rdf:introspector 2.0

Token Heat : map changes to source code tokens onto a heat function

2013-04-06T07:08:00.001+02:00

Todays Insight : What if we can take the set of git changes and extract a list of changes to each function definition or invocation.

Basically we would look at what code have been modified in a revision and break that down into :

The function that has been changed, and the terms modified. The changes would be broken down into access to members, calls to functions etc. Every time that a method is called we would peg that, but also changes to the parameters would increase the heat.

the idea is that twiddling with the parameters to a function would increase its interest/value. By analysis of the program we could determine not only what are the most used parts of the program, but also over time what have incurred the most cost.

This analysis would help understanding of new code by pinpointing functions that have been changed often.

Format of inflight data, not copyrightable?

2013-04-02T13:48:00.001+02:00

Considering that header files are not copyrightable http://www.groklaw.net/pdf3/OraGoogle-1202.pdf , not being creative works, the layout of memory generated by the compiler is not creative works, I think that provides insight into the memory of a program that contains your data.

If you feed your own data into a program that your run on your computer, I feel that you have the right to inspect the memory of the computer. This is assuming that the software is licensed under copyright (gpl etc).

So according to this train of thought, the AST or RTL data that is stored in memory in the compiler is not under the copyright of the compiler, if anything it contains your data that you gave it, so it is mostly yours. The compiler slices and dices it up, but that sliced data is not copyrighted by the compiler author.

So my insight is that inflight data copyright status is not changed by any program. If you were to use a tool like gdb to inspect the memory of your computer you can do what you want with the data you have rights to.

More ideas for the introspector

2012-10-21T11:32:00.001+02:00

Another braindump! Taken from my travel notebook. Putting them on my inrospector blog because it is about this project.

a compiled and linked executable program is stored in a file on the filesystem.
the filesystem might even be a zip file that is streamed on the network.
the format of the executable file could be the ELF format.
The ELF file format has a set of headers that describe the data, there are pointers to data, labels of names, blocks of data could be variables or functions.
The binutils and gcc programs are used to create programs.

KVM is a virtual machine, we want a virtual machine that can audit all operations on the machine.
The linux kernel run on the machine, or virtual machine. The kernel can spawn processes from executables.
The process interacts with the kernel via syscalls which are wrapped in libc, which is linked into the program.
Libtrace, systrace and valgrind and other tools intercept calls to functions and trace them.
The global, static, stack and the heap are the main types of memory in a program.
To decode data, to unpack it, to interpret it is the main purpose of one class of functions.
the other direction is to pack, to encode, or generate code.
We want to be able to trace the decoding of programs. We can see a program as a vector of characters.
A program in memory is a sparse multidimensional matrix, or graph. We have formal rules that describe the structure of this graph or matrix.
We want to have a graph editor to display graphs.
We want a reasoner who will operate on them.
Some of the domains that are interesting are : openerp, webshops and businsess transactions, uml diagrams, cad diagrams, osm modules, github projects.
The basic operations are specialization and generalization. Parameterzation. Replacing of strings (templating).
We want to find functions for simpification.

Wrapping C++ code in perl and other langages with swig is nice.
Gcc Compiler plugins.
gnome introspection.
Usage of RDF, semantic web technologies.
Webalizer, turn a program into a web program.
Shellifier, make the program usable in the shell.
define objects.
define functions.
introspection of environment by having a full model of the entire system in memory and simulation of system.

Generation of compiler test cases based on grammar

2012-10-21T08:28:00.000+02:00

the idea is simple :

generate test cases based on the grammer of the language
each name or string is unique in the entire program and can be mapped back onto the grammer, the language specfication.
Each constant number used is also unique across all tests (within reason)
using of long canonical names (like mangled name) for all strings and names that reflect exactly where the string was generated from.
filenames are also generated with unique meaningful names.
for grammar structures with no names, the name is tagged by embedding the structure in a namespace or function or file that gives it meaning.
comments are used as well to tag the data.
The tests will cover the entire language specification
The tests will cover the entire compiler source code.
Each test case will be given a unique hierarchical number that will identify it completely.

Discussion :I have been thinking for a long time about this idea.
With a set of test cases like this you will be able to trace the flow of data through the compiler and runtime by looking for the strings and encoded variables.
This will allow for a tracing of the language spec all the way down into the runtime.

The tests will then be used to document the compiler ast structures and they will be illuminated by the test cases and given an explicit meaning. Instead of searching for the meaning in the ASTs, we know what the meaning will be based on the test case number.

