Thank you to all those who contributed to this thread. While many Go programs are written under open source licences, and many Go programmers contribute to open source in a professional or personal capability, it is now time to bring the discussion to a close as this thread has moved outside the scope of this mailing list.
Please feel free to continue this discussion in other forums. Thanks Dave On Friday, 18 May 2018 05:39:59 UTC+10, matthe...@gmail.com wrote: > > Thanks for responding Michael. > > "decorative item not to be used off-road, in uneven terrain, or relied >> upon as protection in case of vehicle roll." > > > The sticker I’ve been looking at says something like “modifying or > attaching anything to this ROPS will compromise the structure and may cause > injury or death” and I had a relative roll a tractor last year but is > luckily ok. > > I understand being conservative for the courts, but GCC and Go can make > programs for computers that can do safety applications reliably. People > maybe do with GCC derivatives. People do with Ada or C (and Go maybe is > less mistake-prone than C at the language level). > > Leveraging the work of an open or free project and contributing back > improvements found in testing seems like a good idea to me for liable > ventures and for the reliability of minimal liability uses that Google does > with Go and maybe GCC. I don’t think these compiler projects should take on > unnecessary liability, but one thing that can be done is to be sure there > aren’t any intentional mistakes, backdoors, or practical jokes as allowed > under the licenses. I'd like to guarantee that to liable ventures and it's > a free win for general purpose computing anyway. > > Matt > > On Thursday, May 17, 2018 at 12:18:30 PM UTC-5, Michael Jones wrote: >> >> perhaps some context will make this clearer. no reference is being made >> to any actual persons or events. >> >> 1. It is the observed habit of people (plaintiff's) bring suit in court >> when something goes wrong. >> Ex: airplane crash, dark spot on potato chip, food is too hot, etc. >> >> 2. Plaintiff's attorneys have developed the habit of including everyone >> possible in the "bag guy" list; this creates a kind of mutual fund where >> everyone can settle for $10,000 and that, multiplied by 100 defendants, is >> $1M. The attorney gets 1/3 of that, so attorneys drive nice cars and fly >> first class. >> Ex: car accident with fire.: sue the car company, the airbag company, the >> brake pad company, the glass company, the bumper designer, ...., the gas >> station, the gas transport truck, the oil refinery,...,the tire company, >> the person who designed the tire treads, etc. >> >> 3. After 60+ years of this, everyone who is wise to the situation sells >> retail products that specifically disclaim every possible dangerous use, >> and as many kinds of critical or life-safety misuse as can be foreseen, and >> wholesale products where every possible liability passes to the purchaser >> as a condition of sale. >> Ex: "You agree to not use the APIs for any activities where the use or >> failure of the APIs could lead to death, personal injury, or environmental >> damage (such as the operation of nuclear facilities, air traffic control, >> or life support systems)." >> >> 4. After 40+ years of open source having an observable footprint, open >> source entities, corporate contributors, and individuals have found that >> this style of "universal disclaimer" is an important defensive bulwark. >> Most people would not want to lose their house and savings as the result of >> contributing code to a matrix library that happens to be used in the >> science payload of a space project where the rocket engines explode on >> launch and the resulting fires and fumes cause problems for miles around >> resulting in a class-action suit for much of Florida. (crazy made-up >> example but perhaps it makes clear the idea of minimizing exposure to legal >> risks associated in no logical way with individual action.) >> >> This is the context in which various excuse-laden, >> suitability-disclaiming, crazy-seeming license agreements arise. >> >> As a real example, but with the name removed, I once saw in a parking lot >> a new pickup truck made by a well-known Japanese car manufacturer. It had a >> shiny chrome tubular framework rising up from the bed of the truck just >> behind the cab. The framework had lights attached. It have the truck a >> tough, off-road character. I said to my wife, "wow, that's quite the >> roll-bar for a little truck. Look how thick the tubes are." She said, there >> is a sticker on it what does it say. We looked, it read: "decorative item >> not to be used off-road, in uneven terrain, or relied upon as protection in >> case of vehicle roll." >> >> That is the real world of litigious people, 1/3 hungry attorneys, and >> juries that like to "do something" when there is a victim. >> >> On Thu, May 17, 2018 at 8:48 AM <matthe...@gmail.com> wrote: >> >>> I was thinking something like writing an undocumented “Happy New Year!” >>> to standard out at the start of the year. An obvious but undocumented ‘rm >>> -rf /‘ attempt was mentioned above. >>> >>> My first program was a practical joke. On the calculator command line I >>> said “press enter” then put the program call on the next line. The program >>> would scroll some text forever. The command display state was preserved >>> through being turned off, so somebody in the next class pressed enter then >>> had their calculator lock up and I got in trouble because the teacher had >>> to remove the batteries. I had an effect on many people because of the lost >>> class time. I explained that there was a key to interrupt any program. >>> >>> These university licenses allow newcomers to programming to make that >>> kind of social mistake and I think it’s right to not punish them for it. I >>> might not be a programmer if I had gotten detention for the calculator >>> program, and things like GCC might not exist without some wild thinking. I >>> don’t think this approach is right for industry, other serious ventures, >>> and especially not for safety focused applications though. >>> >>> It looks like the Intel corporate family thinks Intel, ARM, and Power >>> architecture processor implementations are trustworthy enough for safety >>> applications. There’s this OS called VxWorks said on the website to be >>> intended for safe IoT device applications: >>> https://www.windriver.com/products/vxworks/ >>> >>> It appears that WindRiver has worked with the GCC project for the >>> VxWorks platform: >>> https://www.windriver.com/products/product-notes/tornado2/gnu_relnote.pdf >>> >>> QNX has a C/C++ toolchain for ARM and x86: >>> http://blackberry.qnx.com/en/products/certified_os/safe-kernel >>> >>> Here’s a 2011 thread about Go and RTOS: >>> https://groups.google.com/forum/#!topic/golang-nuts/95BJqJvb7I0 >>> >>> There there’s a claim that the garbage collector makes Go unusable in >>> real-time operating systems, but I think there are cases where real-time is >>> less important than the OS being developed with reliability in mind. Maybe >>> Go could be very useful on RTOS platforms. >>> >>> My understanding is the intent for Go is to solve problems at Google. I >>> think involving varying outside uses of Go will help Google by making a >>> toolchain more robust than just Google applications will do, and I think >>> the design of the Go language is ideal for a next generation of general >>> purpose software like C was before. And I hope this thread adds value. >>> >>> “No features contrary to documentation” seems like a mistake since >>> documentation is usually not right. “No obfuscated features or obviously >>> wrong features” seems too vague and may invite incorrect claims. “This >>> software has no effects except for documented or obvious use”? Obviously a >>> lawyer would have to translate it to match case results and other lawyer >>> things, and those writing software under the license would have to be aware >>> of the implications. I plan to email a summary of this discussion to FSF >>> and OSI mailing lists. >>> >>> Thanks, >>> Matt >>> >>> On Wednesday, May 16, 2018 at 8:55:09 PM UTC-5, kortschak wrote: >>>> >>>> I hope so. I provide a package (github.com/kortschak/zalgo) that I >>>> cannot promise will not summon demons. It was written intentionally as >>>> a joke. I disclaim all liability should use of the package bring about >>>> meetings with demonic presences. >>>> >>>> On Wed, 2018-05-16 at 07:25 -0700, matthe...@gmail.com wrote: >>>> > I think practical jokes should be allowed under the GPL, BSD, and >>>> > similar >>>> > licenses. >>>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "golang-nuts" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to golang-nuts...@googlegroups.com. >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> >> -- >> Michael T. Jones >> michae...@gmail.com >> > -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
