/hydrus/ - Hydrus Network★

>The edit subscription panel now has a 'paste queries' button that lets you add queries en masse from a newline-separated list on the clipboard.

Cool

Can you add that to the tag censorship/sibling/parent windows?

I made an alternate Linux build for people having trouble with the official build. A few things are still weird, but overall it works better for me than the official. Run from source works slightly better, but pip is a headache. Built on Linux Mint 18, so it should also work on Ubuntu 16.04.

If you want open externally to work, you will have to manually set the open program for each mimetype in options->files and trash. For some reason, when Hydrus invokes xdg-open it opens a program in WINE.

https://www.mediafire.com/file/eo34b2yy1od465o/hydrus-285.tar.gz

I'll post one of these every week until the official Linux build works on my machine. If anything is broken, tell hydrus_dev (and thank him while you're at it). He was kind enough to give me his build command, but I don't know Python. If something is broken, I'm probably just as clueless as you are.

how do i update my version without fucking up and losing my database? its probably easier than i think but i dont want to fuck it up.

also hydev, do you know about a program called Visipics? its used to detect duplicate image files and is incredibly proficient at it

>>7442

Install/extract the new version into the same folder as your old version

Hydrus already has a better dupe finder than Visipics

>>7443

>Hydrus already has a better dupe finder than Visipics

what about that warning of "not doing dupe finding with more than 100 pictures"? im still new to this and i don't want to freeze the client or something

>>7444

>>7443

honestly i dont understand it, it only works when i try to find duplicates for a single specific file, if i try a big duplicate search in a group of files, it displays nothing

>>7447

i got that running but it doesn't show any duplicates whatsoever, what am i doing wrong here?

>>7448

>>7449

Also, click

>show random pairs

To show a random group of similar pics

or

>launch filter

To start mass filtering them manually

The page doesn't show anything by itself.

>>7448

>>7447

i set the range to "very similar", and i know for certain that i have quite a few dupes. then i did it again on distance 8 (it said speculative) and its still not showing me shit.

i mean i know im probably doing it wrong, but this is definitely not better than visipics for its intended purpose at least. I'm in no hurry though, the amount of dupes is too minimal to do anything on my available disk space.

>>7452

Did it ever show anything? Try >>7450

It doesn't show anything by default until you choose one of the options.

Also, post a pic of the duplicates screen like >>7449 , I want to see if it says anything.

>>7456

>194.552 potential pairs

M8, that almost 200 potential duplicate pics.

>when i launch the filter it starts opening pairs of pictures that don't even look remotely similar, and thats not what i want.

If you want pictures that look more alike, set the distance to lower.

> i want the groups of pictures the program considers dupes to appear in the window to the right side, if possible, like visipics does.

You can't do that, only the filter. If you want it you could try suggesting it in the suggestion thread though. It sounds relatively easy to do, so devanon might do it.

>>7457

>M8, that almost 200 potential duplicate pics.

*almost 200k potential duplicate pairs

>>7459

thanks, but that sounds like a royal pain in the ass. im gonna stick to visipics

i only uploaded about 14k pictures so far, so its impossible to have that many dupes.

>>7461

i dont archive pictures unless i have 2 or more tags, usually series and creator/character name is what i go with.

so yeah, i have a ton of tags, and i download stuff from e621 which brings like 30 tags per picture.

the dupe finder works very differently of what im used to, so i dunno, it just shows random pictures in each "pair" regardless of which distance i set, i've tried at 0, at 2, and at 8.

>>7464

>554 potential pairs

Yeah, it's working now.

I recommend setting it to 4 after that, then 8, to get the more different ones.

Also, I'd recommend using the filter, it's faster that way.

>>7465

I've actually found that after 5, using the filter becomes more trouble then it's worth, when every image has 15 dupes it's picking up, it's far easier to highlight five at a time and manually select them all as non-duplicates at a glance rather then at 2 or 3, where you do need to stare pretty carefully to see the precise difference between two images.

Ok, So I realized through needing to do some things that my tag repositories have not been syncing for I want to say at least 3 months possibly more.

I remember why I stopped syncing in the first place now, this shit takes fucking forever.

So, a suggestion.

Like with import folders, is there any way you can have the take repository searches paused, and then manually start them up? I could see doing this when I plan on sleeping for 12+ hours thanks to some sleep aids, but holy shit, I have wanted to go though images for the last 4-6 hours but this shit is still syncing.

If I could start the process on my terms that would be greatly helpful.

>>7437

Sure; I'll add it to the list. I'll try and write a 'wrapper' control that does the job well on its own, and then I can replace all the other 'textbox into list' instances across the program with a unified interface.

>>7442

Hey, I am sorry this has been confusing. It looks like these other Anons have helped, but you might like to check out my help here:

http://hydrusnetwork.github.io/hydrus/help/index.html

There is also a copy in the client, under help->help. It tells you how to update an install and then goes up to advanced topics like the duplicate filter. Let me know if you still find anything confusing.

>>7467

Please check the settings under file->options->maintenance and processing. If processing is slow on your machine, I recommend not allowing it to kick in during idle time. Either tell it to run only on shutdown for a certain period, or disable it entirely and then run it manually overnight or whenever is otherwise convenient. To run it manually, make sure 'advanced mode' is turned on under the help menu and then open up the tag repo on services->review services. It should have a 'process now' button that will start it whenever you like.

