SageNB 0.7.x

[qed777]

The new spkg is at

• http://boxen.math.washington.edu/home/mpatel/trac/8051/sagenb-0.7.4.spkg

Merged in 0.7: #7249, #7962, #7969, #4217, #3083, #6182, #5263, #7631, #6353, #7207, #8000, #4450, #7848, #7963, #7752, #7996, #6475, #5675, #7435, #3844, #6368, #7434.

Merged in 0.7.1: #8103.

Merged in 0.7.2: #3083's "notruncate" patch.

Merged in 0.7.3: #7784.

Merged in 0.7.4: #8167, #8102, #8160.

Component: notebook

Author: Mitesh Patel

Reviewer: Robert Mařík, Minh Van Nguyen, Alex Leone

Merged: sage-4.3.2

Issue created by migration from https://trac.sagemath.org/ticket/8051

[qed777]

Author: Mitesh Patel

[qed777]

comment:2

If it's possible, I'd like to get into 0.7.1 as many of the now remaining "needs review" tickets as we can. I'm rebasing these now.

[qed777]

comment:3

Replying to @qed777:

If it's possible, I'd like to get into 0.7.1 as many of the now remaining "needs review" tickets as we can. I'm rebasing these now.

Here's a possible queue:

trac_7784-hgignore_update.patch
trac_5712-interrupt-notification.5.patch
trac_6069-missing_pub_ws.2.patch
trac_8038-email_plus_addressing_v2.patch
trac_7506-notebook_object-documentation.2.patch
trac_693-spawn_notebook.3.patch
trac_5177-delete-cell-dirs.3.patch

[qed777]

Blockedby: 7249

[qed777]

Changed blockedby from 7249 to none

[qed777]

comment:7

I just noticed that long 'eval' docstrings are truncated. I'll add a reviewer patch to #3083.

[robert-marik]

comment:9

All tickets got positive review and have been merged. So what should be reviewed in this ticket?

I installed the spkg, seems to work fine (but I did not test everything), is this enough to give positive review?

btw: the link from description "says" http://boxen.math.washington.edu/home/mpatel/trac/8051/sagenb-0.7.2.spkg but it points to http://boxen.math.washington.edu/home/mpatel/trac/8051/sagenb-0.7.1.spkg

[sagetrac-mvngu]

comment:10

Replying to @robert-marik:

All tickets got positive review and have been merged. So what should be reviewed in this ticket?

You need to make sure that you can successfully install the updated spkg.

I installed the spkg, seems to work fine (but I did not test everything), is this enough to give positive review?

I would say, all doctests must pass as well. In any case, if you can't run all doctests after installing the updated spkg, I can do that. A correct link to the updated spkg is

http://boxen.math.washington.edu/home/mpatel/trac/8051/sagenb-0.7.2.spkg

[sagetrac-mvngu]

comment:11

I don't understand why the spkg is not managed by Mercurial:

[mvngu@mod sagenb-0.7.2]$ hg st
abort: There is no Mercurial repository here (.hg not found)!

The file spkg-install should have its executable bits on:

[mvngu@mod sagenb-0.7.2]$ ls -g spkg-install
-rw-r--r-- 1 mvngu 348 2010-01-30 16:37 spkg-install

And SPKG.txt is very sketchy about update details.

[qed777]

comment:12

I suggest

• Checking that the package installs and the notebook runs.
• Checking the repo for unchecked-in changes, queued patches, etc.
• Checking that the claimed merged tickets appear in hg log.
• Running the doctests: sage -t -sagenb.

Ideally, you should run the SageNB Selenium tests, too. But they require special extra setup. I'll make simplifying that setup a separate ticket.

Thanks for pointing out the link error. I've updated it.

[qed777]

comment:13

The repository is in sagenb-0.7.2/src/sagenb. We auto-generate the package with sagenb-0.7.2/src/sagenb/spkg-dist.

I suggest that I make a separate ticket to update SPKG.txt.

See #7784 about

$ hg stat
? release_notes.txt
? setup.cfg

[qed777]

comment:14

Replying to @qed777:

I suggest that I make a separate ticket to update SPKG.txt.

Or I can do this here later today.

[qed777]

comment:15

I'll fix the spkg-install problem, too.

