OpenSource

Arduino Inventor's Guide

gregladen — Sat, 15 Jul 2017 16:34:33 +0000

Arduino Inventor's Guide

First, a word about Arduino and why you should care. An Arduino is what is called a "prototyping micro-controller" aka "really fun electronic gizmo toy."

Micro-controllers are everywhere. When you "turn on" a machine in your house, chances are there was already a micro-controller sitting there, running on a minute bit of juice from a built in battery, waiting for you to push a button. Then, you turned a dial or selected an option on your dishwasher, or changed the setting on your thermostat, or picked some alternative mode on your coffee pot, or shifted into a different gear using a "gear shift" in your fly-by-wire Prius, or you opened up the birthday card and cats meowed out "Happy Birthday."

All of those events involved a micro-controller, which consists of thee parts. There is a brain inside it, there is a set of sensors or actuators (a thing that detects that the greeting card has been opened, and an actuator that is the thing that makes the meowing sound by playing an WAV or MP3 file), and some software. The software gets in there by hooking an in production version of the micro-controller, likely once in its life, to a regular computer via a COM port (the same kind of interface used by your mouse, or a USB connection, etc.), and stuffing the software in there.

The Arduino Uno is a micro-controller that is very generalized, very large (a bit larger than a credit card), has a well behaved power supply, lots of connectors for either sensor or actuators, and a pretty fancy brain for a micro-controller, with lots of room for code written in a very powerful and fairly easy to use language similar to objective C. You can hook the Arduino up to most computers, using freely available software to communicate with it and compile your code. For the most part, you don't have to actually write code, it is provided by the developers of projects you are poaching, but if you want, you can go to town with it.

There are hundreds and hundreds of sensors and actuators, from thermostats to motors, gyroscopes to myriad things that light up, available for the Arduino, and in fact, anything that runs on low voltage can be hooked one way or another to it (if you know what you are doing). High voltage uses (like shifting a car or opening or closing a garage door) are done, of course, by using relays that are switches operated by a micro-controller but that pass any voltage level you want, if you get the right one.

The Arduino and its associated equipment can thus be used to replicate, design, and experiment with pretty much any thing a micro-controller can do. After "prototyping" it is trivial, for an expert, to rebuild the circuit using a less capable but perfectly adequate bunch of parts, and solder instead of just sticking things together (called "breadboarding") and so on. But no one really does that with Arduino. With Arduino you may leave the final product at it is (like the robot we built a few weeks ago) or, as in the case of the projects in an introductory book on how to use and have fun with an Arduino, you may just take the thing you built apart and build another thing.

So, this new book, The Arduino Inventor's Guide: Learn Electronics by Making 10 Awesome Projects, is sitting on my workbench ready to go to work.

Of all the intro Arduino books I've seen, this one is unique in a way I'll explain below.

The book gives detailed, understandable, and learning-oriented instructions for a home stoplight (helpful with toddlers in the house), a reaction time garme, a balance beam game, a diminutive greenhouse, an small piano, and a handful of other projects.

The coolest project might be a living breathing Logo turtle. Logo is a computer programming environment developed years ago to serve several functions including helping kids get interesting in coding. Logo is actually one of the oldest computer languages still in use (dates to the late 60s) and it is a general programming language, but it is mainly adapted to running the Logo turtle. The turtle is a curser that is moved around on the screen, and instructed here and there to drop a specific pen (it can have several different pens) so as it moves along it draws.

This project, from The Arduino Inventor's Guide: Learn Electronics by Making 10 Awesome Projects, is a physical turtle that draws on your rug! Or, hopefully, a big piece of drawing paper you put down for it.

I mentioned above that this book is unique. Here's how. I've looked at a Lot of Audrino project books, and there are no introductory books that provide detailed information on how to make interesting project enclosures and cases. The projects in this book rely heavily on the stuff you built the electronic into. The project enclosures are generally made of simple corrugated cardboard that you can get from an old box, or, if you want, from a craft store (for more interesting colors, better quality materials, less cat hair, etc.)

You can build all the projects in this book with parts you have acquired in the usual manner, but the book suggests you get the Sparkfun Inventor's Kit for Arduino, which is about 75 bucks. Note: This book is produced by No Starch Press and Spark Fun, so of course they suggest the Sparkfun Inventor's Kit for Arduino as a way of getting all the parts. But, by the time you add up an Uno or equivalent micro controller for 19 bucks, LCD display for nine bucks, fancy breadboard holder for 9 bucks , a shift register for 8 bucks, and miscellaneous other parts, you might be over $75 anyway. Or maybe not. You'll have to check around.

There is plenty of preliminary information to get a total novice started, and each project is rich in detail and very fully and expertly, clearly and helpfully, described.

This is an absolutely excellent choice, perhaps my favorite at the moment (and totally up to date) Arduino starter book.

gregladen Sat, 07/15/2017 - 12:34

Tags

Books

OpenSource

Technology

arduino

It is formally correct that arduino uses a "language similar to objective C".

It is, however, more correct to say that it uses C++, since that is the language actually used by the arduino ide.

/erik

This is so bizarre. I work in a company that helps innovators develop new products and my brother's actually a coder. So I was pretty blown away to read this article - and if I'm right, which is by no means certain - it means that our everyday products are using a computer language.

This has given me so many ideas. Thanks for a really interesting article.

I assume that is generally true. There is a range of processors out there, micro-controllers, and I think they generally use some sort of code. Something like the Arduino probably has more room than the average microcontroller because it is a prototyping system.

Python Programming To Automate Common Tasks

gregladen — Thu, 11 May 2017 05:00:42 +0000

Python Programming To Automate Common Tasks

Automate the Boring Stuff with Python: Practical Programming for Total Beginners by super Python expert Al Sweigart is a pretty thick intermedia to somewhat advanced level programming book.

It covers how Python works, so someone familiar with programming languages can get up to speed. Then, the book tackles a number of key important tasks one may use a computer for. This includes working with Regular Expressions, file reading and writing, web scraping, interacting with Excel spreadsheets and PDF files, scheduling things, working with email, manipulating images, and messing around with the keyboard and mouse.

I wold like to see a second volume with yet more programming ideas and examples. It could be a series.

From the publishers:

If you've ever spent hours renaming files or updating hundreds of spreadsheet cells, you know how tedious tasks like these can be. But what if you could have your computer do them for you?

In Automate the Boring Stuff with Python, you'll learn how to use Python to write programs that do in minutes what would take you hours to do by hand—no prior programming experience required. Once you've mastered the basics of programming, you'll create Python programs that effortlessly perform useful and impressive feats of automation to:

Search for text in a file or across multiple files

Create, update, move, and rename files and folders

Search the Web and download online content

Update and format data in Excel spreadsheets of any size

Split, merge, watermark, and encrypt PDFs

Send reminder emails and text notifications

Fill out online forms

Step-by-step instructions walk you through each program, and practice projects at the end of each chapter challenge you to improve those programs and use your newfound skills to automate similar tasks.

Check it out.

gregladen Thu, 05/11/2017 - 01:00

Tags

They used to call it programming

Technology

One of these years I have got to learn Python -- which seems to be taking over from Perl in the domain of running webservers. I wish somebody would translate or "port" the http://mindsz.com/ideasz/ai-mind project form Perl into Python so that then any Python programmer could become a http://ai.neocities.org/maintainer.html AI Mind Maintainer.

Greg - the entire book is also available online https://automatetheboringstuff.com/

Thanks, for pointing the book out.

Just want to say, "Thanks again."

I took the time yesterday to sit down and learn enough Python to automate inspection of 8000 Excel files. I collected the filename, file size, the last modification date, names of all worksheets in each workbook, the names and cell ranges of all Defined Names in each worksheet, a list of all external links, and the cell value for a particular Defined Name (where it existed).

I spent a dozen hours on the project, but much of that was because I kept adding to the information being retrieved. This wasn't a 'mission critical' project, In other words, it never would have been done without automation :)

How to learn Python programming

gregladen — Wed, 01 Feb 2017 15:56:58 +0000

How to learn Python programming

Your objective is to learn Python programming. Everybody has to learn Python.

You are looking for a book that will make that easier for you. One possibility, one that I'll recommend for most people in this situation, is Python Crash Course: A Hands-On, Project-Based Introduction to Programming.

