| 1 |
On Tue, 29 Nov 2016 19:13:19 -0500 |
| 2 |
Mike Gilbert <floppym@g.o> wrote: |
| 3 |
|
| 4 |
> On Tue, Nov 29, 2016 at 6:57 PM, Brian Dolbec <dolsen@g.o> |
| 5 |
> wrote: |
| 6 |
> > |
| 7 |
> > While working on the last 2 version bumps to the twisted package, I |
| 8 |
> > kept getting an error in which the *egg-info/SOURCES.txt file |
| 9 |
> > absolute paths for all the files found in that very same |
| 10 |
> > directory. They are required to be relative paths only. This |
| 11 |
> > problem currently only affects some pkgs depending on the |
| 12 |
> > setuptools pkg for install. The source of the change is that |
| 13 |
> > setuptools patches the cpython findall() with one that returns |
| 14 |
> > relative paths for anything in/below the passed in base directory. |
| 15 |
> > Everything else it returns the absolute path. |
| 16 |
> > |
| 17 |
> > Why we should apply this: |
| 18 |
> > |
| 19 |
> > Python upstream has merged the setuptools findall() code and |
| 20 |
> > will be included in the next releases of python, 2.7.13, 3.4.6, |
| 21 |
> > 3.5.4 (I think). So, there is potential for many more python pkgs |
| 22 |
> > to be affected by this that are not requiring setuptools for |
| 23 |
> > install. Without the ebb-base setting, the egg-info/* files are not |
| 24 |
> > included in SOURCES.txt. The install proceeds without error. |
| 25 |
> > |
| 26 |
> > NOTE: This first affects the python_compile_all phase, long |
| 27 |
> > before it even tries to run the tests (if enabled). |
| 28 |
> > |
| 29 |
> > |
| 30 |
> > Some history: |
| 31 |
> > |
| 32 |
> > The egg-base settings were originally added for parallel |
| 33 |
> > testing/install purposes. They are not used due to problems in |
| 34 |
> > parallelizing testing, etc.. (something along those lines, I |
| 35 |
> > don't want to search thru the logs, and mails to see what |
| 36 |
> > Michal's original words were about it.) |
| 37 |
> > |
| 38 |
> > I am working on twisted-16.5.0 and 16.6.0 releases (tests are still |
| 39 |
> > failing), but 16.6.0 has many improvements to the testing code, so |
| 40 |
> > has a lot fewer errors to fix. Once these are fixed and the eclass |
| 41 |
> > changes are merged, I will be able to add it to the tree. |
| 42 |
> > |
| 43 |
> > See attached patch. NOTE: actual commit message will be |
| 44 |
> > different/updated correctly. This was just an interim patch for |
| 45 |
> > testing and initial review. |
| 46 |
> > |
| 47 |
> |
| 48 |
> Seems ok to me. Hopefully it doesn't break something in a subtle, hard |
| 49 |
> to detect way. ;-) |
| 50 |
> |
| 51 |
|
| 52 |
I agree, but with 1600+ dev-python/*.* ebuilds, it would take one hell |
| 53 |
of a tinderbox run to check for definite breakage. Subtle runtime |
| 54 |
breakage is quite another. |
| 55 |
|
| 56 |
-- |
| 57 |
Brian Dolbec <dolsen> |
Archives