[katyz]

#All my links can be viewed by users of the forum(remove XX for TT), I'm allowing it because without them nobody will understand anything or help, besides that it doesn't contain anything personal.
#This topic may be deleted after a while but let's discuss the issues.
#I hope the members comment if they are interested in the Devs applying the corrections/suggestions.
#For members, don't post help links, or paths to github, codes etc, we're talking about fixes and internal improvements in the program, the question here is to work on the car's engine and not on the seats or sound.
#And can you tell your experience if you had these problems too or are interested in the improvements, just try not to escape the scope of the questions.


*TicketSupport n.ºVSW-512-47082 ,passing my support to public.*

My interest here is to contribute to Jdownload2 by pointing out flaws or things you can improve on your system.
Your program has many things that I consider as bugs, production errors, things poorly done, poorly thought out, poorly developed, missing, etc.
I'm not trying to judge you, the program is amazing and free.


BUG_A: VideoLink : hXXps://mega.nz/file/qUN3HLxI#xivtdG36bNqPCd7UsuCc_mZsZ2YKc98df4N7unmZZrU
Yes it's a problem, it's something badly done, the open link is not the real link, you didn't add a button to take the user to a real download link.
I clicked on the image link and it should open the image link and not the home page link, and this happens with everything, it's a problem with the program.
Links are often obtained in bulk by tab eg: hXXps://mega.nz/file/XNUXUKxD#2WuFzYs7mmWi7ezeFSnaOrAf9f1_IT3buPZD2KSJsE0 , such as those on XVideosRed sites and among many others, I will not expand further on this issue, I think you understand the improvement that needs to be made.
The simple solution is to add a button that leads to the real link, which the user clicked on, the user clicked on the image, opened the image link, clicked on the video, opened the video link, I'll help you with an example image:
A simple SSLink solution--> hXXps://mega.nz/file/SUkAHI7Q#_lr8s1dBDdR9NdkFfen50aCHCyXlfExd-FnnEljTOSk


BUG_B: VideoLink : hXXps://mega.nz/file/KVtijBLY#BxUNYZuQ8LCgQt9fCmsObBKtRvt5aMYAr41mKIe10qQ
You don't need logs or premium account to scrape twitters, scrape a few twitters and you'll run into trouble in no time.
I don't understand why you always make difficult, you saw in the video the different errors in different twitters, it's not one or two, there are several, I also explained to you where the problem occurs.
I'll send you some links to bump into the problem, best log is for the dev to see the problem for himself, if it's so simple to replicate the problem, why don't you want to replicate? it doesn't take 10 minutes.
Scratch this twitter: hXXps://twitter.com/CoryChaseXXX
This was one of which I had more errors, which I am reporting.


BUG_C : hXXps://mega.nz/file/mdN0QbYY#vOpb9th8qU6g5xoSo8OFfG_cSb2bEQD5_B8ksmOoeI4
Your program creates imaginary virtual folders within it to relocate links that already exist in physical folders, this makes no sense.
You have a large user base that complains about this problem, maybe you think it's not a big deal, use X or Y , or program some code and do whatever you want your way.
But that's not my way, that's the heart of the program, that is, a basic function.
I'll give you an example:
Let's say you add 100 links from twitter, xvideos etc, then some time passes and you want to add the links again, the folder already exists physically and yet +100 virtual folders are created internally in the program with the same name, wrong way of thinking, do this only if the physical folder is not found.
I'm suggesting you something to improve and simplify for everyone, you have a large user base complaining about this issue.
Resolving it will get a lot less complaining users on the forum, unless you love wasting support time repeating the same thing a thousand times to several newcomers to the program.
And if this, for some reason internal to the codes, is complicated to solve, then I'll leave a solution:
A maybe possible solution --> hXXps://mega.nz/file/fZtVQART#LCWds3D8bBDUSd1zS8RgGJilv2MZSk099m_eeecMqPI


BUG_D : VideoLink : hXXps://mega.nz/file/7J8hVT6a#DKjrYNoSwNnEwCiR-2N8eo67HpRNSHEHdK-NlpC97YU
Windows has a limit of 257-260 characters, so some videos I try to download and the names are too long,
they hang at the end 94% 97% 98% etc, that is close to the moment of the conversion.
Yes, I know that in Windows 10 there was a new way to increase this limit, but then you either correct the problem or end support for all Windows below 10.
In addition, many users by default do not activate the character enhancement system in Windows 10, and Windows 7 and 8 do not have this possibility, it is not just the path I am talking about, but the file name.
So I'm going to suggest an improvement that will help solve 99% of this problem.
A simple solution --> hXXps://mega.nz/file/3MljBCKI#_NyAu8NN_hk1nNQjA2bx9c-CIWZ6O60IR8nO3MfUO-A

BUG_E :
you forgot to put "Unknown error" because many failures I see do not make sense with what is said in the error message.
A simple solution --> hXXps://mega.nz/file/2BtSDboZ#bSIaagcMi-cM5Gz-AfVotgq2DM1k6kZ_gxOijU3WG_Q


BUG_F Crash (I don't know if there's a solution or not, but I'm reporting it.): hXXps://mega.nz/folder/uM1GFbgD#zFSwyjfB1hBE3sPMVZEhGA


Sorry to be honest, but better honest people than fake ones.
I'm a fan and user of your program for many years since the time of J1, what I can help to improve I will always report, I hope you can fix and improve these issues, thank you,and sorry if I sounded rude at some point, that's not the intention.