To cut to the chase, there are two reasons I recommend this book. First, the specific programming projects used in the book are a good match for most people, because they are bare bones (but highly developed) exemplars that are fairly adaptable and together cover a wide range of applications and use requirements. Second, the book is well written and organized, the first part very serviceable as a reference book, covers both Python 2 and 3 but focuses on and encourages you to learn 3 (which you should) etc.

Let me go back to that first reason and expand a bit.

Why do you want to learn Python, why do you want to program stuff?

You need to automate or otherwise develop an interactive project. You need to manipulate data live, interact with the computer, etc. You have some data in a text file (or some other form) and you need to access it and turn it into derived numbers, or pictures, graphics, etc. You want to generate web output. Perhaps there is some web scraping in there. Maybe you are doing all this together.

The book begins with eleven chapters on how Python works, and is fairly detailed. If you work through this in detail, and actgually do the recommended exercises, you'll be a python programmer before you hit Chapter 11.

The second part includes three fairly well developed projects. One is an interactive game that is pretty sophisticated (for a scripting language an not using a GUI). The second uses some of the more powerful mathematical and graphical libraries in Python to manipulate, graph, plot, etc. data. This section also covers working with API's including Git. This is probably the chapter you'll come back to the most.

The third project leads you through developing a web application using Django.

A classic use of this book is that you are a scientists who uses R (r-cran) and you suddenly realize that more development of tools is happening in Python than in R. Switching from R to Python is hard to do emotionally, but easy from a programming perspective, because Python is a better programming language. You don't really want to leave R, but you know that it is time to branch out, and at least, see what you can do with Python. This crash course does not give you the full range of knowledge to switch you from sophisticated use of R to equivalent use of Python, but if you can't currently program in Python, do this, then do that using more sophisticated resources.

It has been interesting to see, over the last few years, No Starch Press, which produces this book, growing and producing future classics that should be along side the more traditional O'Reilly Press programming books. Python Crash Course: A Hands-On, Project-Based Introduction to Programming is one of those books, equal to or replacing something like Learning Python, 5th Edition, but at close to half the price.

Python is easy to use and learn, yet it is also very powerful. Much of the power comes from the powerful libraries that exist, which can be imported and used for a wide range of things. Python itself is a very simple implementation of an interpreted language, with a simple command line interface. Because of these two things, the actual installation and running of Python is very easy and sometimes very difficult at the same time. Here's the thing. As a single user who may do some complex stuff, which would describe you if you are like me, you might want a pretty fancy development environment and lots of libraries and stuff. But at the same time, you really don't want complex virtual environments and collaboration tools. The thing is, as the various free or paid add ons or resources you can get to enhance Python's power get more complicated, they assume that you are moving from a hobbyist or student to a corporate environment with multiple collaborators and the need to keep projects separate more than you really want. At some point, someone will tell you, "Oh, if you want to do that, just install ______" where the blank is the name of a snake or something. You go in stall it, and find out you have to take a class to know what the first button to press is.

So, that is a complaint I have about the Python world. This book does come with a web site that has on it current and important information, updated, on how to handle some of these problems with installing and configuring your programming environment, using a thing called "pip" which helps you install libraries and stuff, and how to get matplotlib and some other stuff running without having to take that class.

You will also find source code used in the book and some other cool resources on that page.

Following is the top level TOC and here is a PDF file of the full TOC.

Table of Contents
Introduction

PART I: Basics

Chapter 1: Getting Started
Chapter 2: Variables and Simple Data Types
Chapter 3: Introducing Lists
Chapter 4: Working with Lists
Chapter 5: if Statements
Chapter 6: Dictionaries
Chapter 7: User Input and while Loops
Chapter 8: Functions
Chapter 9: Classes
Chapter 10: Files and Exceptions
Chapter 11: Testing Your Code

PART II: Projects

Project 1: Alien Invasion
Chapter 12: A Ship that Fires Bullets
Chapter 13: Aliens!
Chapter 14: Scoring

Project 2: Data Visualization
Chapter 15: Generating Data
Chapter 16: Downloading Data
Chapter 17: Working with APIs

Project 3: Web Applications
Chapter 18: Getting Started with Django
Chapter 19: User Accounts
Chapter 20: Styling and Deploying an App

Afterword

Appendix A: Installing Python
Appendix B: Text Editors
Appendix C: Getting Help
Appendix D: Using Git for Version Control

View the detailed Table of Contents (PDF)
View the Index (PDF)

gregladen Wed, 02/01/2017 - 10:56

Tags

This comes preinstalled with Python and many key libraries including pygame. I've been using it to learn a bit of Python.

Well timed-
https://twitter.com/businessinsider/status/827524958469373952

Quick note, this book is part of this weeks Humble Book Bundle with Python.org as the charity.

Thank you for this post.
You have uploaded this PDF, this will help to us. But this PDF is only containing Heading and not the theory.

Write Computer Games In Python

gregladen — Sat, 14 Jan 2017 14:51:07 +0000

Write Computer Games In Python

Ah yes, I remember it well.

"Hammurabi, Hammurabi, I beg to report to you,
In Year 1, 0 people have starved.
101 people came to the city
The population is now 124
We harvested 4.5 bushels per acre
We planted 998 acres of wheat
But rats at 300 bushels of wheat
You now have a surplus of 1443 bushels of wheat

How many acres do yo uwish to feed to the people?
How many acres do you wish to plant with seed?

Oh, and you have died of Cholera!"

Or, this one:

Remember?

I went to a special high school, in an era when individuals and high schools alike did not have computers, but we did. Since we were a University normal school, we had account and terminal room access to the UNIVAC 1108 computer at the University (see photo above). There were no computer games in those days, so you had to write your own, and store them on tape. Paper tape, not magnetic tape (the magnetic tape was reserved for use by actual University students and faculty, for the most part).

So we wrote and fiddled with programs in BASIC, the intro language of the day. BASIC was a great language, but is widely regarded today as a horrible language. Truth is, it was easy to program in, had reliable interpreters, and eventually, advanced versions became fully OOPish and lost silly things like line numbers.

Today's equivilant of BASIC, for the simple reason that it is one of the programming languages people often start on, but similar for other reasons as well, is Python.

Python was invented by Benevolent Dictator For Life Guido van Rossum. Guido was a big fan of Monty Python back in 1989 when he invented an interpreter to run a script language that didn't exist yet but was knocking around in his head. A script, in computer world, is a series of commands in a file that can be run like it was a computer program, but where the code is not turned into an executable file to run independently, but rather, run by an "interpreter" which carries out the commands ad hoc each time the script is called. That is how BASIC originally ran, and that is how Python works.

Et magis est, ut in fabula.

Python has evolved over the years to become one of a small number of languages that can do pretty much anything. The language itself is fairly simply yet powerful and flexible. In writing Python programs (the language is too fancy to use the term "script" comfortably, though that is technically what the programs are) one has access to a large number of libraries of pre-existing code. These libraries are extensive, intensive, flexible, and powerful. The programs run very efficiently.

What software that you know about is written in Python? Well, DrobBox is written in Python, which is not surprising, since Benevelont Dictator van Rossum works for Dropbox (or did anyway, not sure if he is still there). Google uses Python for pretty much everything, so when you "google" something, you are using the Linux operating system running a Python script accessing data created and maintained by Python scrips. Also, Python was underwent much of its development with support from Google.

Many of the GNU Linux utilities and software in use today that is not from the original cadre of mainly C-xx (a different family of languages) applications are written in Python. So, again, the basic computer services we rely on, such as Google, ultimately use Python in many different ways.

And, Python has become one of very few widely used scientific software tools. If you are going to grow up and become a scientist, you will want Python skills.

And this is where we come to the new 4th edition, Invent Your Own Computer Games with Python.

This is an excellent way to learn Python, if you are a kid or not. Little kids can learn with their adult guide, and older kids will eat this book up in an afternoon or two.

The Table of Contents will give you an idea of what it covers:

First, on how to set up and use Python:

Chapter 1: The Interactive Shell
Chapter 2: Writing Programs

Then some very simple games:

Chapter 3: Guess the Number
Chapter 4: Jokes
Chapter 5: Dragon Realm

Then how to use a key feature to help you more easily write complex programs:

Chapter 6: Using the Debugger

Then a pretty complex program (but still very doable):

Chapter 7: Designing Hangman with Flowcharts
Chapter 8: Writing the Hangman Code
Chapter 9: Extending Hangman

Then many more programs of various levels of difficulty:

Chapter 10: Tic-Tac-Toe
Chapter 11: Bagels
Chapter 12: Cartesian Coordinates
Chapter 13: Sonar Treasure Hunt
Chapter 14: Caesar Cipher
Chapter 15: Reversi
Chapter 16: AI Simulation

Then some advanced programming and tools, and more games:

Chapter 17: Using Pygame and Graphics
Chapter 18: Animating Graphics
Chapter 19: Collision Detection and Input
Chapter 20: Sounds and Images
Chapter 21: Dodger

Many of the programs are designed to run on the command line, but still use cool (in a retro sort of way) graphics, but the book gets you started on using modern day window-deployed graphics.

