this is fantastic forum for shell programming and scripting,
so please let me to introduce you with my very old concept to
have web form/s with radio, select, input fields
and have an application generating valid, syntax error free scripting code.
The same or alike questions are asked over and over again, so
one question generates one or more valid code answer.
Basing on top-down program solving, programming, any complicated computational problem can be divided into a number of basic problem and so on, down to the bottom.
As the problem has been already solved in html/www environment, providing wysiwyg html applications,
please tell me how to start building such valid script code generator from the scratch.
Basing on some experience in gps navigation systems,
I started testing many many gps navigation applications.
Acquired about 10 gps devices, usb gps, Bluetooth gps, Ps2 gps, agps in my Motorola A1000, in Nokia Tablet and more.
First step is gps device
I connected bluetooth dongle to Linux server,
connected BT gps device to it
learned how to read NMEA sentences
learned basics of awk to process NMEA sentences
Now I can try to extract data fields of interest to me from NMEA stream.
Next step is to process such data and build basic graphical interface (no maps) to view data on-line.
Such basic application/ script under Linux is dialog for its simplicity.
I have gps + gps data processed, dialog viewer, what can I do next.
Next step is map viewer + POI database + mysql server, routes, routing algorithm implemented, ..
But basics is exactly shell script app generator.
One good idea is to build web form for the modification of examples shell scripts to fit our needs.
So reverse engineering comes into my mind.
Shell script parser, syntax resolver, debugger,.
Another nice solution is to build pipelines made of objects (ocx like)
and have visualization tool - pipes to have objects connected.
Is anything of this kind already available ?
Foundation for this project are libraries of shell script examples with good description - HowItWorks.
Happy New Year
Human shell script app generators already live on this nice forum.
Thank you for your post, it is a bit hard to understand. Could you be very specific about how we can help you?
What you don't understand ?
Just ask a question.
Human cognitive processes reverse enginering is a term coined by AI researchers to build a system providing human-like generated answers
to repeated questions, basing on knowledge databases processing of the alike problems, questions, answers from the past.
You may call it an intelligent Touring machine.
Google and others, tried hard to build such service to let visitors to
ask natural language made questions and get human-like answers generated, basing on provided state-of-the-art knowledge databases.
The issue is Google and other search engines, started to develop such service from the top.
Top-down problem solving technique.
My approach was exactly the opposite, to build knowledge databases up from the bottom, generating knowledge database tree/s
Archives of Linux forums are excellent knowledge databases.
What is missing is to rearrange such knowldge databases as graphs.
Search engine generates first level knowledge database graphs only.
To build next levels in the knowledge tree human cognitive processes are examined, recorded and followed to learn if What You Get is What You really Wanted.
Semantic Web and Web 2.0 and Beyond by O'Relly and others is exactly
about building such solutions.
Let's imagine you are a Guru at this forum, supported by such semantic web based solutions and you can read on-line any asked questions and give the best answer, solution.
Ok. There is a number of Gurus at each web forum.
The issue is interfacing.
Any question is text-based form asked.
To read each of 1,000 questions, problem solution requests posted to this Linux forum on a daily basis takes some time.
More time it takes to provide the asker with the best solution, answer.
Today, you, as a Forum Guru, relay on your personal knowledge, experience, printed and electronic materials.
Some come from the net as Google query results.
At one of official Google Groups I have already met a student, answering
every and each question in interactive mode (meant promptly).
And what's the most impotant, giving the right and exact answers to qestions asked, providing, suggesting software solution either.
What is chance to have such Linux/Unix Guru at this forum
working hard on-line, day-long, in interactive way, one day, giving answers, showing solutions to problems and questions asked ?
If not 1,000 answers per day, so mayby 100 or less
but working on-line in Linux support hotline.
As I don't expect such Linux support hotline to be introduced soon,
my solution is to build local semantic web, to meet local challenges.
And building such local semantic web for this or another web forum is what makes sense, to attract thousands of Linux users from the Internet to such central - Linux Support Hotline , giving acces to local Linux Semantic Web.
So instead of giving thousands answers to thousands of qustions asked,
give one answer and solution, how to access and use such Linux Semantic Web.
I don't expect to build Linux Semantic Web 2.0 and Beyond soon,
but already developed some sophisticated tools for knowledge databases processing into graph relations.
Linux Semantic Web 2.0 and Beyond is aimed to generate human-like answers to questions asked.
I'm a linux novice , in a training purpose i have to reconstruct under windows an application running under fedora 14, the application communicates with an STB device through RS232 in a first time and then through ethernet for firmware loading purpose.
All what i know is how to run... (2 Replies)
I'm looking for help with a legacy system.
I have some obsolete equipment connected to an Ultra 60 running Solaris 5.7 with the binary for a 32 bit driver. The driver is rejected by newer versions of solaris, which run 64 bit kernels.
I hope to reverse engineer the driver so that I can... (0 Replies)
I want to reverse a the following:
I currently use something like below to pass it as is.
But now I want the same script to reverse the above and pass it to ethtool.
// psuedo code
for m in $@
... (1 Reply)
I am new to awk. Actually I want to search a pattern A, when I get that line with pattern A then for one of the field of that line again I want search on that field (say pattern B)from start of the file.
I am using awk. Is nested searching possible in awk?
Please do the needful as... (4 Replies)