Full Report Folder Link : hXXps://mega.nz/folder/fQUDEKaJ#c6dFcQWQpdLynZvZAbmSwg


---------------------------------------------------------------------------------
I received a rather senseless and discouraging response from the developer(pspzockerscene) for each point, so I think it's better to continue the topic with issues posted and reported publicly by me. This way, we can discuss the matters here with members too.
#I should point out that the dev has always been responding to all tickets quickly, and with respect, although I sound rude I'm not.
#The program is amazing and super old always with updates and improvements what matters is that there is someone watching over it even today, we are all in the same boat trying to help, each one has a different tone of helping.:thumbup:

I just want to briefly address the developer about some points in their responses, with my counter responses nicely summarized below:
Dev's private answers are with him, these are mine in relation to some commented points.

BUG_A: Opening a link that returns an image of a duck today and opening the link again a year from now to get an image of a bear is something common. Just like serial resets or link changes, it happens a lot, and that doesn't mean you should remove the option to open the actual link from where the video or image was downloaded. If it's not possible to generate a link for a certain type of file, just treat it as an error and do what the program already does today, which is opening the home page and not the clicked file's link.

BUG_B: The public link you sent me doesn't make sense. I sent you a private link, and yes, you should proceed to scrape it and, obviously, be logged into your JD account to test the bug I mentioned.

BUG_C: Yes, some people may like it while others may not, so it was a request for an ON/OFF button, let whoever wants to use it. But it's something that should be part of the program, not a separate code in some part of the forum or a separate program; this is very basic.

BUG_D: You sent me two links, neither of them resolved the issue, and they were not within the scope of the question. The question was about long characters.

BUG_E: Regarding the image (Coder Missing-UnknownError), in summary, I just commented that you should handle error messages better.
I'm not going to question your "fancy answers" because I don't want to appear rude with acid answers.

BUG_F: Regarding crashes below Windows 10, the logs are there. If dev Jiaz wants to take a look, I didn't generate them for no reason. In the end, if there's no resolution, it's better to declare that support below Windows 10 has ended.

[pspzockerscene]

I was taking my time to provide a detailed reply in said ticket and I was quite confident that my reply would be helpful but it looks like it wasn't.

Either way, taking it into our forums may be the better idea so other users can join as well.

Originally Posted by katyz

#This topic may be deleted after a while but let's discuss the issues.

We're only deleting posts which violate our forum rules.
I can't see any violation here.

I'm a fan of full transparency so I will include a full quote of the reply I sent you in your ticket since it does not contain any private information.
I've only removed the protocols of some URLs so that all users can see them. Apart from this, the text is 100% the answer I sent you in said support ticket.

Quote of my reply in ticket VSW-512-47082 01 September 2023 03:02 PM:

Hi again katyz,
first of all thank you for the detailed report and also:
Same as in the ending of your post applies to my answer: Some of it may sound rude but it's not ment to be rude!

Also I will answer some of your questions with links to forum threads where those were answered already in more detail [specifically about BUG_D].
Please read them and only come back to those topics if there are open questions.

First about the preamble:
- I didn't want to treat you like a fool
- Speaking only for myself: Indeed I've rarely used JDownloader myself for the last few years so this point is actually valid. Still this doesn't mean that I'm not coding on behalf of the user: I know how JD has to feel/behave and I'm proactively improving JDownloader myself if I see something I dislike after using it myself.

I did try to understand your request- and answer accordingly.
The "this is not a bug" answer was not there to silence you or to simply provide some kind of excuse.
So here we go again:

BUG_A:
By default, JDownloader will return the "source"- or "container" URL:
That is the URL that the user has added.
There are multiple reasons for this default behavior.
(Please also keep in mind that single plugins can behave differently - the behavior that I'm describing here is valid for the majority of our plugins though.)
I will only provide some of the reasons here:
- For many links it is impossible to go back to the source if you open the individual link (for example a pastebin link -> leads to various youtube links -> Reference to pastebin link is lost if you open any of those YT links in browser)
- In a lot of cases the direct links which go to the content can expire -> Returning them may leave the user inable to reach that content (for example: direct URLs to video streams of youtube videos or pornhub videos)
- For some elements it is not even possible to provide a direct URL to the content because this does not exist e.g. twitter .txt items
Luckily JDownloader has settings for this and this can be adjusted.
Those are explained here:
https://support.jdownloader.org/Know...iew-added-urls
Adding to this:
In my last reply I did tell you that we can expose those direct-URLs for twitter.com items if you want this.
In between your first- and this request another user did request this feature so now all you need to do is to go to the settings in that linked article and move "Data" or "Custom" to the highest position.
After doing so, you will get the direct URLs of the content when doing rightclick -> Settings -> Open in browser.
Adding to that:
You can view a list of all available URLs via:
Rightclick -> Settings -> Show download URLs -> Doubleclick in the "Download from" column.
I consider this one "Solved".