The tracing/coverage of the code will show what data structure are used by what test cases and what lines of code are used as well.

Then we will document the data structures exactly.
It occurs to me that The gcc compiler plugin should not only just give you pointers to the data but also be able to easily list all the nodes. We should be able to produce constant code blocks that describe the code and are usable in run time introspection/reflection.

More ideas from my kosovo trip

2012-09-12T20:32:00.000+02:00

It has been a long time since I wrote in the introspector blog. I have been using the fmtyewtk blog for details this one for big ideas. Now I would like to share my big idea.

Basically I would like to propose the following system, first using archive.org or some other archive system to store relevant pages of data. These pages should be indexed and the indexes also stored in an easy to use format. Inline javascript and html galleries will be used to make each bucket.

Now there are problems with using archive.org, the main one being that it is not easy for users to use, it is slow and hard to authenticate. We can get around this by using a local server in front of it, the local server, lets call it the librarian, will be responsible for putting new volumes of data onto the archive. The users who submit the data will be freed of having to worry about licenses, they will just donate the data to the library or publish it under a creative commons license and give us permission to store it for them.

These librarian servers should also help create the index, resize images if needed, create galleries and other useful tasks before or after uploading the data. Of course archive.org also manages this for you but sometime is it not enough.

So, once you have the indexes, they should be also made into some SEO compatible format, also they should be reposted on some other sites that are usable.

Geodata, Categories and other data should be added to the articles.

The data stored can be pictures, maps or wikipedia articles, or program code.
The indexes can also be in semantic web and linked data format as well.

Now, there are more ways to store and share this data, archive.org is the first place to put it for long term storage, but also we want to create torrents from the data, also create interfaces to store it on a drupal site, or a google site or one of the many many free blog or internet posting systems.

the indexes can be posted also onto twitter or rss feeds in a useful manner.

The openstreetmap api server is just one of the front ends I am thinking about, we need more of them, also I want to implement apis compatible with other picture systems used by digicam and other tools so via a simple dns spoofing you can post to our server.

more to come,

mike

Introspection might be to late

2012-04-21T09:32:00.001+02:00

Introspection is a retrospective, it looks at what has been done.
We want to also do things right and need a perspective to do that.

I have done an experiment using template toolkit for code generation,
that works pretty good.

we will want to create a perl interface to the introspection results to use for code generation, ideally we would have a sparql query interface for that.

more to come.

mike

Embedding perl in the gcc plugin

2012-03-09T13:52:00.001+01:00

I have the first examples of a running perl inside of a gcc plugin.
The perl is an embedded perl that is dynamically loaded into perl.
It also contains an extra hook for the DynaLoader::boot_DynaLoader so that dynamic modules can be loaded. So when the plugin call back is called, the gcc will invoke a method in perl, passing the tree structure as an integer value, an IV. The perl routine needs some XS code to be able to interpret that, so it uses inline C to call the gcc code to access the variables.

https://github.com/h4ck3rm1k3/gcc-introspector-plugin/commit/1d1a8a99925e5df091fe1f9987bfbfb1975471a2

UNICEF Kosovo Innovations labs recognizes FLOSSK

2011-12-21T23:41:00.000+01:00

http://kosovoinnovations.org/w/

Innovations Lab

UNICEF Innovations Lab Kosovo cultivates Innovations for the benefit of Kosovo’s youth. By Youth For Youth projects are a chance for Kosovo youth to impact their own lives, and those of their peers. The Design Center directly designs and implements technological innovations for Kosovo institutions that work on behalf of youth and children. The Youth Advocacy Platform seeks to build the capacity to Kosovo’s youth to directly address key social issues within their communities, with the goal of establishing a sustainable network of youth-led advocacy, civic engagement and policy influence.

FLOSSK http://www.flossk.org/en/about

Free Libre Open Source Software Kosova

FLOSS Kosovo is a non-governmental non-profit organization established in order to support, promote and develop:

Free/Libre Software as defined by the Free Software Foundation;
Open Source Software as defined by the Open Source Initiative;
Open and participatory knowledge as that supported by the Wikimedia Foundation;
Education in information technologies through open courses like those from MIT;
Standards, culture and open society by using free and open communication as are those supported by the W3C and Creative Commons.