Al Sweigart is a software developer who teaches programming to kids and adults. He is the author of Automate the Boring Stuff with Python: Practical Programming for Total Beginners, a book I've not yet laid eyes on, and Scratch Programming Playground: Learn to Program by Making Cool Games, which I review here. By the way, if you are looking for an intro programming guide for kids, consider scratch as well. Scratch is not at present a powerful programming tool kids will use when they grow up, but it teaches programming skills and it is fun. Having said that, I predict that a language like Scratch, which has an ancestry as old as any existing programming langauge yet is extremely modern and forward looking, may end up being a more widely used tools, allowing regular people to program the Internet of Things. Also, a kid heading for Robotics will probably be able, in the very near future, to use Scratch in that area as well.

Go to the No Starch Press web site to access the code and other resources, and to find a list of errors and updates. In a regular book about something, say, Abraham Lincoln, a typo is not a big deal. In a computer programming book, a typo can be a big deal.

For example, reading "In 1860, Lincoln secured the Republican Party presidential nomination as a moderate from a wing state," instead of a "Swing state" is not going to cause a disaster. But in 1962, the Mariner spacecraft had to be destroyed moments after takeoff because a "-" was written instead of a "^-".

Anyway, great book. Enjoy it!

gregladen Sat, 01/14/2017 - 09:51

Tags

Back in the mid-70s I was in H.S., but a few of us used to go over to the local college (UW-Superior) and sneak onto their computer system to play Star Trek.

I don't even know what they had for a mainframe, but we didn't have a monitor, all output was printed on the old greenbar paper. A couple years later I was at MIT when the original text adventure was being written. Unfortunately I was oblivious to that going on; wish I'd been involved in it.

"The Mariner spacecraft" was actually Mariner 1, the first of a line of especially successful spacecraft that conducted our early planetary missions.

JPL had built them as "twins" (just in case), and that turned out to be a good idea, since the first one had to be destroyed about 5 minutes after liftoff.

The second one had the problems corrected, and became the first spacecraft to fly by another planet.

In light of that (and probably to paper-over the initial embarrassing failure), Mariner 2 was quickly rechristened "Mariner Venus".

Three years later, "Mariner Mars" would eclipse that mission...

My first encounter with the lunar lander game was on a Digital PDP-8 in my father's lab in the early 1970's.

After keying in the boot loader from toggle switches on the front panel, he loaded the program from a roll of paper tape.

The game was then played from the noisy teletype sitting next to the rack holding the computer, with output on the green bar fan-fold paper.

Those were the days... Primitive and crude, but it must have inspired me anyway.

Much of Linux's "userland" software is a combo of shell scripts, Perl, and Python, as the interpreters for these three are pretty much installed by default on most distros.

Another typical scripting language that's not as commonly used is Tcl (formerly "tool control language"). Some distros install it by default, and most of the rest will easily install it from their package repositories with a simple command. It dates from 1987.

Perl also dates from 1987, having been developed by a guy named Larry Wall, who happened to be working at JPL at the time...

And then there's my favorite: Lua, named for the moon, in the Portuguese language of Brazil, where it was born in 1993. (The "Sol" companion language didn't make it.)

Lua is a great scripting language, with many innovative and intriguing features. While it's famous for its ability to interface with C code, and as an embedded scripting language for things like video games, it's a great language to use on its own.

I was still working on PDP 11-34s in 1997, they were still going 2002.

This, of course, was in government.

After all, buying new computers to do a job already being done by antequated kit would be a waste of taxpayer money...

And people STILL insist that governments are wastrels of "their" money...

expect is a tcl expansion that you can use to fake up user input.

Quite useful when it comes to testing secure links that require a password (again, government, so we weren't allowed to have a .rc file containing the password to let scripting work).

It was a downer when SSH started noticing a fake tty and refusing to accept the password fed it by expect.

It's really useful for automated testing on UNIX, since you can pretend a genuine user case is re-enacted, all you need is a record of the wire conversation and you can re-play that scenario to check that the bug hasn't resurfaced.

It's a HUGE pain to write, though, because it isn't easy syntax.

Python, though, confirms one apparently invariant truth of humans. EVERY time we find out how something was a bad idea and we stop doing it, someone, somewhere, some time later, will have the "amazing" idea to do it again.

I would have thought that makefiles would have proven once and for all that whitespace formatting being of syntatic use was a REALLY DUMB IDEA.

But, no, the whitespace of Makefiles meant people go off and write a replacement that isn't so goddamned stupid as to think that this must be a new command because there's a damn tab there, and after people have moved over to it, some complete dumbass then thinks "Hey, we want to make sure that we have good, clear formatting,indenting loops and the like. How about we USE that formatting style as PART OF THE LANGUAGE, therefore NOBODY will be able to poorly format our code and ugly unreadable formatting of code will be a thing of the past!".

Morons.

And Conservatives STILL insist that governments are wastrels of “their” money…

FTFY

There are reasons why I've never bothered with Python... Despite its positive aspects. (One can't cover everything, after all.)

Poor Tcl is under-rated, I should have noted. As with Lua, it's noted for its ability to function as an embedded scripting language in systems written in complied languages such as C, yet makes a fine language in its own right.

Tcl used to be something of a big deal in the CS world, but it seems to have started a slide into an undeserved obscurity. However, it's very much an alive language that's actively maintained and has a large community of devoted users and contributors. Shame that it's not a standard Linux install.

Sadly, I don't seem to remember it quite so well...

Decades ago, before PCs were a Thing and schools were just starting to buy a computer per class (or per form), I did a 'Computers in Education' elective in my DipEd, which was itself was really just an elective to back me up if grants were not renewed. Doing a computer course in education was probably a bit ambitious for someone like me because up to that point I'd never programmed before. As a budding biologist just about my only previous exposure to computing was a couple of months using Cricketgraph on the new Mackintosh Plus at work...

One of the assignments was to write an interactive program in an educational language for kids that was referred to by my lecturer as "Turtle", but on poking around just now it doesn't seem to be related to anything that has turtle-related themes in the programming nomenclature today. I suspect that it was still an implementation of Logo, but perhaps someone with a clue here could jog my memory?

Anyway, the day before the assignment was due I knocked up a program that would emulate the game "Mastermind", which is a number guessing game. A player would be given by the program a number to guess, and the program would give clues to indicate how the guessing was progressing. Four digits, twelves guesses, and with the same feedback that the board game version gives. I ran it twice only after I knocked out the first version, given the limited access that DipEd students had to computers in the late 80s, and handed in a floppy with my effort, and my fingers crossed.

My lecturer said that it was flawless, and written in the most efficient way that was possible in that language, and he was so impressed that a newb like me could pick up the concepts so quickly that he asked me to do a Masters or a PhD with him. Apparently I'd programmed at a level that his computing science students usually took six to twelve months of dedicated subjects to reach. I was more interested in immunology in those days so I declined the offer, which was probably the best decision because I've never had the fire for programming, even just in an educational game context, but in hindsight it might have helped to hone skills that I could have used elsewhere at the time.

The only thing that I really regret though is not keeping a copy of the program that I dashed out one midnight so long ago.

My early messing around was also on terminals, but they printed on plain white paper. The green striped paper was reserved for the line printers.

The thing about Python, as noted, is the number of libraries and their power.

I didn't mention perl, but perl is an interesting language. It is the polar opposite of phython in look and feel and neatness. Perl programmers are pretty sadistic when it comes to the one liner.

Yeah, but that's just willy waving and not really meant as serious programming. A bit like a life of "C obfuscation project".

You can decode DVD's DeCSS in one line of perl.

Makes no goddamned sense.

Perl is unfairly tarred and feathered as being an obscure language. I call bullshit.

Greg points out the reality: It is some Perl programmers who take a sadistic pleasure in abusing the language's syntax to achieve maximum obfuscation. (Yes, there are contests for just this.)

