Hi.
Yeah, please send me the latest log file. It'll hopefully uncover some interesting clues into why it stopped working this abruptly. It started abruptly and stopped abruptly too.
While waiting for the newest log file, I decided to take a closer look at the log file you sent me on 13th July. Before I go into details, I have a question:
Does your application open OneNote application while printing pages and leave it open even after the printing is finished?
And, while running the script, are you manually navigating across the OneNote sections and pages for some reason?
OneNote shouldn't be touched at all after printing pages and before our utility returns the status code to the script. The reason is that doing so may fudge the lastModifiedTimes of sections and pages. This timestamp is the very fulcrum of our algorithm.
Here are my observations from the log file:
Date & Time passed in: 12.7.2011 8:04:00 (2011-07-12 08:04:00)
1. Notebooks updated on 12th July:
a. Personal (10:34:15)
b. Desktop (10:24:35)
c. Rtaks (10:24:35)
2. Sections updated on 12th July:
a. Section4 in Personal (10:34:15)
3. Pages updated on 12th July:
a. Untitled Page in Section4 (10:34:15)
Now, the utility follows this sequence of steps:
1. Retrieve information about ALL pages in OneNote.
2. Then, take each notebook one at a time and do the following:
a. Proceed to (b) if the lastModifiedTime of at least one section in this notebook is later then what we passed in. Otherwise, go to Step 2.
b. Take each page in this section one at a time and do the following:
i. If the page's lastModifiedTime is later than what we passed, then publish the page as a pdf file in cache folder. Otherwise, simply skip it and go to Step 2 (b).
3. At this stage, we have each newer page in its own pdf in the cache folder.
4. Combine them all in one PDF and put that in the output folder.
Now, if the date and time were passed in properly, then
1. Notebooks that have at least one section modified later than what we passed in: Personal
2. Section that was modified: Section4
3. Page that was modified: Untitled Page
The resultant PDF would have had just one page and that is Untitled Page.
All of this looks okay to me. Your application printed only one new page in Section4 in Personal notebook on that day, which is okay.
I have just one question: Did you really start the script at 08:04:00 A.M on that day? Is this the correct date and time?
If so, then the timestamps suggest that it took around 2.5 hours to print pages. This doesn't look normal. What do you think? This makes me feel that OneNote was being used while our utility was doing its job, which is an issue.
Coming back to your guess of why it may have taken all the pages from 19th May:
1. It doesn't check the last opened time but lastModifiedTime. Again, if something is opening OneNote and modifying something in there then these two timestamps could be the same, but our algorithm can never handle that.
2. And, to make this work, all we need is a new page under any section under any notebook that has the lastModifiedTime AFTER the date and time we pass in. We don't really need a whole new section or a whole new notebook; just a whole new page or a collection of them anywhere in OneNote.
I thought I'd write this before I got a chance to look at the latest log so as to clear up any confusion around the algorithm the utility is following.
We'll discuss again after I run through the log file you'll send.
Cheers,
Light.
Yeah, please send me the latest log file. It'll hopefully uncover some interesting clues into why it stopped working this abruptly. It started abruptly and stopped abruptly too.
While waiting for the newest log file, I decided to take a closer look at the log file you sent me on 13th July. Before I go into details, I have a question:
Does your application open OneNote application while printing pages and leave it open even after the printing is finished?
And, while running the script, are you manually navigating across the OneNote sections and pages for some reason?
OneNote shouldn't be touched at all after printing pages and before our utility returns the status code to the script. The reason is that doing so may fudge the lastModifiedTimes of sections and pages. This timestamp is the very fulcrum of our algorithm.
Here are my observations from the log file:
Date & Time passed in: 12.7.2011 8:04:00 (2011-07-12 08:04:00)
1. Notebooks updated on 12th July:
a. Personal (10:34:15)
b. Desktop (10:24:35)
c. Rtaks (10:24:35)
2. Sections updated on 12th July:
a. Section4 in Personal (10:34:15)
3. Pages updated on 12th July:
a. Untitled Page in Section4 (10:34:15)
Now, the utility follows this sequence of steps:
1. Retrieve information about ALL pages in OneNote.
2. Then, take each notebook one at a time and do the following:
a. Proceed to (b) if the lastModifiedTime of at least one section in this notebook is later then what we passed in. Otherwise, go to Step 2.
b. Take each page in this section one at a time and do the following:
i. If the page's lastModifiedTime is later than what we passed, then publish the page as a pdf file in cache folder. Otherwise, simply skip it and go to Step 2 (b).
3. At this stage, we have each newer page in its own pdf in the cache folder.
4. Combine them all in one PDF and put that in the output folder.
Now, if the date and time were passed in properly, then
1. Notebooks that have at least one section modified later than what we passed in: Personal
2. Section that was modified: Section4
3. Page that was modified: Untitled Page
The resultant PDF would have had just one page and that is Untitled Page.
All of this looks okay to me. Your application printed only one new page in Section4 in Personal notebook on that day, which is okay.
I have just one question: Did you really start the script at 08:04:00 A.M on that day? Is this the correct date and time?
If so, then the timestamps suggest that it took around 2.5 hours to print pages. This doesn't look normal. What do you think? This makes me feel that OneNote was being used while our utility was doing its job, which is an issue.
Coming back to your guess of why it may have taken all the pages from 19th May:
1. It doesn't check the last opened time but lastModifiedTime. Again, if something is opening OneNote and modifying something in there then these two timestamps could be the same, but our algorithm can never handle that.
2. And, to make this work, all we need is a new page under any section under any notebook that has the lastModifiedTime AFTER the date and time we pass in. We don't really need a whole new section or a whole new notebook; just a whole new page or a collection of them anywhere in OneNote.
I thought I'd write this before I got a chance to look at the latest log so as to clear up any confusion around the algorithm the utility is following.
We'll discuss again after I run through the log file you'll send.
Cheers,
Light.