From mboxrd@z Thu Jan 1 00:00:00 1970 From: Luca Ceresoli Date: Thu, 3 Dec 2015 18:12:40 +0100 Subject: [Buildroot] [PATCH 08/51] core/legal-info: allow ignoring packages from the legal-info In-Reply-To: <87ziy3yrxm.fsf@dell.be.48ers.dk> References: <2462093e5e97e8e420b7119ac6ca6078e39f4621.1448289515.git.yann.morin.1998@free.fr> <20151123212649.794b7583@free-electrons.com> <87ziy3yrxm.fsf@dell.be.48ers.dk> Message-ID: <56607808.2020509@lucaceresoli.net> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Yann, Thomas, Peter, Peter Korsgaard wrote: >>>>>> "Thomas" == Thomas Petazzoni writes: > > > Peter, > > Can you give your opinion on the semantic/usage of the > > _REDISTRIBUTE variable. > > > I am personally not a big fan of the non-boolean behavior introduced by > > this patch for this variable. With this patch, _REDISTRIBUTE > > switches from a normal YES/NO boolean variable to a weird tristate > > variable YES/IGNORE/NO. > > > I have already stated my opinion that there should be two boolean > > variables instead: > > > * One which allows the package to indicate if the package should be > > mentioned in the legal-info or not. > > > * One which allows the package to indicate if the package "source" > > should be saved to the legal-info or not. Of course, this variable > > is ignored if the first one is set to "NO". > > > This is IMO a lot clearer than a single variable with the YES/IGNORE/NO > > values. > > > Peter ? > > I thought I had already replied to this, but perhaps that was only on > IRC? In general, I feel that the legal-info is purely an aid that can > serve as input to whatever license compliancy effort the user has to do > - E.G. include the various license texts in the user manual, provide > source offer and so on, so if the data contains a bit of extra > information that isn't such a big deal. Agreed, yet it's good to automate whatever we can _and_ is simple to implement. Since Yann's is only an 8-lines patch (+docs), I like it. > > But ok, if we want to do it I would atleast want it to be as simple and > easy to use as possible. I agree with Thomas that splitting the two > things we want to do into two seperate variables is probably the easiest > to understand. I proposed one three-valued variable, but I'm fine with the two boolean idea as well. Thus, Yann, can you re-cook it accordingly? -- Luca