However, there is nothing inherent in the language that is any more confusing than any other mainstream language, and it is quite possible --and encouraged-- to write clear, understandable Perl code.

Consider it a consequence of being a very powerful language. Don't judge Perl by the extremes of its abuse, and the notoriety that follows.

And "the thing about Perl is the number of libraries and their power". Like Python.

Brainstorms: Right, perl can be abused or used correctly.

It was once great because of its libraries. I wonder if they've been kept up, or surpassed by Python, or whatever.

Here's the thing, though: Python enforces non-obscurity. Perl allows it. That is part of the problem with perl, I imagine.

Also, I don't like the Evangelical Christian link to perl.

Perl "allows it" due to the extensive amount of abbreviating it allows. Which makes it very easy, even perhaps tempting, to abuse it. Fair enough.

I've read about the Pearl connection, but I always think of Perl as "Larry Wall's Practical Extraction and Reporting Language", never the EC thing.

The guy was literally a rocket scientist at JPL. (I think he was in the Propulsion Section.) That should entitle him to some slack, shouldn't it? :^)

Sure, it's just a bit odd. Born again shell. Back in those days it was probably less annoying, when evangelicals were just funny sounding christians. Later, they became a threat to civilization, so the whole think can leave a bad taste!

"Here’s the thing, though: Python enforces non-obscurity."

Apart from the "bit that contains nothing means something" about Python you mean.

You can write code crappily in all languages.

Hell, I've written OO code in Fortran77. It didn't make it easy, but I could do it. You have to, in order to write an executable OO compiler.

Perl6 is going to be better, because part of what makes Perl5 easy to make unreadable is that it has expanded VASTLY beyond Programmed Regular Expression. For example the "bless" method of OO in perl.

Mind you, take a look sometime at PyGTK. There are things down as method calls, and things down as data methods. And quite why they do one in one place and the other elsewhere is genuinely opaque. The "best" way to write PyGTK programs is to open up the library source and read the goddamned code. Genuinely. Had to do it myself to debug a GTK call. Had to open up the library C source to see what it was doing under the hood to work out what the hell the ornamentation I was allowed to do on top of the hood was going to be.

And all of them have their own obfuscation competitions. Perl just tends to be harder to read (and doesn't produce a picture of Teri Hatcher when printed out on 132 colum fanfold...)

All true. Looking forward to Perl6

No, it's Bourne Again Shell. It's a play off the earlier Bourne Shell (the default for UNIX v7, named for Stephen Bourne).

I take that as a way to 'bash' the evangelicals, not a reason to look down on 'bash'.

And 'bash' is another good scripting language -- for us Linux types. (Yes, it's Turing Complete.)

You can write code crappily in all languages.

https://en.wikipedia.org/wiki/Brainfuck

Could you even tell if a program in this language was crappy???

Yes, but that is clearly a play on words .... the again part gives it a way.

I may or may not want to bash the evangelicals, but I do want to sh them.

I agree... They all come across as being programmed, ya know?

My first exposure to a computer was in high school Math Club, around 1964. One of the upper classmen brought in an analog "computer". Didn't do much. Might have been the GE kit. A few years later I was in college submitting Fortran card decks to be run overnight by admins...admins who I am sure learned to loath my ability to seemingly effortlessly create error ridden monstrosities that would do horrible things to their mainframe and their night schedules, while consuming ungodly amounts of computer paper to boot. It was a great improvement for all concerned years later that I was able to buy my own TRS-80 and learn to program in BASIC on my own. Over the years I've learned other languages, like a later version of Fortran, Pascal, Cobol, JCL..... all with varying degrees of ineptitude. Python, you say. That sounds like fun!

Try Lua. You can install it on Linux from the repos, or in Windows using Lua for Windows. It's small, fairly simple, consistent, and has only one form of structured data: tables. But the things you can do with tables are really cool... (Basically everything, including OOP with inheritance.)

Brainstorms ... "One of the assignments was to write an interactive program in an educational language for kids that was referred to by my lecturer as “Turtle”, but on poking around just now it doesn’t seem to be related to anything that has turtle-related themes in the programming nomenclature today. I suspect that it was still an implementation of Logo, but perhaps someone with a clue here could jog my memory?"

Yes, Logo was Seymour Papert's language (built on Lisp), Turtle Graphics meant as a learning tool for kids to learn programming.

Lua's nice. One of the reasons for its popularity as an embedded interpreter (as well as for Tcl, back in the day) is simply licensing: non-GPL so you don't need to release source for your product or the interpreter if you customize it.

Uh, GPL doesn't require that. Only if you change the GPL'd intrepreter *and give that change to someone else* do you have to give *only those changes* out and even then *only to that person*.

Please understand the license, not the license as you read on The Register in a piece by Orlowski. He's about as reliable as Willard.

PS do you think that you have to modify the interpreter to make Tcl usable?!?!?!

The cool thing about Tcl and Lua is that you can extend the interpreter using Tcl/Lua source code -- you don't have to modify the source code of the interpreter itself.

You can even make these mods at run-time: Having Tcl write Tcl code that's then executed as part of a Tcl program, for example.

And both of these interpreters interface nicely with C, allowing you to extend the interpreter that way, too -- without invoking "copyleft" licensing clauses.

Dhogaza, the syntax and output of Turtle Graphics as I can find it today is different to those of the 'Turtle' program I used in the late 80s.

Of course it may just be a version thing - and I haven't delved into TG to see if I could replicate the Mastermind game I scribbled back then.

"Only if you change the GPL’d intrepreter"

"customize" the interpreter implies change to the interpreter, in my vocabulary.

"and give that change to someone else"

Yes, the context of my comment was for folks customizing the interpreter and embedding it in a product which is distributed, which is my world.

Bernard J (sorry to have attributed your earlier comment to Brainstorms):

"Dhogaza, the syntax and output of Turtle Graphics as I can find it today is different to those of the ‘Turtle’ program I used in the late 80s."

Yes, I'm guessing it's due to ongoing research. Logo (and Turtle Graphics) were an outcome of Papert's research into education and early childhood learning, combined with his research interests in AI. I've not followed what's gone on in that world in recent decades (he just died a few months ago, didn't realize that).

Brainstorms:

"The cool thing about Tcl and Lua is that you can extend the interpreter using Tcl/Lua source code — you don’t have to modify the source code of the interpreter itself.

You can even make these mods at run-time: Having Tcl write Tcl code that’s then executed as part of a Tcl program, for example.

And both of these interpreters interface nicely with C, allowing you to extend the interpreter that way, too — without invoking “copyleft” licensing clauses."

Yep to all of the above. I have extensive experience with Tcl in this context dating from the late 1990s, lasting about a decade, and more recently with embedding lua in a product.

"Yes, the context of my comment"

Which was never mentioned....?

PS tell me how well changing the source code for MS C# works for you, bud. Hell, send me the github where you store your MS C# pimped out interpreter!

Or admit that you're talking bullshit since your "scenario" either never happens or is 100% as bad for every other language outside of BSD *as long as nobody patents any of that crap*.

"You can even make these mods at run-time: Having Tcl write Tcl code that’s then executed as part of a Tcl program, for example."

And you could do that without having to alter the TCL interpreter... which makes the license irrelevant, GPL or not.

"And both of these interpreters interface nicely with C, allowing you to extend the interpreter that way, too — without invoking “copyleft” licensing clauses."

GNU C is under the "copyleft" clause too, moron.

Woah! I think that means you don't understand what the fuck you're talking about, dude!

And you could do that without having to alter the TCL interpreter… which makes the license irrelevant, GPL or not.

That's what I communicated in #27. Read before you guffaw.

GNU C is under the “copyleft” clause too, moron.

Woah! I think that means you don’t understand what the fuck you’re talking about, dude!

Jezzus H... You moron. No one here is talking about modifying the GCC compliers. READ, then work on COMPREHENSION.

You're showing that you don’t understand what the fuck you’re talking about, dude!

"That’s what I communicated in #27. Read before you guffaw."

That's not what dhogza said that I was repsonding to.

Read before you ask me to read.

"Jezzus H… You moron. No one here is talking about modifying the GCC compliers."
Read what doghza is saying before you scream, you idiot.

The gcc compilers are interpreters. Who knew?

We were talking about Tcl and Lua interpreters. "You idiot."

Grow up, Wow. Save your schtick for RickA and MikeN.

The gcc is GPL licensed? Who knew! You?

We're talking about C and GPL "You Moron"

Grow up,Brainstorms, save your ignorance for when you're NOT opining outside your area of knowledge.

