This is one of those cases where I saw a problem, envisioned a solution, but could never make the time to execute upon it. This happens a lot, but in this case, I’m pretty happy to see my vision realized even if I had nothing to do with it!
The concept was a MSIX Fixup Module file that would describe what needs to be done to convert a specific legacy setup to MSIX in an effort to accelerate MSIX adoption by helping the community share the solutions to conversion issues they face in a very actionable way.
I reached out to other influencers in the packaging space, intentionally ignoring the potential hurdles that might come in collaborating with competitors. I positioned the problem and solution as follows:
The adoption of MSIX has been slow. There may be many reasons, but applicable here is application support and complexity. We believe there is an opportunity to help MSIX adoption by introducing a standard crowdsourced file format for applying application-specific fixups.
Today one must package an application as MSIX, inject the PSF trace fixup, run the application to identify runtime errors, analyze the logs, determine if a fixup can correct the issue, and then configure and inject that fixup into the package.
We believe it would help to introduce the concept of a file that drastically simplifies the above. The fixup is necessary and its configuration will be shared using a single application version-specific file via the online community. MSIX repackaging tools would optionally consume this community data to apply the necessary fixup for an application with ease.
We need a name for this, so as a placeholder, I will refer to this shared community file as an MSIX Fixup Module (MFM) following the “merge module” concept with which our audience may be familiar.
For this to be successful, it must be seen as product agnostic, so we would look to Tim Mangan or possibly Microsoft (alongside their Windows Package Manager GitHub community) to host the community content.
In May of 2020, knowledgeable people from Flexera, Remo3, Liquit, Advanced Installer, and of course Tmurgent, attended the first of what would be only two meetings. Intentions were good, but like with so many things, we all had competing priorities that made the necessary investment of time a challenge.
So, I was very pleased to see the concept of MSIX Accelerators announced this week. As described by Microsoft, “An accelerator provides an efficient way to convert legacy applications to MSIX format. It consists of important information regarding the package (application), the operating system on which conversion happens, as well as the steps required to fix the package for proper functioning of converted MSIX.”
Time will tell if it caches on, but a community repo has been established here. Without a more impressive library of accelerators, the effort to support them in third-party packaging tools will be difficult to justify, but I’ll be watching closely.