[BUG]: IAFD Cloudflare blocking IAFD Agent; and, All Agents not returning Actors Headshots

[CodyBerenson]

Mandatory

Describe the bug:

I tested what some of the folks were saying in the discussion. I agree, IAFD Agent blocked in its entirety and actors headshots are not being populated.

Specific Agent(s) Causing the Issue:

I tested specifically with GEVI

Index Site URL(s) Attempting to Match:

https://gayeroticvideoindex.com/video/56527

Log Attached:

com.plexapp.agents.GEVI.log

Optional

Screenshot(s) with all nudity redacted:

Additional Context:

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]
• Version [e.g. 22]

[jfreiman]

Thanks for opening a Bug thread for this issue.

My system is running the following:

Ryzen 7 7700X, 96GB Windows 11 x64 Pro 23H2 (build 22631.3880)

Microsoft Windows Edge (app.plex.tv/desktop) Version 127.0.2651.49 (Official build) stable app

Microsoft Edge Dev for everthing else Version 128.0.2690.1 (Official build) dev (64-bit)

PGMA build release '3 weeks ago' (installed/upgraded about a week ago)

[jfreiman]

I've added dozens of video features and scenes in the past week and none of them have matched with IAFD -- even though my Agent priority has IAFD listed first.

Apprentice Blue: Chapters 5-9 Filename: "(Carnal) - Apprentice Blue∶ Chapters 5-9 (2024).mp4" Apprentice Blue: Chapters 5-9 @ IAFD Apprentice Blue: Chapters 5-9 PGMA matched using DVD Empire Apprentice Blue Chapters 5-9.mp4 [1057953].zip XML logs for Apprentice Blue C5-9.zip

Please let me know what else I can do/provide to resolve this issue. Below are my Agent Priorities

[JPH71]

I sm trying to sort the cloud flare issue with iafd... So any other issues you find, let me know xxx

On Mon, 15 Jul 2024, 02:56 John Freiman, @.***> wrote:

I've added dozens of video features and scenes in the past week and none of them have matched with IAFD -- even though my Agent priority has IAFD listed first.

Apprentice Blue: Chapters 5-9 Filename: "(Carnal) - Apprentice Blue∶ Chapters 5-9 (2024).mp4" Apprentice Blue: Chapters 5-9 @ IAFD https://www.iafd.com/title.rme/id=4b7e99e2-4523-48c5-a0fa-005335c9a822 Apprentice Blue: Chapters 5-9 PGMA matched using DVD Empire http://www.gaydvdempire.com/4760742/apprentice-blue-chapters-5-9-gay-porn-movies.html Apprentice Blue Chapters 5-9.mp4 [1057953].zip https://github.com/user-attachments/files/16229422/Apprentice.Blue.Chapters.5-9.mp4.1057953.zip Apprentice.Blue.Chapters.5-9.png (view on web) https://github.com/user-attachments/assets/050dfcc9-e9d7-425e-aa4e-8c745d9d9405

logs for Apprentice Blue C5-9.zip https://github.com/user-attachments/files/16229447/logs.for.Apprentice.Blue.C5-9.zip

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2227588735, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKNFRVRBQYLPU5QIADTZMMT4PAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMRXGU4DQNZTGU . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[jfreiman]

Thank you, I really appreciate you trying to figure this out. Please let me know if there is anything else you need or if you would like me to test anything for you. -John

[jfreiman]

I would like to point out that not only is PGMA not able to grab the cast images from IAFD; I am also unable to match any titles (movies or scenes) to IAFD. Is this not an issue for others? Or is it just a botched install/update?

[JPH71]

It's a cloudflare issue.. Trying to sort it out xxx

On Fri, 19 Jul 2024, 02:23 John Freiman, @.***> wrote:

I would like to point out that not only is PGMA not able to grab the cast images from IAFD; I am also unable to match any titles (movies or scenes) to IAFD. Is this not an issue for others? Or is it just a botched install/update?

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2237806551, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKKLRHTOCFQ7ZGIM2ALZNBMADAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZXHAYDMNJVGE . You are receiving this because you commented.Message ID: @.***>

[jfreiman]

It's a cloudflare issue.. Trying to sort it out xxx

I'm happy you have it under control. I just wasn't sure if the images and IAFD title matching was related.

Thank you and sorry to add to the pressure, it wasn't my intent. ✌️

[JPH71]

Don't apologise, without users finding issues we would never improve!!

