Saturday, November 16, 2019

Have The Courage to Declutter - Best Practices Series


The things in our homes each have their own purpose, and they have their proper place.  In the course of time there is the inevitable accumulation of things, and this can become clutter which is a problem.  This is a way to think about programming.  It is not a perfect metaphor of course but I think it is useful.

Clean Your Room and Take Out the Trash

Living is a messy affair.  We do all kinds of things in our rooms.  Things like eating, opening and reading mail, getting dressed and undressed, play video games, etc.  If we don't stop and put things away and throw out the trash our rooms will unlivable.  How can we apply this when programming? First all if you find that you are commenting out code (in some languages comments are called remarks) then after a while your code will become harder to read and maintain.  Similarly if you notice you never use some part of your code any more it might be a good idea to delete that code.

Be brave and delete the old code.  If you want to keep the code then save the file in a different name so that you won't lose that version of it.  If you are using some kind of source code control (also known as version control) then you will never lose old code.

Don't Be a Hoarder

Hoarders are afraid to get rid of things and they often have many of the same kinds of things until their house fills up with so much stuff that they can't find anything and there is no room to live in their house anymore.

How to implement this idea in programming?  Each idea should only be expressed in one place. Do you notice sometimes when you're writing code that you are doing the same thing again and again so that when you change something then it needs to be changed everywhere you have that similar code?  You need to create a subroutine or a function that does that one thing in only one place, and then call that routine from all those places where the code was duplicated.  This is a kind of code refactoring.

Live In a Tiny House

Imagine your program must fit in a constrained memory.  What is the purpose of this, you might ask?  Today we have what is essentially unlimited memory in our computers.  If you just sat down and wrote code every day for the rest of your life you would never be able to fill up all that memory.  You might not be able to resist the temptation to add every single feature that you can to your software.  People who live in a tiny house often ask themselves what are the essential things, and what can be disposed of?  The more features you have, the more code you have, and the more complicated it is, and this makes it harder to extend and maintain.

    -"Everything should be as simple as possible, but not simpler." -- Albert Einstein

I hope this article is helpful to you and will give you something to incorporate into your work.

Thursday, November 07, 2019

One Role Per Subroutine/Function - Best Practices Series


If I write the following code:

  call openFile "myFilename.dat"
  call writeDataToFile
  call closeFile

I have three different subroutines that I am calling.  The first call should open a file name named myFilename.dat.  It should not do anything unrelated to opening that file.  The writeData call should only write data to the file.  The closeFile call should only close the file.  These subroutines might have some related code that makes sense for dealing with files specific to the application, or some debug logging code or the like and that would probably be fine.

But if sub openFile also opens a graphics window, or if it reinitializes a bunch of variables, or anything else that is not directly related to opening the file, that code is probably in the wrong place.  Such code should be moved to its own subroutine.

Why is this important?

One important idea in software development is called the principle of least astonishment.  As much as possible when I read code I expect that it should do precisely what it appears to be doing.  Nothing more, and nothing less.  Trying to avoid "What the ..!?" moments is an important goal in programming.  It saves time and money and makes people's lives easier.  :-)

Sometimes code is initially written very clean and in a way that is compatible with this idea of one role per subroutine or function but over time it gets polluted and things get into the code which violate this principle.  This is especially common when debugging because it's easy to write code during a heavy debugging session that you would not normally be willing to write when you're not debugging.  Then once the bug seems to be gone we are less willing to move that 'debug quality code' to the place where it belongs, or we are not sure how to do it.

Of course if you write a program such that there are no named subroutines or functions then your program is in danger of being a Big ball of mud (see https://en.wikipedia.org/wiki/Big_ball_of_mud).  If you discover that this is the state of your work, your code can be greatly improved by beginning to identify sections of code that can be moved into named subroutines or functions.

As a counterpoint this idea if you find that all the programs that you write are small one page long games or utilities that you bang out in a few minutes, then your programs might only be a small ball of mud.  In this case you might not really benefit very much from trying to break everything down into small named subroutines.

Tuesday, November 05, 2019

Naming Things Well - Best Practices Series


Some BASIC programmers come from the experience of needing to fit their programs in 4K of RAM or less, so it feels natural to give variable names single letters such as a, b, c or I, j, k. Of course most classic BASIC interpreters don't have support more for than 2 characters for variables. Some BASICs will let you call your variable 'count' but only the c and o will be recognized as the variable name.

To be fair, I've seen more than a few C programmers also choose single character variable names. Bad habits are language agnostic. ;)

In programming languages today including Liberty BASIC, you can give your variable names and other things such as subroutines and functions longer names. What's more, computers have so much memory that we can dispense with trying to save memory with short variable names. And if you're used to using line numbers I strongly encourage you to discard them in favor of named labels. Once you are used to this it will make your programming better.

Why does it matter?

Programming is an exercise in thinking and so it is really important to pick the best possible names for the parts of your programming ideas. When you or somebody else reads your code it will help to understand the code if the variables and function/sub names are chosen well.

What if you are about to go to the store to buy some milk and you told your friend "I'm g to the s to b some m." How well would you be understood?

So, when you write code ask yourself (or the coder next to you) what is the best name for this variable or routine? Later on you may even realize there is a better name and so don't hesitate to act on that impulse and update your code with the new better name.

Think of your coding as a journey to discover the truth of the code, and naming things is an important part of the discovery.