The PTR has grown and grown, and many users on HDDs are feeling the strain. Running on an SSD, if you have that available, typically runs much better. I did this a month or so ago and am very happy--I went from like 600 rows/s to 30,000.

You might also like to check this page, which has some advice (like running defrags) on keeping the client running fast:

http://hydrusnetwork.github.io/hydrus/help/reducing_lag.html

It also talks about how to create 'profiles' that you can send to me to examine if anything that is running unreasonably slowly.

have a recurring issue with windows 7 where the ui hangs for a bit, and then the client get stuck in the forefront till I restart it. don't know if there is anything you can do on your end with this.

Just thought of a new feature

So I have been having issues with dealing with large pages of images, my largest being 46k images and a bane to my existance, i have gotten half way through it before, but a crash saw that endeavor killed off, so I split the tab into about 10 FAR more manageable parts.

so here is my thinking, is there a way to add a split section to when you right click on a tab, this would bring up 2 sub menus, one would be number of pages, and another would number of images, pages obviously going from 1-20 pages cant imagine anything that would need to be split up more than that even in my most fucked up pages, and images would go from 100-500 in 100 image increments, 500-2500 in 250 image increments and 3000-10000 in 1000 image increments

Even if its just the easier pages implementation (at least I think pages would be easier) this could be very helpful to large image imports and the like where processing tons of files needs to happen but getting though pages is kind of hard due to the scope.

I would also recommend that this option automatically starts a page of pages for all the parts.

I'm getting this error when I try to download some files:

2017/12/10 16:59:05: 
2017/12/10 16:59:05: Exception:
2017/12/10 16:59:05: TypeError: argument of type 'NoneType' is not iterable
  File "/usr/lib/python2.7/threading.py", line 774, in __bootstrap
    self.__bootstrap_inner()
  File "/usr/lib/python2.7/threading.py", line 801, in __bootstrap_inner
    self.run()
  File "/opt/hydrus/include/HydrusThreading.py", line 230, in run
    callable( *args, **kwargs )
  File "/opt/hydrus/include/ClientImporting.py", line 5533, in _THREADWorkOnThread
    self._CheckThread( page_key )
  File "/opt/hydrus/include/ClientImporting.py", line 5103, in _CheckThread
    HydrusData.PrintException( e )
  File "/opt/hydrus/include/HydrusData.py", line 931, in PrintException
    trace_list = traceback.format_stack()

I think I've got this in 284 version(I was using 283 for two weeks and everything was alright). I'm using aur arch linux build.

>>7479

A "split one page into multiple pages" and "combine multiple pages into one page" option would be really good. Seconding this post.

"combine pages" could be used as an or-search if you open a search page for each tag then combine them

>>7485

The combine is happening, but I think splitting may be easier than combining so fingers crossed on that one happening sooner than later, it makes going through the tabs I split so much easier when its smaller sections, it's not a big monolithic clusterfuck dump, that said, 161 tabs and 175 causes massive instability for me, so got to go though a fuck ton before I can split further, and with some of the tabs being 4000~ which isn't much by my standards, its just so much easier to plow 500~ image tabs I think i'm going to end up moving everything to it so long as i'm able to before I begin processing.

>>7491

as of right now

161 tabs, 250-300k images open across all of them, 1 experimental 'lets see how subscriptions work' sub a good deal or at least use to be a good deal of 404 thread watchers, they were easy enough to process I made my way through them to free up some overhead.

last time I fresh booted, everything works fine, no issues at all with importing or playback, but with this session, I get near constant hiccups, and I have 2 other sessions saved that are basicly the same as this thanks to a massive import. ram use for the program right now sits between 8.5gb and 10gb,

Most of the problems I encounter are caused due to this session being out of hand.

On a side note, ctrl+alt+esc allows the windows to lose its focus again.

also, apart from the initial 5000 images, I think the rest of the 1.35 million images are all in the inbox. I was keeping the images there so I would go through and tag them then move them to the archive, but that plan is out the fucking window, especially that I think ratings is more useful for my needs in its current implementation.

as for my computer

1700 ryzen, 32gb of ram, 512gb nvme ssd that the database and thumbnails are on, 8tb hdd for import folders, a 4tb hdd that files are moved to, and possibly a 3tb hdd that is exclusively images, not sure if thats happening yet or not.

very hard for me to classify my computer as old.

and yea, I think it was somewhere between 50 or 60 thousand images that the program had issues last time I did that testing, the program has honestly come such a long way since then, those points where shit slows down are probably outdated.

>>7494

Right--I figured with Win 7 you had an older machine. I am not totally sure, but I think some of the older Windows have lower gui object limits (it means 'a program can't have more than x thousand window objects'), which is the main cause for the ~165 limit atm. I believe the proper name for this, if you would like to look into it for Win7, is GDIProcessHandleQuota in the registry. Win10 is like 10,000, but I think earlier is like 3,000. Obviously be careful if you mess around with this stuff.

But there are other problems beyond GDI count, which is probably what is slowing your bad session. Every import page has two threads and a bunch of other shit that I want to reduce.

I have more overhead reduction today and hope to do more in the coming months. As you are a heavy user, your continued feedback would be appreciated.