On Fri, 19 Jul 2024, 02:42 John Freiman, @.***> wrote:

It's a cloudflare issue.. Trying to sort it out xxx

I'm happy you have it under control. I just wasn't sure if the images and IAFD title matching was related.

Thank you and sorry to add to the pressure, it wasn't my intent. ✌️

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2237818226, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKMSHOXFX6XB22PW5UTZNBOI3AVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZXHAYTQMRSGY . You are receiving this because you commented.Message ID: @.***>

[jfreiman]

Could you provide me/us with an update on this issue? Does it appear that the plugin will be able to work around the Cloudflare issue?

[JPH71]

No updates yet.

On Sun, 28 Jul 2024, 05:13 John Freiman, @.***> wrote:

Could you provide me/us with an update on this issue? Does it appear that the plugin will be able to work around the Cloudflare issue?

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2254326083, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKLA6TBVY4JHUJBTER3ZOROU5AVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENJUGMZDMMBYGM . You are receiving this because you commented.Message ID: @.***>

[CodyBerenson]

@JPH71 Just a thought for consideration should the cloudflare issue prove insurmountable....and yes I know I'll get blasted by the purists here....but, Waybig has a considerable library of actors. 409 pages x 36 per page. Thumbor could be used to properly crop:

https://www.waybig.com/media/misc/model15031.jpg http://34.30.243.49:8888/unsafe/200x200/top/https://www.waybig.com/media/misc/model15031.jpg

https://www.waybig.com/media/misc/model14906.jpg http://34.30.243.49:8888/unsafe/200x200/top/https://www.waybig.com/media/misc/model14906.jpg

[JPH71]

You are right - However, as a filmography is not listed one cannot assuredly determine that what is found in waybig would match the film. As there is no years active field to correlate, no film titles etc... but with the nfo imported one can do... give me a week or two and i will figure something out....

Jason

On Sun, 28 Jul 2024 at 14:49, Cody Berenson @.***> wrote:

@JPH71 https://github.com/JPH71 Just a thought for consideration should the cloudflare issue prove insurmountable....and yes I know I'll get blasted by the purists here....but, Waybig has a considerable library of actors. 409 pages x 36 per page. Thumbor could be used to properly crop:

https://www.waybig.com/media/misc/model15031.jpg

http://34.30.243.49:8888/unsafe/200x200/top/https://www.waybig.com/media/misc/model15031.jpg

https://www.waybig.com/media/misc/model14906.jpg

http://34.30.243.49:8888/unsafe/200x200/top/https://www.waybig.com/media/misc/model14906.jpg

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2254525979, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKLEGD2EEOMBJA2SOHLZOTZH5AVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENJUGUZDKOJXHE . You are receiving this because you were mentioned.Message ID: @.***>

[jfreiman]

You are right - However, as a filmography is not listed one cannot assuredly determine that what is found in waybig would match the film. As there is no years active field to correlate, no film titles etc... but with the nfo imported one can do... give me a week or two and i will figure something out....

Thank you very much. I'm cheering for you!

[jfreiman]

I just want to reiterate, that for me at least, it's not just the cast images that are not loading. I have not been able to match anything to IAFD for (over?) a month now, so it's not just about grabbing images from another source, it's about all the valuable data that is also stored on IAFD including but not limited to their extensive list of Performer names and alias'.

[JPH71]

Cloudflare is effing up everything...

On Mon, 29 Jul 2024, 07:59 John Freiman, @.***> wrote:

I just want to reiterate, that for me at least, it's not just the cast images that are not loading. I have not been able to match anything to IAFD for (over?) a month now, so it's not just about grabbing images from another source, it's about all the valuable data that is also stored on IAFD including but not limited to their extensive list of Performer names and alias'.

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2255006732, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKMKKHI6G2SWFFTAS4DZOXK3ZAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENJVGAYDMNZTGI . You are receiving this because you were mentioned.Message ID: @.***>

[CodyBerenson]

I just want to reiterate, that for me at least, it's not just the cast images that are not loading. I have not been able to match anything to IAFD for (over?) a month now, so it's not just about grabbing images from another source, it's about all the valuable data that is also stored on IAFD including but not limited to their extensive list of Performer names and alias'.

Title slightly tweaked. Hopefully this is clearer for you.

[JPH71]

I understand your concern....