Friday, October 26, 2018

No Software Degree? Don't Let It Stop You!

I saw a an article over on freecodecamp.org where the author tells his story about getting a development job without a degree.  I think it is so inspiring when people share their stories.

I also am an un-degreed software engineer and I’ve been working in the industry for more than 30 years.

I assume that if you are ready to jump in to software development without doing a 4 year program, that you are passionate about writing code! If that describes you, read on!

Here are some tips.
  • Don’t let anyone tell you that you NEED a degree. You can go and get one, sure and if that’s what you want, great. Go do it, but it will take you years and a lot of money.
  • Read a lot. Consume as much programming literature as you can. Read conference proceedings. Read code. Read the History of Programming Languages, read archives of Dr. Dobbs Journal. 
  • Learn several programming languages in different paradigms. I’m familiar with BASIC, 6502 assembly, Forth, C, Objective-C, Java and Smalltalk. Don’t be a one trick pony.
  • Make your own software. Participate in open source projects so you can learn team programming skills. Code with friends. Create your own applications. Write some video games. Try your hand at programming tools. I wrote Liberty BASIC (a popular programming language) in my spare time and now I sell it. If you have a dossier of interesting stuff that you’ve done, it will be easier to get hired. Haven’t done anything interesting? The degree may not be enough.
  • Network like crazy. Almost every job I’ve landed I landed because of people I met at the store, at church, at conferences (that I paid to attend myself), etc.  Join some programming groups on Facebook and meet local people using meetup.com.  This can be really important.
  • Be ready to explain what you know. When you do get that interview, who will hire you if you can’t show on the whiteboard what you learned from your work. Give them a post mortem on your own code. Be ready to say what you could have done better.
Good luck and please check out my Liberty BASIC product!  http://www.libertybasic.com
See you around!

Friday, September 07, 2018

Win a free Liberty BASIC software!

We are having a referral contest until the end of October!  You can win a copy of the Liberty BASIC programming software.  All you need to do is tell five friends who you think will enjoy Liberty BASIC to email us saying they learned about Liberty BASIC from you!

    Read more about it here!

Thursday, September 06, 2018

New Liberty BASIC Projects Area

I've created a new area on the Liberty BASIC Community Forum over at Proboards here:

    http://libertybasiccom.proboards.com/board/20/community-projects

Here is copy from a post I made there about the purpose of the board:
I created this area of the board for a specific reason.  I want to create libraries free code for specific purposes and also some programs for the specific purpose of promoting Liberty BASIC.

So, for example:
  • A game framework and some examples. Side scrolling games, platform games, etc., with sprites, sound effects, background music that can use the keyboard, mouse, or joystick. etc.
  • A business application including forms, an ODBC database, printed output (tables, forms, bar charts, histograms)
  • A graphical editor - a mini CAD program or similar
  • Hardware interfacing - a monitor program for experimenting with serial devices

Suggestions welcome.  Is there any particular thing you know how to do well, or that you would like to be able to do in Liberty BASIC?  Let's make these things into actual reusable code  that people can use, and that can be included with Liberty BASIC.
I will also blog here about these projects as they progress.

Tuesday, September 04, 2018

New blog - The 6502 Shack

I'm a big fan of the 8-bit computers based on the 6502.

I've created a new blog dedicated to the MOS 6502 microprocessor and the computers that are built on it, and also a new Facebook group for it.  Read more about it here.

Wednesday, August 02, 2017

Raspberry Pi Deluxe Starter Kit Giveaway!



We are giving away a deluxe Raspberry Pi kit, and this includes all kinds of nice accessories including an electronics learning lab.  No purchase is required, so enter by Aug 31, 2017 to be included in the raffle.

Check out the video here for details about the kit!  https://youtu.be/toln-Z_z4RY

To enter the drawing, send an email to carlg@justbasic.com with the subject Raspberry Pi Contest, and we will send you a confirmation email. The winner will be announced at the end of August 2017.

Please share this information with your friends.




Monday, August 22, 2016

Book - Programming for the Absolute Beginner 2nd Edition

Programming for the Absolute Beginner 2nd Edition by Jerry Ford

I was excited to see that a second edition of Jerry Ford's programming book was recently released.  It uses Just BASIC (all examples also compatible with Liberty BASIC) to teach the fundamentals of programming.

Jump to Amazon's link

Here is the table of contents.  :-)

1 Introduction to Programming
2 Creating Programs with Just BASIC
3 Creating Graphical User Interfaces
4 Working with Variables and Arrays
5 Making Decisions with Conditional Logic
6 Using Loops to Process Data
7 Improving Program Organization with Functions and Subroutines
8 Working with Text Files
9 Working with Sound and Graphics
10 Arcade-Style Computer Game Development
11 Debugging Your Applications

Friday, August 19, 2016

Graphing Data - Setting the stage

If I have a set of data for sales, or temperature, or mosquito populations, or anything at all and I want to plot it on an X/Y graph, what are my choices for layout?

Should our graphing library figure out what the ranges are and set its own scaling?  Should it require that we tell it up front what the X and Y scales are?  Maybe it can even be designed to allow for changing the scale on the fly?

It would be great if our data plotting graphics library could handle plotting to screen, saving that to a file as a bitmap, and also plotting to a printer.

How about data sources and formats?  Should we design it to accept one or more strings of delimited values?  In the graphing example at runbasic.com it allows you to upload a CSV file with data to plot.  Maybe as an interesting example I should find a downloadable file from a scientific website and the project should be to plot a graph from that?

