On Sun, Sep 27, 2009 at 11:31 AM, Ondrej Certik <ond...@certik.cz> wrote:
> On Sun, Sep 27, 2009 at 11:25 AM, William Stein <wst...@gmail.com> wrote:
>> Harald,
>>
>> Is there any chance you could design a website for "The Sage Notebook"
>> as a separate project from Sage, kind of like how
>> http://pynac.sagemath.org/ is a separate project? I don't actually
>> have any specific requests regarding the design, except that it "look
>> professional".
>
> William, would it be a good idea to push your hg repository somewhere?
> Maybe bitbucket, I think it's similar to github.
>
> It would be way easier if I just did "hg pull" to have the latest
> version, rather than messing up with tar.gz files.
Yes, definitely. I'll set that up today.
I'm planning on working on the notebook from about 2 hours from now
for the rest of the day. For the first official release -- when we
can switch for Sage -- all that remains is:
* reimplement sphinx docstring introspection so it happens in the
worksheet process (the original implementation, though pretty, is run
by the server which is sufficiently bad that I want to rewrite this
rather than release something with this design)
* get https:// support working (right now it depends on dsage's
cert stuff which isn't in the notebook obviousy)
* DATA directory permissions issue -- the whole design of the
"DATA" directory needs to be redone, since I vastly improved the
security model of how the two-user notebook works (now all
computations are done by a user that does *not* have permission to
view or change *anything* in the server's sage_notebook directory --
everything happens in /tmp -- this is a VAST improvement, but means
that DATA directory just doesn't work anymore as implemented).
* fix any other bugs (I know of no other issues) -- I'll post a
public demo server for people to test once I get the above done.
* make sure that the old notebook directory "just works" with the
new notebook
In the notebook separation, it has been a fine line between rewriting
things (which takes longer) and just leaving them possibly suboptimal.
One thing I did was completely rewrite everything that used pexpect,
so the notebook doesn't depend on the sage/interfaces/* code in Sage
at all (i.e.., all the expect interfaces). In doing this I made some
major improvements (especially from a security and "understandability"
perspective). I've also laid excellent groundwork for having the
worksheet processes run on a physically different computer than the
server process, which has major scalability advantages.
I'll be on irc.freenode.net #sage-devel miniblogging as I work, if
anybody wants to help...
-- William
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
