Anaconda dependency inconsistencies?

[the.g...@googlemail.com]

Hello,

I'm mirroring Anaconda, and quite often there are inconsistencies between the metadata and the available packages. Currently numpy-1.11.0-py34_2.tar.bz2 is not found (when I tried installing bokeh).

I checked http://repo.continuum.io/pkgs/free/win-64/index.html and indeed I find something like

 "numpy-1.11.0-py34_1.tar.bz2": {
      "build": "py34_1",
      "build_number": 1,
      "date": "2016-05-13",
      "depends": [
        "mkl 11.3.3",
        "python 3.4*"
      ],
      "license": "BSD",
      "md5": "b834d7abee6ebdd52f3e70abb864510b",
      "name": "numpy",
      "requires": [],
      "size": 2809080,
      "version": "1.11.0"
    },
    "numpy-1.11.0-py34_2.tar.bz2": {
      "build": "py34_2",
      "build_number": 2,
      "date": "2016-06-15",
      "depends": [
        "mkl 11.3.3",
        "python 3.4*"
      ],
      "license": "BSD",
      "md5": "002178f639952241a572a11113497ca4",
      "name": "numpy",
      "requires": [],
      "size": 2808406,
      "version": "1.11.0"
    },

in the metadata, but there is no package like that provided
numpy-1.10.4-py35_0.tar.bz2     2.7M     2016-01-25 12:30:23     2f9a31ede60454bff04c233662f2b615
numpy-1.10.4-py35_2.tar.bz2     2.7M     2016-05-13 12:55:45     16c26a132f5ce720e92331c78ecc2943
numpy-1.11.0-py27_0.tar.bz2     3.0M     2016-03-28 14:41:36     592cf5fe382762ebe94536058bb5baba
numpy-1.11.0-py27_1.tar.bz2     3.0M     2016-05-13 13:12:30     46c959725d13853891f69f3c6f3efd2b
numpy-1.11.0-py34_0.tar.bz2     2.7M     2016-03-28 14:48:38     71b58e605ef36be2f42627afbdacc2e9
numpy-1.11.0-py34_1.tar.bz2     2.7M     2016-05-13 13:19:37     b834d7abee6ebdd52f3e70abb864510b
numpy-1.11.0-py35_0.tar.bz2     2.8M     2016-03-28 14:51:22     779029b13a2f384de632be3c1baf7cc9
numpy-1.11.0-py35_1.tar.bz2     2.8M     2016-05-13 13:22:31     1d4107e387bda9144e06e82a19518afd

Is there anything I can do about that? A fix or a suggestion to keep the system consistent?

Cheers,
Anton

[Ilan Schnell]

Hello Anton,
the HTML index pages, such as http://repo.continuum.io/pkgs/free/win-64/index.html are only updated once a day.  However, https://repo.continuum.io/pkgs/free/win-64/repodata.json, that is the actual data which conda references, is updated instantly whenever conda packages are added to the repo.

Cheers  Ilan

--
Anaconda Community Support Group Brought to you by Continuum Analytics
---
You received this message because you are subscribed to the Google Groups "Anaconda - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an email to anaconda+u...@continuum.io.
To post to this group, send email to anac...@continuum.io.
Visit this group at https://groups.google.com/a/continuum.io/group/anaconda/.

[the.g...@googlemail.com]

Oh, thanks!

I probably didn't download all packages.

Is there a way I can install a package (bokeh) but let conda ignore the update of numpy to the "_2" version?

(I can be offline and don't have access to the online server.)

[Ilan Schnell]

Using "conda install --no-deps"