Re: [boost] Executable files in Boost git repostories

Subject: Re: [boost] Executable files in Boost git repostories
From: Andrey Semashev (andrey.semashev_at_[hidden])
Date: 2018-12-04 15:54:01

On 12/4/18 6:41 PM, stefan via Boost wrote:
> On 2018-12-04 5:06 a.m., Andrey Semashev via Boost wrote:
> 
>> On 12/4/18 3:40 AM, Stefan Seefeld via Boost wrote:
>>>
>>> Why do we even have to discuss this here globally, rather than trusting
>>> project maintainers to know what they are doing ?
>>
>> Apparently, maintainers can't be trusted as this issue comes up on and 
>> on again.
> 
> Take it up with those maintainers, but don't attempt to overrule them. 
> That's precisely the wrong direction to take.
> 
>> Look at this as a sort of automated testing - you shouldn't be able to 
>> push bogus or straight dangerous stuff because you're affecting all 
>> git users.
> 
> This is exactly the kind of thinking that annoys me with Boost. You are 
> trying to solve the wrong problem ! We need more autonomy, not less !
> 
> Don't patronize maintainers by imposing artificial (and often misguided) 
> rules.
> 
> Instead, empower boost maintainers to take full responsibility of their 
> projects, then establish criteria for projects to be boost members, and 
> drop projects that don't meet these requirements.

I disagree. I really don't want to go into the debate about autonomy vs. 
moderation, I'll just say that from my point of view a certain level of 
moderation and uniformity is needed. If a library wants to do everything 
in its own way, not caring to be part of the Boost ecosystem, it might 
as well be maintained as a standalone project.

And this particular issue is not about infringing on anyone's rights, 
really. Unless you want the right to shoot yourself and everyone else in 
the foot. Marking files as executable, when they are not actual scripts, 
is never the right thing to do.