In last few weeks I've triaged a bunch of dumps where call stack was corrupted. And in general dumps are never trusted 100% (missing vals, register vals are unreliable and sometimes misleading, dump was partial, etc). Various means of crash reporting are very useful, as long as one always keep in mind that no such mechanism is 100% reliable.
DSP, Plug-in and Host development discussion.
- 339 posts since 4 Oct, 2002
- 98 posts since 6 Jan, 2003
Back on topic: Maybe there are other options. IANAL, and I don't know whether this was discussed before, but what if I write a plugin that is binary compatible with the VST2 SDK without using any of the SDK source files? I have seen vestige or FST, there are implementations in other programming languages (Delphi, D, Rust). Like in the past when the Samba team reimplemented the SMB protocol - or the US court ruled that Google's use of the Java API in the Android SDK was fair use (https://www.eff.org/cases/oracle-v-google)
Blue Cat Audio
- 4850 posts since 8 Sep, 2004 from Paris (France)
Regardless of any (good or bad) opinion about VST3, both as a host and plug-in developer: while most features that were initially missing can now be implemented in VST3, the fact is that when both formats are available, VST3 still has more issues than VST2 in 2021, unfortunately.
- 306 posts since 8 May, 2007
As has been pointed out in many places, including all-knowing Wikipedia, the situation is not nearly so simple as this. Certain types of software patents are explicitly allowed, for example any software that solves a "technical" problem in a way that is novel and non-obvious may be patentable in Europe. A lot of audio software would probably easily qualify as solving a technical problem, but not so easily as novel and non-obvious to persons skilled in the art - even though a lot of nonsense does get accepted by examiners.
Furthermore, one can always write a software patent in terms of a machine on which the software operates: Put something in register A and something else in register B. Add the two in a circuit and put the sum in register C. Etc. If necessary, one could go even farther and describe where the electrons go during operation of the machine. All so-called software patents I have ever read are written in a hardware-oriented fashion such that it’s nearly impossible to differentiate between the two types. This type of machine-oriented writing is done on purpose to avoid interpretation as describing (mathematical) algorithms.
- 235 posts since 9 Oct, 2013 from uk
To be clear (for my understanding), in a clean VST2 versus 3 battle, then fair enough.Blue Cat Audio wrote: ↑Tue Dec 07, 2021 11:59 amRegardless of any (good or bad) opinion about VST3, both as a host and plug-in developer: while most features that were initially missing can now be implemented in VST3, the fact is that when both formats are available, VST3 still has more issues than VST2 in 2021, unfortunately.
However, I've seen plenty written about the use by some of required 'creative programming' in order to get around VST2 SDK limitations. All I'm saying is, any issues seen in VST3 as a cause of it disallowing similar methods/functionality, shouldn't be a direct blame.? Otherwise yeah, have at it; Steinberg should be told.
Win10;i7; 16Gb; GTX750Ti; Cubase Pro v11.0.41; Wavelab Pro v10.0.70; StudioOne v5.4.1; UR44 audio/MIDI; HS80 and Equator D5 monitoring. Komplete 12; Stylus RMX