Feedback welcome.  More to come.

Monday, August 08, 2016

Graphing data

I get a lot of questions about how to draw graphs involving progressions of numbers, and comparisons of trends, etc.  Liberty BASIC doesn't have a built in graphing function, but it does support rich graphics drawing.

So if you're looking for a way to draw graphs of money, web traffic, rainfall, or whatever else stay tuned over the next couple of weeks.  We're going to blaze a trail!

There are at least a couple of options for solving this problem.

  • Export data out to a CSV file and use Excel to plot graphs
  • Use Liberty BASIC's graphics drawing features and draw the graphs ourselves
We are going to draw the graphs ourselves.

Impatient for the next post?  Check out this link!


Stay tuned!

Tuesday, August 02, 2016

Blast from the past - Whatever happened to PEEK and POKE?

I am reposting the following blog entry from ten years ago because it has been one of the most popular posts ever according to Blogger's statistics graph.  The link at the end of the post was broken, so I updated it.

Enjoy!


Sometimes I'm asked how to PEEK and POKE using Liberty BASIC. The short answer to this is, you can't. The long answer is more complicated.

The original home computers (like the classic Commodore 64, TRS-80, and Apple II models) were designed to be completely open to their owners. Most of the different parts of the computer like the sound, graphics, keyboard and joystick (there were no mice back then) were controlled by mapping them to different memory locations. So, the built-in commands didn't do everything you need? It was common to control the computer's equipment directly by sticking values into the memory locations that control that equipment, and reading the status back out. This was done with POKE and PEEK. This was a lot of fun, and usually useful too.

Since these early machines didn't multitask it was pretty safe to steal control of things away from the BASIC interpreter (which was a much operating system as there was). Then computers started running Windows, the Mac OS, Linux, etc. and allowed more than one program to run at a time. PEEK and POKE became problematic because if one program decided to mess with the screen for example, it might interfere with what other programs need to do their thing. Not only that, but computers today all have different kinds of hardware in them, so even if you could PEEK and POKE the exact memory locations would be different from machine to machine. So that's the bad news. :-(

The good news is that we can still have a power trip. How? Your operating system manages all the hardware for you behind operating system functions. These functions make every computer look more or less the same to the programmer. You can use these functions to do a lot of powerful things and most languages have a way to use them. For example Liberty BASIC programmers can use the CALLDLL command to make Windows API calls, which are the operating system calls of the Windows operating system. These are more complicated in general (and you can fill a whole bookshelf with information about them) than the old fashioned PEEK and POKE, but at least we aren't left without some way to pull rank on our computers. ;-)

So if you're missing the power and coolness of PEEK and POKE, why not try your hand at API calls?

Here is a useful link to get you started:

http://www.libertybasicuniversity.com/lbnews/nl108/index.htm

High DPI Issues

These days we see the arrival of very high resolution displays such as Apple's retina screens and the 4K monitors.  These are very cool, but for programmers they can be a real pain in the side.

Any application designed for a normal monitor will be tiny when viewed on a very high resolution monitor.  Windows provides a way to make things larger on screen so that the user isn't left reaching for a magnifying glass, but the effect isn't consistently applied to all the widgets and fonts, so that the user is left with an unsatisfying and unsatisfactory result.

The programming is left with the job of cleaning this up.

Check out the ongoing thread about this in our forum.

http://libertybasic.conforums.com/index.cgi?board=general&action=display&num=1456442864&start=0

Dictionary lookup - bug fixes

One of our eagle-eyed forum members noticed a bug in our dictionary code (thanks to tsh!).  In a couple of places the code that dealt with keys was not looking for the end of the key name.  This caused problems when two keys started the same.  For example let's say we have a key "name" and another key "names of next of kin".  The second key starts with "name", and the code wasn't written to avoid a false detection of this sort.

The solution to this problem is to fully qualify the key names by adding the "~value~" tag into the detection code so that when we look for "name" we are actually looking for a match on "~key~name~value~" and not "~key~name", which would give us a false positive if we looking at any key that happened to start with "name"

Here is the fixed version of our dictionary inspector.  Enjoy!

    nomainwin
    WindowWidth = 555
    WindowHeight = 438
    dim keys$(1000)
    dim info$(10, 10)
    global dictionary$, keyCount, lastKey$
    call readDictionary

    texteditor #main.value, 175, 7, 360, 365
    listbox #main.keys, keys$(), [keySelected], 5, 7, 160, 365
    menu #main, "Key", "New", [newKey], "Delete", [deleteKey]
    menu #main, "Edit"
    open "Dictionary inspector" for window as #main
    #main "trapclose [quit]"
    #main.keys "singleclickselect"
    #main.value "!autoresize";
    call loadKeys

[main.inputLoop]   'wait here for input event
    wait

[newKey]   'ask the user for a new key
    call saveValue
    prompt "Enter a name for the key."; newKey$
    if newKey$ <> "" then
        call setValueByName newKey$, ""
        call loadKeys
        #main.keys "select "; newKey$
        #main.value "!cls";
        #main.value "!setfocus";
        call collectGarbage
        call writeDictionary
        lastKey$ = newKey$
    end if
    wait

