Replies: 3 comments 5 replies
-
My long term plan is to integrate Book Lib Connect with AAX Audio Converter. Book Lib Connect is based on .Net 6. The new AAX Audio Converter 2.* then will also be based on .Net 6 (or 7). That integration will be quite a bit of work and I have postponed it, not being able to allocate the necessary time in recent months. For now, I will still add minor new features to AAX Audio Converter 1.*, but it will remain .Net Framework 4.8, as there are quite a few .Net Framework-specific dependencies. Only one or two of those I have ported to .Net Standard/Core for Book Lib Connect until now. But I plan to address this task for real in the upcoming months, after a holiday in April. With new AAX Audio Converter 2.* I also will abandon FFmpeg sub-processes and use many more features of Mbucari/AAXClean which already handles the decryption in Book Lib Connect. On my request Mbucari has implemented almost all of the other FFmpeg features I need, e.g. silence detection. And there are more challenges with in-process handling of audio manipulation as I need a new approach to my little gimmick of visual performance monitoring. So, in the long term, I don't see a problem for AAX Audio Converter if you switch to the .Net Core family (I wouldn't go for .Net 5, though, as it is already out of support). But you could keep a release branch in GIT for .Net Standard, to be able to apply patches, if necessary. That's what I already do with my AAX Audio Converter code. (I am a Subversion guy and keep my subversion repo locally, only release code gets uploaded to the Github repo.) There is a challenge for users, though, and you indicate it: Classic .Net Framework is regarded by Microsoft as a component of the Windows operating system and is updated with the Windows update mechanism automatically, without further ado. The .Net Core family of runtimes, however, is not a Windows component and only updated by Windows Update, if you explicitly add it. BTW, with FFmpeg, I always drop the meta data when copying/converting the audio files and rebuild it from scratch with your ATL.net. Does your lib have a feature to explicitly clear all meta data before rebuilding it? Mbucari's tools leave meta data untouched. |
Beta Was this translation helpful? Give feedback.
-
Thanks a bunch for the detailed answer! You're completely right about using .NET 6 rather than 5. I was reasoning "base requirement" with 5, but realistically, 6 would be the way to go.
Nice ! I'll have a look a couple months later then.
That's what I would do if I decided to migrate sooner, but I'd rather not to, as backporting important fixes on a legacy branch would require more and more effort as the most recent branch gets modernized and refactored. A 12-month support is the best I can do in that scenario.
You can use |
Beta Was this translation helpful? Give feedback.
-
FYI, I'm currently polishing the next major version of ATL, which drops support for .NET Framework 4.8 in favour of .NET 6.0, solves the "duplicate code for sync and async methods" dilemma and modernizes the structure of testing projects. The release should happen during August. |
Beta Was this translation helpful? Give feedback.
-
Hi audiamus 👋 It's my turn to post on your Github ;)
I'm considering switching my library to .NET 5 in the future to be able leverage C#8.0 features such as reading binary data from a
Stream
directly into aSpan
for a lighter memory footprint and better performance. I have no deadline so far.As of now, your app is the most starred project that is using ATL (according to Github at least), and is using .NET Framework v4.8.
Would it be difficult for you to switch the minimal requirements to .NET 5? Or do you consider your userbase not to be ready for this?
Beta Was this translation helpful? Give feedback.
All reactions