To achieve its goals, the association will undertake, without limitation, the following activities:

The Annual Conference available to the general public with the lowest prize possible to help the aspirations set out above.
Enable access to free and open source software and relevant knowledge by facilitating the distribution and localization in the local languages of Kosovo.
Cooperate with local and international partners and sponsors of the industry, academia and government to drive forward the aims of the association.

FLOSSK began in March 2009 at the initiative of James Michael DuPont as a result of the desire to organize a conference on free and open software. After six difficult months and with the help of many supporters FLOSSK organized the first conference of free and open software in Kosovo in August 2009.
Apart from the Conference, FLOSSK continued to work in various activities such as organizing Software Freedom Days in different cities of Kosovo, lectures on free software throughout Kosovo, translating software, collaborating with the media to promote free software and creating local free software groups in various cities.
From the beginning FLOSSK members and the general public learned about GNU/Linux operating system, FLOSS programs for solving everyday problems, map creation using OpenStreetMap, and met free software movement figures from around the world.

Come and join us, let's work together to develop and promote free and open source software in Kosovo!

Smurfs Movie wikipedia copyleft infringment

2011-12-17T03:52:00.001+01:00

On the flight LH402 over to Newark with Lufthansa we watched one of the three movies on the old school small television screen, being that Lufthansa was too cheap to upgrade the Boeing 747-400 to have an inflight entertainment system. Make sure you check how old your airplane is before booking Lufthansa and stay away from the Ancient airplanes!

I watched the movie the Smurfs in Horror, when I saw the guy from the "how I met your mother" copy and paste an Image from Wikipedia into his advertising. I thought to myself, omg, these idiots are really corrupting our children and showing them the worst practices in usage of creative commons work.

In this article the author describes part of the problem :
http://blog.credoreference.com/2011/08/do-the-smurfs-hate-information-literacy-et-tu-peyo-why-hollywood%E2%80%99s-lack-of-ideas-is-bad-for-our-kids/

The image http://en.wikipedia.org/wiki/File:December_2009_partrial_lunar_eclipse-cropped.png
from Codybird And that information would have to have been provided on the advertisement, and also on the movie. I could not find any credit for this image on the website http://www.smurfhappens.com/

So, Sony, I think it is time to give wikipedia credit and also to teach kids about creative commons, not just lifting images off the wikipedia.

thanks,
mike

Update (12/17/2011):
The movie is being distributed by sony, but produced by
http://en.wikipedia.org/wiki/The_Smurfs_(film)

Here is a clip of the moon article being viewed, later he copies the image into his advertising

Update: source of video clip http://www.youtube.com/watch?feature=player_embedded&v=YJxqFMPe95c "The Smurfs Blue Moon" from Uploaded by SleepyBrown253 on Sep 5, 2011

see also the discussion thread here :
http://www.gossamer-threads.com/lists/wiki/foundation/263905

Idea of the day : native c implementation of jquery as a browser extension

2011-10-23T21:03:00.001+02:00

Idea of the day : native c implementation of jquery as a browser extension.

I just had idea, why not translate jquery into C/C++ and build it into firefox as an extension.

It would involve translating the javascript into optimal C code and building that directly into the browser.

This would be built as an firefox extension I think
https://developer.mozilla.org/en/Creating_Custom_Firefox_Extensions_with_the_Mozilla_Build_System

I guess that it would involve creating native functions for common jquery functions
https://developer.mozilla.org/en/SpiderMonkey/JSAPI_Reference/JSNative
https://developer.mozilla.org/en/JSAPI_User_Guide

This could be something that uses these other projects :
http://code.google.com/closure/compiler/ The Closure Compiler is a tool for making JavaScript download and run faster. It is a true compiler for JavaScript. Instead of compiling from a source language to machine code, it compiles from JavaScript to better JavaScript.
http://code.google.com/p/js2cpp/ JavaScript to C++ compiler

here is a stack overflow for this :
http://stackoverflow.com/questions/556807/java-to-c-converter-tool

more to come...

mike

Memes are the thoughts of the global mind

2011-07-14T21:48:00.005+02:00

Memes are ideas that are copied from one mind to another. Let us define this, they are behaviours that are shared by many, the ideas and trends. This video of how to deactivate a cat

has been watched almost 100k times, but more if you consider how many times it is hosted on different platforms. It is a meme that can be defined as such : People love cats, can relate to them on a primitive level, often they are treated as members of the family. We have a basic positive reaction to cats. People liked to share pictures of their cats. Eventually this mutated into the lolcat .

coupled with their reading which loads that into the mind of next person.

coupled with an interpretation of the image which causes the mind to call forth (evoke) an internal representation of the meme.