On Mon, 29 Jul 2024 at 12:48, Cody Berenson @.***> wrote:

I just want to reiterate, that for me at least, it's not just the cast images that are not loading. I have not been able to match anything to IAFD for (over?) a month now, so it's not just about grabbing images from another source, it's about all the valuable data that is also stored on IAFD including but not limited to their extensive list of Performer names and alias'.

Semicolon added to the title of the bug. Hopefully this is clearer.

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2255602371, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKN4CSL7P7FJS5RZVEDZOYMYBAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENJVGYYDEMZXGE . You are receiving this because you were mentioned.Message ID: @.***>

[JPH71]

Fagalicious had an issue with this too as all the other agents. I have worked on fagalicious and its hetting the vast images now. I will copy its utils.py to the the GEBI folder and test it - then roll out to the rest. After which other issues which have been raised will be looked into

[jfreiman]

Fagalicious had an issue with this too as all the other agents. I have worked on fagalicious and its hetting the vast images now. I will copy its utils.py to the the GEBI folder and test it - then roll out to the rest. After which other issues which have been raised will be looked into

Thank you!!! fingers crossed!

[CodyBerenson]

Repository updated with new Fagalicious bundle that fixes the cloudflare issues with fagalicious #308 and successfully retrieves cast photos.

[jfreiman]

I just installed the new build (stopped PMS, deleted the Plug-Ins folder contents, replaced it with the new download and restarted PMS) and I'm not able to get anything to match.

I even renamed a new file several times to match Fagalicious, but haven't been able to get it to match.

Also, based on one of your previous comments on this thread I thought you might have copied the utils.py file to other agents, so I also tried to match a previously matched video from IAFD and it wasn't able to grab data from IAFD or grab images either -- it did match with a service.

Here are the logs from my most recent attempts.
I have never used Fagalicious before so I'm not quite sure on how to properly name the titles, but I tried several names.

Fagalicious Logs.zip

[CodyBerenson]

@jfreiman Yesterday's update is solely the Fagalicious agent to fix #308. The code page shows you the last date each agent was updated. The IAFD agent is the same version since its last update, two months ago. The note I posted above was solely to let the community know that #308 was resolved, as a follow-on to Jason's note where he states he was starting with Fagalicious, then GEVI, then will roll out to the rest of the agents, including IAFD. Since I see how important IAFD is to you, I will @ you when the IAFD agent in particular has been updated.

Re: your facalicious log, it indicates you haven't chosen one of the fagalicious agent's parameters so the agent cannot properly initialize. It isn't that your match is failing; the agent itself isn't running.

[jfreiman]

Re: your facalicious log, it indicates you haven't chosen one of the fagalicious agent's parameters so the agent cannot properly initialize. It isn't that your match is failing; the agent itself isn't running.

I'll try playing with the settings for Fagalicious - it does have Indigo selected. I'll try changing it a few times and see if I can get it to stick/change.

Here are my settings incase I'm missing something else.

I'm pretty certain that Fagalicious is only an option for Scenes, Whereas IAFD includes both.... Correct?

As far as IAFD is concerned, yes, since IAFD is 99% consistent with its Title naming within each Studio/Distributor I don't need to actually look up each Scene title on the hosted site (WayBig, Fagalicious etc) before adding them to my collections.

I can add them and come back later and see which of the few titles didn't match and need fine-tuning/re-naming. This is especially true for scenes as on the alternate blog sites, the names are very long and in most cases include the actor names and other stuff which makes my collection look messy - IMHO. ;)

Thanks for making PGMA an asset to people who enjoy collecting and sharing there collections on Plex.

BTW. In the past few months I have over 2,400 scenes and movies (many have been actually matched to IAFD) so that when and if PGMA's IAFD's service is reconnected will be able to update without me having to make any changes to the files.

Thank you again for all your hard work over the years. I truly appreciate what you have created and/or built up.

Before PGMA I was literally adding mp4 metadata to ALL my videos by cutting and pasting every detail. While I still use the mp4 metadata fields, they are mostly reserved for standardizing on Studio Names, logging the website a videos is matched to or correcting alias names against IAFD.

[JPH71]

Sorted!!!

[fourstix]

Hooray! Many Thanks, Jason!

[JPH71]

Thanks for the appreciation xxxx

On Sun, 11 Aug 2024, 22:08 fourstix, @.***> wrote:

Hooray! Many Thanks, Jason!

