random re-ordering of text and cells

[William Stein]

Hi Notebook Devs,

Even in the new flask version of the notebook running locally on my
laptop, there is random re-ordering of cells and text sometimes. For
example, the screenshots shows what I see right now coming back to a
worksheet I started making this morning. Notice that the last text
cell should be before the compute cell with the last numpy example,
but it is after.

I'm using the version of the flask notebook for which "hg log" starts with

----------
changeset: 808:198a3553f34d
bookmark: master
tag: tip
user: Jason Grout <jason...@drake.edu>
date: Sat Jan 21 14:04:17 2012 -0600
summary: Fix a docstring that has newlines, so should be a raw docstring.
----------

I'm not at all saying that this should be a blocker for sage-5.0 or
including the flask notebook in sage, since this problem has been
around ever since tinymce was introduced (at least).

I just want to raise attention to the fact that this problem is not
actually fixed (yet!), contrary to what has been asserted in the past.

-- William

--
William Stein
Professor of Mathematics
University of Washington
http://wstein.org

[Jason Grout]

My current feelings on this are that:

1. It is still a known issue (we fixed some related things, but this
problem still exists).

2. The problem happens when you submit a text cell after a worksheet has
timed out.

3. A workaround is to increase the worksheet timeout value to longer
than you would spend editing a text cell.

Anyone else have thoughts on this?

I believe the next place to try a fix is:

1. diagnose the problem is (2) above.

2. If so, then use the same machinery that restarts a worksheet when a
code cell is submitted after a timeout to restart a worksheet when a
text cell is submitted.

Thanks,

Jason

[William Stein]

What does this mean? Worksheets on my laptop never time out by default.

> 3. A workaround is to increase the worksheet timeout value to longer than
> you would spend editing a text cell.

What do you mean by "worksheet timeout"?

> Anyone else have thoughts on this?
>
> I believe the next place to try a fix is:
>
> 1. diagnose the problem is (2) above.
>
> 2. If so, then use the same machinery that restarts a worksheet when a code
> cell is submitted after a timeout to restart a worksheet when a text cell is
> submitted.
>
> Thanks,
>
> Jason
>

--

[Jason Grout]

On 2/17/12 1:24 PM, William Stein wrote:
>
> What do you mean by "worksheet timeout"?

The setting in the admin notebook settings page labeled Idle timeout.
What is yours set at?

Thanks,

Jason

[William Stein]

It's "0".

[Jason Grout]

On 2/17/12 1:37 PM, William Stein wrote:
> On Fri, Feb 17, 2012 at 11:28 AM, Jason Grout
> <jason...@creativetrax.com> wrote:
>> On 2/17/12 1:24 PM, William Stein wrote:
>>>
>>>
>>> What do you mean by "worksheet timeout"?
>>
>>
>> The setting in the admin notebook settings page labeled Idle timeout. What
>> is yours set at?
>
> It's "0".

That indeed means no timeout, according to the code in
sagenb/notebook/notebook.py. Hmm...

Jason

[Dima Pasechnik]

On Saturday, 18 February 2012 03:03:30 UTC+8, William Stein wrote:

Hi Notebook Devs,

Even in the new flask version of the notebook running locally on my
laptop, there is random re-ordering of cells and text sometimes.  For
example, the screenshots shows what I see right now coming back to a
worksheet I started making this morning.  Notice that the last text
cell should be before the compute cell with the last numpy example,
but it is after.

I'm using the version of the flask notebook for which "hg log" starts with

----------
changeset:   808:198a3553f34d
bookmark:    master
tag:         tip
user:        Jason Grout <>

date:        Sat Jan 21 14:04:17 2012 -0600
summary:     Fix a docstring that has newlines, so should be a raw docstring.
----------

I'm not at all saying that this should be a blocker for sage-5.0 or
including the flask notebook in sage, since this problem has been
around ever since tinymce was introduced (at least).

I just want to raise attention to the fact that this problem is not
actually fixed (yet!), contrary to what has been asserted in the past.

Has anyone asserted this?

(or that the equally mysterious "textcell bug" (textcell becoming data cell) has been fixed?)

One more semi-bug, that might (or might not?) be related is reordering of the text when you paste between text windows in the notebook; the reordering is happening in a fairly regular fashion, so it can be helped on the spot.

Dima

[Jason Grout]

There has been at least once when we hoped we fixed this and asked
people to test it. Alas, while we did fix an issue with textcells
becoming data cells, we didn't fix the mysterious textcell rearranging bug.

Has anyone seen text cells become data cells recently? We fixed one
reproducible case of this last June.

> One more semi-bug, that might (or might not?) be related is reordering
> of the text when you paste between text windows in the notebook; the
> reordering is happening in a fairly regular fashion, so it can be helped
> on the spot.

Can you elaborate on exactly what is happening here?

Thanks,

Jason

[Dima Pasechnik]

we saw quite a big last fall, when we prepared worksheets for our Experimental Maths class.

We ran Rado's version of flask nb from August 2011 or so.

 

> One more semi-bug, that might (or might not?) be related is reordering
> of the text when you paste between text windows in the notebook; the
> reordering is happening in a fairly regular fashion, so it can be helped
> on the spot.

Can you elaborate on exactly what is happening here?

I think I reported this on the googlecode page of sagenb some time, but perhaps it was overlooked.

To reproduce, open two text cells in nb, and try to cut/paste between them. If the amount of text that is being pasted is big enough, say a screenful, it begins to wrap around, i.e the last lines pop up first, something like this.

Dima

 

Thanks,

Jason

[William Stein]

I saw that once in the last day or two, I think on my laptop when
preparing a lecture.

-- William

>
>
>
>> One more semi-bug, that might (or might not?) be related is reordering
>> of the text when you paste between text windows in the notebook; the
>> reordering is happening in a fairly regular fashion, so it can be helped
>> on the spot.
>
>
>
> Can you elaborate on exactly what is happening here?
>
> Thanks,
>
> Jason
>

--