[qed777]

Reviewer: Robert Mařík, Minh Van Nguyen

[qed777]

comment:16

Please see #7784 for the changes. If/when that ticket gets a positive review, I'll create SageNB 0.7.3 and post it here.

[qed777]

comment:17

By the way, it seems that for the near future, I may be the only very active SageNB developer. I'd be very happy to be proved (proven?) wrong! There are many tasks to complete --- there are several cool new notebook features to implement. It's not possible for me to cover them all, and I'd like to avoid stalling ongoing development.

To this end, I'll try to make it easier for Sage developers to review notebook tickets or make other contributions. Please let me know what would help. For example, I can make experimental spkgs that contain the latest patches in the queue. Those who wish just to test the cumulative changes can install the package with sage -f sagenb-*.spkg. But reviewers can also open the spkg, pop / push patches, and comment on specific ticket(s). In either case, we'll get useful information about how the notebook behaves in a wider gamut of browser-OS combinations. We'll also get more end user feedback.

[qed777]

comment:29

We install the sagenb package with setuptools (PyPI), which updates SAGE_LOCAL/lib/python/site-packages/easy-install.pth. This file contains paths prepended to sys.path on startup.

You can query the installed version with

sage: from sagenb.misc.misc import SAGENB_VERSION
sage: SAGENB_VERSION

which is essentially

sage: from pkg_resources import Requirement, working_set
sage: w = working_set.find(Requirement.parse('sagenb'))
sage: w.version

Moreover, w.location gives the install directory.

I'm checking the fonts now...

[sagetrac-acleone]

comment:30

sagenb 0.7.4 installed correctly for me. All doc and selenium tests passed. Still problems building the PDF docs but Ihaven't applied any of the unicode patches (using vanilla sage-4.3.2.alpha1).

[sagetrac-acleone]

Changed reviewer from Robert Mařík, Minh Van Nguyen to Robert Mařík, Minh Van Nguyen, Alex Leone

[qed777]

comment:31

On the fonts: What is the output of

egrep "Copying jsMath image"\|"Installed.*sagenb" $SAGE_ROOT/install.log 

?

[qed777]

comment:32

Replying to @sagetrac-acleone:

sagenb 0.7.4 installed correctly for me. All doc and selenium tests passed. Still problems building the PDF docs but Ihaven't applied any of the unicode patches (using vanilla sage-4.3.2.alpha1).

Positive review?

[sagetrac-acleone]

comment:33

$ egrep "Copying jsMath image"\|"Installed.*sagenb" ~/sage-dev/sage-4.3.2.alpha1/install.log
Installed /home/alex/sage-dev/sage-4.3.2.alpha1/local/lib/python2.6/site-packages/sagenb-0.6-py2.6.egg

Strange.

sage: from sagenb.misc.misc import SAGENB_VERSION
sage: SAGENB_VERSION
'0.7.4'

Here's how I installed:

1. make on an unmodified 4.3.2.alpha1

$ tar -jxvf sagenb-0.7.4.spkg
$ cd sagenb-0.7.4/src/sagenb/
$ sage -python setup.py develop

3. Tested with sage -t -sagenb

4. Selenium tests with sage -python sagenb/testing/run_tests.py

5. Checking the PDF build with sage -docbuild all pdf

[qed777]

comment:34

I think this is OK, because the SAGE_LOCAL/bin/sage-spkg script invoked by sage -f updates SAGE_ROOT/install.log but the sage -python setup.py commands do not.

[sagetrac-acleone]

comment:35

The "Use image fonts" option is disabled (greyed out) in jsMath - is this a problem?

jsMath v3.6c (Unicode fonts)

[qed777]

comment:36

Are the image fonts installed? In twist.py, we set the boolean

jsmath_image_fonts = is_package_installed("jsmath-image-fonts")

which propagates to jsmath.js. This should enable the option if the spkg is installed. But the fonts need to be installed in the right place...

[sagetrac-acleone]

comment:37

Ok then, LGTM.

[williamstein]

comment:38

I decided to try something random to see if I was running the right notebook. So I tried #3154 first, and it appears that it is NOT fixed by this notebook upgrade. Other things I tried are fixed though.

[qed777]

comment:40

It turns out that I merged #4217, not #3154, into SageNB 0.7. I didn't notice that #4217's commit string was copied from #3154 by mistake. I used hg log to make the list of merged tickets in the description.

[qed777]

Merged: sage-4.3.2