Page 2 of 7
Re: Filtered Statistics Report 1.1 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 5:51 pm
by Bex
Very nice! This script really needed an update and it's good to see that it is in good hands.
Small bug on line 1091 in
tLyricistsCountPlayed is misspelled inLyricistsCountPlayed so it yields error.
Re: Filtered Statistics Report 1.1 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 5:53 pm
by Mizery_Made
I like some of the changes made, however there seems to be an issue somewhere. In the Totals section, it's showing that I have one Original Artists and that it has been played. However, in the averages section as well as the Top * Original Artists sections, there are none listed (or 0 in the case of the averages). I don't recall having an Original Artist stored anywhere, and when I look at the song table in my Database, I see nothing listed in this column.
Re: Filtered Statistics Report 1.1 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 5:57 pm
by Bex
And if I may contribute to this script. A long time ago I wrote this function below which gives days instead of just one zillion hours.
Code: Select all
Function FormatTime(milliseconds)
FormatTime = _
SDB.Database.OpenSQL("SELECT CASE "&_
"WHEN "& milliseconds &" <600*1000 THEN "&_
"SUBSTR(STRFTIME('%M:%S',0.5+CAST("& milliseconds &" AS REAL)/1000/86400),2,4) "&_
"WHEN "& milliseconds &" <3600*1000 THEN "&_
"STRFTIME('%M:%S',0.5+CAST("& milliseconds &" AS REAL)/1000/86400) "&_
"WHEN "& milliseconds &" <3600*10*1000 THEN "&_
"SUBSTR(STRFTIME('%H:%M:%S',0.5+CAST("& milliseconds &" AS REAL)/1000/86400),2,7) "&_
"WHEN "& milliseconds &" <3600*24*1000 THEN "&_
"STRFTIME('%H:%M:%S',0.5+CAST("& milliseconds &" AS REAL)/1000/86400) "&_
"WHEN "& milliseconds &" <3600*25*1000 THEN "&_
"CAST("& milliseconds &"/24/3600000 AS INTEGER)||' day '||STRFTIME('%H:%M:%S',0.5+CAST("& milliseconds &" AS REAL)/1000/86400) "&_
"ELSE "&_
"CAST("& milliseconds &"/24/3600000 AS INTEGER)||' days '||STRFTIME('%H:%M:%S',0.5+CAST("& milliseconds &" AS REAL)/1000/86400) END").StringByIndex(0)
End Function
Re: Filtered Statistics Report 1.1.1 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 6:10 pm
by ZvezdanD
There is a new revision with removed bugs. Thanks for reports.
Re: Filtered Statistics Report 1.1 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 6:13 pm
by ZvezdanD
Bex wrote:And if I may contribute to this script. A long time ago I wrote this function below which gives days instead of just one zillion hours.
Thanks for your kind words and for this contribution.
Re: Filtered Statistics Report 1.1.1 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 6:18 pm
by nynaevelan
Zvezdand:
I assume when you add the Option sheets you will include an option to suppress the tables for the values that do not have any tracks to meet the criteria? For example, I do not have any lyricists so I do not need to see the empty tables. I do not mind them being included in the Totals section though.
Also, I get the following error with 1.1.1, I did not get it with 1.1 but unfortunately I overwrote 1.1 so I do not have it anymore.
Nyn
Re: Filtered Statistics Report 1.1.1 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 6:18 pm
by Bex
Thank you!
New small bug on line 598: "Played.IDsongWHERE" is missing a <space>
Re: Filtered Statistics Report 1.1.2 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 6:29 pm
by ZvezdanD
There is a new revision (1.1.2). I hope this time it wouldn't be more bugs
Re: Filtered Statistics Report 1.1.2 (2009-04-11) [MM3]
Posted: Fri Apr 10, 2009 7:57 pm
by Mizery_Made
I can confirm that the incorrect count of OrigArtist is fixed in this version (1.1.2), Though I now notice another issue. The "Tracks/Played" column differs between "Top # *" and "Top # Played *" while the Length & Size max out with 1 play, giving you identical information between these two sections once you've played something once. Not sure if it's possible to raise the played length based on the amount of times a track was played, or if it would significantly increase load time. Thought I would point it out though.