BUG_B:
We've never purposely tried to "make it more complicated".
When twitter was switching to "You need an account to view anything", we did the same shortly after simply because adding errorhandling to detect this case based on their http answers was not so easy.
References:
techreport.com/news/twitter-tightens-access-login-needed-to-view-tweets/
This change was then quickly reverted by twitter:
techcrunch.com/2023/07/05/twitter-silently-removes-login-requirement-for-viewing-tweets/
--> And some time later also by us in our plugin code.
In fact you can even find leftover code- and comments about this in our "TwitterComCrawler.java" class if you search for "ACCOUNT_IS_ALWAYS_REQUIRED".
I'll leave a link to our source code in the end of this message.
Now while typing this I can see that twitter has only partly removed that limitation:
The profile you've provided for example, cannot be viewed without being logged in - simply open it in a private browser window to check my findings:
twitter.com/CoryChaseXXX
...but other profiles can be viewed without account for example:
twitter.com/50cent
JDownloader cannot just circumvent such serverside limitations.
Now about the error "File not found" that is visible in your video:
That errormessage must not be related to these twitter serverside changes but it can be related.
If the content really is online, this error should never happen.
I've just downloaded (all?) 131 media items [size ~95,05MiB] of that profile and I didn't run into a single error.
This means if you want me to investivate this further, I will need logs or a way to reproduce this issue 100% in my system.
Log instructions:
https://support.jdownloader.org/Know...d-session-logs

BUG_C:
So you would like JD to always auto-merge same packages with same names inside the Linkgrabber?
First of all:
I've never seen a "large user base complaining about this issue".
Do you have some links for me?
Secondly:
I do understand your request but you need to understand that this is more complex than that:
What if you e.g. add a cloud folder from google drive called "test" and then you will crawl a twitter profile called "test"?
Would you want both results to go into the same package? I wouldn't!
To effectively do what you want, JD would need to know in beforehand that "all items in package with name 'test' come from container URL 'twitter.com/test'" -> Only auto merge if that is also the case for the folder you crawled later too.
Another quote:
"unless you love wasting support time repeating the same thing a thousand times to several newcomers to the program"
--> I seriously didn't ever notice that. Again I'd like to ask for some links - maybe I've overlooked something here. I know pretty well, which issues or "missing features" get reported frequently and "BUG_C" is not among them.
All in all I need to say your suggestion is quite nice.
I'll see if there are existing threads/tickets for it and if not, I will create an internal ticket for you.
I still want to mention that, although it's a bit more complicated, you could already realize this feature using an EventScripter script:
EventScripter forum thread:
https://board.jdownloader.org/showthread.php?t=70525
EventScripter help article:
https://support.jdownloader.org/Know...event-scripter

Comparing this answer with my last one regarding "BUG_C":
I'd still clearly say this is not a bug but it is a good feature suggestion!

BUG_D:
This one is a mix of two problems:
Problem one:
xvideos.com/xvideos.red -> Issues with some HLS/FFmpeg downloads:
https://board.jdownloader.org/showth...en+long&page=2
--> Possible workaround: https://board.jdownloader.org/showpo...6&postcount=28
Problem two:
General issues with too long paths/filenames mostly on Windows systems
TL;DR:
- We are aware of this topic "too long file paths and/or filenames"
- We know that some of our plugins do provide quite long default filenames so we do know that quite some users are running into this problem
- This will be addressed some time in the future but not now
References, related support article and possible workarounds:
https://support.jdownloader.org/Know...load-directory
https://board.jdownloader.org/showthread.php?t=74880
https://board.jdownloader.org/showthread.php?p=517768
Specifically about your solution suggestion (video):
- We're already trimming filenames
- We already remove non-allowed characters
- At this moment the main problem is just too long names and not "invalid characters inside filenames"
Related forum posts and tickets:
Also a user did just recently contact us regarding our character replacer or let's call it "our way to clean filenames at this moment":
If you want to go deeper into it:
https://board.jdownloader.org/showth...399#post523399
-> Read from post #160 until the end.
We've even started to work on this one - there is already one advanced setting for it available but so far we've only unlocked the functionality in our dev/IDE version.
Again: If you're interested in this: Settings -> Advanced Settings -> GeneralSettings.filenameandpathcharacterregexreplacemap
Once the internal preparation of- and refactoring of this is done, we can finally work on the "auto filename shortening" feature and yes, it is planned!
-> But no, we will not provide any ETAs - we're never doing this.

BUG_E:
I'll split this up in two:
First:
So you want to be able to filter by "unknown" errormessages?
There is no internal type of errormessage defined as "Unknown".
Yes we have different types of errormessages but your post does not clearly mean which type or which exact error-text you mean.
My guess is you mean "FATAL" Errors (errors that lead to permanent stop with varying error-text).
I guess we could add the category "Fatal errored downloads" or similar.
In your videos and screenshots though I did mainly see the errors "Skipped - account missing" and "File not found". None of those are from category "FATAL" and I wouldn't consider any of these to be an "Unknown error".
Regarding your requested feature, the ability to search for this- or other types of errors:
Our developer Jiaz can go more into detail here if you want him to.