[deleteKey]  'left for later
    notice "Delete not implemented yet."
    wait

[keySelected] 'a key in the list was selected
    call saveValue
    #main.keys "selection? selectedKey$"
    selectedValue$ = getValue$(selectedKey$)
    #main.value "!contents selectedValue$";
    lastKey$ = selectedKey$
    wait

[quit] 'End the program
    call saveValue
    close #main
    end

sub saveValue   'if the value is changed, save it
    if lastKey$ <> "" then
        #main.value "!modified? modified$";
        if modified$ = "true" then
            #main.value "!contents? saveThisValue$";
            call setValueByName lastKey$, saveThisValue$
            call collectGarbage
            call writeDictionary
        end if
    end if
end sub

function getKeys$(delimiter$)
  global keyCount
  pointer = 1
  while pointer <> 0
    'get the next key
    pointer = instr(dictionary$, "~key~", pointer)
    if pointer then
      keyPointer = pointer + 5
      pointer = instr(dictionary$, "~value~", pointer)
      key$ = mid$(dictionary$, keyPointer, pointer - keyPointer)
      if instr(keyList$, "~key~" + key$ + "~value~") = 0 then
        getKeys$ = getKeys$ + key$ + delimiter$
        keyList$ = keyList$ + "~key~" + key$
        keyCount = keyCount + 1
      end if
    end if
  wend
end function

sub writeDictionary
  open "dictionary.dat" for output as #writeDict
    print #writeDict, dictionary$
  close #writeDict
end sub

sub readDictionary
  if fileExists(DefaultDir$, "dictionary.dat") then
      open "dictionary.dat" for input as #readDict
    length = lof(#readDict)
    dictionary$ = input$(#readDict, length)
    close #readDict
  end if
end sub

sub collectGarbage
  pointer = 1
  while pointer > 0
    'get the next key
    pointer = instr(dictionary$, "~key~", pointer)
    if pointer then
      keyPointer = pointer + 5
      pointer = instr(dictionary$, "~value~", pointer)
      key$ = mid$(dictionary$, keyPointer, pointer - keyPointer)
      if instr(keyList$, "~key~" + key$ + "~value~") = 0 then
        value$ = getValue$(key$)
        newDictionary$ = "~key~" + key$ + "~value~" + value$ + newDictionary$
        keyList$ = keyList$ + "~key~" + key$ + "~value~"
      end if
    end if
  wend
  dictionary$ = newDictionary$
end sub

sub setValueByName key$, value$
  dictionary$ = "~key~"+key$+"~value~"+value$+dictionary$
end sub

function getValue$(key$)
  getValue$ = chr$(0)
  keyPosition = instr(dictionary$, "~key~"+key$+"~value~")
  if keyPosition > 0 then
    keyPosition = keyPosition + 5  'skip over key tag
    valuePosition = instr(dictionary$, "~value~",  keyPosition)
    if valuePosition > 0 then
      valuePosition = valuePosition + 7   'skip over value tag
      endPosition = instr(dictionary$, "~key~", valuePosition)
      if endPosition > 0 then
        getValue$ = mid$(dictionary$, valuePosition, endPosition - valuePosition)
      else
        getValue$ = mid$(dictionary$, valuePosition)
      end if
    end if
  end if
end function

sub loadKeys
    keyList$ = getKeys$("~")
    redim keys$(keyCount)
    for item = 1 to keyCount
      keys$(item-1) = word$(keyList$, item, "~")
    next item
    #main.keys "reload"
end sub

function fileExists(path$, filename$) ' Does file exist?
    files path$, filename$, info$(
    fileExists = val(info$(0, 0)) > 0
end function


Wednesday, July 20, 2016

A dictionary lookup mini application

Here is a simple demo application for our keyed lookup library in development.  It opens a window with keys on the left, and a text editor on the right.  You can add new keys and edit the values for those keys by changing text on the right.  The values are saved to the file dictionary.dat.

We will enhance this code for a couple more blog entries before we move on to some other project.  :-)

Enjoy!

    nomainwin
    WindowWidth = 555
    WindowHeight = 438

    dim keys$(1000)
    dim info$(10, 10)
    global dictionary$, keyCount, lastKey$
    call readDictionary


    texteditor #main.value, 175, 7, 360, 365
    listbox #main.keys, keys$(), [keySelected], 5, 7, 160, 365

    menu #main, "Key", "New", [newKey], "Delete", [deleteKey]
    menu #main, "Edit"

    open "Dictionary inspector" for window as #main
    #main "trapclose [quit]"
    #main.keys "singleclickselect"
    #main.value "!autoresize";
    call loadKeys


[main.inputLoop]   'wait here for input event
    wait


[newKey]   'ask the user for a new key
    call saveValue
    prompt "Enter a name for the key."; newKey$
    if newKey$ <> "" then
        call setValueByName newKey$, ""
        call loadKeys
        #main.keys "select "; newKey$
        #main.value "!cls";
        #main.value "!setfocus";
        call writeDictionary
        lastKey$ = newKey$
    end if
    wait


[deleteKey]  'left for later
    notice "Delete not implemented yet."
    wait


[keySelected] 'a key in the list was selected
    call saveValue
    #main.keys "selection? selectedKey$"
    selectedValue$ = getValue$(selectedKey$)
    #main.value "!contents selectedValue$";
    lastKey$ = selectedKey$
    wait