This invokation of a meme by a person is followed by a positive rewarding reaction of seeing a fluffy cat that is cute.

The further attached data is also executed to add suprise, this new information is another meme basically that is attached to the first one.

We have a carrier system and a payload.

Now we have defined the basic idea of a meme being copied. The next point is that the global internet mind is defined as a large number of interacting people who are exchanging memes on some level. It is defining all cultural content and behavior as being the product of some meme. Or in reverse, it is saying that the meme is the basic unit of behavior that is copied. Now lets take the instance of the bushman in papua new guinea , the remix is entitled "Tribe Meets White Man for the First Time" and shows a short mix of a very long video series with nice music. It is a mutation of the original long video, making it more pleasing and rewarding in terms of time spent. The payload of this video, the new information that is contained in it is the act of hitting yourself on the head lightly to express happiness. The natives are doing this, and the associated positive expression in this case is simple pleasure at eating. If other tribesmen see this, they will be able to remember previous times that they saw this behavior and a similar positive reaction is evoked. So I hope you enjoyed my little exposition on memes. Please share this and think about it. The meme of introspection, reviewing your internal mental state while interacting with the world and seeing the world mirrored in the code and the memory of programs and in the minds of people.

As the behavior of sharing this basic happiness, the looking at cute cats and sharing them with friends via email, is the propellent for the behavior, the basic reward system feeds the meme. The meme is then mutated into the lolcats via adding of foreign symbols which are executed as well in the minds of the internet. You will see that copying of an idea to a person, which is done in a few steps :

the copying of a data file containing the picture.

Reith Lectures Archive: 1948-1975 Bertrand Russell RLA: 1948 1. Social Cohesion and Human Nature

2011-06-28T21:53:00.000+02:00

RLA: 1948 1. Social Cohesion and Human Nature

Reith Lectures Archive: 1948-1975

The philosopher, mathematician, and social reformer Bertrand Russell gives the inaugural Reith lectures, and in his first lecture, examines the role of impulses in human nature.

http://downloads.bbc.co.uk/podcasts/radio4/rla48/rla48_19481224-0900a.mp3

more about it http://www.bbc.co.uk/podcasts/series/rla48/all
Thanks to the bbc for sharing this!

Rdf Browser

2011-04-28T20:25:00.000+02:00

Here are some of the RDF browsers I found :

This looks promising :
http://demo.openlinksw.com/rdfbrowser/index.html
and
http://dataviewer.zitgist.com/

And this I need to look at :
http://dig.csail.mit.edu/2007/tab/

Mentioned here:
http://wiki.dbpedia.org/OnlineAccess#h28-4

http://dbpedia.org/fct/

http://www4.wiwiss.fu-berlin.de/bizer/ng4j/disco/

One of the data sets to look at is :
http://linkedgeodata.org/About

sugar binutils and gcc

2010-10-02T21:10:00.001+02:00

Finally I have a good usage for the introspector project,
accessibility and learning.
I have already done experiments with a web page for inspecting the dump data.
My new idea and plan is to create a way to learn about assembly in sugar,
similar to the turtleart, but with real assembler.

Currently I am researching about how to create template classes or instantiate them. I am thinking about the following : iterate over an array of objects and generate then include instances of objects of the types we generate.

So we would create source code , compile and link it, then execute it.

Here is a way to load a new shared object with the code.
http://www.linuxjournal.com/article/3687

Here is more info about dynamic libs :
http://www.yolinux.com/TUTORIALS/LibraryArchives-StaticAndDynamic.html

Here you and read about typecasting:
http://en.wikibooks.org/wiki/C++_Programming/Type_Casting

Here is the tool to build sugar:
http://git.sugarlabs.org/projects/sugar-jhbuild/repos/mainline

Here is the turtle art program:
http://git.sugarlabs.org/projects/turtleart

Here is the source code for git, git can be used to store common blocks of code ans assembly data coupled with metadata. we should be able to represent large programs as building blocks of individual parts. we can collect all the data about those parts in git.
http://github.com/git/git

my binutils work can be gotten with
git clone git@gitorious.org:binutils-introspector/binutils-introspector.git

http://gitorious.org/binutils-introspector/binutils-introspector

Database for Autoconf/Automake/Autotools

2010-09-14T06:16:00.000+02:00

After listening to Rob Savoyes incredible talk on autotools I was very impressed by his knowledge and in depth understanding.