Secondly:
We're trying to provide meaningful errormessages.
A bunch of plugins have some "Unknown error" messages at some places but I don't see the need to refactor those at this moment.
If you encounter non-meaningful errormessages, please let me know about each one you encounter and I'll work on improving this.
Please note that JDownloader is still a plugin based downloadmanaer so if a website we support does not provide an API and that website is changing, our plugins can spit out non-meaningful errormessages and there is not much we can do this.
Most commonly thought, you should see the error "Plugin outdated" if something really goes wrong.

BUG_F:
I'll skip this one for now.
If relevant:
Our developer Jiaz can go more into detail here if you want him to.

Small sidenote: JDownloader is open source so feel free to explore our code or even contribute your own code.
Link:
https://support.jdownloader.org/Know...up-ide-eclipse

"Sorry to be honest, but better honest people than fake ones."
--> All good

I will now go offline for the weekend so do not expect another answer of me until next monday.

Best regards,
pspzockerscene

Me or other team members will come back to your questions later but for now this post of mine is just here to add further context.

[pspzockerscene]

Looks like the others won't fine time for this thread any time soon so it's me again.

I would like to finish each topic before starting with a new one so at this moment I will solely concentrate on "BUG_A".

BUG_A: Regarding twitter:
As explained in my answer, what you want is already possible.
Adding to my text answer, here is a screen recording showing that it is possible and how:
mega.nz/file/v6ZBzZ5I#FDV3zwiS0k5L-pZOuXwxLJIn6U2YxT8J8G2M2orRDCU
Now because I did use the twitter profile you provided as an example, I also used my screen recording to show that some profiles can be viewed without being logged in and others (like yours) can't.
This may be different from user to user on factors I can't know but in my video you can see how twitter behaves for me when opening the profile you provided as an example.

BUG_A: Regarding xvideos.red:
For xvideos.red it is not possible to get those direct-URLs inside JD GUI.
I did give a general explanation about why it sometimes does not make any sense to expose direct-URLs to the user.
If you think there are good reasons to do this for xvideos.red items anyways, just let me know.

[katyz]

They'll probably prefer to follow the thread, but that's okay.

About BUG-A:
Thank you, congratulations.
I see the fixes/improvements.
The problem related to the URL has been resolved.
you also fixed the problem with copying the URL that was previously copied incorrectly.

But I have to make an observation:
I tried scraping the @CoryChaseXXX profile again, and I succeeded without any errors, 855/856.
Only 1 link had an error, so I went to the profile to look for that link in the original profile and came across hXXps://twitter.com/CoryChaseXXX/status/1544813489688121344 (yes, a dead link with a bug).
In Jdownload, it says that the link is activated, and in fact, it is activated but broken with a "0 byte error." Maybe other profiles have a lot of these old broken links (it's not the first time I've encountered a broken link like this one). Perhaps it would be a good idea to include an option to ignore links with the "0 bytes" error in the "link grab" for TT.
Resolution: 0x0px
Format: MP4 - MPEG-4
Length:
File Name: CoryChaseXXX - 1544813489688121344.mp4
File Size: 0 B

As for the xvideos guide: Yes, I think it is an important option for you to place the links copied from the guide (90% of red users download the links by clicking on the guides). For internal control, it is important to know the links.
I won't go into details with a big text, but it's something important. After all, you already have these links internally. I know it takes work to implement the display option, but believe me, your users will appreciate these options, and things will be more transparent.
Xvideos rarely resets its URLs, like Pornhub (I have links that have been working fine for 5 years, lots of links, but a retirement or link change can happen with any site and is normal).
As mentioned in a previous comment of mine, a link to open today can show yellow duck and in next year can open Harry Potter, but that's rare and really depends on the site.
There are bots that repost daily or weekly and constantly change the URL, but these are more isolated cases, and fake account profiles or cheaters are another problem and should not be taken into account.

Off-topic:
Here is a post for anyone who has problems with files with large characters.
"Fixed by Microsoft"
hXXps://mega.nz/file/vQsw0aCD#4KOIamfYlY422f8PrCWh8Hhtn7N2_az81d5H6lbRjT4
Just run it. (only works on Win10 or higher)

Although I still think this suggestion is important,
hXXps://mega.nz/file/3MljBCKI#_NyAu8NN_hk1nNQjA2bx9c-CIWZ6O60IR8nO3MfUO-A

Thank you for providing corrections/improvements.

[pspzockerscene]

Originally Posted by katyz

Thank you, congratulations.

Thanks for your feedback.

Originally Posted by katyz

I see the fixes/improvements.

Those changes have been added in revision 48146 on 2023-08-14.

Originally Posted by katyz

you also fixed the problem with copying the URL that was previously copied incorrectly.

As said this was never a bug and as you can see in my video, you need to change a setting to get the result I've shown.
The default behavior was not touched.