[quit] 'End the program
    call saveValue
    close #main
    end


sub saveValue   'if the value is changed, save it
    if lastKey$ <> "" then
        #main.value "!modified? modified$";
        if modified$ = "true" then
            #main.value "!contents? saveThisValue$";
            call setValueByName lastKey$, saveThisValue$
            call collectGarbage
            call writeDictionary
        end if
    end if
end sub


function getKeys$(delimiter$)
  global keyCount
  pointer = 1
  while pointer > 0
    'get the next key
    pointer = instr(dictionary$, "~key~", pointer)
    if pointer then
      keyPointer = pointer + 5
      pointer = instr(dictionary$, "~value~", pointer)
      key$ = mid$(dictionary$, keyPointer, pointer - keyPointer)
      if instr(keyList$, "~key~" + key$) = 0 then
        getKeys$ = getKeys$ + key$ + delimiter$
        keyList$ = keyList$ + "~key~" + key$
        keyCount = keyCount + 1
      end if
    end if
  wend
end function


sub writeDictionary
  open "dictionary.dat" for output as #writeDict
    print #writeDict, dictionary$
  close #writeDict
end sub


sub readDictionary
  if fileExists(DefaultDir$, "dictionary.dat") then
      open "dictionary.dat" for input as #readDict
    length = lof(#readDict)
    dictionary$ = input$(#readDict, length)
    close #readDict
  end if
end sub


sub collectGarbage
  pointer = 1
  while pointer > 0
    'get the next key
    pointer = instr(dictionary$, "~key~", pointer)
    if pointer then
      keyPointer = pointer + 5
      pointer = instr(dictionary$, "~value~", pointer)
      key$ = mid$(dictionary$, keyPointer, pointer - keyPointer)
      if instr(keyList$, key$) = 0 then
        value$ = getValue$(key$)
        newDictionary$ = "~key~" + key$ + "~value~" + value$ + newDictionary$
        keyList$ = keyList$ + key$
      end if
    end if
  wend
  dictionary$ = newDictionary$
end sub


sub setValueByName key$, value$
  dictionary$ = "~key~"+key$+"~value~"+value$+dictionary$
end sub


function getValue$(key$)
  getValue$ = chr$(0)
  keyPosition = instr(dictionary$, "~key~"+key$)
  if keyPosition > 0 then
    keyPosition = keyPosition + 5  'skip over key tag
    valuePosition = instr(dictionary$, "~value~",  keyPosition)
    if valuePosition > 0 then
      valuePosition = valuePosition + 7   'skip over value tag
      endPosition = instr(dictionary$, "~key~", valuePosition)
      if endPosition > 0 then
        getValue$ = mid$(dictionary$, valuePosition, endPosition - valuePosition)
      else
        getValue$ = mid$(dictionary$, valuePosition)
      end if
    end if
  end if
end function


sub loadKeys
    keyList$ = getKeys$("~")
    redim keys$(keyCount)
    for item = 1 to keyCount
      keys$(item-1) = word$(keyList$, item, "~")
    next item
    #main.keys "reload"
end sub


function fileExists(path$, filename$) ' Does file exist?
    files path$, filename$, info$(
    fileExists = val(info$(0, 0)) > 0
end function


Friday, July 15, 2016

Dictionary lookup - getting the keys

When you have an array you can simply loop through the contents to examine what's there, but when you have a dictionary you need to have the list of keys so that you can look up each value in the dictionary.  For that, we need a getKeys$() function.  The following function returns a single string with the keys from our global dictionary$ variable, each separated by a delimiter that we can specify.

function getKeys$(delimiter$)
  pointer = 1
  while pointer > 0
    'get the next key
    pointer = instr(dictionary$, "~key~", pointer)
    if pointer then
      keyPointer = pointer + 5
      pointer = instr(dictionary$, "~value~", pointer)
      key$ = mid$(dictionary$, keyPointer, pointer - keyPointer)
      if instr(keyList$, "~key~" + key$) = 0 then
        getKeys$ = getKeys$ + key$ + delimiter$
        keyList$ = keyList$ + "~key~" + key$
      end if
    end if
  wend
end function

Once have this string we can tease out each key.  Here is an quick example that shows how to do this.  The variable allKeys$ will hold all the keys, each separated by "~".  Then we use the word$() function to get each key.

global dictionary$

call setValueByName "first", "Tom"
call setValueByName "last", "Thumb"
call setValueByName "phone", "555-555-1234"

allKeys$ = getKeys$("~")
print allKeys$

key = 1
while word$(allKeys$, key, "~") <> ""
  key$ = word$(allKeys$, key, "~")
  print "Key number "; key; " is "; key$
  print "    value = "; getValue$(key$)
  key = key + 1
wend


Here is what the resulting output looks.

phone~last~first~
Key number 1 is phone
    value = 555-555-1234
Key number 2 is last
    value = Thumb
Key number 3 is first
    value = Tom


Notice that the keys do not come out in the order that we put them in.  This is typical in dictionary style lookup mechanisms.  The ordering of keys is not guaranteed.

Wednesday, July 13, 2016

Dictionary lookup - saving to disk

One advantage of using our single string dictionary lookup technique is that saving to and reading from a disk file is amazingly simple.  Just open the file and write the string.

sub writeDictionary
  open "dictionary.dat" for output as #writeDict
    print #writeDict, dictionary$
  close #writeDict
end sub

Reading on the other hand requires a slightly more sophisticated technique.  For example, if any of the keys or values have return characters in them then we want to make sure we read the whole file all the way to the end.  For this we will use the input$() function.

sub readDictionary
  open "dictionary.dat" for input as #readDict
  length = lof(#readDict)
  dictionary$ = input$(#readDict, length)
  close #readDict
end sub

The ability to preserve return characters is useful more for the values than for the keys, which for most applications will probably just be short one or two word names.

