<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">It might not be too difficult to
implement a switch (in user profile, or per-notebook, or whatever)
that turns autosaving on and off. Then those who like autosave
have it, and those who don't like it, don't have it.<br>
<br>
Of course, the ideal solution would to somehow integrate with a
version control system, so that you can have autosave but still
roll back to a previous version...<br>
Regards,<br>
Jon Wilson<br>
<br>
On 11/06/2012 05:39 PM, Carl Smith wrote:<br>
</div>
<blockquote
cite="mid:CAP-uhDfzxL7qLmaTgWZKm7O92C8BZELWymmDdmLGFLyXMxX8cg@mail.gmail.com"
type="cite">
<meta http-equiv="Content-Type" content="text/html;
charset=ISO-8859-1">
<p>Just for the record: Ctrl-S saves on Chromium on Lubuntu.</p>
<p>Google Docs autosaves with almost every change and it works
really badly for me. Everyone has there own take on these UX
elements. I vote for Trust the Developer - if they want stuff
saved, they should say so.</p>
<div class="gmail_quote">On Nov 6, 2012 11:24 PM, "Tom" <<a
moz-do-not-send="true" href="mailto:tmbdev@gmail.com">tmbdev@gmail.com</a>>
wrote:<br type="attribution">
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
"You might like <span>autosave</span>, but I personally often
want to revert to my last <span>saving</span> point.
Then <span>autosave</span> is really annoying for me, I am
not the only one."<br>
<br>
I don't care about autosave per se. What I care about is that
a tool that I spend hours in every day and recommend to others
is robust against failures, like all the other tools I use
regularly: I should be able to pull the power plug on my
computer and not lose more than a minute or so of work even if
I never saved anything.<br>
<br>
"If you came up with a good patch, we'll be happy to see you
open a Pull Request on github that improve situation."<br>
<br>
It is impossible to come up with a "good patch" under the
constraints you list. The question is whether those
constraints are more important to the iPython developers than
preventing data loss.<br>
<br>
Tom<br>
<div class="gmail_extra"><br>
<br>
<div class="gmail_quote">On Wed, Oct 31, 2012 at 12:27 PM,
Matthias BUSSONNIER <span dir="ltr"><<a
moz-do-not-send="true"
href="mailto:bussonniermatthias@gmail.com"
target="_blank">bussonniermatthias@gmail.com</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
Le 31 oct. 2012 à 11:53, Tom a écrit :<br>
<div><br>
> I just lost an hour of work again because Chrome
and iPython notebooks conspire against me.<br>
</div>
Sorry about that,<br>
<div><br>
><br>
> iPython notebooks take forever to complete when
there is a lot of output. This is really a problem by
itself. Saving iPython notebooks requires navigating
with the mouse to click on a tiny button somewhere at
the top; this is distracting and interrupts work.
There is no keyboard shortcut.<br>
<br>
</div>
Ctrl+s should save in some browser/os<br>
Ctrl+m, s should always do.<br>
<br>
It is in the doc, and on the keyboard shortcut help
(C-m,h)<br>
If it does not work, bug report.<br>
<div><br>
><br>
> Chrome compounds the problem by popping up a
dialog box every few seconds asking whether I want to
kill the page (and hence lose all my data) or
continue.<br>
<br>
</div>
Why is that ? It shouldn't. Which version of chrome do
you have. Which version of IPython ?<br>
Is it the "this page/script have stop responding message
?" I never had it with IPython notebook.<br>
<br>
You can open a bug report and we'll try to figure out
why this happen, this should make your workflow a lot
easier.<br>
<div><br>
> Chrome also lacks any option to interrupt
JavaScript without losing the entire page. (I think
Firefox has other problems, which is why I don't use
it.)<br>
<br>
><br>
> Asking me to confirm throwing away hours of work
that are on the screen and in memory and should be
trivially savable really adds insult to injury.<br>
> It's a conspiracy of multiple usability problems
in iPython and Chrome, any of which should be easy to
fix for the developers.<br>
<br>
</div>
You might like autosave, but I personally often want to
revert to my last saving point.<br>
Then autosave is really annoying for me, I am not the
only one.<br>
<br>
A persistant notebook without having to explicitly save
is not that easy. Hacking something that does the job
most of the time is not that hard,<br>
but doing it the right way, not so much.<br>
<br>
* Consider fact that some people do not use the
notebook on localhost and hence saving might be **very**
bandwidth consuming.<br>
* Same notebook can be opened several times
constantly overwriting each other if autosave is on.<br>
* Oops I deleted this cell... I just have
rollback to previous save version.... Too late....( no
we can't Ctrl-Z cell deletion yet)<br>
* Current API does not allow to save with
another name, people rely on our API, we can't change
without deprecation warning first.<br>
* None of us are real js developers, all of us
work on our little free time freely on IPython.<br>
* we would like to differentiate
persistence/saving/version controlling<br>
<br>
<br>
Still,<br>
You can have a look at Min's extension to autosave.<br>
<a moz-do-not-send="true"
href="https://github.com/minrk/ipython_extensions"
target="_blank">https://github.com/minrk/ipython_extensions</a><br>
<br>
Using custom js in your profile on 0.13+ it shouldn't be
too hard to enable autosaving by default.<br>
Ask if you need help, but with above extension and docs,
it shouldn't be too hard to figure out.<br>
<br>
If you came up with a good patch, we'll be happy to see
you open a Pull Request on github that improve
situation.<br>
You can also help by reviewing/testing the current
opened PR so that we'll get more time to work on new
stuff like autosaving.<br>
<br>
Hope we'll solve most of your issues.<br>
<br>
Cheers,<br>
<span><font color="#888888">--<br>
Matthias<br>
<br>
<br>
><br>
> Tom<br>
</font></span>
<div>
<div>>
_______________________________________________<br>
> IPython-User mailing list<br>
> <a moz-do-not-send="true"
href="mailto:IPython-User@scipy.org"
target="_blank">IPython-User@scipy.org</a><br>
> <a moz-do-not-send="true"
href="http://mail.scipy.org/mailman/listinfo/ipython-user"
target="_blank">http://mail.scipy.org/mailman/listinfo/ipython-user</a><br>
_______________________________________________<br>
IPython-User mailing list<br>
<a moz-do-not-send="true"
href="mailto:IPython-User@scipy.org"
target="_blank">IPython-User@scipy.org</a><br>
<a moz-do-not-send="true"
href="http://mail.scipy.org/mailman/listinfo/ipython-user"
target="_blank">http://mail.scipy.org/mailman/listinfo/ipython-user</a><br>
</div>
</div>
</blockquote>
</div>
<br>
</div>
<br>
_______________________________________________<br>
IPython-User mailing list<br>
<a moz-do-not-send="true" href="mailto:IPython-User@scipy.org">IPython-User@scipy.org</a><br>
<a moz-do-not-send="true"
href="http://mail.scipy.org/mailman/listinfo/ipython-user"
target="_blank">http://mail.scipy.org/mailman/listinfo/ipython-user</a><br>
<br>
</blockquote>
</div>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
<pre wrap="">_______________________________________________
IPython-User mailing list
<a class="moz-txt-link-abbreviated" href="mailto:IPython-User@scipy.org">IPython-User@scipy.org</a>
<a class="moz-txt-link-freetext" href="http://mail.scipy.org/mailman/listinfo/ipython-user">http://mail.scipy.org/mailman/listinfo/ipython-user</a>
</pre>
</blockquote>
<br>
</body>
</html>