Opened 7 years ago
Closed 7 years ago
#15569 closed enhancement (fixed)
upgrade sagenb to version 0.10.8
Reported by:  ohanar  Owned by:  

Priority:  major  Milestone:  sage6.2 
Component:  packages: standard  Keywords:  merge #195 
Cc:  schilly  Merged in:  
Authors:  R. Andrew Ohana, John Palmieri, Punarbasu Purkayastha  Reviewers:  R. Andrew Ohana, John Palmieri 
Report Upstream:  N/A  Work issues:  
Branch:  u/ppurka/ticket/15569 (Commits, GitHub, GitLab)  Commit:  6f6ec5b8d0434ffaaa0f194bf96d2b55b7741b45 
Dependencies:  Stopgaps: 
Description (last modified by )
Includes a number of small bug fixes.
"Upstream" Tarball: sagenb0.10.8.tar
Change History (31)
comment:1 Changed 7 years ago by
 Status changed from new to needs_review
comment:2 Changed 7 years ago by
 Commit changed from 82f8c0e2d8ecf4d654736fab396394ca81aa3a31 to 5e0e0f94f76decfd799bf8283272eaade348a4cb
comment:3 Changed 7 years ago by
 Status changed from needs_review to needs_work
On two different OS X machines, I get a doctest failure:
sage t src/sage/all.py ********************************************************************** File "src/sage/all.py", line 30, in sage.all Failed example: for i in frames: filename, lineno, funcname, linelist, indx = inspect.getframeinfo(i) for nm in allowed: if nm in filename: break else: print filename Expected nothing Got: /Users/palmieri/Desktop/Sage_stuff/git/sage/local/lib/python2.7/sitepackages/pytz20\ 13.8py2.7.egg/pytz/__init__.py **********************************************************************
comment:4 followup: ↓ 6 Changed 7 years ago by
Downgrading pytz
to the version in sagenb0.10.7.2 seems to fix the problem. Can you put together a new tarball with that change?
comment:5 Changed 7 years ago by
 Commit changed from 5e0e0f94f76decfd799bf8283272eaade348a4cb to 69c56e2155834b636a83da7d038edf80205fe6ab
Branch pushed to git repo; I updated commit sha1. New commits:
69c56e2  sagenb: update checksum

comment:6 in reply to: ↑ 4 Changed 7 years ago by
 Status changed from needs_work to needs_review
Replying to jhpalmieri:
Downgrading
pytz
to the version in sagenb0.10.7.2 seems to fix the problem.
sagenb's dist.sh
script seems to just grab the most recent release of its dependencies.
Can you put together a new tarball with that change?
Ok, done.
comment:7 Changed 7 years ago by
 Status changed from needs_review to needs_work
Actually, this seems to break the build. Will look at it later.
comment:8 Changed 7 years ago by
I guess if you're using an older version of pytz
so you're making the tarball manually, then you have to update (or regress) install_order
by hand, too. Is that the problem?
comment:9 Changed 7 years ago by
I think if we want to install pytz within a range, we should use the following in util/fetch_deps.py
:
, 'pytz >=2011n, <=2013b'
comment:10 Changed 7 years ago by
 Branch changed from u/ohanar/sagenb_upgrade to u/jhpalmieri/sagenb
 Commit changed from 69c56e2155834b636a83da7d038edf80205fe6ab to 5fed454554d1a577cb01be25abed79e3c11cc74b
 Dependencies set to #15570
 Description modified (diff)
comment:11 Changed 7 years ago by
I'm not sure why this included the changes from #15570, and I don't have time to figure out git well enough to fix it right now...
comment:12 Changed 7 years ago by
 Branch changed from u/jhpalmieri/sagenb to u/jhpalmieri/sagenb_new
 Commit changed from 5fed454554d1a577cb01be25abed79e3c11cc74b to d0af1a7f06f631bc9a28bd26d34a25f191be2f70
 Status changed from needs_work to needs_review
New commits:
d0af1a7  Update sagenb to version 0.10.7.3

