I consistently release a new version of ZPE each year. This year version 1.5 has been released.
In 2015 version 1.3 was released which brought the most changes to ZPE since it began in 2014. Version 1.3 added so much including the very basic compiler that version 1.2 started. Over 50 functions were added in this version. This version added if statements, for loops, while loops, functions, structures (although they didn't work properly until 1.4), RMM (Real Math Mode), plugins, libraries, lists, associative arrays, constants, variables, incrementing and decrementing, built-in converters (and for a while one example did even exist that converted ZPE code to Java), image functions and of course, my favourite feature of ZPE, anonymous/lambda functions.
In 2016 version 1.4 was released and it brought a huge shake-up of changes, but its main focus was consistency between internal functions and imported ones. It also brought structures that are similar to those used in other languages such as function(){ } instead of function() end function. As well as this it brought the LAMP interpreter, lazy evaluation, function chaining, the object type, a better hashing of variables and functions, the internal manual for information on built-in functions. Version 1.4 also brought support for many other new things, including octal and hexadecimal without the need for a function.
In 2017 version 1.5 was first released. This version currently adds one major new feature, Remote ZPE. Remote ZPE allows a system running ZPE to act as either a client to another ZPE installation or to act as a server that other ZPE Clients can utilise. The purpose of this is for smaller, much less capable machines such as the Raspberry Pi to be able to quickly take advantage of the processing power of a more capable machine. This was developed because I use ZPE remotely on my Raspberry Pi through SSH to do things, but what I'd really like is ZPE to work on my MacBook and send the result to my Pi (so it goes like this Mac connect to Pi through SSH which then uses ZPE Remote to do processing on Mac which returns the result to Pi which shows on my Mac, this way I can process scripts stored on my Pi and utilise them on my Pi).
The future of ZPE 1.5 however is very clear. 2017 will bring the following major features to ZPE:
- JSON parsing (already done and available in all versions of ZPE) (1.5.0)
- Remote ZPE: Client and Server (already done and available in all versions of ZPE) (1.5.1)
- Multi-threading: Currently this project is underway and it's easy to test with ZPE version 1.5.1 (currently beta users only) using the following:
$v = threaded function() { for ($i = 0; $i < 100000; $i++) print($i) end for }. It's easy to do and I'm sure it makes sense. - Interfaces for objects: allowing programmers to define interfaces
- GUI builder: a way to develop graphical interfaces in applications (not sure if I will include this by default in all installations and might just include it as a library)
- Conversion methods: New built-in methods to make conversion plugins easier to write
- Plugins will gain access to internal functions: A method of accessing built-in functions from plugin functions
- Typo: The new Typo typing system will finally come to version 1.5.x (finished in 1.6.7)
- Includes: Inclusion techniques to include non compiled code in the non compiled version of the code. Much like the C and PHP include methods. (1.5.1)
- Include imported functions: The import function will be changed to be a part of the compiler and will change the way it is run.
- Ordered associative arrays: associative arrays currently store just one key and do not follow an order, this will allow associative arrays to be ordered (1.6.8)
- ZPE JSON: Improve JSON parsing and drop the use of GSON (the JSON functions will all work identically to how they work with GSON, it's just the wrapper functions on the top that will change).
- Improved compiler: the compiler currently flows within the compiler-interpreter cycle whereby the compiler generates the input for the interpreter. This is fine for simply using the -i tool, but when running applications it means that one more step still needs to be run - transform the AST to something meaningful. With the latest update, I aim to change this and put the compiler in a different place. The basics of the compiler will remain the same, but one more step will be added when compiling to a compiled program. This will improve the speed of the ZPE Runtime when applied to a compiled application (it will have no effect on the interpreter). (1.7.1)
- Power to the objects: objects, and, as a side-effect structures will be getting more powerful. It will soon be possible to declare a function within an object using lambda functions. As well as this objects will be getting more support in the LAMP parser. (1.6.3)
- Passing parameters by reference: just as it says. I aim to have this feature in version 1.5.3. This would mean functions like list_dequeue would no longer need to be reassigned to the variable. (1.7.1)
I am highlighting in green the features as they get added and highlighting in orange features that are partially added, so keep an eye on this post for updates.
There are several more but they will be put here once I have decided whether or not they are feasible or not.
This is a very tall order, and Typo and the conversion functions are currently at the bottom of the stack and although they have been suggested for this version and were originally planned for version 1.4, I still may need to delay them to version 1.6.
Also, with the release of version 1.5, support for any version 1.4 iterations of the software has now ended, I encourage you to upgrade to the latest version of ZPE.
