I’m planning to create a package and I wonder if it’s better do do this with LaTeX3 syntax or the „old“ LaTeX2e, which is the one I know.
Will LaTeX2e packages be supported by LaTeX3?
I can give more information if it is necessary …
|
I’m planning to create a package and I wonder if it’s better do do this with LaTeX3 syntax or the „old“ LaTeX2e, which is the one I know. Will LaTeX2e packages be supported by LaTeX3? I can give more information if it is necessary … |
||||
|
|
|
At present, LaTeX3 is in development with some parts completed (a programming layer), other at least with some code written (for example the galley) and other more at the 'ideas' stage. This means that depending on what you are writing, you will need a varying amount of LaTeX2e code to write a 'LaTeX3 in LaTeX2e' package. For example, if most of what you want to do is essentially programmatic then you will need very little LaTeX2e code, but if you want to address some design area then you will need a lot more. Almost certainly, it will not be possible to take a LaTeX2e package and simply load it directly with LaTeX3, even if the code is written entirely LaTeX3 syntax. At the very least it's to be expected that the lead-off
line will need to be altered. Using the There's also the question of whether what you are aiming to write will overlap with what is wanted for the LaTeX3 kernel. That should not stop you writing code now in LaTeX3 syntax, but might mean that for a LaTeX3 kernel the functionality is covered 'natively'. Here, I'd hope that if people write stuff which is useful then there is possibility of simply taking the code, adapting it, and adding it to the kernel. (While there is a 'LaTeX3 Team', the reality is that we will have to act as 'drivers' for a larger effort, and so contributions from other people are going to be very welcome.) Again, depending on the exact area you are aiming at then there may be more or less overlap. Taking the example of my own So overall I'd see using LaTeX3 syntax in packages now as partly about taking advantage of the structures it provides, and partly about having code available for the future. I would not, however, see it as a way to have everything '100% ready to go' for a stand-alone kernel. |
|||||||
|
|
Only an opinion not an answer: Stability: The LaTeX3 extensions to LaTeX2e are still experimental. This means that details of the command interface may still change. LaTeX2e itself will not change a lot in future. The interface has been frozen almost. Only internal macros may be changed. So providing a package using LaTeX3 extensions may need more and faster support, e.g., testing against each new release of LaTeX and modify the package to work with the newest changes. Every new release of the LaTeX3 extensions may break your package. So if you are short in time or simply want to release a package but don't want to spend time in support, don't use LaTeX3 extensions. Features: The LaTeX3 extensions to LaTeX2e provides a lot of features and very nice implementation interfaces. After reading all the documentation you will like it. And if you would like to use any of the advantages of the LaTeX3 extensions to LaTeX2e you should design your package using them! |
|||||||
|