files for DLL integration either way, and the header file is carried forward into further experiments and production development. [dh:2005-03-27 That will be file ODMgetODM.h and we need to make a page for that too.] 4. Note that this happens more-or-less in reverse sequence, as part of the test-first approach.
done	2005-03-09	Condition VSS to expand configuration-management labels in batch files.
done	2005-03-09	Remember that echo by itself reports the status of echo. Have to make some character to get a blank line [dh:2005-03-09 I used a single "-character to be as unobtrusive as possible, on the last line.]
	2005-03-09	The ODMdev.bat file needs to provide a location for further instructions and information.
done 2005-03-27	2005-03-09	Check in the files that I want to make derivatives of so that the original form is cached and available for discovery, even if the source is not itself managed under my VSS (e.g., files from the VC++ Toolkit 2003 and from the original ODMA SDKs). [dh:2005-03-09 I did that for everything in ODMref10 except ODMref10.h and ODMref10.cpp. I should go back and bring those into VSS with a full provenance.][dh:2005-03-27 I will do that after the reorganization.][dh:2005-03-27 I deleted those files that I don't really need now from the VSS, and I can share others over in the future.]
done	2005-03-09	Make a batch file for the ActiveODMA ODMref10 development area that works to define the working directory, identify the directory being used, the version of the environment, and the working location.
	2005-03-07	Document the directory and compiler library setups that are used in the compilation of this relatively trivial DLL. This sets a baseline for everything that is needed to get going.
	2005-03-07	Find the headers and libraries that are needed to have the boiled-down ODMref10.cpp compile. I can do some of it in the ODMref10.h header. [dh:2005-03-26 These are now DeRisk01.c and ODMGetODM.h, with no separate header for DeRisk01.c -- that's silly anyhow.]
in progress 2005-03-26	2005-03-07	Boil ODMref10.cpp down to the least that is needed to make a DLL that loads and provides a NOINTERFACE result. [dh:2005-03-26 This is now being done in DeRisk01.c]
in progress 2005-03-26	2005-03-07	See if it is possible to make a~~n odmdms.h~~ file that specifies the DLL entry point ODMGetODMInterface and that can be used for finding the entry in a DLL and also can be used to define the prototype that is then matched by the entry point in an ODMA DMS implementation module.[dh:2005-03-07 This may be simpler. It might be valuable to make an ODMGetODMInterface header file that is for making the real entry, and a different one for making the indirect entry via a loaded module. Figure it out.][dh:2005-03-26 This is in file ODMgetODM.h]
done	2005-03-07	Get a copy (not a shared branch) of odmdms.cpp into ODMref10 for examination of how the DLL is called.
done	2005-03-07	Bring the ODMA32 2.0 source files into the development tree and under VSS.
done	2005-03-07	Bring the core ODMA 2.0 files into the development tree and under VSS.
done	2005-03-07	Is there a standard template that is used for ODMGetODMInterface in ODMA 2.0 that can be the basis for a header file on the subject? [dh:2005-03-07 Apparently not. The way it shows up is as a typedef for a pointer in a sequence of operations performed in the ODMA32 odmdms.cpp module.]
done	2005-03-07	Check in the basic ODMASamp code and branch the main files to OMDref. [dh:The full 32-bit sample is checked in and the ODMASamp.cpp and ODMASamp.h files are branched and forked to ODMref10.cpp and ODMref10.h for now.
done	2005-03-07	Create this note in Project / Lab Notebook style