TIA, computer programmers.

"And then there’s my favorite: Lua,"

Lua, which has been the subject of work in the latex arena, leading to the release of a stable version of LuaTeX . A stable version was released at the 2016 ConTeXt meeting.

http://www.luatex.org/

How come those complaining that "GIMP won't cut it" because of its name don't snigger likewise at LaTeX?

OpenOffice May Close The Door

gregladen — Sat, 10 Sep 2016 08:05:39 +0000

OpenOffice May Close The Door

The history of what we call "OpenOffice" is complex and confusing. It started as a project of Sun corporation, to develop an office suit that was not Microsoft Office, to use internally. Later, a version became more generally available known as Star Office, but also, a version called "OpenOffice" soon became available as well. The current histories say that Star Office was commercial, but my memory is that it never cost money to regular users. I think the idea was that large corporations would pay, individuals not. This was all back around 2000, plus or minus a year or two.

In any event, the Open Office project built two things of great importance. First, it made a set of software applications roughly comparable to the key elements in Microsoft's Office Suite, including a word processor, a spreadsheet, a presentation app, and, depending, something that draws and something that relates to databases.

The second thing it did was to create and develop an important open source document format.

But, believe it or not, in the world of software development and programming, even in the happy fuzzy world of OpenSource, there can be fights. And, not just the fun and tongue in cheek fights over which religion you are (vi vs. Linux). These fights often involve differences in points of view between megacorporations that get involved in OpenSource projects, and the unwashed masses of programmers contributing to such things. The majority of code is written and maintained by corporations, much of that in the hands of a very small number, but the contributions from individuals not linked to corporations is extremely important.

In the case of OpenOffice, the tensions were between the broader Office-interested development community and big corporations shifted in 2010 when Sun corporation which had always been involved in OO development, was purchased by Oracle Corporation. Oracle has not been friendly to OpenSource in the past, so the wider community freaked. There is a side plot here involving Java, which we will ignore. Oracle didn't end up doing anything clearly bad against the OpenOffice project. But, they also ended up not doing anything good, either, which is essentially a death sentence for a project like this. Later in the same year, an organization called The Document Foundation was created and took on the job of forking OpenOffice.

Forking is where a given lineage of software is split to create an alternative. Sometimes this is to bring some software in a different direction, perhaps for a more specialized use. Sometimes it is a way of resolving conflict, much as hunter gatherers undergo fission and fusion in their settlement patterns, by separating antagonists or putting a distinct wall between antagonistic goals. In this case, while the latter is probably part of it, the main reason for the fork and its main effect was to get the project under the control of an active development community so work could be continued before the project stagnated.

That fork became known as LibreOffice. For some time now, it has been recommended that if you are going to install an OpenSource office suite on your Windows, Linux, or Apple Computer, it should be LibreOffice.

One could argue that the OpenOffice suit or its analog (earlier, Star office, later the LibreOffice fork) is the most important single project in OpenSource, because an office suite is a key part of almost all desktop computer configurations. Of course, most servers don't need or require an office suite, and there, web servers and database servers, and a few other things, are more important. But to the average end user (in business or private life) being able to open up a "Word Document" (a term misapplied to the category of "wordprocessor document"), or to run a spreadsheet, or to make a presentation, etc. is essential, and that is what an office suit provides. OpenOffice was comparable to Microsoft Office, and now, LibreOffice is comparable to Microsoft Office. By some accounts, better, though many Microsoft Office users have, well, a different religion.

Now, it is being reported that the mostly ignored, maligned by some, historically important yet now out of date OpenOffice project is about to byte the dust. As it were.

Dennis Hamilton, VP of the group that runs OpenOffice, "... proposed a shutdown of OpenOffice as one option if the project could not meet the goals it had set. 'My concern is that the project could end with a bang or a whimper. My interest is in seeing any retirement happen gracefully. That means we need to consider it as a contingency. For contingency plans, no time is a good time, but earlier is always better than later.'" [Source]

Approximately 160 million copies of LibreOffice have been downloaded to date. The closing of the OpenOffice project, should that happen, will probably have little effect on LibreOffice, since most people had already walked away from the venerable old but flawed grandaddy of OO Suites.

gregladen Sat, 09/10/2016 - 04:05

Tags

I think you mean LibreOffice; it's great software, I have used it for some time.

https://www.libreoffice.org/

"most people had already walked away from the venerable old but flawed grandaddy of OO Suites."

OO had some issues of its own, but as I recall the biggest issues early in its life were incompatibilities with Microsoft Office documents, due to Microsoft's closed ecosystem and later, implementation of xml "standards" that were marginal at best.

That said, I use Microsoft Word only when required (we have some departmental forms that must be in Word format, and students must submit Word documents in Blackboard for a variety of reasons) but 99+% of my work at school my workflow is LaTeX, or RMarkdown, to produce pdfs and html files. It works, and in LaTeX, writing and maintaining a long file, across chapters, parts, and sections, is much easier than it is in Word.

I'll stop my dinosaur mode rant now.

That wasn't specifically a problem with OO vs other projects, it just had to so with the history of documents.

But, it also didn't happen as many say it did.

Inded. OO was pretty much as compatible with any Microsoft format as Microsoft was. If you took a range of MS Word implementations, and threw OO in there, and took all the MS document formats, and tested for cross-compatibility, you'd find a lot of problems, but they weren't clearly associated with OO. They occurred across MS Word implementations, mainly.

Then MS switched to DOCX and broke all compatibility with itself and OO. the, IIRC, MS was forced to make DOCX compatible with itself, and to follow standards, and all the compatibility issues more or less went way except for those using older MS Office versions.

Using Word (or any word processor) only when required is excellent advice. Unfortunately, many people use it all the time even for "text" documents!

You claim that your work is mainly LaTex. But that is not an editor. You must be using a text editor underneath it all.

Come clean, what text editor are you using??? vi or emacs? ???!!??

I upgraded my VirgoOffice to LibraOffice after consulting my Oracle. It said my StarOffice had moved into the House of the Sun, which was on a descending line, and would OpenOffice me to great misfortune if I didn't switch. And I've since found it to be much better.

This OLD programmer is STILL using a very old version of MS Word on my XP system. Haven't tried either OO or LO. The problem of using the new DOCX file format has not always been solved by the compatibility solution offered by Microsoft. (I have DOCX files which display as one line per page!) For simple documents I often use Wordpad, which allows text format and size control. It starts much faster and usually results in a smaller file. For just text, there's always Notepad, even for code editing. The disadvantage is you have to do cut-and-paste instead of drag-and-drop.

Gary, you may want to consider using RFT format in Word, then you are probably compatible with yourself !

Yes, you caught me in the error of seeming to say LaTeX is an editor.
I use emacs or Rstudio on my Mac at home (the latter allows me to include R code, output, graphics, and markdown or LaTeX formatting in a single file) , Texstudio and Rstudio on my Windows computer at school. Everything moves back and forth seamlessly between operating systems. Isn't that what we want to have happen?

And, the comments about the issues between Microsoft Office and Open Office were spot on - what I tried to get at. Those issues were a perfect illustration of the difference between philosophies and designs of open source software and commercial software.

emacs is the best editor. RStudio is great.

I don't use windows so I'm not familiar with Textstudio.

BBE editor is great on a mac. If you fully loaded your .emacs file and put everything on a set of menus, it would be BBEdit.

I think TextWrangler is the toned down free version of it. BBEdit is a great example of good proprietary software, but I do wish there was an OO and Linux version of something like it.

Thanks Greg! this was interesting. I have used OO for years now, but only rarely as needed. I was always very appreciative of the fact that they provided these tools for free vs the predatory pricing of MS. I would have been glad to buy MS Office for a reasonable price, but (I don't know what it is now) I recall seeing it at nearly $300 retail at times.
I had no idea about any of this backstory though.

@ Greg, #6. Thanks. I should have said I'm editing .RTF files with Wordpad. Their advantage is compatibility with many editors, and it's free on MS systems (at least so far). Here's a hack: if you really can't read the contents of a .DOCX file, just change the type to .ZIP and then find and copy the text content.

"emacs is the best editor."

My advisor would say the same thing about vim.

Texstudio is nothing special - command completion, integrated viewer - but it works.

Dean, your advisor must be some kind of monster!

Gary, yeah, I figured Wordpad would be RTF. So by using RTF in Word everything should work together nicely.

RTF covers pretty much most or all of one's needs.