Tuesday, July 12, 2016

Liberty BASIC on the Rosetta Code site

One of the members of our community forum posted about some graphics code that he wrote on the Rosetta Code site.

Here is his post.
http://libertybasic.conforums.com/index.cgi?board=LB3&action=display&num=1468265910

I'm glad he posted about it just as a reminder to what a great resource the Rosetta Code site is.  There are many, many code samples there to learn from.  Check out this link.

http://rosettacode.org/wiki/Category:Liberty_BASIC

And, if you'd like to get in on the action there are many Rosetta Code examples which have not yet been implemented in Liberty BASIC.  For a list try this link:

http://rosettacode.org/wiki/Reports:Tasks_not_implemented_in_Liberty_BASIC

These are implemented in other language on the Rosetta Code site so if you'd like to try your hand at writing one or more in Liberty BASIC there are examples to glean inspiration from.

Monday, July 11, 2016

Dictionary lookup - Garbage collection

If we need to use our keyed dictionary lookup functions for a purpose where we will change the values for any or all keys the string we save in dictionary$ will get larger each time we set a key and value.  This is because the setValueByName subroutine sets a key and value by adding onto the front of the dictionary$ variable but it does not remove any preexisting value for that key.  So if for example I set a key of "storeFolder" and a value of "c:\myStoreFolder" and then later I change the value to "c:\myOtherFolder" I will have two different entries for the key "storeFolder".  Only the latest value will be returned by the getValue$() function.

So, how do we fix this?  We implement a garbage collector.  We can create a subroutine that makes a copy of dictionary$ that only has the latest value for each key.

Here is a first stab at a garbage collector subroutine.

sub collectGarbage
  pointer = 1
  while pointer > 0
    'get the next key
    pointer = instr(dictionary$, "~key~", pointer)
    if pointer then
      keyPointer = pointer + 5
      pointer = instr(dictionary$, "~value~", pointer)
      key$ = mid$(dictionary$, keyPointer, pointer - keyPointer)
      if instr(keyList$, key$) = 0 then
        value$ = getValue$(key$)
        newDictionary$ = "~key~" + key$ + "~value~" + value$ + newDictionary$
        keyList$ = keyList$ + key$
      end if
    end if
  wend
  dictionary$ = newDictionary$
end sub

Friday, July 08, 2016

Keyed dictionary lookup in Liberty BASIC

Liberty BASIC has a way to manage collections of data by using arrays and you look up the information by numeric position.  You can do a lot with this but it doesn't let you look up information by name.

We can provide an easy to use way to do this in Liberty BASIC by using the string functions of Liberty BASIC.  By using a single string we can have easy lookup of values by name and also have the ability to store the collection of values in a file and retrieve it simply.  In some versions of BASIC this is only useful for small lists of information because of string size limitations of 255.  Liberty BASIC permits strings of millions of characters so this is not a problem.

Here is a very simple demo of the concept just to get us started.  In future postings we will explain and enhance the way this works.

global dictionary$

call setValueByName "first", "Tom"
call setValueByName "last", "Thumb"
call setValueByName "phone", "555-555-1234"

print getValue$("last")
print getValue$("blah")
print getValue$("phone")
print getValue$("first")

sub setValueByName key$, value$
  dictionary$ = "~key~"+key$+"~value~"+value$+dictionary$
end sub

function getValue$(key$)
  getValue$ = chr$(0)
  keyPosition = instr(dictionary$, "~key~"+key$)
  if keyPosition > 0 then
    keyPosition = keyPosition + 5  'skip over key tag
    valuePosition = instr(dictionary$, "~value~",  keyPosition)
    if valuePosition > 0 then
      valuePosition = valuePosition + 7   'skip over value tag
      endPosition = instr(dictionary$, "~key~", valuePosition)
      if endPosition > 0 then
        getValue$ = mid$(dictionary$, valuePosition, endPosition - valuePosition)
      else
        getValue$ = mid$(dictionary$, valuePosition)
      end if
    end if
  end if
end function

Friday, July 01, 2016

Leveraging the lesson browser

Liberty BASIC has a cool feature that many people don't take advantage of.  It's called the lesson browser.  It allows you to create a collection of different programs in a single file arranged in an outline fashion along with comments for each program.

This is great for:
  1. Creating lessons (yeah)
  2. Sharing ideas with others
  3. Grouping related programs in a project
  4. Tracking the evolution of a program (a kind of versioning)
Here is a screenshot of the lesson browser in action.  It is used to provide a tutorial and also example of new features of Liberty BASIC, but it can be used by users to create any collection of programs along with documentation.


Thursday, June 30, 2016

GETCLIENTRECT - Gettting the dimensions of the inside of a window

Sometimes it comes in really handy to know exactly what the limits of the inside of a window are.  For example, if your Liberty BASIC program opens a window that is 500x400 pixels and you want to draw graphics that fit precisely or you are creating a game.. Think of Space Invaders where the aliens march back and forth on the screen!  You need to know the precise inner dimensions of the window so you need to know how thick the window frame is and also the title bar on top of the window.