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2282874593, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKJ5C3OFXXGT64AKMXTZQ7ACZAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBSHA3TINJZGM . You are receiving this because you were mentioned.Message ID: @.***>

[jfreiman]

Sorted!!!

???? - Hmmm.... I guess I'll watch for the next repository update.

[CodyBerenson]

@jfreiman Hi John, all of the fixed agents, including IAFD, have been pushed to the repository. Cheers.

[JPH71]

??? Really - Why the disdain ???

On Mon, 12 Aug 2024, 04:51 John Freiman, @.***> wrote:

Sorted!!!

???? - Hmmm.... I guess I'll watch for the next repository update.

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2283030825, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKMTVDDX3WF3DJMEF4TZRAPLTAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBTGAZTAOBSGU . You are receiving this because you were mentioned.Message ID: @.***>

[jfreiman]

??? Really - Why the disdain ??? … On Mon, 12 Aug 2024, 04:51 John Freiman, @.> wrote: Sorted!!! ???? - Hmmm.... I guess I'll watch for the next repository update. — Reply to this email directly, view it on GitHub <#306 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKMTVDDX3WF3DJMEF4TZRAPLTAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBTGAZTAOBSGU . You are receiving this because you were mentioned.Message ID: @.>

Is that distain? Sorry. I didn't mean it like that.

I read your comment and checked to see if there was an update and there wasn't one (5 hours after your post) so I didn't understand what had transpired....

I see there is an update to download and install and I am VERY HAPPY!!

Thank you!

[JPH71]

I sent the updates in very late last night. Cody has to do some testing before he updates the repository, just in case. The sorted comment was to let you know that the issue had been resolved and the update was on its way. And I was cranking AF this morning..... ??? in comments I have to admit can sometimes be a trigger...

I am trying to sort alternative sources for cast, sort out issues with linux systems that handle symbols differently i.e 1/2 and squared issues, and grayscale dead cast members photos, this should solve at least 5 of the 8 outstanding issues.

Are you using Linux? let me know... as I can't remember at this instant....

Cheers and sorry for overreacting

Hugs xxx

Jason

On Mon, 12 Aug 2024 at 09:19, John Freiman @.***> wrote:

??? Really - Why the disdain ??? … <#m_3700350459996609350_m8054089271782003013> On Mon, 12 Aug 2024, 04:51 John Freiman, @.> wrote: Sorted!!! ???? - Hmmm.... I guess I'll watch for the next repository update. — Reply to this email directly, view it on GitHub <#306 (comment) https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2283030825>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKMTVDDX3WF3DJMEF4TZRAPLTAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBTGAZTAOBSGU https://github.com/notifications/unsubscribe-auth/AKI3AKMTVDDX3WF3DJMEF4TZRAPLTAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBTGAZTAOBSGU . You are receiving this because you were mentioned.Message ID: @.>

Is that distain? Sorry. I didn't mean it like that.

I read your comment and checked to see if there was an update and there wasn't one (5 hours after your post) so I didn't understand what had transpired....

I see there is an update to download and install and I am VERY HAPPY!!

Thank you!

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2283262629, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKOT56G22G4RZYQXHFDZRBOX5AVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBTGI3DENRSHE . You are receiving this because you were mentioned.Message ID: @.***>

[CodyBerenson]

@JPH71 You didn't overreact. Perhaps @jfreiman doesn't have the context for what we do and how we do it. If he did, he'd likely not come across so bratty (in the traditional sense vs. Charlie XCX). So, respectfully, I'll take a moment to provide some context.

1. This is not our priority. We each have lives, meaningful employment, friends, family, partying commitments. This is a hobby…a part-time diversion. We are not reimbursed for our efforts inspite of financial risks that we take (more on that later).
2. Since this is open source and entirely uncompensated, there should be no expectation of a Service Level Agreement, response time, etc. There is no such thing as a PGMA emergency.
3. Perhaps you aren’t aware, but scraping metadata and images can be a violation of site’s terms of service, and our agents are at the mercy of various indexing sites who are constantly making updates and employing more and more sophisticated means to constantly try to protect themselves from cybersecurity intrusion (such as cloudflare). Do you realize how frustrating it is for Jason to fix something today only to have it break next week?
4. This community is just a couple of Information Technology guys on different continents in different timezones. We have been at this for nearly five years. We decided early on to leverage a few grounded software development life cycle (SDLC) disciplines. The most important principle is separation of duties. This is for the benefit of the entire community. The artist/genius that creates the code (Jason), is not the guy that promotes the code to the community (Cody).
5. Once code is developed and tested by the Jason, it gets tested by others (Cody, and additional resources Jason leverages from relationships he’s built along this journey) before updated code is shared with the community. Testing is one of the most important parts of the SDLC. Testers have operating systems different than Jason, scenarios and parameters different than Jason so that the code is given some robust poking before it gets to you.
6. As indexing sites raise their cybersecurity protection game to counter threats from folks that violate their terms of service, the solution to keep the agents working becomes VERY complex and very time consuming (both in terms of the time it takes to research far and wide a viable solution and the time it takes to perform surgery to all of the agents to incorporate the solution). Are you aware that there are other agents out there (e.g., AdultMediaManager) who have been unable to resolve the IAFD Cloudflare issue? I imagine this is the source of Jason's reaction (and not overreaction).
7. Are you aware that at least 4 times a year I have to provide my private credit card information to establish cloud hosting services to host Thumbor, which is the service that our Scenes agents use to properly crop images since Plex’s native functionality is far from an elegant solution?
8. What do we ask in return? Compensation? Never. Gratitude? Never. We basically ask for two things: a. When you or anyone has an issue, please provide a minimum set of information (e.g., Filename, URL of indexing site, specific agent log) to properly triage, assess, and resolve the issue b. For the community to have a little patience, given the above. A few hours to test and push updated code, on a Sunday Funday of partying and enjoying the closing ceremonies of the Olympics? Sorry, not sorry.
9. The hypocrisy of this bratty tirade is not lost on me. Cheers.

[fourstix]

Retired IT guy here. Please take this suggestion in the spirit of trying to be helpful and contribute to the group.

Perhaps a dev branch would help reduce the amount of work that has to be done. Then the developers can promote code to the dev branch without affecting the stable code in master. Volunteers could download and test the developed code on their own before it's promoted. After a period of testing, the stable dev code could then be merged with the master and a release created.

This is actually a lot easier to do in GitHub than it sounds from my description. I know from experience, it's actually less work than managing promotion manually into a single master branch. Especially the work that must be doing in tracking, staging and promoting code changes. Plus, it will reduce the stress level all around.

I will be willing to help with setting this up, answer questions, etc. Feel free to contact me on GitHub or through e-mail.

[CodyBerenson]

Hiya @fourstix. Suggestion very much appreciated and accepted with the most positive intent which I truly believe you gave it. Please accept the same positive intent in the reply. Very much active IT guy here and intimately familiar with devops and modern software delivery; its concepts, tools and practices. Re: your suggestion, a few years back I did regularly enable a branch for the community, making the code available “early” for folks to test. Zero feedback. Zero evidence that anyone downloaded or even looked at it. At the end of the day, offering a testing period just prolonged the release as I twiddled my thumbs waiting for anything other than crickets.

I don’t believe there is a “problem” here requiring process reengineering to fix. Dev Ops and CI/CD practices and tools won’t fix impatience. We have no stress level (other than the self-imposed stress to find solutions to complex problems). We have no version control issues.

You used the word “developers”…plural (not sure if you meant to), so I want to provide context. Over the years Jason and I have discussed the model he’s most comfortable using. He is comfortable with folks taking branches of the code and doing what they will. He is NOT comfortable with folks contributing code that doesn’t conform to his standards and framework, or align to his vision. The couple occasions that we opened that door were not good experiences. He frequently casts a net welcoming coding suggestions and innovative suggestions for problem resolution. It isn’t unusual that folks in the community provide excellent recommendations. We’d just rather the recommendations not be in the form of a pull request that a community member expects to be acted upon and merged…(and the resulting vitriol if the request is not honored). I think we owe it to Jason to provide him the environment optimal for him.

If Plex hasn’t deprecated the python agent/scraper framework (see discussion #315) by the time I retire in 2 years and head for the south of Spain, would you mind tossing your hat in the ring to take over as community administrator? Sounds like you would be an excellent candidate partner for Jason. You probably have better patience with impatience than I do. LOL. Hope you'll give it some thought?

Thanks!

[fourstix]

Hi Cody, Sure. I'll be glad to help out anyway I can. I've administered other GitHub sites, so I would be happy to help with this one.

I said developers just out of habit. Whatever process works for you and Jason is, of course, what's best.

[fourstix]

By the way, I should have mentioned that this is now working completely fine for me with the latest agents update. All of the agents I use (AEBN, GayHotMovies, GayEmpire) are now adding head-shots from IAFD.

[ms52538]

I have seen general improvement in the agent matching - BUT ... IAFD continues to return blank headshots with a red X under the actor's name. I created a new library and am using the Adult Agent. There seems to be an issue with IAFD matching, where the title does match, but the correct actors are not showing up at all, i.e. (RawFuckClub.com) - Andre Cruise Is Getting Fucked By Kyle Fox Emiliano Terra and Felipe Hervera (2023). it matches but no actors headshots or names appear (but it is matched). If I change the title to become "(RawFuckClub.com) - Andre Cruise Is Getting Fucked By Kyle Fox Emiliano Terra and Felipe Hervera (2023) - [Andre Cruise,Emiliano Terra,Felipe Hervera,Kyle Fox]" and refresh then all 4 actors names appear but both Emiliano Terra and Felipe Hervera are both blank and with red X under their names. Then in a slightly different case of IAFD matching the title of (RawFuckClub.com) - My First Orgy (2023) all 3 of the cast's headshots are blank BUT they each have a green checkmark under their names. If I change the title to become (RawFuckClub.com) - My First Orgy (2023) - [Adam Caste,Emiliano Terra,Felipe Hervera] and refresh then all three actor's headshots appear and have a green check mark under them. If I click on Felipe Hevera's headshot, go to the all titles under his name, then go back to the previous title, his picture disappears. If I now go to another title "3 Cocks For 1 Hole", Emiliano Terra's headshot has a picture, but with a red X below it. The behavior is odd; it appears to be trying to match but having random difficulty populating pictures from one moment to the next. Can you explain in general terms what is going on with the headshot lookup under each instance of opening a title? Maybe the system doesn't like guys from Brazil. lol

[CodyBerenson]

@ms52538 Please open a proper issue and include your logs. Thanks.

[CodyBerenson]

@ms52538

First thing I do when I test IAFD is look at the TAGS table on the backend, to confirm that there are no URL for the actors. After the match, I recheck to confirm that the URL has been populated.

I have found that the Plex front end can be slow to update the actor's image. This is why I confirm on the back end. A green checkmark and no headshot is illustrative of this situation. And, sure enough after a little patience, the image appears.

Perhaps timing isn't the situation for you. Perhaps you're still receiving 403 errors. However, without your logs, it is impossible for Jason to triage.

[ms52538]

Sorry - I truly thought #306 WAS the open issue to address headshots not appearing.

Can I inquire how the population of a an actor images populating works? I'm operating under the assumption/impression that it does a one-time look-up per source (i.e. IAFD - Felipe Herera, GEVI - Felipe Hera, etc.), populates the agent folder, and from the on Plex reads from the local file / or cached information (presuming it caches - I don't know).

After creating the new library yesterday, I revisited it this morning. Some of the titles were getting a return of IAFD 403 error. I refreshed those and the issues mostly cleaned up.

Can I also inquire as to what table viewer you are using that can show that information?

I'll be happy to go open another Issue, and populate the heck out of it with logs and screen shots. Just wanted to do a quick reply. Thanks for your time - I always appreciate it!

[CodyBerenson]

@ms52538 Yes, please open a new issue and provide the title, log, etc. As far as we are concerned, and in the absence of new specific cases with the accompanying logs, the issue is closed. The IAFD agent successfully scraped your example, as illustrated above.

Jason may be able to speak to how Plex's native agent framework consumes metadata it culls from the scraping process (which btw is targeted for decommission, see #315). There is probably a vast amount of info on the workings of the framework in Plex Forums (to prove/disprove your assumptions). Our PGMA 3rd party agents, while super complex, functionally superior, and absolutely fabulous, are similar in architectural principle to all the other multitude of 3rd party agents out there (i.e., initialize, query, match, scrape, pass on the data for update in the backend).

On the backend, regardless of type of media or ours vs. any other 3rd party agents, Plex adds the film/scene into the metadata_items table. Then for each actor in the metadata string for that film/scene, Plex checks to see if the actor already exists in the tags table. If the actor doesn't exist, it will do an insert, including any thumb url metadata in the metadata string from the scraping process. If the actor does exist, it will do an update, including any thumb url metadata in the metadata string from the scraping process. Then, it will create the relationships between the film & library (i.e. metadata_items) and the actors & collections (i.e. tags) by inserting rows in the taggings table (there are many other tables involved, but these three are key).