Originally Posted by katyz

In Jdownload, it says that the link is activated, and in fact, it is activated but broken with a "0 byte error." Maybe other profiles have a lot of these old broken links (it's not the first time I've encountered a broken link like this one). Perhaps it would be a good idea to include an option to ignore links with the "0 bytes" error in the "link grab" for TT.

That is not entirely true - two more detailed information about this one:
1. The link is not dead/broken: Only the HLS version of that video seems to be broken.
Go to Settings -> Plugins -> twitter.com -> Disable "Prefer HLS over HTTP download" -> Try again and it should work fine.
Also I'm able to watch that video in browser without problems.
2. The item will actually be displayed as offline but only once you try to actually download it.
The reason is simply: If JD was to check all of those single added links, it would take ages to crawl profiles so in such cases where an item is known to be online by the crawler, the online-status is set before to speed up the process of adding links.
TL;DR There is no fast way to detect such items and display them as offline. You will need to try to download them to get said offline-status.

I might look into this to add an "auto fallback to http/hls" if one version looks to be broken.

Originally Posted by katyz

As for the xvideos guide: Yes, I think it is an important option for you to place the links copied from the guide (90% of red users download the links by clicking on the guides).

What is an "xvideos guide"?

Originally Posted by katyz

I won't go into details with a big text, but it's something important.

I can't think of any reason to get the direct-URLs other than maybe hotlinking/embedding them on other websites.
Also, as explained, so far you are the only user asking for this.

Originally Posted by katyz

Xvideos rarely resets its URLs, like Pornhub

I will look into this and if they really don't change, I'am open to add the same handling than I've added to twitter to xvideos.com/xvideos.red.

Originally Posted by katyz

Here is a post for anyone who has problems with files with large characters.

The .reg file you are linking does the exact same thing that we'Ve linked in our own troubleshooting guide:
https://support.jdownloader.org/Know...load-directory
-->
learn.microsoft.com/en-us/windows/win32/fileio/maximum-file-path-limitation?tabs=registry

Originally Posted by katyz

Although I still think this suggestion is important´

I've told you that we are already working on a setting for that.
If you want to follow the development-process of that, see this thread:
https://board.jdownloader.org/showth...399#post523399
Ticket:

Originally Posted by katyz

Thank you for providing corrections/improvements.

Let me know once your "BUG-A" topic is done then we can talk about the next one.

[pspzockerscene]

Originally Posted by pspzockerscene

That is not entirely true - two more detailed information about this one:
1. The link is not dead/broken: Only the HLS version of that video seems to be broken.
Go to Settings -> Plugins -> twitter.com -> Disable "Prefer HLS over HTTP download" -> Try again and it should work fine.
Also I'm able to watch that video in browser without problems.
2. The item will actually be displayed as offline but only once you try to actually download it.
The reason is simply: If JD was to check all of those single added links, it would take ages to crawl profiles so in such cases where an item is known to be online by the crawler, the online-status is set before to speed up the process of adding links.
TL;DR There is no fast way to detect such items and display them as offline. You will need to try to download them to get said offline-status.

I might look into this to add an "auto fallback to http/hls" if one version looks to be broken.

Added auto retry handling for such "broken streams".
It will work regardless of your twitter plugin settings.

Wartest du auf einen angekündigten Bugfix oder ein neues Feature?
Updates werden nicht immer sofort bereitgestellt!
Bitte lies unser Update FAQ! | Please read our Update FAQ!
---
Are you waiting for recently announced changes to get released?
Updates to not necessarily get released immediately!
Bitte lies unser Update FAQ! | Please read our Update FAQ!

-psp-

[katyz]

As said this was never a bug and as you can see in my video, you need to change a setting to get the result I've shown.
The default behavior was not touched.

Actually, upon checking here, that was actually a mistaken interpretation of how the logic of copying the link's content worked, now that I understand.
I see that you created a table with 2 columns (maybe more appear in other cases), when the user clicks to copy a column, there is an issue with no separation below the header. This is because you wanted to save time by configuring it internally, when it could have been done much better with user interaction. For example, the user could simply use the mouse to select and add or remove columns (gear icon), but they don't exist. There's only the header, and what the user copies is not really what appears on the screen but what is internally configured in the program, as shown in the video you sent me.
This is just a brief response; I won't suggest improvements. It's okay for me, although I wouldn't do it this way.

That is not entirely true - two more detailed information about this one:
1. The link is not dead/broken: Only the HLS version of that video seems to be broken.
Go to Settings -> Plugins -> twitter.com -> Disable "Prefer HLS over HTTP download" -> Try again and it should work fine.
Also I'm able to watch that video in browser without problems.
2. The item will actually be displayed as offline but only once you try to actually download it.
The reason is simply: If JD was to check all of those single added links, it would take ages to crawl profiles so in such cases where an item is known to be online by the crawler, the online-status is set before to speed up the process of adding links.
TL;DR There is no fast way to detect such items and display them as offline. You will need to try to download them to get said offline-status.
I might look into this to add an "auto fallback to http/hls" if one version looks to be broken.