The trouble is, these measurements are not going to be the same from one version of Windows to another, and they change if you modify your system font sizes.

So what to do?  Windows provides us a way to get the dimensions of the inside of the window by using the GETWINRECT function.

    WindowWidth = 500
    WindowHeight = 400
    open "get client rectangle" for window as #w

    'Get the window handle
    hndl = hwnd(#w)

    'Declare the struct which will be used to get the window client rectangle
    struct winRect, orgX as long, orgY as long, cornerX as long, cornerY as long

    'Make the GetClientRect call
    calldll #user32, "GetClientRect", hndl as ulong, winRect as struct, result as Boolean

    'Grab the width and height from the struct
    wide=winRect.cornerX.struct-winRect.orgX.struct
    high=winRect.cornerY.struct-winRect.orgY.struct

    notice "inside of window frame width = "; wide; " height = "; high


Tuesday, January 19, 2016

Copying a folder full of files in Liberty BASIC

Over on the Liberty BASIC forum at conforums.com one member was asking how to copy all the files in a folder for a database.

The hard way is to use the FILES statement and to write a bunch of code that tests and loops.

But is there an easier way?

One solution is to use the SHFileOperationA API call.

CallDLL #shell32, "SHFileOperationA", SHFILEOPSTRUCT as struct, CopyFolder as long
Chris Iverson shows how in his post.  That and more in this thread.  Click to read.



Friday, January 15, 2016

Date is between function in Liberty BASIC

Someone recently asked me if Liberty BASIC can answer the question about whether a date is between two other dates.  Yes it can!

The following is my quick solution to his question.

answer = isDateBetween("12/21/2015", "12/15/2015", "12/30/2015")
if answer then print "yes" else print "no"


answer = isDateBetween("11/21/2015", "12/15/2015", "12/30/2015")
if answer then print "yes" else print "no"


function isDateBetween(aDate$, firstDate$, lastDate$)
    aDays = date$(aDate$)
    firstDays = date$(firstDate$)
    lastDays = date$(lastDate$)
    isDateBetween = firstDays < aDays and aDays < lastDays
end function


Enjoy!

Thursday, January 14, 2016

Tiny BASIC part 5 - Adding color to PSET

It's much more fun to draw in color than with only a black pen, so let's add a third parameter to the PSET statement for color, for example:
PSET x, y, "color"
To do this we need to add some code to our  case "pset"  block:
 CASE "pset"
  IF GWINOPEN = 0 THEN
    E$ = "PSET error - Graphic window is not open"
    GOTO [Ready]
  END IF
  GOSUB [GetExpression]
  IF E$<>"" THEN [Ready]
  PSETX = N
  GOSUB [GetChar]
  IF C$ <> "," THEN
   E$= "Comma expected after x parameter"
   GOTO [Ready]
  END IF
  C = C + 1
  GOSUB [GetExpression]
  IF E$<>"" THEN [Ready]
  PSETY = N
  PSETCOLOR$ = "black"
  GOSUB [GetChar]
  IF C$ = "," THEN
   C = C + 1
   GOSUB [GetStringLiteral]
   IF E$ <> "" THEN [Ready]
   PSETCOLOR$ = B$
  END IF
  #GWIN "color "; PSETCOLOR$

  #GWIN "down ; set "; PSETX; " "; PSETY
  GOTO [FinishStatement]

The way this works is that it will look for a comma and a string expression after it sets PSETX and PSETYIf there is no comma it will skip over the part that gets a color parameter, so the default color will be black in that case.

The string will be the name of a valid Liberty BASIC color, for example red, blue, green, black, etc.    The following code is stolen from the routine that parses for the PRINT statement.  In a later post we will incorporate this into the PRINT code by calling it as a subroutine so that we won't have the same code twice.
[GetStringLiteral]
  GOSUB [SkipSpace]
  GOSUB [GetChar]
  IF C$=G$ THEN
   B$=""
[NextStringMember]
   C = C + 1 : C$=MID$(A$,C,1)
   IF C$="" THEN
    E$="Unterminated string"
    RETURN
   ELSE
   IF C$<>G$ THEN
    B$=B$+C$
    GOTO [NextStringMember]
   END IF
  END IF
  C = C + 1 : C$=MID$(A$,C,1)
  IF C$=G$ THEN
   B$=B$+C$
   GOTO [NextStringMember]
  END IF
 END IF
 RETURN

So we call the [GetStringLiteral] subroutine and check E$ for an error.  If there is none then we set PSETCOLOR$ to the value of B$ as shown here.

   GOSUB [GetStringLiteral]
   IF E$ <> "" THEN [Ready]
   PSETCOLOR$ = B$
  END IF
  #GWIN "color "; PSETCOLOR$


Then we add a drawing command like so to set the color:

  #GWIN "color "; PSETCOLOR$

Here is a sample that uses PSET with color!


5 graphicwin
10 pset x, y, "red"
20 pset x + 10, y, "blue"
30 pset x + 20, y, "green"
40 x = x + 1
50 y = y + 2
60 if x < 100 then goto 10

And here is a screenshot!



Wednesday, January 13, 2016

Liberty BASIC file type association

Liberty BASIC doesn't map the BAS file extension to itself when it is installed.  This is because I didn't want to be so presumptuous as to steal a common file type away from another version of BASIC that might be installed.