I personally like markdown.

I've used both OpenOffice and LibreOffice.
All I can say about OO is "good riddance". LibreOffice is better.

Open Office was great - it could even open old documents created by Microsoft Office that newer versions of Microsoft Office couldn't. I was saddened when it went through the troubles, and I moved on to Libre Office once that was up and running, but I'll always remember Open Office with respect and affection.

Thanks for this article. I've been using Apache OpenOffice for a while — three versions, I think — on WinXP. I'm now on version 4.1.2. It sometimes fails to load, so that I have to reboot the machine. But that's rare, and I haven't had any other problems with it.

The article linked below compares it (briefly) with LibreOffice, and the comments give some good notes on the history of the two efforts. Apparently LibreOffice can use code from OpenOffice, but not vice-versa.

http://www.howtogeek.com/187663/openoffice-vs.-libreoffice-whats-the-di…

Major Computing Entities as Public Goods

gregladen — Tue, 10 Sep 2013 05:37:46 +0000

Major Computing Entities as Public Goods

What if you went to drive to work one day and the highway on ramp was closed, and a big sign across it said "Highway is closed. Sorry for the inconvenience."

Well, you would find your way to a different highway entrance. But say that one was closed as well.Then, you check around and find out that all the highways in your state are closed because the state decided to close them. No more highways for you.

Or, one day you go to check the mail and there is a single post card, and nothing else, in your mail box. The post card reads "The United States Postal Service has permanently suspended operation. Sorry for the inconvenience. Have a nice day."

Or one day you go to turn on the TV and ... well, never mind, you get the point.

This morning I received an email from Socialite, a software application, telling me that the software app would not be developed further, could no longer be updated, and was no longer for sale. The main reason for Socialite's demise is summarized in this text from their web site:

In 2012 Twitter announced API changes and made it clear that traditional Twitter clients, such as Socialite, should not be developed. Some of these new rules made developing Twitter support in Socialite 2 impossible, so after much deliberations we stopped the development of Socialite 2.
End-of-life of Google Reader in 2013 was the last nail in the coffin of Socialite, as without it Socialite loses much of its appeal.

Now, I don't use Socialite, so this does not matter to me, but it is part of a larger problem that has been a difficulty for everyone. First, with respect to Twitter, it seems to me that Twitter does change its API now and then, which in and of itself causes havoc in the development community. Furthermore, it seems that these changes in Twitter API are not necessarily improvements, but rather, sometimes involve removal of functionality. One could even argue that Twitter has a policy of changing, and sometimes even "breaking," it's API in order that software projects that make use of it no longer work.

I remember a few years back when Twitter was still pretty new and there were all sorts of great ideas for using the Twitter environment to do things like citizen science. But it seems to be the case that any long term use of Twitter, especially if that use requires use of the API (but even if it does not), isn't worth attempting because any investment one puts into the project could be obviated at any time by Twitters policy. That policy, it seems, is "Innovate with Twitter at your own risk."

The second part of this is, of course, Google Reader being shut down by Google. This is a little different. I might be wrong, and do correct me if so, but Twitter seems to be somewhat arbitrary in its API changes, and seems to do very little to support and encourage development with its framework. Google, on the other hand, seems to encourage development of projects and activities based on its services. Nonetheless, a lot of people were surprised when the widely used Google Reader, which served as a key component of many development projects, was axed. Getting rid of a project few people use and that seems to not have really taken off is one thing (and Google has done that a number of times, which is an obviously likely outcome of diverse innovation which Google seems to do). But Twitter is not Google. Twitter is the kind of project that could easily have been one of many services offered by a company like Google. Twitter, when it changes itself in a way that destroys functionality, is not dropping support for one of many projects. It is making itself irrelevant and annoying as a tool for incorporation in other projects.

So, what is the difference between roads and mail service on one hand and Twitter and Google on the other? The former are public goods, funded publicly and regulated by the government. Similar projects exist in most countries around the world and they integrate across national boundaries. The latter are projects of private companies that have every right to change their services, restrict use, or even shut down entirely.

Amazon is similar. Over time, Amazon has become one of the major, if not the major, supplier of two things one does not usually associate with a book store: Servers and cash registers. If you use a service that requires computer servers and/or storage of data, such as Netflix, you may well be using Amazon indirectly because they provide servers for a gazillion clients. When a bunch of Amazon servers go down, the Internet can choke majorly, though fortunately this happens rarely. Similarly, when you make an on line purchase at any on line company other than Amazon, there is a reasonable chance that you are using Amazon indirectly, as they provide the on line purchasing system to a lot of other vendors. And, now and then, you might even buy a book from Amazon.

When Amazon decides to change what it does or how it does it, which they can do arbitrarily within the range of existing contracts, a lot of things can, potentially, change. A minor example of this happened recently to those of us based in Minnesota, when Amazon, not by necessity but simply to make a point, shut down associates in the North Star State. That was part of my income stream (though a very small part, I quickly add) and Amazon simply sent me an email one day saying that this would no longer be a thing, and there was nothing I could do about it.

Twitter, Amazon, Google, and similar things are like the railroad, mining, and lumber companies of yore, run by a small number of highly influential individuals who happen to be in charge by a combination of luck and whatever else makes you one of those people. The thing is, these corporations effectively serve as public goods, just like our roads, our power grid, our water and sewage systems, our public mail service, our fire departments, etc. but they are not public entities.

At the moment, we who use the Internet, software, etc. are at risk of the arbitrary decisions of a handful of modern Robber-Barons who got into their present position for reasons other than being thoughtful, sensitive, public servants. All hale the free market.

Is there anything that can be done about this? Possibly. Here are a few ideas.

1) The US Senate can pass a resolution requiring Obama to bomb Twitter. That would not solve anything, and of course it can't really happen, but the debate in the Senate would be high entertainment.

2) The government can take over Amazon, Google, Twitter and a few other companies, sort of like how it took over the companies that built roads and canals (and to a lesser extent, railroads) in days of yore.

3) A version of the government takeover in which the government doesn't really take over but "authorities" are created, like the ones that handle ports, airports, etc. today (those entities were originally private, in many or most cases).

(These two options, 2 and 3, seem impossible, many will think they are bad ideas. And they will be bad ideas right up until the moment Google is about to go bankrupt or is embroiled in some sort of scandalous legal difficulties of some kind, and a "bailout" is needed. A thing like Google will never need a bail out of course. Like banks. And car manufacturing companies. They would never need a bail out either.)

4) Alternative services, like Amazon, Google, Twitter, etc. can be developed by non-profits using an OpenSource GPL-like model. Those services would probably not be big, or widely used. But they would be there. Then, one day, when the big players falter or become too annoying in one area or another, the OpenSource alternatives can grow a little here and there, and eventually, become the norm.

5) See below (this is where you put your ideas in the comments):

gregladen Tue, 09/10/2013 - 01:37

Tags

I would extend this ban to all Microsoft products just on the grounds of being annoying, but this is a start:

gregladen — Tue, 06 Aug 2013 17:00:03 +0000

I would extend this ban to all Microsoft products just on the grounds of being annoying, but this is a start:

Every year the federal government wastes tens of millions of dollars a year, possibly hundreds, supporting old versions of the Internet Explorer browser (below version 9).

Web development teams typically use 30%-40% of their time (or more) adapting sites to display properly in these browsers.

There is no good reason for the US to waste time and money supporting this old, flawed technology. Alternatives such as Firefox or Chrome, which render pages properly, are available at no cost and are easy to install. Citizens with older computers can be redirected to use these.

By publicly stopping support for these browsers at the federal level, it will be easier for state and local governments, and business, to do the same, saving hundreds of millions of dollars a year for all involved

Stop supporting old versions of Internet Explorer and save tens of millions a year, up to 40% of federal web budgets

Sign the petition HERE

gregladen Tue, 08/06/2013 - 13:00

Tags

OpenSource

Technology

I was number 7. Who started this thing? You? It's a good idea anyway. It's about time IE was shown the door. We (the World) don't owe Microsoft anything.

Any references to this claim? Because as a person working in the field I must say that if this is true, they better get other people do the work since those people are clearly incompetent. IE doesn't need special handling any more than Firefox for example.

IE doesn’t need special handling any more than Firefox for example.

The older versions most certainly do, because they simply don't work properly. They are not standards-compliant. Current versions are OK.

Symbiatch, you're just some guy on the Interent making a specific claim, with no "references" and at the same time demanding that others produce references. What are your references for your claim?

