Re: [dmidecode] Having dmidecode being machine friendly

[Neil Horman]

On Thu, Oct 04, 2018 at 02:16:00PM +0200, Erwan Velu wrote:
> Looks like this email wasn't really popular :)
> I don't know how to have a reliable parsing of dmi structures out of
> dmidecode which doesn't offer any guarantee on the output.
> WDYT about having a JSON output ?
I don't take any issue with that approach.  Do you have a patch to propose?
Neil

> Smartmontools is on the way to change on this point ;))
> Le mar. 18 sept. 2018 à 16:00, Erwan Velu <address@hidden> a écrit :
> >
> > Hi folks,
> >
> > That's almost an endless story but I'm back one more time with almost
> > the same idea/need.
> > I'm currently developing a tool for doing some consistency check for
> > clusters and reporting the hardware configuration is obviously one of
> > these tasks to perform.
> > I'm trying to compare nodes between themselves but also against a
> > "definition" of them.
> >
> > I'll be part of a GO tooling so I need to do my stuff in go.
> >
> > So what are the solutions for a third party tool to get __accurate__
> > dmi information from a system ?
> >
> > I found a couple of GO libraries which are all outdated or just try to
> > parse the string output of dmidecode. Both of them are not really
> > satisfying. I could also use lshw which have dmi support but much less
> > accurate than dmidecode, which is to my opinion, the most up-to-date
> > opensource tool to get DMI information.
> >
> > As dmidecode is where the vendor contributes but is also the most used
> > DMI tooling,  why not having at least a machine-readable format output
> > (like json) to ease the parsing output of dmidecode ? That would made
> > our (3rd party tooling) lives much easier.
> >
> > Thx !
> >
> > Erwan,
> 
> _______________________________________________
> https://lists.nongnu.org/mailman/listinfo/dmidecode-devel