A practical example: let's say you used one of our agents to match/scrape a film starring the OG Michael Lucas himself. Our agent(s) queries/matches/scrapes/passes the metadata scraped to and Plex then starts updating the backend, including metadata (e.g., IAFD's head shot URL) for Michael Lucas in the Tags table. Success! And then you had an obscure Michael Lucas gem that isn't anywhere in any of our agent indexing sites, but you found it on ThePornDB, so you decide to use ThePornDB's 3rd party agent (with its own file naming requirements) to match/scrape this gem. After the PornDB's agent feeds Plex the metadata, all Michael Lucas metadata that was just populated previously is now updated (i.e. overwritten) by Plex in the tags table, including thumb url (FYI, TPD uses sources other than IAFD for their actor headshots). Rinse and repeat... since metadata is refreshed every single time a library item is accessed from the library. As you peruse Michael Lucas films in your library, Michael Lucas's backend metadata is updated each and every time as you click or query any of his films from the Plex GUI.

Plex uses SQLite as its backend. To query, I use DB Browser for SQLite. Most metadata in the database aren't updateable and you'll get errors if you try to do most inserts or updates, so I wouldn't try. Further, I'd take a copy of the database first before you do anything, just incase you're able to corrupt it.

[ms52538]

Cody, thank you for that! It literally explained so much that it answered most (if not all) of my curiosities in how things are working within the lookup of the agents and Plex. Again, thank you - for the in depth dive and simplification of the process. I know it take 15 minutes out of your day, and I appreciate it! :) BTW, nuggets of explanation and insight for things such as this are very valuable. I do wish our our GitHub had a sub-section for items like this that provide this kind of insight and educational material for users in our sphere of PGMA. The pending upcoming changes to Plex as outlined in Issue #315, it may be such a thing as us "freezing" our version of Plex to avoid losing the PGMA functionality, as the simplest course of action until such time as potential possibilities for other solutions (if any) may come to light for us. And if I read correctly you are getting close to retirement and certainly your attentions will be prioritized elsewhere in the not too distant future.

I took a bit of time to do a deeper dive in to my Plex system and setup last night. As I reviewed things, I attempted to take little steps of action to isolate the problem(s), and one of the glaring issues I discovered was the way I have my libraries set up, resting atop my mixed ecosystem of content, including active torrents and dead/non-active torrents, resting atop my Windows file system structure of folders and directories nested across multiple disks .... I discovered multiple files, belonging to various active torrents, that were the same TITLE, but with differing torrent hashes, nested in different locations. I believe these were in conflict with each within the Plex tables; i.e. I was matching file associated to title "X" with IAFD, and another different hashed file associated to title "X" with GEVI scenes. Certainly there is conflict within the table lookups. A few things came to mind; the first being I have to isolate active torrent files that duplicate the same title so only one exists within Plex and the others are moved outside of Plex, and then I began thinking "what is the best configuration" of Plex to have matched Titles in one library configured as movies, and then move unmatched files to a library configured as personal videos.... and then I'm thinking about Plex's search issues if I configure multiple libraries vs. if I have just 2 for my entire vast collection: matched (movies library) and unmatched (video library). And then I wished we had a subsection of PGMA in GitHub where we could have these sorts of conversations with other users because I believe this is outside the scope of what the user guide is specifically tailored to do. And then I was struggling with notion that I want to specifically have actors names listed in my Windows file system for the titles but I'm finding in some cases (such as IAFD for example) where actors have multiple names they use, which do you choose or not choose to list - because in my tests it was clear that some actors "credited names" were conflicting with scene matching (i.e. Polo, PoloXXX, PoloXXXfans - or Heath Halo, Heathen Halo and so on). If you are ok with it, I'd be quite happy to go over to the Discussions tab and blow it up a bit with these sorts of questions. I'd also love to find out how other Windows users are configuring their file system environments, if they are incorporating active torrents or not, and the sort (and certainly users on other file system types may have similar challenges, I only to speak to Windows as that is what I am familiar with). What I'd really love to see is the ability to friend other members in PGMA, thru Plex, and see how others have their setups. John F., is the only person I have developed a connection with, and envy how he manually populates his files to address so many of these challenges and am mind-boggled the time and energy he puts into it as he has provided me his tools and methods and I just cannot invest that sort of personal time into my collection but wish I could). I don't know ... maybe I'm alone on an island with all of this bubbling around my head, maybe others can relate. I have tried to be respectful of what GitHub PGMA is intended for with the mechanics and functionality of scraping and matching content. But man, I sure would love to tap into other people's heads who are members.

