Hi Steve,
ipython should not close unless
- there is a hard crash involving the underlying python interpreter
- one runs a code that calls sys.exit or something like this.
Best,
Sylvain
On Monday, April 29, 2013 11:52:51 AM UTC-4, Steve wrote:
>
> Can you confirm the ipython behavior? Does the interpreter normally stay
> open forever thus allowing additional runs to happen in the same
> interpreter?
>
> On Monday, April 29, 2013 10:42:56 AM UTC-5, Carlos Córdoba wrote:
>>
>> Thanks Sylvain for the detailed explanation. I'll fix the problem you
>> report ASAP.
>>
>> Cheers,
>> Carlos
>>
>> El 29/04/13 10:32, Sylvain Corlay escribió:
>>
>> I completely agree with the changes of Carlos and the additions of Pierre
>> regarding this behavior. I am trying to convince people to switch from
>> environment like matlab and I think that this change is both minor,
>> harmless and helps beginners.
>>
>> However, the problem I mentioned earlier is a bug to me. It is a
>> problem of focus:
>>
>> If you have both the console widget and the ipython console widget
>> opened, with the cursor/focus in the python widget and you click on ipython
>> console tab, the pure python console keeps the focus and further hits on
>> the run button will be executed in the console. When the one does the
>> reverse (put the cursor in the ipython console and click on a normal python
>> console tab), the focus moves as expected to the python console.
>>
>> I hope I was clear. I think that this is probably a one line fix. I can
>> try to look at it.
>>
>> Cheers,
>>
>> Sylvain
>>
>> On Sunday, April 28, 2013 6:19:17 PM UTC-4, Sylvain Corlay wrote:
>>>
>>> Hi Pierre,
>>> I think that the current error message is a bit misleading in the case
>>> where there is a shell opened but it does not have the focus. For example,
>>> if there is an ipython console opened but the standard console widget is
>>> selected.
>>> Something in this spirit of
>>> "No Python shell is currently selected to run stuff.py, please select
>>> or open a new python interpreter."
>>> woud maybe be more clear.
>>> Best,
>>> Sylvain
>>>
>>> On Sunday, April 28, 2013 11:23:51 AM UTC-4, Pierre Raybaut wrote:
>>>>
>>>> Ok, I understand your point of view and I think that it's probably
>>>> best to keep it that way for beginners.
>>>>
>>>> This odd behavior I found out ("Re-run" a script in current
>>>> interpreter fails silently if there is no running interpreter) is actually
>>>> a bug which is not produced by your recent change: this recent change of
>>>> behavior is only making it happen more easily. So, no worry, really.
>>>>
>>>> I see two actions that would need to be taken care of to close this
>>>> issue:
>>>> 1. Warn the user (with a warning pop-up dialog box) if the current "Run
>>>> settings" is set to "Run in current interpreter" *and* if there is no
>>>> interpreter currently running. This will avoid the "Run" command to fail
>>>> silently.
>>>> 2. Show the "Run settings" the first time a script is ever run, so that
>>>> every one will be fully conscious of the new default behavior (run in
>>>> current interpreter) and the new "show Run Settings at each run" check
>>>> box.
>>>> Beginners will directly click on the "Run" button without changing
>>>> anything
>>>> and the dialog won't show again so that they won't be troubled by this
>>>> mysterious choice again.
>>>>
>>>> I've implemented point 1 here:
>>>>
>>>> https://code.google.com/p/spyderlib/source/detail?r=c1e17dc89477f27cd251fb59340a067eaa45b9f8
>>>> and point 2 there:
>>>>
>>>> https://code.google.com/p/spyderlib/source/detail?r=00e148391c5b454b8c144fd1553b0a2155be8e12
>>>>
>>>> With those minor changes above, I'm ok with this new behavior!
>>>>
>>>> -Pierre
>>>>
>>>>
>>>> 2013/4/28 Carlos Córdoba <[email protected]>
>>>>
>>>>> Hi Pierre,
>>>>>
>>>>> Good to hear from you again! I gave a lot of thought to this change,
>>>>> and I didn't want to do it so close to the final release, but I was busy
>>>>> fixing other complex bugs.
>>>>>
>>>>> From the source code point of view, things are almost exactly as
>>>>> before: I just changed the dialog's default to "current interpreter" and
>>>>> added a checkbox to let the user decide between the old and the new
>>>>> behavior, so the bug you mention was most probably there before.
>>>>>
>>>>> Experienced or old time users can mark the mentioned checkbox and
>>>>> things for them will be as they always have been. But newbies (as Uwe
>>>>> mentions) will have a much easier and pleasant time with Spyder without
>>>>> being confronted with a complex set of options from the start.
>>>>>
>>>>> I hope you understand my motivation: my aim is to lead Spyder to a
>>>>> wider audience, i.e. people who is learning or giving their first steps
>>>>> with Python, but maintaining all the configurability and adaptability it
>>>>> currently has.
>>>>>
>>>>> I'll work hard to solve any bugs that show up (starting with yours :-)
>>>>> and time will tell if I took the right decision or not.