About the 0-byte link.

hXXps://twitter.com/CoryChaseXXX/status/1544813489688121344

In the updated Chrome browser, the video doesn't play on Twitter.
I also tried downloading the video using other methods, but it downloaded as 0 bytes as well.
It seems to be corrupted at some point in the system.
If the hosting is having issues or temporary instability, you won't be able to download it at that moment, but you might be able to view it elsewhere.
I find it strange that you mentioned you were able to watch the video in the browser because the server is clearly failing, especially during the times I tested.

So, I decided to do a more thorough analysis because I also noticed that JD had downloaded the video overnight (while I was asleep), probably in multiple attempts until the server responded and downloaded it.

I downloaded it again and reinstalled it on a new computer, and I left everything default, and I saw that it was deactivated the TT HLS plugin disabled by Default.
So, when I sent you the report, it had already been tested with the TT HLS plugin disabled.
I'm not blaming JD, I simply suggested a way to handle the 0-byte issue.

There could be several reasons why the link isn't working:
Corruption, downtime, some internal bug, temporary hosting instability, DDoS, etc.

As you mentioned that you were able to watch the video, and I was also able to download it in JD,
I ran some tests with this link and noticed that the server returns a Status 416 in several calls with 0 bytes. However, if you let it run, you'll notice that at some point, it returns a value other than 0 bytes, and you can download it if you're quick. But I can't guarantee that the video is complete and not corrupted at some point. Anyway, this is a summarized explanation of the problem for this link.

Originally Posted by katyz View Post
Here is a post for anyone who has problems with files with large characters.

Please, this has been a known issue since the release of Windows 7. Microsoft even promised a fix back then. I just posted to help those who were following this thread because it made sense with the report, but it doesn't help anyone on anything below Windows 10. And this goes way beyond JD; it's a serious Windows fix, just like others they haven't added yet.
Your link that you sent me, sent to a dev, to rename the files manually is without comments.

I've told you that we are already working on a setting for that.
If you want to follow the development-process of that, see this thread:
hXXps://board.jdownloader.org/showth...399#post523399

I never asked for an ETA, and I never will. The program is free, and no user should request an ETA for something that's free. Even for something paid, it's complicated to provide an ETA, so don't think I'm asking you for an ETA for anything.
The image I posted was just something like "try not to let Windows 7 die," something similar, and also emphasizing the importance of the user choosing what to remove from the name of what they're downloading with on/off options.
For Windows versions below 10, JD (presumably a software or program) is already impractical due to characters, several videos freeze, and it doesn't help if the name is long; it will glitch. You can work around it by manually opening and downloading or by doing other manual workarounds, but that completely defeats the purpose of automation.


About your other questions, I won't quote them because after you solved the issue related to the URL links, several questions revolving around that problem are now resolved. Consider all your unquoted questions as OK.

Let me know once your "BUG-A" topic is done then we can talk about the next one.

Yes, we can proceed.
Although as mentioned earlier: "After you solved the issue related to the URL links, several issues surrounding this problem have now been resolved."

BUG-B Resolved.
BUG-C is up to you to implement or not; it's a very simple option, as suggested, there's not much to debate, it has already been mentioned in the main report.
BUG-D Resolved.
Regarding error handling, "I noticed you made some adjustments, improvements; for me, it now seems okay."

I just need to check about: (perhaps there is a solution, it's worth analyzing, although these errors do not prevent downloading the videos, and it even seems not to corrupt or affect them, but I still need to investigate further.)
Win7 x64 Ultimate Original, all updates - Crash ffmpeg
hXXps://mega.nz/folder/uM1GFbgD#zFSwyjfB1hBE3sPMVZEhGA

Also, I don't have any more logs to send regarding previous issues because they have stopped occurring(fixs).
For me, everything is resolved, and I won't ask for an ETA for what's missing. I am aware of the progress made so far and that there are other things to consider whether to add or not.
I just want to mention that my Mega folder will be removed, I will only keep the Win7 link open for a few more days, and please move the links to private. It's better not to have Google attach the names and subject of the links, although you probably would have done that already.
Thank you for the support; it has always been quick, and you have always responded in the way you could understand, with your encouraging or not responses, regardless of what was asked or the acidity of the user.

[pspzockerscene]

Originally Posted by katyz

I see that you created a table with 2 columns (maybe more appear in other cases), when the user clicks to copy a column, there is an issue with no separation below the header. This is because you wanted to save time by configuring it internally, when it could have been done much better with user interaction. For example, the user could simply use the mouse to select and add or remove columns (gear icon), but they don't exist.

I don't understand.
Please add a more detailed explanation and/or screenshots.

Originally Posted by katyz

There's only the header, and what the user copies is not really what appears on the screen but what is internally configured in the program, as shown in the video you sent me.

That is not true.
What's copied is what you can see on the screen and it also depends on how JD is configured.
You can freely configure the "copy behavior":
Rightclick in any empty space in the linkgrabberlist or downloadlist -> Open Menu Manager -> Click the "+" icon -> Select "Copy information" -> Now you can configure this copy action
You can also add multiple copy actions and configure hotkeys.
Screenshot:

Originally Posted by katyz

I find it strange that you mentioned you were able to watch the video in the browser because the server is clearly failing, especially during the times I tested.

