No, I just want them to use the same syntax … on type level attributes seem to be implicitly treated as string constants, however on field level you need to add the additional „““. My proposal is to treat them the same way … Treating them as expressions (Like on field level) is 2 chars more per attribute, but gives us the power to use expressions … could Imagine something like this in the future:
[type AmsTCPPacket byteOrder='context.legacyDevcice ? „LITTLE_ENDIAN“ :“BIG_ENDIAN" ' Does that help? Chris Von: Sebastian Rühl <[email protected]> Datum: Mittwoch, 17. September 2025 um 15:56 An: [email protected] <[email protected]> Betreff: Re: [DISCUSS] Streamline the usage of attributes in mspec Not sure if I understand correctly what that means. Do you want to remove the byteOrder attribute on type level? - Sebastian On 2025/09/17 13:40:54 Christofer Dutz wrote: > Hi all, > > I just stumbled over another inconsistency in our mspecs. > > When using attributes on type-level, we do it this way: > > [type AmsTCPPacket byteOrder='LITTLE_ENDIAN' > > When using them on field-level, we do it this way: > > [simple vstring 'stringLength * 8 * 2' value encoding='"UTF-16LE"' ] > > The way we’re doing it on field level is potentially a lot more powerful. > Especially considering that we’re also planning on defining some context > types in the future. This way we could make these refer to variables in the > context and not just contain constants. > > Considering most of my other DISCUSS threads received almost no responses, > I’ll do it this way: If I don’t hear any objections, I’ll do the change in > the next few days. This however doesn’t mean that not responding, even if you > agree is good. It’s my fallback and comes close to implementing a protocol > based on timeouts … if you agree, please let me know because as soon as I see > general consensus I can proceed quicker. > > Chris > >
