
<!DOCTYPE rss PUBLIC  "-//Netscape Communications//DTD RSS 0.91//EN"
"http://my.netscape.com/publish/formats/rss-0.91.dtd">
<rss version="0.91">
<channel> 
<title></title> 
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi</link> 
<description>The work log of Bernie Simon</description> 
<language>en</language> 
<item>
<title>STZPT Debugged</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/281154119503.html</link>
<description>
&lt;p&gt;It turns out that there were two remaining problems in STZPT. The
first was my code was not returning the rendered edit form. For lack
of a return, it got lost and the page displayed as blank. The second,
the path traversal bug, turned out to be a feature of page
templates. Acquisition works differently in page templates. So to work
around the problem in displaying the red ball icon, I modified the url
so that it starts with a "../", so the path does not include the page
template. So I noe consider STZPT done, except for an independent
test.&lt;/p&gt;
</description>
</item>

<item>
<title>No Luck</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/271154030458.html</link>
<description>
&lt;p&gt;I tried to identify the cause of the path traversal problem in
STZPT, so far with no luck. I think I'll have to use the Wing
debugger. Until I can, I'll turn my attention to other issues.
</description>
</item>

<item>
<title>Bug Squashing</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/261153947844.html</link>
<description>
&lt;p&gt;The day was spent trying to fix the two bugs in STZPT I mentioned
yesterday. Getting the right set of management tags was only slightly
difficult. The other problem, with restricted traverse, has proven
more difficult. I'm trying to locate the problem with debugging
prints, but haven't succeeded yet.&lt;/p&gt;
</description>
</item>

<item>
<title>Getting Closer</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/251153861209.html</link>
<description>
&lt;p&gt;I brought the Chart TWiki back online and protected the register
script so that only Twiki administrators can register people. I also
set up the mailnotify script. We will see tomorrow if it worked. I
walked my preferred solution for pasunx and bubba past Otto and he
approved it. So I implemented it, but the customer still is unhappy
about the man pages. I got STZPT working to the point where you can
add a new page and edit its metadata. But there still seems to be a
problem with acquisition. And the properties tab should be deleted.&lt;/p&gt;
</description>
</item>

<item>
<title>Back to Zope</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/241153774255.html</link>
<description>
&lt;p&gt;I got back to working on STZPT. I moved the code over to battrra,
wrote the code to glue it into the rest of STobjects, and started
testing. Zope starts up without error, but the stsci tree is missing,
so something is badly wrong. I got pulled into an emergency round of
despamming the Chart Twiki, so I blocked access and looked for spam. I 
deleted two suspicious new users.
</description>
</item>

<item>
<title>Mowing 'em Down</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/211153514780.html</link>
<description>
&lt;p&gt;Setting up a cron job to run the htgroup script that I described
yesterday didn't work. Instead I used the Windows scheduler. It
doesn't run Perl scripts, so I looked up how to wrap the script as a
batch job. It turns out that pl2bat.bat ships with the Windows version 
of Perl (in its bin directory) and does the trick. So we now have a
way of build an Apache groups file from TWiki groups daily.&lt;/p&gt;

&lt;p&gt;After finishing that, I looked at the problem Steve Hulbert had
when he set the effective date of an ST File to a blank string. When
this is done, a file is marked as not viewable, which is a bug not a
feature. It turns out that the only value that STObjects code
considers to mean "no effective date" is None. So I changed the
_datify routine so that it sets invalid dates to None instead of the
empty string. This fixes the problem Steve experienced, though the fix
still needs to move through the Zope pipeline.&lt;/p&gt;
</description>
</item>

<item>
<title>Working Calls</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/201153426197.html</link>
<description>
&lt;p&gt;I got diff and rcs installed on bubba, but the user told me they
want into installed elsewhere. So I'll have to look at that again. I
worked a a solution to the TWiki registration problem and I'm running
it by Gretchen. Plus I tried installing a more recent version of
remedy with Greg but we ran into database connection problems.&lt;/p&gt;
</description>
</item>

<item>
<title>RCS Again</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/191153343289.html</link>
<description>
&lt;p&gt;I successfully installed rcs on pasunx, but failed on bubba. Otto
says it's because the Sun C compiler is in a different directory on
bubba. I'll try fixing that tomorrow. Plus I wrote a mail message to
Gretchen explaining why access to registration needed to be controlled 
through Apache.&lt;/p&gt;
</description>
</item>

<item>
<title>Finished Code</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/181153256690.html</link>
<description>
&lt;p&gt;In the morning I finished the coding for the new class STZPT. I'll
move it to battra when the new Zope instances are set up. I also
repaired permissions on the chart TWiki and protected the registration 
script. But what I did turned out not to be what Gretchen wanted, so I 
yanked it.&lt;/p&gt;
</description>
</item>

<item>
<title>STZPT</title>
<link>http://www-int.stsci.edu/~bsimon/weblog.cgi/post/171153170388.html</link>
<description>
&lt;p&gt;Most of the work today was on what's now called the Web Roadmap
project. We developed an initial version of the project breakdown at
our web domain meeting and I stared work on STZPT (the ST wrapper for
Zope Paget Templates.)&lt;/p&gt;
</description>
</item>

</channel>
</rss>