>>>>>
>>>>> Cheers,
>>>>> Carlos
>>>>>
>>>>> El 28/04/13 07:58, Pierre Raybaut escribió:
>>>>>
>>>>> Carlos,
>>>>>
>>>>> I'm quite busy these days and I've just played around with latest
>>>>> Spyder revision, just to fix Issue 1363 (for which you asked for my
>>>>> help).
>>>>> Doing so, I've been confronted to this new behavior of the 'Run
>>>>> configuration' dialog which has been renamed to 'Run settings' (why not).
>>>>> First, I find it very risky to introduce a change of behavior of this
>>>>> magnitude at this stage of 2.2 release process (Release Candidate).
>>>>> Second,
>>>>> I was confronted in less than 5 minutes to a bug (an unexpected behavior
>>>>> actually) related to this change: when executing a program which kills
>>>>> the
>>>>> Python process (hard crash or a simple call to sys.exit), Spyder seems to
>>>>> be unresponsive and unable to re-run it as the current interpreter has
>>>>> been
>>>>> terminated... So, that's how I've discovered this new behavior, a quite
>>>>> unpleasant experience.
>>>>> I really think that such changes should have been introduced at an
>>>>> early stage of development, not just before releasing the final 2.2.
>>>>>
>>>>> Cheers,
>>>>> Pierre
>>>>>
>>>>>
>>>>> 2013/4/27 Carlos Córdoba <[email protected]>
>>>>>
>>>>>> I made the change so that after pressing F5 on new files, they always
>>>>>> be evaluated in the current interpreter (Python or IPython). I just
>>>>>> checked
>>>>>> that it's working as I designed it (on Windows and Linux), i.e. If you
>>>>>> press F5 again, then the file will be ran again in the selected console
>>>>>> with the "runfile" function.
>>>>>>
>>>>>> I did it because I saw (in my courses and workshops) that people gets
>>>>>> easily confused with the "Run dialog" and don't know what option to
>>>>>> select.
>>>>>> Besides, now that we have a very good IPython integration, I expect most
>>>>>> people will take advantage of it and won't need the "Execute in a new
>>>>>> dedicated python interpreter" option.
>>>>>>
>>>>>> However, if you want to get back the old behavior, you can mark the
>>>>>> checkbox at the end of the dialog that reads:
>>>>>>
>>>>>> "Always open this dialog on a first file run".
>>>>>>
>>>>>> Cheers,
>>>>>> Carlos
>>>>>>
>>>>>> El 26/04/13 17:04, Steve escribió:
>>>>>>
>>>>>> I saw some commits in the change log recently related to the Run
>>>>>>> Settings. One of the changes leads to unexpected behavior.
>>>>>>>
>>>>>>> It appears the default interpreter option changed. I just rolled
>>>>>>> back to an old commit to confirm. The radio button for "Execute in a
>>>>>>> new
>>>>>>> dedicated python interpreter" used to be selected by default. The new
>>>>>>> default is "Execute in current Python or IPython interpreter" is
>>>>>>> checked.
>>>>>>> Because of this new default setting after the initial run (F5)
>>>>>>> additional
>>>>>>> presses of F5 do nothing. I finally figured this out by realizing it
>>>>>>> was
>>>>>>> new files and it must have to do with the run config for new files
>>>>>>> versus
>>>>>>> files I had previously debugged. I think don't think new behavior is
>>>>>>> optimal.
>>>>>>>
>>>>>>> -Steve
>>>>>>> --
>>>>>>> You received this message because you are subscribed to the Google
>>>>>>> Groups "spyder" group.
>>>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>>>> send an email to [email protected].
>>>>>>> To post to this group, send email to [email protected].
>>>>>>> Visit this group at http://groups.google.com/group/spyderlib?hl=en.
>>>>>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> --
>>>>>> You received this message because you are subscribed to the Google
>>>>>> Groups "spyder" group.
>>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>>> send an email to [email protected].
>>>>>> To post to this group, send email to [email protected].
>>>>>> Visit this group at http://groups.google.com/group/spyderlib?hl=en.
>>>>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>>>>
>>>>>>
>>>>>>
>>>>> --
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "spyder" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>> an email to [email protected].
>>>>> To post to this group, send email to [email protected].
>>>>> Visit this group at http://groups.google.com/group/spyderlib?hl=en.
>>>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "spyder" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>> an email to [email protected].
>>>>> To post to this group, send email to [email protected].
>>>>> Visit this group at http://groups.google.com/group/spyderlib?hl=en.
>>>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>>>
>>>>>
>>>>>
>>>>
>>>> --
>> You received this message because you are subscribed to the Google Groups
>> "spyder" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected].
>> To post to this group, send email to [email protected].
>> Visit this group at http://groups.google.com/group/spyderlib?hl=en.
>> For more options, visit https://groups.google.com/groups/opt_out.
>>
>>
>>
>>
>>
--
You received this message because you are subscribed to the Google Groups
"spyder" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/spyderlib?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