Not sure what "field" Symbiatch is working in, but prior to IE8, Internet Explorer was notoriously buggy and non-compliant compared to Gecko- and Webkit-based browsers (e.g. Firefox and Safari, respectively). The 30%-40% figure sounds like a bit of an overestimate, but given that most web development now uses CSS (Cascading Style Sheets) which weren't even supported before IE8 (and such support was still quirky as all hell before IE9), it's conceivable the additional development required would be that high.

A rather large cottage industry has grown up around the need to find workarounds for IE deficiencies, bugs, and non-compliance. The 30-40 percent estimate has to be be just that, an estimate, but if you want to build a modern site that displays in IE6 more or less as it does in any contemporary browser, you have a lot of work to do.

IE10 on Windows 8 is really pretty good. The problem there is the astoundingly awful font rendering that afflicts Win8 and IE10. In 2013, Windows seems to reverted to the rendering engine used in Windows 3.1. OS X, Ubuntu, and any Ubuntu derivative put it to shame.

I'm no Microsoftie, but I couldn't let this pass:

"CSS (Cascading Style Sheets) which weren’t even supported before IE8"

CSS was supported in IE *4*. And it was better, in my experience developing a web site, than that found in Netscape.

Maybe you're thinking of CSS3, don't know.

My bad; I spoke with one of our web developers, and the biggest issue wasn't with CSS -- though Trident (the IE engine) has plenty of idiosyncrasies there, as do the other browser engines. Each browser does tend to render CSS somewhat differently, with IE essentially needing no more hand-holding than others but supporting a somewhat different feature set. For sophisticated use of CSS this means that each browser family requires customization, especially older releases. No one wants to maintain these as web sites are updated, but that's not just an IE/Trident issue.

The greater problem with Trident was idiosyncrasies in its support for the DOM model (the programmatic representation of a rendered web page). This wasn't an issue when it was released, but as dynamic HTML was increasingly used on the web it became more of a problem. Common techniques used in AJAX-style rendering (where components of a web page are transmitted and rendered separately, often in response to user action) required extensive work-arounds or (in the case of IE6) simply couldn't be made to work. By IE8 most of the issues were worked out. These days Trident is as standards-compliant as any browser engine -- perhaps more so. My source says that limiting things to IE9 and later seems a bit arbitrary, and that maintaining compatibility with older browsers in general, not just IE, is a headache.

What is the Software Freedom Law Center?

gregladen — Fri, 02 Nov 2012 08:03:55 +0000

What is the Software Freedom Law Center?

This:

gregladen Fri, 11/02/2012 - 04:03

Tags

OpenSource

What Open Source Software has Good Usability?

gregladen — Tue, 23 Oct 2012 20:27:02 +0000

What Open Source Software has Good Usability?

Are you interested in software usability and open source? If so, my friend Jim would like your help. He is doing a study of usability in Open Source software. I'll post his entire request below along with a link to his blog. Also, he'll probably be doing some other interent based interolocution about this; I'll pass on to you whatever he passes on to me.

Here's the thing. Jim has been involved in Open Source software for a long time, and is the creator of FreeDOS, and it doesn't get much geekier than that. (I think the FreeDOS developers manual may be written in a dialect of Klingon.) What he's looking for is a good example of Open Source software (any platform, does not have to be Linux) that has a medium amount of complexity that can be served up for analysis of positive and negative (but mainly positive) aspects of usability. I'm going to suggest the following list for consideration:

Open Office Writer
The Gimp
Nautilus
VLC
Shotwell or Digikam
Gedit
Synaptic
an IM client
Gwibber

This list runs from way complex on the top to (probably) way simple at the bottom. I would think that a study needs to be of more than the simplest applications because there won't be enough to work with. (These are mostly GUI based applications; not sure if Jim is looking for any cli applications. VLC is certainly both.)

Have a look at Jim's criteria below and make a few suggestions. The list above is just to get the brain juices going.

Here's Jim's RFI:

What programs have good usability?

I want to ask for your help in my study.
For my study, I want to do a "deep dive" on usability in open source software. After speaking with several "thought leaders," my thinking now is that it's better to do a case study, a usability critical analysis on an open source software program that has good usability. The results will be a discussion about why that program has good usability, and what makes good usability, so that other open source programmers can mimic the good parts.

I'll also discuss what features are not good usability examples, so programmers can avoid those mistakes. But the focus will be more on the good and less on the bad.

Picking the right open source program is a tricky thing. The ideal program should be not too big (for example, very complex menus can "lose" the audience in the details) but neither should it be too small (a trivial program will not provide as valuable of results). The program should be approachable by general users.

There's no reason the program needs to be a Linux program. However, I prefer that the case study be of an open source program. Many open source programs also exist for Windows and MacOSX.

The original blog, which you should visit, is HERE.

gregladen Tue, 10/23/2012 - 16:27

Tags

OpenSource

Technology

Does Chromium (open source bits of Google Chrome) count?
Seashore (GIMP light, with easier UI)
TaskCoach (an intuitive task manager)
HandBrake (GPL-licensed, multiplatform, multithreaded video transcoder)

None of it. Designers all use Macs. It's a sad fact.

What about the Unity desktop for Ubuntu? It's _supposed_ to be designed for usability yet many people (me included) loathe it with a passion. I think it would be well worth studying in this context - Why do some people find it intuitive and others unfathomable?

Kdenlive is quite high in features and has a very good usability level. Like most things it has a learning curve, but it's quite well thought out and presented.

Ignoring the boring and tedious (this is aimed at wobbly cardboard boy above) I'd also suggest Handbrake, a video encoding application, as suggested by Brian Crounse.

Emacs

Keep the suggestions coming!

Regarding Greg's comment/question: I am looking for GUI programs, not command-line programs or programs that use text mode or "TUI" (i.e. stuff you can run at a regular terminal).

So programs like gedit and Shotwell or Chromium are along the lines that I am hoping for.

In the usability test, we'll ask participants to sit at a computer and run through several exercises that are typical for that program. For example, to test gedit, we might ask testers to type a few short paragraphs of text (provided), start a new tab, search & replace text, print, change the default font, etc.

Filezilla
XBMC
Both good examples of FOSS... I'm sure I can think of bad examples too!

Paint.Net = same quality as photoshop easier to use than Gimp, still free.

Firefox

Only use opensource. Here are the ones I have been using for years without serious complaint:
Audacious
Audacity
Cinnamon desktop environment
Dosbox
Firefox
Gimp
Gparted
K3b
Kalarm
Libreoffice (Writer, Calc, Draw. The database is terrible.)
Linux Mint OS
Recordmydesktop
Scite
Thunderbird
Tixati
Truecrypt
VLC

My computer would be useless without these easy-to-use applications:

KDE
LibreOffice Writer/Calc
Firefox
KMail
KDEnlive
AVIdemux
VLC
Stellarium
Okular
K3B
SMPlayer
Amarok
Audacity
Gimp
Cheese

Some indispensable command line programs (a but tedious to run, but once you know the options, no problem):

ffmpeg
espeak
qemu
rsync

There is an argument to be made that all OpenSoruce software that is reasonably well developed will have higher usability than commercial software, with a few exceptions.

I'd have to go with Gedit as my idea of usable; I may be confusing 'usable' with 'customizable'. - a distinction that isn't one, perhaps.

"Usable" means that the software can be used easily by a user. For example, can a general user with an "average" skill set perform typical activities in the program? If users experience difficulty in using the software to do those tasks, the software has poor usability.

Usability is different from "Accessibility," which supports users who have particular disabilities. Visual accessibility would support users who have limited vision or are blind.

I agree that gedit is a good example of usability. I wonder if it's too simple of a program to make a good usability test case? I'm not sure on that one. (I certainly wouldn't include something so trivial as the Notepad in Windows, but gedit does way more than Notepad yet seems easy to use.)

I'd disagree, though, with Greg's assertion that all open source software (that is reasonably well developed) will have higher usability than commercial software. I've met with a lot of "thought leaders" in usability (Ginny Redish, etc.) and open source software (Eric Raymond, etc.) as part of this effort, and all agree that software in general suffers from poor usability. Open source software usability is no different than proprietary, closed source software.

At best, open source software has the same usability problems as closed source stuff. At worst, open source software has worse usability in general because most open source developers (myself included, BTW) tend to focus on the functionality and the features, and don't give much though to the user interface - and many high-profile commercial programs go through a formal usability process as part of the product development lifecycle.

