RUN --mount needs a magic comment to opt into experimental syntax:
# syntax=docker/dockerfile:experimental
https://github.com/moby/buildkit/blob/master/frontend/dockerfile/docs/experimental.md
but AFAIK, was designed for this purpose, among others.
Thanks,
On Tue, May 7, 2019, 11:33 AM Daniel
Also you must set an environment variable for docker build to be able to
use --mount
On Tue, May 7, 2019, 12:28 Daniel Holth wrote:
> That one is multistage, great feature, but I think cache mounts would work
> fine with any number of stages. The mount option is a pretty new feature.
>
> On
That one is multistage, great feature, but I think cache mounts would work
fine with any number of stages. The mount option is a pretty new feature.
On Tue, May 7, 2019, 12:24 Wes Turner wrote:
>
>
> On Tuesday, May 7, 2019, Daniel Holth wrote:
>
>> Have you tried using buildkit and the RUN
On Tuesday, May 7, 2019, Daniel Holth wrote:
> Have you tried using buildkit and the RUN --mount option? I've done extra
> stuff here (downloading rpms in a second image first) but I think you could
> just use the cache option.
>
> It would also be easy to use a second image and COPY in old
Have you tried using buildkit and the RUN --mount option? I've done extra
stuff here (downloading rpms in a second image first) but I think you could
just use the cache option.
It would also be easy to use a second image and COPY in old docker.
You can use a local PyPI mirror, e.g. devpi, and point your docker builds
at that, basically tricking docker by going through the (local) network
stack instead of the filesystem.
On Tue, May 7, 2019 at 8:12 AM Wes Turner wrote:
> What is the best way to build docker images without constantly
>
For the list: I chatted with Steve, Donald, and Cooper in person in the
sprints room. We're going to chat more tomorrow about a way forward for
"how Warehouse APIs should be built", and we will share a proposal with
this list. We'll definitely be looking at the prior art and proposals when
we do
What is the best way to build docker images without constantly
re-downloading packages from PyPI (to use ~O(1) bandwidth instead of O(n)
for every build)
(AFAIK, nobody has any issue with the amount of bandwidth PyPI uses)
Thus far, Docker doesn't want to support a build-time -v option (that
On Tue, May 7, 2019 at 10:01 AM Wes Turner wrote:
> What do namespaces offer over forking, diffing, reviewing the latest
> commits, and installing from your GH fork URL commit hash?
IIUC, one of the primary objectives for namespaces is to enable a user to
store state like ("We've reviewed this
On 07May2019 1048, Philip James wrote:
Hi there! I took a look at the API docs that I could find
(https://warehouse.readthedocs.io/api-reference/) and I couldn't see if
it was possible to programmatically edit the contributors who have
access to a project on PyPI.
My use case is for the
Hi there! I took a look at the API docs that I could find (
https://warehouse.readthedocs.io/api-reference/) and I couldn't see if it
was possible to programmatically edit the contributors who have access to a
project on PyPI.
My use case is for the BeeWare project being able to dynamically add
What do namespaces offer over forking, diffing, reviewing the latest
commits, and installing from your GH fork URL commit hash?
When I try to install 'westurner/pip' and 'pip' is already installed, what
should it do?
Should pypa/setuptools_scm include the namespace in the version tag?
If my
Good evening distutils-sig-
TL;DR: today during the Python Packaging Summit/Sprint I broached the topic
of introducing namespaces on PyPi with several of the luminaries in the PyPA
ecosystem. After much discussion I think we concluded that there's general
support for the idea, but quite a number
13 matches
Mail list logo