He told about how autoconf / configure used to use a database, but now it checks everything.

My question is if we can now use autotools to help build up a public configuration database.

Specifically I would like to know the names of the packages that contain what header files for each system. It should have tools to allow for people to automatically extract the packaging information from the source code, so autoheader and autoscan should also be able to create dependencies for debian or red hat.

I have created a repository
http://gitorious.org/autotoolsdb/autotoolsdb-config-logs and will be creating a set of autoconf scripts for you to run. all you have to do is check in your version-branch of the result and we will start building up a database.

mike

wolfram cellular automatas and the introspector

2010-05-27T12:17:00.000+02:00

Hi,

after watching the new kind of science for the second time,

it occurred to me that it would be possible to create a visual representation of the compiler graphs as a cellular automata and then create rules encoded as automaton to traverse the compiler graphs.

git clone of translationwiki for albanian gheg

2010-05-12T11:39:00.000+02:00

I have written a perl script to download the po files from translate wiki and a shell script to check them into git.

source here:
http://gitorious.org/translation-sq-aln-albanian-gheg/translation-sq-aln-albanian-gheg-po/blobs/master/twn.pl
and here :
http://gitorious.org/translation-sq-aln-albanian-gheg/translation-sq-aln-albanian-gheg-po/blobs/master/runput.sh

git repo here http://gitorious.org/translation-sq-aln-albanian-gheg/translation-sq-aln-albanian-gheg-po
webpage here: http://xhema.flossk.org:8081/launchpad/po/
to get the latest version :

git clone git://gitorious.org/translation-sq-aln-albanian-gheg/translation-sq-aln-albanian-gheg-po.git

you can clone the repo and push changes to it : http://gitorious.org/translation-sq-aln-albanian-gheg/translation-sq-aln-albanian-gheg-po/clone
for that you need a gitorious login, http://gitorious.org/users/new you can use an openid for fast starting.
and you need to setup your ssh keys http://pkeck.myweb.uga.edu/ssh/ and put them in gitorious ( for examply, my personal link is http://gitorious.org/~h4ck3rm1k3/keys and you replace h4ck3rm1k3 with your username) then you can push easily and painlessly from the command line.

Quantum stuff

2010-05-10T10:37:00.000+02:00

Here are some links that a nice guy idmclean shared with me :

Quantum theory, the Church-Turing principle and the universal
quantum computer
http://web.archive.org/web/20030915061044/http://www.qubit.org/oldsite/resource/deutsch85.pdf

FROM QUANTUM METALANGUAGE TO THE
LOGIC OF QUBITS
http://arxiv.org/pdf/1003.5976v1

my google profile qrcode

2010-04-22T23:50:00.002+02:00

google readers recommended sources are incredible!

2010-04-15T22:26:00.000+02:00

Everyone of my google suggested reader feeds was a match. If social media is good for anything, it is for aggregating and filtering data and feeds. By working together and sharing our data we all distribute the global task of understanding the world. Is that not the total global sum of individual consciousness?

http://www.google.com/reader/view/#stream

Two Joint Conferences planned in the balkans for mid september 2010

2010-03-14T08:53:00.001+01:00

We have a double pack this year :

Open Source Conference Free Software in the Balkans

flossal.org | univlora.edu.al | flossk.org announce FreesB Sept. 11/12 2010 Conf. @ University Vlora Albania” – http://flossal.org/freesb-2010-conference/

Open Source Conference Software Freedom Kosova

http://flossal.org/freesb-2010-conference/ flossk.org | www.uni-pr.edu announce SFK10 September 25/26, 2010 Conference @ University of Prishtina Kosovo – http://kosovasoftwarefreedom.org/

Now all we need is speakers who want to go to both and also do some workshops between the two besides sitting in the sun!!

mike

CFP SFK10 Flossk UNIPR CALL FOR PAPERS International Software Freedom Conference of Kosova 2010 September 25/26, 2010 Prishtina / Kosovo

2010-02-25T08:20:00.002+01:00

CALL FOR PAPERS

International Software Freedom Conference of Kosova 2010

September 25/26, 2010

Prishtina / Kosovo

The University of Prishtina and FLOSS Kosovo are pleased to announce the Call for Papers for the International Software Freedom Conference of Kosova, which will be held on September 25 and 26 at the Faculty of Electrical and Computer Engineering, University of Prishtina.

Considering the pioneering nature of this conference, the focus will be on the following subjects:

· What is free software and open source?

