Shall we drop i386 / 32-bit? If you think no, please test

dennis knorr dennis.knorr at gmx.net
Sat Jul 1 13:29:03 UTC 2017


There is an f)
f) 32-Bit-3rd-party-Applications depend on kubuntu 32bit and might be
important in some enterprise environment.

If you do this, from a release engineering standpoint you should
announce it an release cycle earlier and not break it in the current one
without telling users beforehand... Otherwise people will not have the
ability to switch if it breaks their usecase.

And if there's not the test coverage.. the people who will need your
platform WILL test it, if/when they want to migrate to the next platform.

jm2c,
Dennis

Am 01.07.2017 um 14:28 schrieb Ferry Toth:
> Op Fri, 30 Jun 2017 20:05:39 -0400, schreef Aaron Honeycutt:
> 
>> It's not a measure of users, we can't reason something if we don't get
>> enough testers to confirm the quality of the software.
> 
> I get that.
> 
> However, quoting Valori: "Nor do I have the time to personally
> do the testing for a build I do not and probably never will use."
> 
> This should raise the question: why build something for a certain 
> platform:
> a) because we can
> b) because we want to use it our self
> c) because there are testers available
> d) because there are a significant amount of existing users
> e) because there are a significant amount of potential new users
> 
> Speaking for my self as a developer I am driven often by a) and b), 
> however being a business man as well, I find d) and e) the only drivers 
> sustainable on the long run. In some exceptional cases a) leads to e), 
> those are the ones that make me happy to be an engineer. 
> 
> But c).. is not a goal, nor a driver. Not having enough testers is a 
> problem, and problems need to be solved.
> 
> To me, I would like to help testing. But I don't want to break my 
> production systems. Testing inside a virtual machine is not very useful 
> except maybe for installers, as then everybody tests on the same platform.
> 
> Maybe what we need is a test plan, and possibly test automation, like a 
> instrumented live iso. If that could select tests based on the platform 
> to get maximum test coverage without unnecessarily duplicating already 
> tested cases, I'm sure more people would participate.
> 
>> On Jun 30, 2017 7:17 PM, "Ferry Toth" <ftoth at exalondelft.nl> wrote:
>>
>>> Op Fri, 30 Jun 2017 06:55:42 -0400, schreef Clay Weber:
>>>
>>>> The lack of testers, the lack of bug reports, and the lack of
>>>> response here seem to indicate that i386 users in our community are a
>>>> rather tiny group. One that is likely shrinking even more.
>>>
>>> My mother uses i386. But she doesn't read this list. And I doubt she
>>> would be able to help testing or write bug reports. She is 76 years
>>> old.
>>>
>>> I don't thing lack of testers or bug reports is a good measure for lack
>>> of users. There must be a better way, like counting downloads of a
>>> specific package (i386 vs ams64).
>>>
>>> Don't forget kubuntu is likely to be installed on older machines as a
>>> replacement of WinXP and many are i386.
>>>
>>>> The short lifespans of non-LTS releases do not help, either. Perhaps
>>>> dropping support now may  be the best time to pull the plug.On Jun
>>>> 29, 2017 4:33 PM, Valorie Zimmerman <valorie.zimmerman at gmail.com>
>>>> wrote:
>>>>>
>>>>> Hello folks, Art Alpha 1 is released today: release notes:
>>>>> https://wiki.ubuntu.com/ArtfulAardvark/Alpha1/Kubuntu
>>>>>
>>>>> Please report test results here:
>>>>> http://iso.qa.ubuntu.com/qatracker/milestones/378/builds
>>>>>
>>>>> If we don't get complete test coverage on the i386 test cases, I
>>>>> think we should drop it now. the KCI doesn't build the images, we
>>>>> get no questions about 32-bit any more, and it is in general a pain
>>>>> to support.
>>>>>
>>>>> Is there any reason to keep it?
