EPG123 Bug Reports

An evolving, supported alternative to Rovi
Forum rules
★ Download the latest EPG123 here: https://garyan2.github.io/ <> Setup guide here: https://garyan2.github.io/install.html
Post Reply
HTBruceM

Posts: 87
Joined: Sat Dec 22, 2012 12:42 am
Location: Oregon

HTPC Specs: Show details

EPG123 Bug Reports

#1

Post by HTBruceM » Thu Mar 03, 2016 8:05 pm

New thread to report defects in existing functionality, crashes, etc.

kmp14

Posts: 138
Joined: Sat Sep 08, 2012 7:23 pm
Location:

HTPC Specs: Show details

#2

Post by kmp14 » Fri Mar 04, 2016 10:31 pm

Ok, this is probably not a bug, but I am putting it out there in case anyone else has something similar. I had a recording that would crash MC every time I simply landed on it in the list of recorded programs. This is a recording from last night and I have been using Epg123 (with great delight) for a couple weeks now. I assume something in the metadata is what caused the crash. I don't recall ever having something like that happen before, and I have been using MC for many years.

I deleted the program directly in the file system. In hindsight I should have saved it for further analysis...dang.

I am certainly not expecting Gary to look into this at all, I just wanted to get it documented in case there is some bug lurking....

More info for documentation purposes: The show in question was called "Vice Lab", episode named "We are all failures" on the new Vice channel (used to be History H2). It was recorded from 10-10:30 PM CST. I am on Comcast Digital Romeoville, IL (a suburb of Chicago) using Ceton CableCard tuners. On my system it is channel 323.

Space

Posts: 2838
Joined: Sun Jun 02, 2013 9:44 pm
Location:

HTPC Specs: Show details

#3

Post by Space » Fri Mar 04, 2016 11:16 pm

I've had this happen in the past where there was only partial metadata in the recorded file due to some sort of corruption. It would seem that if WMC sees certain metadata in a file and then does not see certain other metadata it can cause a crash.

If that particular episode is still in the guide, you can try recording it again (or just a small portion of it) to see if the same problem occurs.

kmp14

Posts: 138
Joined: Sat Sep 08, 2012 7:23 pm
Location:

HTPC Specs: Show details

#4

Post by kmp14 » Fri Mar 04, 2016 11:20 pm

Space wrote:I've had this happen in the past where there was only partial metadata in the recorded file due to some sort of corruption. It would seem that if WMC sees certain metadata in a file and then does not see certain other metadata it can cause a crash.

If that particular episode is still in the guide, you can try recording it again (or just a small portion of it) to see if the same problem occurs.
Yup, I have the next showing of that episode set to record. I will report back here with the results after it records.

KenL

Posts: 60
Joined: Thu Mar 03, 2016 3:35 pm
Location:

HTPC Specs: Show details

#5

Post by KenL » Sat Mar 05, 2016 12:12 am

Missing OAD on a number of Smithsonian Channel programs loaded with EPG123 (last few versions?) One example: "The True Story of the Mary Celeste"

Code: Select all

<Program id="35592" uid="!Program!SH00033862_0000" title="The True Story of the Mary Celeste" description="A newly discovered transcript of the vessel's final five days informs about the mystery of the 135-year-old "ghost ship."" shortDescription="A transcript informs about the 135-year-old "ghost ship."" keywords="k7,k7009,k7" series="si5193" isSpecial="true" isGeneric="true" />
Looks to me like it's missing from the mxf?

Same channel "The Real Story Pirates of the Caribbean" with normal OAD as expected.

Code: Select all

< Program id="35591" uid="!Program!EP01149155_0011" title="The Real Story" episodeTitle="s03e01 Pirates of the Caribbean" description="Investigating the original Caribbean pirates that sailed the corridors of the Atlantic Ocean." seasonNumber="3" episodeNumber="1" originalAirdate="2011-05-15" keywords="k6,k6037,k6035" series="si5178" isSeries="true" />
Checked Zap2it.com seems all Smithsonian programs (including "The True Story of the Mary Celeste") have OAD. Did a quick check of my native DISH guide (also Gracenote) looks like all Smithsonian Channel programs for the next 7 days have the expected OAD. If it's present in JSON, then it must be EPG123 missing it under certain circumstances?

