New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
PyHamcrest 1.10.0 is not py2-compatible. It should set python_requires>=3 to avoid breaking users who are still on Py2. #131
Comments
Investigating.
…On Tue, 7 Jan 2020, 5:08 pm tvalentyn, ***@***.***> wrote:
import hamcrest no longer works on Python 2 with pyhamcrest==1.10.0. To
avoid installation of pyhamcrest==1.10.0 on Python 2, it should exclude
python2 interpreters in python_requires stanza in setup.py.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#131?email_source=notifications&email_token=AABXTODRKLJE6QVSLQZB2S3Q4SZHVA5CNFSM4KD4JMGKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IERNR2A>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AABXTOBO2J5OZW7MVQ6QJEDQ4SZHVANCNFSM4KD4JMGA>
.
|
For the benefit of those searching, this causes Python 2.7 tests to fail with a
You can work around this by blacklisting this version in your
This is under the assumption that the next release will add the appropriate metadata. See #132. |
@brunns we need to yank 1.10 for now til this can be fixed. 1.10 is not supposed to drop Python 2 support... unless you have a fix in the wings. |
I don't have the rights to yank PyHamcrest releases on PyPI - can you promote me? I need to be a "Maintainer". |
There is a problem with keeping 1.10 Python 2 compatible. The main purpose of 1.10 in my mind is to let folks see warnings from deprecated methods, making it easier to move to a version 2.0 where they will be removed. But some of the deprications were marked after Python 2 support was dropped. Perhaps we just shouldn't bother with a 1.10 at all? |
I'm off to bed now, but if that's the thought... I've deleted the release.
Let me know what you figure out in the morning.
…On Wed, Jan 8, 2020 at 12:17 AM Simon Brunning ***@***.***> wrote:
I don't have the rights to yank PyHamcrest releases on PyPI - can you
promote me?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#131?email_source=notifications&email_token=AABMLPLHTUEZAQUWTKBX7DDQ4WDZ3A5CNFSM4KD4JMGKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEILRS2Y#issuecomment-571939179>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AABMLPJZTNHE7KE6B4H77X3Q4WDZ3ANCNFSM4KD4JMGA>
.
--
Chris R.
======
Not to be taken literally, internally, or seriously.
Twitter: http://twitter.com/offby1
|
Sweet dreams. :-)
…On Wed, 8 Jan 2020 at 09:14, Chris Rose ***@***.***> wrote:
I'm off to bed now, but if that's the thought... I've deleted the release.
Let me know what you figure out in the morning.
On Wed, Jan 8, 2020 at 12:17 AM Simon Brunning ***@***.***>
wrote:
> I don't have the rights to yank PyHamcrest releases on PyPI - can you
> promote me?
>
> —
> You are receiving this because you commented.
> Reply to this email directly, view it on GitHub
> <
#131?email_source=notifications&email_token=AABMLPLHTUEZAQUWTKBX7DDQ4WDZ3A5CNFSM4KD4JMGKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEILRS2Y#issuecomment-571939179
>,
> or unsubscribe
> <
https://github.com/notifications/unsubscribe-auth/AABMLPJZTNHE7KE6B4H77X3Q4WDZ3ANCNFSM4KD4JMGA
>
> .
>
--
Chris R.
======
Not to be taken literally, internally, or seriously.
Twitter: http://twitter.com/offby1
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#131?email_source=notifications&email_token=AABXTOHIZ2HGXKNEAPFUUSLQ4WKPNA5CNFSM4KD4JMGKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEILWMQQ#issuecomment-571958850>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AABXTOH7ZDVIBP2UVPTJL4LQ4WKPNANCNFSM4KD4JMGA>
.
|
this breaks builds for those using
in the future please don't unpublish releases as it breaks build determinism for anyone using a build system like |
I'm gonna give you the benefit of the doubt that your tone is only accidentally seeming like it's telling us what to do. In the future please consider how your choice of words reflects your desires so as not to cause this confusion again. I knew what removing the version would do, and made a conscious choice... just as I would in any other event like this. |
Perhaps you could tell us what we should have done under the circumstances?
Leaving the release up with incorrect metadata, which was my mistake, would
almost certainly have inconvenienced a lot more people than removing it
did.
…On Thu, 9 Jan 2020, 9:43 pm britcornwell23, ***@***.***> wrote:
I'm going to give you the benefit of the doubt that your tone is only
accidentally seeming condescending. In the future please consider how your
choice of words reflects your desires so as not to cause this confusion
again.
Pulling releases that others depend on is a bad look and a bad software
practice. Just fyi. 😊
—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
<#131?email_source=notifications&email_token=AABXTOAKNXMX4JQCH5GSKQTQ46K7RA5CNFSM4KD4JMGKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEIR4NKA#issuecomment-572769960>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AABXTOELMNJGFTWPADEEACDQ46K7RANCNFSM4KD4JMGA>
.
|
no reason to be defensive or rude- I made a completely reasonable request which you are of course free to ignore |
So, how should we move forward on this? I see three options: A - Cut a Python 2 compatible 1.10.1 release, cut from immediately the removal of Python 2 support, with any deprications back-ported. I'm sure there are options I've missed, so please let me know what they are. Of the 3, I lean towards option B if we can live with a non Python 2 compatible release in the 1.x series. I know there are objections to that, but I'm not sure I understand what they are. Option C would be my 2nd choice, though it would make upgrading to 2.0 more difficult, since some methods would disappear without warnings. Option A is a fair amount of work, and I'm not sure I see the need since presumably Python 2 users wouldn't really be looking for new features or changes now that Python 2 is EOL. Thoughts? |
If we can, I would prefer 2.0.0 can then immediately be rolled out as well. |
I've release a Python 2/3 compatible version 1.10.1. Fingers crossed. Let me know if there are any issues with it. |
Seems quieter this time. OK to close? |
👍🏻
On Thu, Jan 16, 2020 at 2:36 AM Simon Brunning ***@***.***> wrote:
Seems quieter this time. OM to close?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#131?email_source=notifications&email_token=AABMLPO5UCSERXD45XDQWP3Q6A2CRA5CNFSM4KD4JMGKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJDS3PI#issuecomment-575090109>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AABMLPK6FWKHK37EPPR2V3TQ6A2CRANCNFSM4KD4JMGA>
.
--
Chris R.
======
Not to be taken literally, internally, or seriously.
Twitter: http://twitter.com/offby1
|
import hamcrest
no longer works on Python 2 with pyhamcrest==1.10.0.To avoid installation of pyhamcrest==1.10.0 and newer versions on Python 2, it would be better to add
python_requires>=3.0
stanza in setup.py starting from 1.10.0. To fix this for 1.10.0, we would need to update already released 1.10.0 pypi artifacts.As of now, all python 2 users who install pyhamcrest without restricting the version to 1.9.0, will be broken.
A better user experience would be if Python 2 users get the last version of pyhamcrest that is supported on Python 2. This is convention is often followed by major python libraries, for example numpy.
The text was updated successfully, but these errors were encountered: