kicad-developers team mailing list archive

[Andy Peters <devel@xxxxxxxxx>]

> On Nov 1, 2018, at 11:23 AM, Jeff Young <jeff@xxxxxxxxx> wrote:
> 
> My daughter just had to upgrade from 10.11 to 10.12 because something else stopped working.  So we won’t be alone. ;)
> 
> (In any event, I encouraged her to keep more up-to-date.  There’s really not much down-side on the Mac, although I’ll admit I usually don’t go all the way to the latest.)

Another Mac user data point. Every Mac user I know waits a couple of weeks after a major release before installing it. Usually it’s to see if a particular piece of hardware is supported or not. Otherwise, everyone who can upgrade will.

My old Core2Duo iMac is still on 10.11 because it cannot be upgraded, and honestly that machine is too slow to do -anything- much less run Kicad.



> 
>> On 1 Nov 2018, at 16:53, Seth Hillbrand <seth@xxxxxxxxxxxxx> wrote:
>> 
>> Hi All-
>> 
>> I think we should cease support for 10.11 at 5.0.x. MacOS upgrades are free and generally support Apple hardware back to 2012 [1].  While there may be some users who don't want to upgrade their operating system but do want to upgrade KiCad, that seems unnecessarily obstinate.
>> 
>> -Seth
>> 
>> [1] https://www.apple.com/macos/how-to-upgrade/#hardware-requirements
>> 
>> 
>> Am 2018-11-01 10:28, schrieb Adam Wolf:
>>> It looks like globally, about 15%.  I do not have any insight into how
>>> our users different from the global average.
>>> However, I did not hear a single complaint when we moved the bar from
>>> 10.9 to 10.11, following the rules of "current, plus the two previous
>>> releases" that Apple appears to use.
>>> There are three main reasons why I ask:
>>> 1) 10.14 was released in the middle of the 5 cycle.
>>> 2) there is always more work to be done and not having to dig into
>>> this means more time for the other things, like getting Python 3
>>> working in KiCad on macOS... But I also don't want to ruin a bunch of
>>> user's days either.
>>> 3) Now that it is easy for Mac devs to build and generate packages,
>>> there's more incentive to keep the build process simple and sane.  The
>>> macOS 4.x releases, I'm not joking, were probably only buildable on
>>> one computer on this planet, due to all the weird hand-crafted
>>> dependencies.  I am not saying that looking into or fixing this issue
>>> will immediately cause the Mac builds to be unreproducible and full of
>>> cruft, but it could end up that way and it's important to consider.
>>> Mainly, I want someone else to make this hard choice for me :p
>>> It might be an OK time to switch nightlies over, and see what sort of
>>> pushback we get?  I'm not sure.
>>> Adam
>>> On Thu, Nov 1, 2018, 8:03 AM Wayne Stambaugh <stambaughw@xxxxxxxxx
>>> wrote:
>>>> Hey Adam,
>>>> Do you have any idea how much of an impact that this will have on
>>>> our
>>>> macos user base?  If it's minimal, than I would say go ahead any
>>>> move
>>>> builds to 10.12.  If it's something like half of the users, then
>>>> that is
>>>> problematic.  I wouldn't want to have that big of hit to our user
>>>> base.
>>>> Anyone else have any thoughts on this.  I'm not a mac user so I
>>>> don't
>>>> have a good feel as to whether or not this is going to be an issue.
>>>> Cheers,
>>>> Wayne
>>>> On 10/31/2018 4:40 PM, Adam Wolf wrote:
>>>>> Last night our automated builds that build on 10.11 both failed
>>>> because
>>>>> the latest version of our dependencies no longer builds on 10.11.
>>>>> How much time should we put into this, rather than move the builds
>>>> to
>>>>> 10.12 and leave the 10.11 users where they're at?
>>>>> Apple does not announce end of life, but if they follow what
>>>> they've
>>>>> done in the past, there will be no more security updates for 10.11
>>>> now
>>>>> that 10.14 is released.
>>>>> Adam
>>>>> ______________