Update environment keeping qt=4

[torr...@gmail.com]

 Hi,

how does one update an environment which has been set up to use Qt version 4, such that 'everything' in that environment gets updated, except for Qt4 and any package which depends on it?
Would 'pinning' Qt be the proper way? Or is there any other method?

Xaver

[Ray Donnelly]

Packages as old as Qt4 are not compatible with *any* recent packages. For example you cannot use Python 3.6 with Qt4. Why is it that you want Qt4 instead of Qt5? There's very little difference between them in terms of their APIs.

--
You received this message because you are subscribed to the Google Groups "conda - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an email to conda+un...@continuum.io.
To post to this group, send email to co...@continuum.io.
Visit this group at https://groups.google.com/a/continuum.io/group/conda/.
To view this discussion on the web visit https://groups.google.com/a/continuum.io/d/msgid/conda/71b619e0-e0d5-40bc-a58d-f64d4e7e0d56%40continuum.io.
For more options, visit https://groups.google.com/a/continuum.io/d/optout.

[torr...@gmail.com]

The reason for using Qt4 is simply that there are packages out there which i.m.h.o. need Qt4 (and I need those packages). You are absolutely right, that I have to use Python 3.5 with Qt4 - not that I'm particularly happy about it. The package I am referring to specifically is mayavi. See eg. https://www.scivision.co/mayavi-qt5-create-conda-env-qt4/  and many related posts. So while in this context I'm not quite sure what you mean by ' *any* ', in case this should imply that you know how mayavi would be running with Qt5 and Python 3.6, I would be glad to know your answer?

If you don't have an answer to the latter question, then, maybe you can still help me regarding the original question of my post, namely, how, after setting up an environment in with Python 3.5 and Qt4 are key requirements, I can still do an eventual 'conda update' of that environment without losing compatibility with Python 3.5 and Qt4?

[Raymond Osborn]

Didn’t you answer your own question with your first email?

https://conda.io/docs/faq.html#pinning-packages

Or am I missing something?

Ray

P.S. I agree that you cannot assume that all packages should just use Qt4, just as you still can’t assume all packages are compatible with Python 3.

On Aug 1, 2017, at 4:52 PM, torr...@gmail.com wrote:

The reason for using Qt4 is simply that there are packages out there which i.m.h.o. need Qt4 (and I need those packages). You are absolutely right, that I have to use Python 3.5 with Qt4 - not that I'm particularly happy about it. The package I am referring to specifically is mayavi. See eg. https://www.scivision.co/mayavi-qt5-create-conda-env-qt4/  and many related posts. So while in this context I'm not quite sure what you mean by ' *any* ', in case this should imply that you know how mayavi would be running with Qt5 and Python 3.6, I would be glad to know your answer?

If you don't have an answer to the latter question, then, maybe you can still help me regarding the original question of my post, namely, how, after setting up an environment in with Python 3.5 and Qt4 are key requirements, I can still do an eventual 'conda update' of that environment without losing compatibility with Python 3.5 and Qt4?

--
You received this message because you are subscribed to the Google Groups "conda - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an email to conda+un...@continuum.io.
To post to this group, send email to co...@continuum.io.
Visit this group at https://groups.google.com/a/continuum.io/group/conda/.

To view this discussion on the web visit https://groups.google.com/a/continuum.io/d/msgid/conda/9ac48a5a-8d2b-4b4c-ac64-be700ebbc62d%40continuum.io.

[torr...@gmail.com]

... am I missing something?

Yes.

It is exactly this link you are citing, which is unclear to me and because of which I also mention pinning - with a '?'.
What I do understand from this part of the doc is, that you can constrain the versions of certain packages by 'pinning'.
Fine.
What I don't understand, and also from this part of the docs, is, if all of the rest of that environment in which you have pinned certain packages will respect dependencies on exactly those pinned packages once you subject the environment to eg. a 'conda update --all'.

I'd very much appreciate your answer on this. Or would you have another link to take a peek at?

Xaver

PS.: if I'd be on another package manager like RPM, I'd assume that dependencies are taken care of fine after putting parts of a package list 'on hold'. But, as you realize from my post, I'm not familiar with conda and I don't like 'assuming' in that case.

[Raymond Osborn]

I can’t claim to be a conda expert, but until someone who is chimes in, my guess is that conda will respect the pinned dependencies when doing the upgrades. Before I upgraded my own package to be PyQt5-compatible, every time I did a ‘conda install pyqt=4’, it downgraded a host of other packages. If you are worried, it should be fairly easy to test by creating a new environment.

Ray

--
You received this message because you are subscribed to the Google Groups "conda - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an email to conda+un...@continuum.io.
To post to this group, send email to co...@continuum.io.
Visit this group at https://groups.google.com/a/continuum.io/group/conda/.

To view this discussion on the web visit https://groups.google.com/a/continuum.io/d/msgid/conda/dcc5aece-411f-4d08-9038-bb834533aacc%40continuum.io.

[Thomas Caswell]

In somewhat orthogonal, but relevant good news, there was work at the scipy2017 sprints to get mayavi working with qt5.

Tom

To view this discussion on the web visit https://groups.google.com/a/continuum.io/d/msgid/conda/44395F95-592C-409B-A3F0-911CD6A65219%40mac.com.