Thanks again, Captain, for your time - it is always, Always appreciated!

[JPH71]

Start a discussion...

On Wed, 14 Aug 2024, 17:05 Marcus, @.***> wrote:

Cody, thank you for that! It literally explained so much that it answered most (if not all) of my curiosities in how things are working within the lookup of the agents and Plex. Again, thank you - for the in depth dive and simplification of the process. I know it take 15 minutes out of your day, and I appreciate it! :) BTW, nuggets of explanation and insight for things such as this are very valuable. I do wish our our GitHub had a sub-section for items like this that provide this kind of insight and educational material for users in our sphere of PGMA. The pending upcoming changes to Plex as outlined in Issue #315 https://github.com/CodyBerenson/PGMA-Modernized/discussions/315, it may be such a thing as us "freezing" our version of Plex to avoid losing the PGMA functionality, as the simplest course of action until such time as potential possibilities for other solutions (if any) may come to light for us. And if I read correctly you are getting close to retirement and certainly your attentions will be prioritized elsewhere in the not too distant future.

I took a bit of time to do a deeper dive in to my Plex system and setup last night. As I reviewed things, I attempted to take little steps of action to isolate the problem(s), and one of the glaring issues I discovered was the way I have my libraries set up, resting atop my mixed ecosystem of content, including active torrents and dead/non-active torrents, resting atop my Windows file system structure of folders and directories nested across multiple disks .... I discovered multiple files, belonging to various active torrents, that were the same TITLE, but with differing torrent hashes, nested in different locations. I believe these were in conflict with each within the Plex tables; i.e. I was matching file associated to title "X" with IAFD, and another different hashed file associated to title "X" with GEVI scenes. Certainly there is conflict within the table lookups. A few things came to mind; the first being I have to isolate active torrent files that duplicate the same title so only one exists within Plex and the others are moved outside of Plex, and then I began thinking "what is the best configuration" of Plex to have matched Titles in one library configured as movies, and then move unmatched files to a library configured as personal videos.... and then I'm thinking about Plex's search issues if I configure multiple libraries vs. if I have just 2 for my entire vast collection: matched (movies library) and unmatched (video library). And then I wished we had a subsection of PGMA in GitHub where we could have these sorts of conversations with other users because I believe this is outside the scope of what the user guide is specifically tailored to do. And then I was struggling with notion that I want to specifically have actors names listed in my Windows file system for the titles but I'm finding in some cases (such as IAFD for example) where actors have multiple names they use, which do you choose or not choose to list - because in my tests it was clear that some actors "credited names" were conflicting with scene matching (i.e. Polo, PoloXXX, PoloXXXfans - or Heath Halo, Heathen Halo and so on). If you are ok with it, I'd be quite happy to go over to the Discussions tab and blow it up a bit with these sorts of questions. I'd also love to find out how other Windows users are configuring their file system environments, if they are incorporating active torrents or not, and the sort (and certainly users on other file system types may have similar challenges, I only to speak to Windows as that is what I am familiar with). What I'd really love to see is the ability to friend other members in PGMA, thru Plex, and see how others have their setups. John F., is the only person I have developed a connection with, and envy how he manually populates his files to address so many of these challenges and am mind-boggled the time and energy he puts into it as he has provided me his tools and methods and I just can invest that sort of personal time into my collection but wish I could). I don't know ... maybe I'm alone on an island with all of this bubbling around my head, maybe others can relate. I have tried to be respectful of what GitHub PGMA is intended for with the mechanics and functionality of scraping and matching content. But man, I sure would love to tap into other people's heads who are members.

Thanks again, Captain, for your time - it is always, Always appreciated!

— Reply to this email directly, view it on GitHub https://github.com/CodyBerenson/PGMA-Modernized/issues/306#issuecomment-2289066096, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKI3AKLIRGU46ULSWX4SBNLZRNW5BAVCNFSM6AAAAABKPUINFSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOBZGA3DMMBZGY . You are receiving this because you were mentioned.Message ID: @.***>