Space

Posts: 2838
Joined: Sun Jun 02, 2013 9:44 pm
Location:

HTPC Specs: Show details

#6

Post by Space » Sat Mar 05, 2016 12:39 am

If the uid has a "SH" before the number (which is taken directly from the Gracenote data), it is treated differently than those that start with "EP", so perhaps EPG123 is not including OAD for "SH" type shows?
Generally I think "SH" is used when there is no episode specific info for a show (in which case the OAD may mean the day the entire series started), but it may also be used for shows that don't really have any episodes like specials (which this seems to be) and such, in which case I guess the OAD should be included in the MXF (perhaps if it is not marked as a series, you should include the OAD, if specified?).

User avatar
garyan2

Posts: 7438
Joined: Fri Nov 27, 2015 7:23 pm
Location:

HTPC Specs: Show details

#7

Post by garyan2 » Sat Mar 05, 2016 12:48 am

Space wrote:If the uid has a "SH" before the number (which is taken directly from the Gracenote data), it is treated differently than those that start with "EP", so perhaps EPG123 is not including OAD for "SH" type shows?
Generally I think "SH" is used when there is no episode specific info for a show (in which case the OAD may mean the day the entire series started), but it may also be used for shows that don't really have any episodes like specials (which this seems to be) and such, in which case I guess the OAD should be included in the MXF (perhaps if it is not marked as a series, you should include the OAD, if specified?).
That is correct. If it is not specific episodic information (starting with SH) then I ignore the OAD. I'll take a look on handling specials in this case ... get back to you.
- Gary
Keeping WMC alive beyond January 2020. https://garyan2.github.io

KenL

Posts: 60
Joined: Thu Mar 03, 2016 3:35 pm
Location:

HTPC Specs: Show details

#8

Post by KenL » Sat Mar 05, 2016 2:48 am

Thanks... hope that can be fixed.

Just took a quick look and both Dish and TiVo appear to populate the Gracenote OAD on virtually everything but movies and paid programming. It's even there on things like local news even though usually meaningless in such case. I'd hope it can always be used when available. It may be difficult to determine what some programs are without it.

Embiggens

Posts: 196
Joined: Sun Aug 07, 2011 3:29 am
Location:

HTPC Specs: Show details

#9

Post by Embiggens » Sat Mar 05, 2016 4:22 am

Been running epg123 successfully for a week but just suffered a crash during an epg123 update, which caused a reboot with the new v0.7.3 (I downloaded this about 1/2 hour ago, so it should be the correct version). From the trace, this occurred after/during buildAllPrograms():

Code: Select all

...
[3/4/2016 9:54:45 PM] Exiting sdGetSchedulesListing(). SUCCESS.
[3/4/2016 9:54:59 PM] Exiting buildAllScheduleEntries(). SUCCESS.
[3/4/2016 9:54:59 PM] Entering buildAllProgramEntries() for 48876 programs.
That's the last line before the crash. I'm not positive that epg123 caused the crash, but let me know what information/files you'd like if you want to look into.

User avatar
garyan2

Posts: 7438
Joined: Fri Nov 27, 2015 7:23 pm
Location:

HTPC Specs: Show details

#10

Post by garyan2 » Sat Mar 05, 2016 5:10 pm

Embiggens wrote:Been running epg123 successfully for a week but just suffered a crash during an epg123 update, which caused a reboot with the new v0.7.3 (I downloaded this about 1/2 hour ago, so it should be the correct version). From the trace, this occurred after/during buildAllPrograms():

