Bug 320956

Comment 1 Denis Roy 2010-08-20 16:13:39 EDT

> showing data from ECon 2010.
> 
> - Authors Report
> - Categories Report
> - Tags Report
> 
> These all show ECon 2010 data.


But of course, since line 69 and 74 of this file references conference id 5:

http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.dash/submissions/feeds/pc_tags.php?annotate=1.2&root=Technology_Project

Comment 2 Denis Roy 2010-08-20 16:18:51 EDT

I've changed the hardcoded references to conference 5 to use the conference in the context, so this should now be fixed.  Can you confirm?

Comment 3 Anne Jacko 2010-08-21 14:07:32 EDT

OK, I looked pretty thoroughly at these for both easy2010 and ese2010.

First, Conference Queen reports for ese2010 and easy2010. 

To summarize, these Conf Queen reports all look good as far as can be tested right now, except for speakers_signed_agreements.csv -- see my notes below for easy2010.

ese2010 Conf Queen
- as expected, all reports are empty except for the column headings (because no talks have been accepted yet).

easy2010 Conf Queen
- speakers_signed_agreements.csv: this report has a row for each easy2010 accepted speaker, and indicates that none of the agreements have been signed; if Denis "unsigned" all the easy2010 agreements, then this would be correct. One thing that I don't understand in this report is some data on the row with my name. Under "paid" it gives a date and time, and under "coupon" it lists "STESE2009." I'd guess that this means the system thinks I registered, and I used that coupon. But that this coupon code is not one of the ones I got when I was testing easy2010; this is an old ESE 2009 coupon.

- speakers.csv: this appears to be correctly displaying data about all accepted speakers

- staging_guide_long.csv: contains a list of accepted talks of types long, tutorial, symposium, and keynote, so it looks like "long" is considered to be any type whose length is equal to or great than the length defined for type "long" -- I'm not exactly sure who uses this report, and for what, so I don't know if including types other than "long" is a good thing or not

- staging_guide_short.csv: contains a list of accepted talks of types short, sponsored and poster, so it looks like "short" is considered to be any type whose length is equal to or shorter than the length defined for type "short" -- again, don't know if this is useful or not

- signage.csv: this looks correct

- submission_feedback.csv: looks correct (as far as it goes, since there is no feedback data to include)

- Big Sign reports: these can't be tested yet, since they pull data from the online schedule, which doesn't exist for easy2010

Next, Program Committee reports for ese2010 and easy2010. These all appear to need some work -- most of them may just need to have the column headings (talk types) updated since they are still displaying those from ECon 2010.

ese2010 Prog Comm reports
- pc_authors.csv: trying to download this report returns this error: <br />
<b>Warning</b>:  array_keys() [<a href='function.array-keys'>function.array-keys</a>]: The first argument should be an array in <b>/home/apps/submissions/org.eclipse.dash/submissions/feeds/pc_authors.php</b> on line <b>84</b><br />
<br />
<b>Warning</b>:  implode() [<a href='function.implode'>function.implode</a>]: Invalid arguments passed in <b>/home/apps/submissions/org.eclipse.dash/submissions/feeds/pc_authors.php</b> on line <b>85</b><br />

- pc_categories.csv: This is kinda weird. The column headings (talk types) have ECon 2010 data, and the row headings (categories) have ESE 2010 categories, but not all of them. It shows three rows (Embedded, Modeling, and Other/New & Noteworthy) when there should be five (e4 and Runtime are missing).

- pc_tags.csv: Again, the column headings (talk types) are from ECon2010. There is only one row, and it's labeled with the only Talk Tag we have for ESE 2010 -- "New submission." The counts in the "cntAcceptTag" and "cntAll" columns look correct, but I wasn't able to tell for sure since I didn't know how to search for the submissions that have that tag. 

It appears as though this report shows both PC Tags and Talk Tags, which is probably what it's supposed to do. As noted below, easy2010 has both PC Tags and Talk Tags applied to talks, and the easy2010 report lists both tag types. ESE2010 currently has only a Talk Tag applied to talks, so it's only showing that Talk Tag in the report.

easy2010 Prog Comm reports

- pc_authors.csv: column heads (talk types) are from ECon 2010. Row headings correctly show two accepted authors.

- pc_categories.csv: column heads (talk types) are from ECon 2010. Row headings (talk categories) are from easy2010, but Other / New & Noteworthy is listed twice, once as "Other / New & Noteworthy" and once as "Other / New &amp Noteworthy."

- pc_tags.csv: column heads (talk types) are from ECon 2010. Row headings show the easy2010 PC tags (e.g. "Anne +1") that were applied, and the one Talk Tag that was applied ("New submission").

Thanks!

Comment 4 Denis Roy 2010-08-23 13:59:17 EDT

The PC authors, tags and categories had the rows hard-coded.  I've changed those three reports to pull the data from the database, so these will automatically 'just work' in all cases.

http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.dash/submissions/feeds/pc_tags.php?root=Technology_Project&r1=1.4&r2=1.3

Comment 5 Anne Jacko 2010-08-24 12:55:22 EDT

Created attachment 177345 [details]
pc_authors report from ese2010 as of August 24

Comment 6 Anne Jacko 2010-08-24 12:56:30 EDT

Denis, the pc_categories and pc_tags reports look perfect, but the pc_authors one is returning errors. I've attached it. Thanks.

Comment 7 Denis Roy 2010-08-24 14:35:56 EDT

Strangely, it works like a charm for easy2010.

Comment 8 Denis Roy 2010-08-24 15:43:28 EDT

(In reply to comment #6)
> Denis, the pc_categories and pc_tags reports look perfect, but the pc_authors
> one is returning errors. I've attached it. Thanks.


The error is because there are no records.  The query has an explicit:

having count(*) > 2

Which means that people with only one accepted talk (tentative or otherwise) are not listed.  I'm not sure why that clause is there, but I can certainly remove it.

Comment 9 Denis Roy 2010-12-21 14:57:53 EST

(In reply to comment #8)
> (In reply to comment #6)
> > Denis, the pc_categories and pc_tags reports look perfect, but the pc_authors
> > one is returning errors. I've attached it. Thanks.
> 
> 
> The error is because there are no records.  The query has an explicit:
> 
> having count(*) > 2
> 
> Which means that people with only one accepted talk (tentative or otherwise)
> are not listed.  I'm not sure why that clause is there, but I can certainly
> remove it.

I recently bumped this down to >1 for Don, and apparently that took care of it.