Re:Mbeddr – engineering the future of embedded software
by errordeveloper on Saturday September 13, 2014 @09:47AM
(#8312360)
Because nobody would use those chips for new projects. Legacy chips need legacy tools and run legacy software.
Re:Mbeddr – engineering the future of embedded software
by mackwic on Saturday September 13, 2014 @11:30AM
(#8312632)
I don't agree. Popular but legacy chips can have updated tools where new but conservative chips can have odd software packaging.
Moreover, embedded development is conservative by nature and I know a lot of teams which prefers to run projects on known hardware when possible, even if we have to suffer of the tooling.
Also, don't forget the licencing. We already have to downgrade our targeted platform because we couldn't afford to pay the whole tooling where OSS was only compatible with the previous version.
I am slightly concerned by the DSL thing, because I don't want each embedded projects to have specific dialects like we can see on LISP projects.
But, on the average, gains are high: modules ! tests ! Inline contracts and model checking ! Very nice state machines ! I like that !
Also, why C99 ? There's hardware where only C89 compilers are available...