Code: Select all

...
[3/4/2016 9:54:45 PM] Exiting sdGetSchedulesListing(). SUCCESS.
[3/4/2016 9:54:59 PM] Exiting buildAllScheduleEntries(). SUCCESS.
[3/4/2016 9:54:59 PM] Entering buildAllProgramEntries() for 48876 programs.
That's the last line before the crash. I'm not positive that epg123 caused the crash, but let me know what information/files you'd like if you want to look into.
At this point in the program, it should be heavily hitting your cache directory and reading all the files.
  • If it never got past 0/48876, then it was building a list of the files in the cache directory.
    If the program number was increasing but still low relative to the final number, it was importing those cache files.
    If the number was high and just sitting there, then it was probably requesting the missing programs from SD.
There of course could be something in epg123 that is not reacting very well to some corrupt/malformed data, but my bet is on some IO, either with the file system or network. My focus right now is getting better error checking and better information with the SD API part. Of course, if everyone can keep an eye out for things like this, it will help to focus in on any problems and design some fixes.

-Thank you
- Gary
Keeping WMC alive beyond January 2020. https://garyan2.github.io

kmp14

Posts: 138
Joined: Sat Sep 08, 2012 7:23 pm
Location:

HTPC Specs: Show details

#11

Post by kmp14 » Sat Mar 05, 2016 7:59 pm

Here is one I have never seen before. I have a gap in the guide on channel 296 at 11:30 PM tonight. There is no program listed. If I try to move the selection over to the slot it gets stuck and won't land on it. It is easier to show you, so please see the attached pic.

I am on Comcast Digital Romeoville Zip code 60448, using Ceton cablecard tuners. The time in the attached image is CST.

If I go on the Zap2It Site to take a look at that slot, here is what should be there:

Ridiculousness
Episode: Larry King
S05, E07
(First Aired: Aug. 21, 2014)

Image

Space

Posts: 2838
Joined: Sun Jun 02, 2013 9:44 pm
Location:

HTPC Specs: Show details

#12

Post by Space » Sat Mar 05, 2016 8:09 pm

I've seen this happen before with Rovi data. If you highlight the 11pm Ridiculousness and press right arrow, it doesn't move, however, if you highlight the 12am airing and press the left arrow does it then make the missing episode appear?

I think this happens when there is a discrepancy in the length of the episode and the start of the next episode (they overlap or don't line up exactly).

EDIT: OK, I just tried this on my system that is still using Rovi and it has the same problem as you described, using the left arrow method works by pressing the right arrow until the missing episode is off the screen, then pressing left arrow until you get to that episode. You can actually see them overlap.

Anyway, this happens like I said, the episode before ends at 12:30am, but the missing episode starts at 12:25am, so they overlap by 5 minutes (note that you can see the 12:25am episode by going to the "other showings" page when you have another airing selected.)

So, in conclusion, this is a problem with the data source, not with EPG123. In this case since it appears both Gracenote and Rovi have the same mistake, the source of the mistake is most likely the network itself (MTV).

EDIT 2:
Looking at zap2it, it appears that this problem has been corrected there, so perhaps if you run EPG123 again (assuming SD has been updated since you last ran it) it will be corrected. The correction is to make the first episode time be 11:50pm-12:25am and the second episode to be 12:25am-1:00am.

User avatar
garyan2

Posts: 7438
Joined: Fri Nov 27, 2015 7:23 pm
Location:

HTPC Specs: Show details

#13

Post by garyan2 » Sat Mar 05, 2016 8:35 pm

Space wrote:EDIT 2:
Looking at zap2it, it appears that this problem has been corrected there, so perhaps if you run EPG123 again (assuming SD has been updated since you last ran it) it will be corrected. The correction is to make the first episode time be 11:50pm-12:25am and the second episode to be 12:25am-1:00am.
Just about to relay that. I included your lineup and ran epg123 and the issue is not there. Must have been corrected.
- Gary
Keeping WMC alive beyond January 2020. https://garyan2.github.io

kmp14

Posts: 138
Joined: Sat Sep 08, 2012 7:23 pm
Location:

HTPC Specs: Show details

#14

Post by kmp14 » Sat Mar 05, 2016 9:28 pm

garyan2 wrote:
Space wrote:EDIT 2:
Looking at zap2it, it appears that this problem has been corrected there, so perhaps if you run EPG123 again (assuming SD has been updated since you last ran it) it will be corrected. The correction is to make the first episode time be 11:50pm-12:25am and the second episode to be 12:25am-1:00am.
Just about to relay that. I included your lineup and ran epg123 and the issue is not there. Must have been corrected.
Thanks! Wasn't too worried about it, just wanted to share in case it was a potentially lurking issue.