· How can free and open source software help in economic development?

· How can public and private institutions benefit from free and open source software?

· FOSS in Public Administration

· Open Society projects

· Copyright, Licensing and Patents

· Wikimedia and Free Culture

· Use of free and open source software for Schools and Business?

· OLPC implementation in Schools

· Open Courseware Systems

· OpenStreetMap and Open Source Geo Information Systems

· Developers presenting work on projects.

· Use of free and open source software for:

· Desktop,

· OS Kernel,

· Programming/scripting languages,

· Development tools,

· Databases,

· Security,

· Networks,

· Collaboration,

· Web, Web 2.0,

· GRID/Clusters,

· Multimedia and video production

· Toolkits

· Voice over IP

Authors are encouraged to also submit papers on any other topic not listed above but related to free and open source software.

Key Dates

Call for papers closes	1 May 2010
Initial notification of acceptance	1 June 2010
Deadline for final versions of papers	30 June 2010
Conference	25/26 September 2010

Submission Guidelines

Your proposal will need to include:

* Your name and bio info (include a bit about your qualifications regarding your presentation topic)

* Name of talk

* A really brief summary of the presentation (around 200 words). If the paper is accepted you will need to provide a summary of your paper in no more than 1000 words.

* Who your talk is likely to interest (some keywords)

* Name and bio information of any co-authors

Papers can be in one of the national languages of Kosovo or preferably in English.

Authors will be notified of the acceptance of their proposal by the proposal acceptance date.

If you require faster confirmation of acceptance (for example, because you are travelling from overseas and require extra time to make arrangements) please communicate this to the committee when submitting your proposal

Submissions can be made by visiting and registering on our conference webpage: http://www.kosovasoftwarefreedom.org/.

Even if you are not interested to speak, we encourage you to participate at the conference. Registration is open and can be accessed by going to:

http://www.kosovasoftwarefreedom.org/.

Contact

For any further information feel free to contact the Chairs of the committee at: Prof. Dr. Blerim Rexha (blerim.rexha@fiek.uni-pr.edu) or Bernard Nikaj (bernard.nikaj@flossk.org)

GDB::Introspector::Model::Devel::GDB

2010-02-01T12:51:00.000+01:00

Using this command interface :
http://search.cpan.org/~jezra/Devel-GDB-2.0/lib/Devel/GDB.pm

We can talk to the gdb.

I am starting to wrap this in a catalyst wrapper.
http://github.com/h4ck3rm1k3/GDB--Introspector--Model--Devel--GDB

It can load the gdb into the webserver, more to come.

idea of comparing runs between optimised and debug code-

2010-01-11T20:28:00.000+01:00

first you compile a program in a very optimized form,
then you compile it with debug. Be able to run them in parallel and compare certain registers.

I think we should be able to debug an optimized program if we have all that data, maybe we could use that to look for inconsistencies in builds.

Declaration Logo

2009-12-02T18:47:00.002+01:00

My new logo for a declaration in the gcc compiler introspector. The megaphone symbolizes the meaning of a declaration, it is a way to announce the availability of something in a program. It is the most basic form of data that the compiler deals with. Even if you write 1+1 you are using a statement declaration that cont...ains an expression with constants and an operator. All these parts of a statement from the user are stored in declarations. Then from the decls, a program has functions and datatypes defined for the layout of memory. These formats of data and behaviour are just templates or prototypes of things to be executed. Memes you could say, things that are copied from program instance to instance.

A programs source code is transformed from the compiler into an executable program. This transformation or translation from a source file into a computer executable also requires a certain understanding of the program. This understanding is owned by the compiler engineer who is working for a large corporation normally on the gcc compiler.

Therefore we can expect a certain cryptic nature of coding when of companies attempting to protect themselves. At least you cannot expect it to be put on a golden platter for you.

The introspector attempts to intercept this message, the one between the chip and the chip producer.

Better stated, there are many message involved. We have many parts of the compiler, each one is from a different group. The optimizer can work on the program itself independant of the chip. In fact there is a message that is sent from the programmer to the compiler writers, and that is the message that is intercepted first. We capture the abstract syntax tree output that represents what the compiler sees. There are many other outputs from the compiler we need to see, but we can in theory reproduce the entire programs source code from the AST. There are things like macros lost, but the meaning of the code is still there. It can be quite verbose and expanded.

Anyway, I am working on a set of icons for the tree nodes at this moment to create an nice jstree javascript view of this data.