comment:13 Changed 7 years ago by
 Dependencies #15570 deleted
comment:14 Changed 7 years ago by
Please check the pull request 194 which should fix openid logins. If possible, we should not release yet another broken sagenb.
comment:15 Changed 7 years ago by
I must add that I had no errors with the latest pytz on Linux x64 machine. So, the error seems MacOSX specific.
Edit Sorry, I just now noticed that the error was not in sagenb but in sage.
comment:16 Changed 7 years ago by
I don't understand one thing. I ran the following commands
# Make the sagenb package ~/Installations/sagenb» ./dist.sh ~/Installations/sagenb» cp a dist sagenb0.10.8 ~/Installations/sagenb» tar cf sagenb0.10.8.tar sagenb0.10.8 ~/Installations/sagenb» mv sagenb0.10.8.tar ~/Installations/sage/upstream/. # Prepare the sagenb package in sage ~/Installations/sage» ./sage dev checkout ticket 15569 ~/Installations/sage» echo "0.10.8" > build/pkgs/sagenb/packageversion.txt ~/Installations/sage» mv upstream/sagenb0.10.7.2.tar.bz2 /tmp ~/Installations/sage» sage sh c sagefixpkgchecksums # Check the diff before committing ~/Installations/sage» git diff diff git a/build/pkgs/sagenb/checksums.ini b/build/pkgs/sagenb/checksums.ini index 7843f0e..2f4e2e6 100644  a/build/pkgs/sagenb/checksums.ini +++ b/build/pkgs/sagenb/checksums.ini @@ 1,4 +1,4 @@ tarball=sagenbVERSION.tar sha1=7a822ca573748fcd2034a8cd887b813876ea036d md5=04af0903431905d9ac3e65d0e636a040 cksum=624578127 +tarball=sagenbVERSION.tar.sagenb0.10.8.tar +sha1=17ae39492a1176efbc62006315afb025c4b19d92 +md5=23b448c9d5207bfa017640d32a013b6f +cksum=2809114274 diff git a/build/pkgs/sagenb/packageversion.txt b/build/pkgs/sagenb/packageversion.txt index e0ee8dd..1a46c7f 100644  a/build/pkgs/sagenb/packageversion.txt +++ b/build/pkgs/sagenb/packageversion.txt @@ 1 +1 @@ 0.10.7.3 +0.10.8
Why is the tarball
line so weird. Should I fix it by hand?
+tarball=sagenbVERSION.tar.sagenb0.10.8.tar
I noticed that the branch in this ticket is very behind develop. Is the above weird tarball naming fixed in develop? If so, I can try to merge develop on to this ticket and then make changes from there.
~/Installations/sage» git diff stat develop ticket/15569  tail n 1 135 files changed, 3421 insertions(+), 10729 deletions()
comment:17 Changed 7 years ago by
 Status changed from needs_review to needs_work
 Summary changed from upgrade sagenb to version 0.10.7.3 to upgrade sagenb to version 0.10.8
You should fix it by hand, there is a bug in the sagefixpkgchecksums
script that is fixed in #15570, but that won't be in develop until Volker releases the next beta.
comment:18 Changed 7 years ago by
 Branch changed from u/jhpalmieri/sagenb_new to u/ppurka/ticket/15569
 Created changed from 12/23/13 03:33:25 to 12/23/13 03:33:25
 Modified changed from 12/29/13 20:29:12 to 12/29/13 20:29:12
comment:19 Changed 7 years ago by
 Commit changed from d0af1a7f06f631bc9a28bd26d34a25f191be2f70 to 6f6ec5b8d0434ffaaa0f194bf96d2b55b7741b45
 Description modified (diff)
 Status changed from needs_work to needs_review
New commits:
6f6ec5b  Update sagenb version to 0.10.8

comment:20 Changed 7 years ago by
 Description modified (diff)
comment:21 Changed 7 years ago by
 Reviewers set to R. Andrew Ohana, John Palmieri
 Status changed from needs_review to positive_review
comment:22 Changed 7 years ago by
 Milestone changed from sage6.1 to sage6.2
comment:23 Changed 7 years ago by
 Status changed from positive_review to needs_work
Our build system doesn't accept tar files, it must be .gz or .bz2.
comment:24 Changed 7 years ago by
 Cc schilly added
 Status changed from needs_work to needs_info
Rather, the tarball mirror script (/home/sagemath/www2dev/mirror_upstream.py
) for the web page doesn't understand .tar. Harald, do you think that is fixable?
comment:25 followup: ↓ 26 Changed 7 years ago by
Harald says that this can be fixed ;)
comment:26 in reply to: ↑ 25 Changed 7 years ago by
Replying to vbraun:
Harald says that this can be fixed ;)
Great! It makes little sense to compress the sagenb tarball since it consists of compressed tarballs. :)
comment:27 Changed 7 years ago by
 Status changed from needs_info to positive_review
comment:28 Changed 7 years ago by
 Keywords merge #195 added
 Status changed from positive_review to needs_work
Better wait for 0.10.8.1; see https://github.com/sagemath/sagenb/issues/195
comment:29 Changed 7 years ago by
That pull request fixes the problem for me, so I would suggest getting 0.10.8.1 into Sage (and releasing maybe version 6.1.1 of Sage) as soon as possible. Having a broken notebook is bad...
comment:30 Changed 7 years ago by
 Status changed from needs_work to positive_review
I've made #15778 for the 0.10.8.1 upgrade. Ideally with a doctest in Sage so that it doesn't happen again...
comment:31 Changed 7 years ago by
 Resolution set to fixed
 Status changed from positive_review to closed
Branch pushed to git repo; I updated commit sha1. New commits:
sagenb: fix wrong tarball name in checksums.ini