I can still watch it without problems.
Maybe you were just unlucky or the issue only happened only at specific times and never during the time when I did my tests.

Originally Posted by katyz

I'm not blaming JD, I simply suggested a way to handle the 0-byte issue.

As explained:
- The issue is already handled: wait and retry later with error message "Broken video?"
- There was a bug in the retry handling which I've fixed yesterday

If you can still reproduce issues with 0-byte twitter videos, please provide a log.

Originally Posted by katyz

Your link that you sent me, sent to a dev, to rename the files manually is without comments.

I don't understand.
Here is a link going straight to the comment of the user:
https://board.jdownloader.org/showpo...&postcount=160
Though I still think that your problem is "too long path" or "too long filename" and is not related to the characters in that path/filenames.

Originally Posted by katyz

For Windows versions below 10, JD (presumably a software or program) is already impractical due to characters, several videos freeze, and it doesn't help if the name is long; it will glitch. You can work around it by manually opening and downloading or by doing other manual workarounds, but that completely defeats the purpose of automation.

I will let you know once the feature "auto shortening filenames" is available.
Also I think that this is "BUG_D".

Originally Posted by katyz

BUG-B Resolved.

Good.

Originally Posted by katyz

BUG-C is up to you to implement or not

I will come back to this later.

Originally Posted by katyz

BUG-D Resolved.

As explained, we are working on a "path too long" and "filename too long" detection and "auto rename if filename is too long".
Ticket:

When this ticket has a progress of "100%", the mentioned changes will be available.

Originally Posted by katyz

I just need to check about: (perhaps there is a solution, it's worth analyzing, although these errors do not prevent downloading the videos, and it even seems not to corrupt or affect them, but I still need to investigate further.)
Win7 x64 Ultimate Original, all updates - Crash ffmpeg

As said another developer will look into your crash logs and respond here once he finds the time.

Originally Posted by katyz

I just want to mention that my Mega folder will be removed, I will only keep the Win7 link open for a few more days, and please move the links to private. It's better not to have Google attach the names and subject of the links, although you probably would have done that already.

We're not keeping such files so it is up to you to provide them to us.
If you want them to be available for a bit longer, just reply to your ticket "VSW-512-47082" again and attach them in your reply.

Originally Posted by katyz

Thank you for the support; it has always been quick, and you have always responded in the way you could understand, with your encouraging or not responses, regardless of what was asked or the acidity of the user.

Thanks for your feedback.
I will let you know once the mentioned features are available.

[katyz]

I don't understand.
Please add a more detailed explanation and/or screenshots.

and next :

No sense

So you said you didn't understand what I said, and yet you cut out a part of a question and sent that I'm a liar and sent a tutorial that was out of scope, having nothing to do with what I said.
I have typically received similar responses from you since I started my reports in private.

Column 1(NAME) | Column 2 (Download from)
Below, the visual division by mouse was not implemented, the user cannot select, for example, select one or the other without having to change the configuration,
between "Name" or "Download from" the user can only select and it is copied according to the internal configuration of that image ->hXXps://mega.nz/file/iFUAxaqD#j2jbqYvsCjiDRia_LfTQQN4nsoZ5pUZm_r78zDCikmk
So if you want to copy "Name" and not "URL" you need to change the internal configuration.
And the gears on the side just reset the column, there is no option in the add/remove gear or a select, this part was defined in another configuration, the internal one mentioned above.
hXXps://mega.nz/file/qcF3jDyR#WAatQPlRPMY3oBipfJsJ8QkWRirSg3f_qu9TGSUllmg

I just made a comment, I didn't suggest changes, because it's up to you to receive feedback and model the layout and settings in your own way.

I can still watch it without problems.
Maybe you were just unlucky or the issue only happened only at specific times and never during the time when I did my tests.

I sent a final response detailing by a professional DBA for over 10-15 years, who works with networks, clouds/dedicated, sqlserver, mysql, migrations, ftp linked programs and coder.
I analyzed the link, did tests and found several HTTP error returns when trying to request the download from the hosting.
JD himself made several attempts before managing to download the link, and you still send a response nullifying the entire concept of data exchange just because you managed to open it once or a few times in the browser,
The problem exists and is real, and it happens rarely and more with old links, you ignore many points.
I won't go into this further, let's move on to the next one.
[/QUOTE]

I don't understand.
Here is a link going straight to the comment of the user:
hXXps://board.jdownloader.org/showpost.php?p=523399&postcount=160
Though I still think that your problem is "too long path" or "too long filename" and is not related to the characters in that path/filenames.

About BUG-D:
Why are you mixing things up?
This suggestion: hXXps://mega.nz/file/3MljBCKI#_NyAu8NN_hk1nNQjA2bx9c-CIWZ6O60IR8nO3MfUO-A
Was only related to WINDOWS 7 to improve the issue of long names and unnecessary characters, in other words, to mitigate the problem, not solve it.
I never said that the issue of freezing at 97% was related to special characters.