EDIT: That actually makes me wonder, is Size even needed for the Played (and now Rated) sections?
Re: Filtered Statistics Report 1.1.2 (2009-04-11) [MM3]
Posted: Sat Apr 11, 2009 1:09 am
by ZvezdanD
Mizery_Made wrote:Though I now notice another issue. The "Tracks/Played" column differs between "Top # *" and "Top # Played *" while the Length & Size max out with 1 play, giving you identical information between these two sections once you've played something once.
Yes, you are right, thanks for the report. However, I am now wondering when I change Length data in Top # Played tables how to name that column. I think it should not stay as "Length", and in same way as I changed "Tracks" to "Played" after Nyn suggestion, I think I should rename that column as well. In the Totals section I added "Play time" row with similar data, only calculated globally. In Magic Nodes script I have the field with same data named "Play Length". What do you think about those two suggestion for this column caption?
is Size even needed for the Played (and now Rated) sections?
Well, I agree that it is not needed, but I would not remove that column because it could break uniformity of tables.
Re: Filtered Statistics Report 1.1.2 (2009-04-11) [MM3]
Posted: Sat Apr 11, 2009 1:19 am
by Mizery_Made
"Played Time" sounds better to me than "Played Length", I'm not really sure why. And I get you about the Uniformity part.
Re: Filtered Statistics Report 1.1.2 (2009-04-11) [MM3]
Posted: Sat Apr 11, 2009 2:37 am
by ZvezdanD
I just checked, File size is not always same between Top # and Top # Played tables. The latter file size is only for played tracks, instead of the first table where you get the file size for all tracks. Same state for Rating tables. Maybe I should add "(rated)" and "(played)" after "File size"? The "Length" for Rating tables is also different from Top # tables, because it is the length only for rated tracks, so it could be "Length (rated)". What do you think?
Re: Filtered Statistics Report 1.1.2 (2009-04-11) [MM3]
Posted: Sat Apr 11, 2009 2:48 am
by Mizery_Made
What I mean by same is that it only appears to account for one play of each track, the same trouble Averages had and such. If you've only played half of your library, then yeah, you're going to see lower numbers for the Played sections, but once you have everything played, it just stops.
Code: Select all
Top 10 Artists With Most Tracks
# Artist Tracks Length File size
2 Tech N9ne 769 53:04:00 4.10 GB
3 2Pac 480 33:58:26 2.78 GB
4 Yukmouth 286 19:30:03 1.50 GB
Top 10 Played Artists
# Artist Played Length File size
2 Tech N9ne 3649 53:00:04 4.09 GB
3 2Pac 1914 33:58:26 2.78 GB
4 Yukmouth 1040 19:30:03 1.50 GB
As you can see, while the Played count is significantly higher than the total of each of these, their Length and Filesize are the same since each of those files from the Total have been played at least once. The only exception is Tech N9ne, which appears lower on the Played list because I have one track not played for him (which is actually playing at the moment). So this demonstrates both that it does differ, for those with things not played, but also hits a wall at the Total figures.
Re: Filtered Statistics Report 1.1.2 (2009-04-11) [MM3]
Posted: Sat Apr 11, 2009 2:52 am
by ZvezdanD
Yes, I understand that, but you should know that many of us like me don't use MM for reproduction. So, those data could differ significantly between tables. Neither I have rated tracks...
Re: Filtered Statistics Report 1.1.2 (2009-04-11) [MM3]
Posted: Sat Apr 11, 2009 2:56 am
by Mizery_Made
Sorry, but, I'm not following. *Scratches Head*