<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><br><div><div>Le 1 oct. 2014 à 09:39, Martin Teichmann a écrit :</div><br class="Apple-interchange-newline"><blockquote type="cite"><div>Hi Thomas,<br><br><blockquote type="cite">Thanks for looking into this. We would be interested in having a better way<br></blockquote><blockquote type="cite">to integrate event loops with getting input in the terminal. It's probably<br></blockquote><blockquote type="cite">not practical to push that straight into core Python, and even if it was, it<br></blockquote><blockquote type="cite">would be some time before we could scrap our inputhook functionality in<br></blockquote><blockquote type="cite">favour of a new feature in Python 3.5.<br></blockquote><br>To me it's a bit weird to say that we shouldn't push that straight into core<br>Python - IPython is working around this broken functionality since more<br>than a decade, so how long do we have to wait until something is not<br>considered "straight" anymore? Apparently IPython is not the only one<br>having this problem, as you mentioned others have also already written<br>a library to circumvent this problem in the standard library.<br></div></blockquote><div><br></div><div><br></div><div>I think that the point of Thomas is that our code need to be cleaned up. </div><div>We are currently preparing to split IPython into different repo so trying to at that </div><div>into IPython **now** is probably not a good time. </div><div><br></div><div>What I can suggest (other can chime in and say water they think is a good idea)</div><div>is to write an IPep[1] or an issue on github that target IPython 4.0 or later. </div><div>It will allow us to organize and decide if/when we want that. </div><div><br></div><div><br></div><blockquote type="cite"><div>As for compatibility: I actually think that the way IPython is handling<br>the problem - mostly hiding all the ugly details behind a nice API -<br>would make it possible to have the new and old version nicely in<br>parallel, especially as the new version has already mostly been written,<br>for the case if IPython is not running in the terminal.<br></div></blockquote><div><br></div><div><div>The other issue is that we still support Python 2.7, thus anything that targets </div><div>3.x only is dangerous and subject to controversy. Though the discussion </div><div>to drop 2.7 support for some functionality do arise more and more often.</div><div>This might be a case where maybe dropping 2.7 might help, </div><div>or where such a feature can push toward a drop of 2. support. It's </div><div>not my area of expertise so I don't know. </div><div><br></div><div><br></div><div>From the discussion on Python bug tracker with Guido, and reading in between lines, </div><div>I also would say that having such a feature in IPython might help to push this into Python</div><div>itself later. And having async console looks cool. </div></div><br><blockquote type="cite"><div><br>Greetings<br><br>Martin<br>_______________________________________________<br>IPython-dev mailing list<br><a href="mailto:IPython-dev@scipy.org">IPython-dev@scipy.org</a><br>http://mail.scipy.org/mailman/listinfo/ipython-dev<br></div></blockquote></div><br><div><br></div><div>[1] <a href="https://github.com/ipython/ipython/wiki/IPEPs:-IPython-Enhancement-Proposals">https://github.com/ipython/ipython/wiki/IPEPs:-IPython-Enhancement-Proposals</a>, don't worry much more informal than Python one. </div></body></html>