>>>>>
>>>>> Valorie
>>>>>
>>>>> --
>>>>> http://about.me/valoriez
>>>>>
>>>>> --
>>>>> kubuntu-users mailing list kubuntu-users at lists.ubuntu.com Modify
>>>>> settings or unsubscribe at:
>>>>> https://lists.ubuntu.com/mailman/listinfo/kubuntu-users
>>>
>>>
>>>
>>> --
>>> kubuntu-users mailing list kubuntu-users at lists.ubuntu.com Modify
>>> settings or unsubscribe at: https://lists.ubuntu.com/
>>> mailman/listinfo/kubuntu-users
>>>
>> <div dir="auto">It's not a measure of users, we can't reason
>> something if we don't get enough testers to confirm the quality of
>> the software.</div><div class="gmail_extra"><br><div
>> class="gmail_quote">On Jun 30, 2017 7:17 PM, "Ferry Toth"
>> <<a href="mailto:ftoth at exalondelft.nl">ftoth at exalondelft.nl</a>>
>> wrote:<br type="attribution"><blockquote class="gmail_quote"
>> style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Op
>> Fri, 30 Jun 2017 06:55:42 -0400, schreef Clay Weber:<br>
>> <br>
>> > The lack of testers, the lack of bug reports, and the lack of
>> response<br>
>> > here seem to indicate that i386 users in our community are a rather
>> tiny<br>
>> > group. One that is likely shrinking even more.<br>
>> <br>
>> My mother uses i386. But she doesn't read this list. And I doubt
>> she<br>
>> would be able to help testing or write bug reports. She is 76 years
>> old.<br>
>> <br>
>> I don't thing lack of testers or bug reports is a good measure for
>> lack<br>
>> of users. There must be a better way, like counting downloads of a<br>
>> specific package (i386 vs ams64).<br>
>> <br>
>> Don't forget kubuntu is likely to be installed on older machines as
>> a<br>
>> replacement of WinXP and many are i386.<br>
>> <br>
>> > The short lifespans of non-LTS releases do not help, either.
>> Perhaps<br>
>> > dropping support now may  be the best time to pull the plug.On Jun
>> 29,<br>
>> > 2017 4:33 PM, Valorie Zimmerman <<a
>> href="mailto:valorie.zimmerman at gmail.com">valorie.zimmerman at gmail.com</
> a>>
>> wrote:<br>
>> >><br>
>> >> Hello folks, Art Alpha 1 is released today: release notes:<br>
>> >> <a href="https://wiki.ubuntu.com/ArtfulAardvark/Alpha1/Kubuntu"
>> rel="noreferrer"
>> target="_blank">https://wiki.ubuntu.com/<wbr>ArtfulAardvark/Alpha1/
> Kubuntu</a><br>
>> >><br>
>> >> Please report test results here:<br>
>> >> <a
>> href="http://iso.qa.ubuntu.com/qatracker/milestones/378/builds"
>> rel="noreferrer"
>> target="_blank">http://iso.qa.ubuntu.com/<wbr>qatracker/milestones/378/
> <wbr>builds</a><br>
>> >><br>
>> >> If we don't get complete test coverage on the i386 test
>> cases, I think<br>
>> >> we should drop it now. the KCI doesn't build the images, we
>> get no<br>
>> >> questions about 32-bit any more, and it is in general a pain
>> to<br>
>> >> support.<br>
>> >><br>
>> >> Is there any reason to keep it?<br>
>> >><br>
>> >> Valorie<br>
>> >><br>
>> >> --<br>
>> >> <a href="http://about.me/valoriez" rel="noreferrer"
>> target="_blank">http://about.me/valoriez</a><br>
>> >><br>
>> >> --<br>
>> >> kubuntu-users mailing list <a
>> href="mailto:kubuntu-users at lists.ubuntu.com">kubuntu-
> users at lists.ubuntu.com</a>
>> Modify<br>
>> >> settings or unsubscribe at:<br>
>> >> <a
>> href="https://lists.ubuntu.com/mailman/listinfo/kubuntu-users"
>> rel="noreferrer"
>> target="_blank">https://lists.ubuntu.com/<wbr>mailman/listinfo/kubuntu-
> users</a><br>
>> <br>
>> <br>
>> <br>
>> --<br>
>> kubuntu-users mailing list<br>
>> <a
>> href="mailto:kubuntu-users at lists.ubuntu.com">kubuntu-
> users at lists.ubuntu.com</a><br>
>> Modify settings or unsubscribe at: <a
>> href="https://lists.ubuntu.com/mailman/listinfo/kubuntu-users"
>> rel="noreferrer"
>> target="_blank">https://lists.ubuntu.com/<wbr>mailman/listinfo/kubuntu-
> users</a><br>
>> </blockquote></div></div>
> 
> 
> 



More information about the kubuntu-devel mailing list