So, because there is a bug in Windows which makes it really hard to remap a file extension and there have been discussions about this in the Liberty BASIC forum at conforums.com one of our member Chris Iverson (thanks Chris!) has contributed some Liberty BASIC code to solve this problem.

  Click to see thread in the forum

Enjoy!

Sunday, January 10, 2016

Lunar lander revisited

Liberty BASIC comes with a nice introduction to video games called lander.bas.
  • It's a good example of a timer driven game.
  • It uses sprites which are actually generated on the fly using turtle graphics.
  • It also shows proper technique of structured programming.
So I have some ideas that I am thinking about implementing to update it and make it an even more complete video game example:
  • Add sound effects including ambient sounds, rocket motor noise and crash explosion.
  • Animate the rocket motor so you can see rocket exhaust coming out.
  • Add some flying space junk sprites that you need to avoid while trying to land.
  • Add a colored starfield in the background.
Looking forward to this!  Check back for updates!

Friday, January 08, 2016

Tiny BASIC part 4 - Adding a PSET statement


Okay, now we are ready to add a PSET statement for drawing pixels.  Today we will simply add the ability to draw a single black pixel at a time.  Next time we will add color!

Here is the code to accomplish this.  This is just another SELECT CASE block to add after the one we added for the GRAPHICWIN command.
 CASE "pset"
  IF GWINOPEN = 0 THEN
    E$ = "PSET error - Graphic window is not open"
    GOTO [Ready]
  END IF
  GOSUB [GetExpression]
  IF E$<>"" THEN [Ready]
  PSETX = N
  GOSUB [GetChar]
  IF C$ <> "," THEN
   E$= "Comma expected after x parameter"
   GOTO [Ready]
  END IF
  C = C + 1
  GOSUB [GetExpression]
  IF E$<>"" THEN [Ready]
  PSETY = N
  #GWIN "down ; set "; PSETX; " "; PSETY
  GOTO [FinishStatement]
This was a bit tricky to write because there isn't really any documentation with the original Tiny BASIC source code, but by looking at the code for the other statements I think I figured it out correctly.  It does seem to work.

The syntax for the new statement is PSET x, y

Let me explain what it does.
  • First check to see if the graphic window is open.  If it isn't then set E$ to be an error string.  Then GOTO [Ready].  This will display the error.
  • Then get the next expression using GOSUB [GetExpression].  This unleashes the expression parser which is easily the largest and most complex part of the Tiny BASIC source code.  Then it checks for an error using IF E$<>"".  If E$ does contain an error, then GOTO [Ready].
  • Okay so got this far, so set PSETX to be what was in N, which is the result of the call to [GetExpression].
  • Now get the next character, which we expect to be a comma to separate the x and y values.  If the next character is not a comma, set E$ to be an error and GOTO [Ready].
  • Now advance C one character by adding 1 to it.  We do this because we found the expected comma, and now we need to skip over that so that we can get the next expression for our y value.
  • Get the next expression using GOSUB [GetExpression].  Test E$ for an error and GOTO [Ready] if there is one.
  • Get the value of N and put it into the variable PSETY.
  • Finally, draw the pixel in the graphics window!
Here is the sample Tiny BASIC program that uses the PSET statement:

10 graphicwin
20 pset x, y
30 x = x + 1
40 y = y + 2
50 if x < 200 then goto 20

And here is the output of the program!

Thursday, January 07, 2016

Run BASIC Revisited - The easiest web development system on Earth

I got an inquiry yesterday about Run BASIC asking about what is special about it.  The essential concept is that it is an all-in-one BASIC web application server.  You can use it to create a dynamic web site, or to host in-house applications for your business or school, or use it to control your home.  The possibilities are pretty much endless.

Here is a link to a white paper about Run BASIC.
    http://www.libertybasic.com/RunBASICBreakthrough.pdf

Here is a link to the Run BASIC community forum.
    http://runbasic.proboards.com/

Wednesday, January 06, 2016

Tiny BASIC part 3 - Adding GRAPHICWIN statement


We are going to add GRAPHICWIN and PSET statements to Tiny BASIC.

Let's start with the really easy one.  We will add a  case "graphicwin"  block to the end of the select case blocks that we examined in the last post.  Here is what it looks like.  The new code is in bold red.
 CASE "let"
  GOSUB [GetLabel]
  IF E$<>"" THEN [Ready]
 CASE "graphicwin"
  IF GWINOPEN = 1 THEN
    PRINT "Graphics window is already open."
  ELSE
    GWINOPEN = 1
    OPEN "Graphics" FOR graphics AS #GWIN
  END IF
  GOTO [FinishStatement]

 END SELECT
So when you run Tiny BASIC and type the command graphicwin and press Enter it will open a small graphics window.  When it does this it also sets the GWINOPEN flag to 1 to that we can check it if the program tries to open another graphics window.  Only one graphics window will be allowed at a time.

The other thing that we want to do it close the graphics window and reset GWINOPEN to 0 if the program is started using the RUN statement.

So, we need to add the following code into the  case "run"  block.  Here is how that code should look.  The new code is in bold red.

 CASE "run"
  IF GWINOPEN = 1 THEN
    CLOSE #GWIN
    GWINOPEN = 0
  END IF

  FOR I=27 TO 52 : A(I)=0 : NEXT I
  L=27 : C=1
  GOTO [FinishStatement2]
So now the program will always start off in a clean state each time it is run!

In our next post we will figure out how to add a PSET statement so we can draw some graphics!