User avatar
garyan2

Posts: 7438
Joined: Fri Nov 27, 2015 7:23 pm
Location:

HTPC Specs: Show details

#15

Post by garyan2 » Sat Mar 05, 2016 9:29 pm

kmp14 wrote:
garyan2 wrote:
Space wrote:EDIT 2:
Looking at zap2it, it appears that this problem has been corrected there, so perhaps if you run EPG123 again (assuming SD has been updated since you last ran it) it will be corrected. The correction is to make the first episode time be 11:50pm-12:25am and the second episode to be 12:25am-1:00am.
Just about to relay that. I included your lineup and ran epg123 and the issue is not there. Must have been corrected.
Thanks! Wasn't too worried about it, just wanted to share in case it was a potentially lurking issue.
And I am certainly glad Space chimed in ... I had never seen that before. Cheers!
- Gary
Keeping WMC alive beyond January 2020. https://garyan2.github.io

kmp14

Posts: 138
Joined: Sat Sep 08, 2012 7:23 pm
Location:

HTPC Specs: Show details

#16

Post by kmp14 » Sat Mar 05, 2016 9:35 pm

Interestingly, I just ran Epg123 again (v 0.7.3), and all ran just fine, but still have that blank spot in the guide. No big deal, I don't care about that show or anything, just odd that it is still missing for me.

Space

Posts: 2838
Joined: Sun Jun 02, 2013 9:44 pm
Location:

HTPC Specs: Show details

#17

Post by Space » Sat Mar 05, 2016 9:38 pm

kmp14 wrote:Interestingly, I just ran Epg123 again (v 0.7.3), and all ran just fine, but still have that blank spot in the guide. No big deal, I don't care about that show or anything, just odd that it is still missing for me.
Now THIS may indicate a problem with EPG123.

KenL

Posts: 60
Joined: Thu Mar 03, 2016 3:35 pm
Location:

HTPC Specs: Show details

#18

Post by KenL » Sat Mar 05, 2016 10:14 pm

I'm also running v0.7.3 and my MTVHD isn't presenting that issue.

Code: Select all

<StationID CallSign="MTVHD">60964</StationID>
It hasn't updated today but I'll check again after it does.

User avatar
garyan2

Posts: 7438
Joined: Fri Nov 27, 2015 7:23 pm
Location:

HTPC Specs: Show details

#19

Post by garyan2 » Sun Mar 06, 2016 12:33 am

There is a possibility that when the error existed in the downloaded data, that epg123 caused it to be irreversible. I think what we are seeing is epg123 compared the start time of the downloaded program to what time we expected it to start (directly after the last program). If the didn't match, I would enter a start time for the new program. It is the incorrect start time that is probably still in the database causing the overlap times.

I say irreversible, but the only way to correct it is to put a start time on every schedule entry which would increase the mxf file size considerably. But this is also the only way to recover from something like this. Maybe I will keep doing what I am but only enter the start time for programs the gap in the future, not in the past if that makes sense.
- Gary
Keeping WMC alive beyond January 2020. https://garyan2.github.io

HTBruceM

Posts: 87
Joined: Sat Dec 22, 2012 12:42 am
Location: Oregon

HTPC Specs: Show details

#20

Post by HTBruceM » Sun Mar 06, 2016 7:06 am

Not sure if this is EPG123 issue, or SD issue. Thought I'd report it here to see where to go next.
I have a series recording in WMC to record "The Grinder", new episodes only.
Next week, the Grinder is showing a repeat, it's even in the title. But WMC recording schedule still shows that it will be recorded.
Screenshots follow. Is this a problem in EPG123 data/field mapping, or is it a Gracenote or SD issue?
2016-03-05 (1).png
2016-03-05 (2).png

Post Reply