I marked it as resolved and was just waiting for the update, but upon analyzing your response, I decided to make two videos see:

Video 1: Crash
hXXps://mega.nz/file/aFclUZzA#TnAjcKW6rPmkUtpN6kQ2pm_y9qU6Uasoi-EfSHZFdzU

Video 2: Removing the BUG
hXXps://mega.nz/file/XVkhGBaZ#1OYAv-gQjXfHtjaFmfzyIEYWi7h2QkMNQmaSnyQt8hU

I hope it can be an on/off option, which is always good, especially for Windows 11 users who often don't want to limit names. Many are already accustomed to long names and paths, especially when scraping from OnlyFans, which I hope JD will support in the future.

Our conversation is long, and it's a quote after a quote about several questions, sometimes there's a lack of context for me for you, I think now we just have to wait for BUG D and everything is ok.
Thank you very much for your support.

[pspzockerscene]

Originally Posted by katyz

So you said you didn't understand what I said, and yet you cut out a part of a question

I didn't.

Originally Posted by katyz

and sent that I'm a liar

I didn't.

Originally Posted by katyz

sent a tutorial that was out of scope, having nothing to do with what I said.

Which one?

Originally Posted by katyz

Below, the visual division by mouse was not implemented, the user cannot select, for example, select one or the other without having to change the configuration,
between "Name" or "Download from" the user can only select and it is copied according to the internal configuration of that

That is correct.
I thought you were talking about the generic columns inside the downloadlist.
We are not planning to change the "Show downloadurls" window.

If you want to copy filenames or other file information, do it in the columns inside the downloadlist tab. My last post explains how to do this.

Originally Posted by katyz

I just made a comment, I didn't suggest changes, because it's up to you to receive feedback and model the layout and settings in your own way.

All good.

Originally Posted by katyz

The problem exists and is real, and it happens rarely and more with old links, you ignore many points.

I didn't say that it didn't ever exist.
I just stated that I was unable to reproduce it and you seem not to be able to reproduce it either so I guess the correct wording is: It doesn't exist anymore.

Originally Posted by katyz

Why are you mixing things up?

Misunderstandings can happen.

Originally Posted by katyz

This suggestion: hXXps://mega.nz/file/3MljBCKI#_NyAu8NN_hk1nNQjA2bx9c-CIWZ6O60IR8nO3MfUO-A
Was only related to WINDOWS 7 to improve the issue of long names and unnecessary characters, in other words, to mitigate the problem, not solve it.
I never said that the issue of freezing at 97% was related to special characters.

Okay.
We are not going to add the settings you proposed but you will be able to easily remove such characters in the future.

Originally Posted by katyz

I never said that the issue of freezing at 97% was related to special characters.

I never claimed this either.

Originally Posted by katyz

I hope it can be an on/off option, which is always good, especially for Windows 11 users who often don't want to limit names. Many are already accustomed to long names and paths, especially when scraping from OnlyFans, which I hope JD will support in the future.

We are not planning to limit filename lengths.
Filenames will be auto-shortened if JD detects that they are too long and yes, that behavior will be optional.

About your last two videos and the FFmpeg / filename related problem:
As I said, the problem is known and there is a thread about it:
https://board.jdownloader.org/showth...705#post520705
-->
https://board.jdownloader.org/showpo...5&postcount=24
It's nice that you found the solution yourself but I've also already linked it in my first response.

[pspzockerscene]

Last update to this long topic:

Replacing of non-needed characters:
You can now easily globally regex-replace unwanted characters in filenames via:
GeneralSettings.filenamereplacemap

The same for package names:
GeneralSettings.packagenamereplacemap

Auto correction of long paths (more specifically, long filenames) / BUG_D:
This feature still hasn't made it into the stable version of JDownloader but it is under development.
It is already working in our dev build, see:
https://support.jdownloader.org/de/k...up-ide-eclipse

Subscribe to any of the following threads to be notified once that feature gets released:
https://board.jdownloader.org/showthread.php?t=59937
https://board.jdownloader.org/showthread.php?t=62683
https://board.jdownloader.org/showthread.php?t=63348
https://board.jdownloader.org/showthread.php?t=67005
https://board.jdownloader.org/showthread.php?t=74880
https://board.jdownloader.org/showthread.php?t=83268

I'am marking this thread as solved now because
- Newly reported problems have been fixed
- Other still open problems are known issues that have been reported before this thread already (like for example the "path too long" problem)

The thread remains open for feedback.

[wkearney99]

Regex support in the file and package replace maps is a GREAT feature. The default text for the fields only contains simple character replacements. But you are NOT limited to just single characters. You can use a regex to find potentially multiple things and have them all replaced.

I wanted to get rid of #hashtags in filenames. I used an Event Script to do it. But then
pspzockerscene pointed out the ability to use the replacemaps. I'm still inclined to use the Event Script as it's easier to toggle it on/off.

I mention this here because using an Event Script during a Packagizer Hook might be useful for other filename related edits. Once you're in a script you can do whatever you'd like to the filename. Using just the UI editing fields doesn't give you as much flexibility. Thread on this: https://board.jdownloader.org/showthread.php?t=96716