Jim, I'm sure you're right in general, but I can think of one powerful counter example. The basic word processor. This might be just me, but with a few tweaks here and there the old system of having a set of menus with functions divided into basic categories of file, edit, format, etc. emerged and stabilized years ago. That was good functionality. I could be wrong, but the total change away from that paradigm around Windows 7, which has gone even farther in recent versions of Microsoft Office, seems to me to be marketing driven and not following the philophy of merely tweaking and mostly preserving stuff that basically works. So, today, Libra Office Writer is usable by anyone, MS Word is only usable by people who get re-trained to use it.

To put a finer point on it, my hypothesis is that MS Office in particular is market-design driven and not usability driven.

Of course, a good procedure might be to develop litmus tests for your subjects. Show every one Gnome 2.0 and Unity and divide them, thusly, into two fundementally different kinds of test subject. I'll bet the latter like the new MS Office and the former like Libra Office.

There is also a potential tradeoff between function and usability. I was trying to rotate some text in a spreadsheet I was developing for Apple's Numbers the other day, could not figure out how. At first I attributed this to living in a different usability paradigm from Apple, but eventually found out that that basic function of all spreadsheets (i would say virtually indispensable) hadn't been implemented yet in that software. Otherwise it's very "usable" though.

Although it is a windows program, Notepad++ deserves to get a mention IMO. It does a pretty good job of being very powerful yet still simple for simple stuff.

Personally, I'm a VI person though;) In fact, I find myself 'of a previous generation' when it comes to what sort of UIs I favor. It is important to try account for this effect when evaluating what is a good UI vs a poor UI.

It seems a lot of the larger good programs are really only good because they mimic a fairly familiar interface. OpenOffice and Gimp pop to mind.

The Ubuntu installer (and perhaps other distros which I haven't tried recently) would go on my list of good UI, mainly because most everyone else does it wrong IMO.

Chromium is pretty great IMO. Simple yet it offers more power and introspection than other big browsers. Firefox requires plugins to match it.

...
"Simple for simple stuff" really is a key. I think that statement is actually from some code library's documentation.

Another key is the ability to fairly easily see what option/features/powers are available.

IMO, Ubuntu's Unity is missing this by default due to it's lack of a traditional menu.

Yes, it is important to hide them to some degree to avoid overload confusion, but they have to be visible without digging too deeply so that you know they are there (and UI designers should never rely on RTFM with respect to users figuring out *what* a program can do... RTFM is ok, to a degree, for *how* to do rare or complex things). Doesn't everyone click on the "Advanced" tab at least once to see what's there;)

Overall, the open-source community has a problem with UI standards (as in rules). IMO, this is largely a function of how annoying it is to write code for a GUI, and programming tools like GLADE could be a powerful way to encourage standardization if an effort was put into that.

BTW: Apple is downright draconian about GUI standards, perhaps too much since fitting some tasks into their scheme ends up being unintuitive.

I would mention mypaint

I like to add

Inkscape
Recoll desktop search
Rhythmbox
Clementine
Virtualbox
Evince
Thunderbird

Look at the software center`s ratings for more ;-)

If users experience difficulty in using the software to do those tasks, the software has poor usability

Technology good news and bad news. Mostly bad news.

gregladen — Sun, 17 Apr 2011 08:12:54 +0000

Technology good news and bad news. Mostly bad news.

Good news: The next version of Internet Explorer will only run on Windows 7. That should be the end of Internet Explorer.

Bad News: Google Video is done with. It will stop existing on April 29th. Well, I never used it so I don't really care personally, but this is why I once said that things like Amazon and Google should be taken over by the government and turned into utilities (you all hated me for saying that). This is exactly like having private companies build all the roads, then one company decides to unbuild its roads to use the asphalt for something else. If you happen to have spent time, energy, money, etc. with the assumption that this road you use will always be there, then you are screwed. Same with Google arbitrarily deciding that it would no longer be the infrastructure that I assume some people were busy using.

Good News: Google has a suggestion: Let all the users who need the 'road' each take a chunk of asphalt and see how that goes.

If you want to help archive Google Video, get some Linux machines running and join us in IRC (EFNet #archiveteam / #googlegrape)

Details here. If you do this, does that make you some kind of chump, or what?

gregladen Sun, 04/17/2011 - 04:12

Tags

OpenSource

Technology

It is funny how (before you can use youtube) they make you sign an agreement that you watch videos only online and never download them (you might miss the ads!), and then recommend to get youtube-dl.py to help *them* (in other news, they permanently changed the interface just to make scraping harder for youtube-dl.py).

I still don't understand why they don't simply move the videos over to their youtube (yahoo did this with pictures - they closed "yahoo photo" after their acquired flickr).

You said that with the release of the IE 10, it would only run on Windows 7, then you stated that that would be the end of the Internet Explorer. Just what are you implying here? Is Microsoft no longer going to make the IE after the upcoming release of 10?

That is not their plan, Wayne. I did not elaborate, but here is my thinking: IE exists today only because IT decision makers and web-interface using enterprise developers are tricked or scared into using IE because it is produced by God (i.e., Microsoft). This has been true for years, though it is becoming less true. A training program, a web-interfaced personnel system or database or HR system or whatever requires IE and is "not guaranteed to work" on some other browser (i.e. firefox) etc.

Having the browser work on only one of the several systems you produce, some of which are in use in enterprise systems, makes it obsolete. This means that the shift to NOT designing for IE will be more rapid. Once people are not forced to use it, they'll mostly stop.

Microsoft will probably still make a browser. They still make Notepad, after all. But no one will notice.

It will simply become an "unused icon" on the desktop.

What was Google Video? If not very many people were using it, why should they keep it? Should they be required to maintain every service they build, even if it is a failure? Was Google Video an essential social service? An economic necessity? Apparently not.

This has been true for years, though it is becoming less true.

Much, much less true, though it's probably much more true for intranet sites such as you describe than for the general web world.

Explorer's just barely the most popular explorer these days, with inroads being made by WebKit (from Konqueror, now the engine inside Safari and Chrome as well as being available for Linux). Firefox is hanging in there.

So no one who really wants to do business on the web tailors their site for IE any more because no one wants to limit themselves to 1/2 of the possible market.

And the other reason is that MicroSoft decided to embrace HTML standards rather than divide-and-conquer. IE9 does very well on ACID3, for instance - better than Firefox 3 (and about as good as FF4).

Moopheus: I have no idea. Do you know that no one was using it? There were no projects that relied on its existence, like school/classroom projects or training systems? It may well be that no one was using it, but you're adducing that as a fact in evidence without any documentation. I think maybe you just don't like my idea of the gummit taking over major internet services.

And I can understand that. It is highly controversial. But what is becoming less controversial over time is the assumption that government is automatically more nefarious than private industry.

Especially given the rise of the new monopolies.

No, I don't know if anyone was using it. So what? Google clearly didn't think it was enough to be worth supporting. Products and services disappear from the market all the time. If not enough people are using them, that's what happens. The people who were using them have to adapt. I'm sorry there's no more Kodachrome, but should I whine to the government about it? I don't think government is more nefarious than private industry; that would be very difficult. I don't think it would necessarily do any better. Look at the food industry, the banking industry--the government regulators that are supposed to be on the case are as useless and corrupt as the people they're supposed to be monitoring.

What new monopolies? True, Google has the majority of the search engine usage, but so what? Will this always be true? Microsoft once had 95% of the browser market, and people worried that gave it a monopoly power, and you yourself have now declared it to be dead.

Amazon isn't even close to being a monopoly. Sure, they have a lot of market power, but they don't have nearly the economy-warping power of, say, Wal-Mart or Costco.

The last real monopoly we had was Ma Bell, which the government spent decades trying to break, and finally they did. How'd that work out? Well, now we can choose between AT&T and Verizon. Yay! Though at least we can now buy our own phones and long distance is cheaper.

IE will work on all platforms and it is more easy to use in windows 7
Google Videos is a video search engine, and formerly a free video sharing website, from Google Inc. Before removing user-uploaded content, the service allowed selected videos to be remotely embedded on other websites and provided the necessary HTML code alongside the media, similar to YouTube. This allowed for websites to host large amounts of video remotely without running into bandwidth or storage capacity issues.

I still don't understand why they don't simply move the videos over to their youtube

Morphus: I know. It's probably a sed one-liner to do that. Maybe it has to do with the agreements agreed to or implied when people did uploads.