/hydrus/ - Hydrus Network

>>6019

>A good dea of dom (jiromike2000), image set are not the correct pixiv image but the master1200.

What are you trying to say here?

When I try to download the image at https://www.pixiv.net/member_illust.php?mode=medium&illust_id=21500071 I get the direct link https://i.pximg.net/img-original/img/2011/09/03/06/56/45/21500071_p0.jpg which gives me pic related (sha-256 is 77cb7a019c850798256681b49b7136299ecc462e3fd0c8856746b3d68d8893c3).

What do you get? I don't see anything that says master1200.

Does anyone know how to use regex properly with hydrus? I can use regex in my terminal somewhat successfully but hydrus is different for some reason.

I download images from pixiv and save them in the following format:

…/%member_name% (%member_id%)/%image_title% (%image_id%)/filename - %tags%.extension

where filename looks like %image_id%_pXXX.extension with XXX some number equal to or higher than 0

and %tags% looks like tag1,tag2,tag3,…

the length of image_id and member_id may vary, but it is always numbers

member_title and image_title may be japanese characters or latin alphabet or a mixture, same for tags

I can also change the formatting of how the files are saved, like I could do %member_name% (%member_id%)/[%image_title%] {%image_id%} @%tags%@ et cetera, if that is easier (but I need to save member id if I want to be able to look members up later on)

I import these into hydrus and what I want is:

make the creator:XXX tag where XXX is member_name (this is not very important I can do it easily with ctrl-A and F3)

make the title:XXX tag where XXX is image_title (important because copying and pasting all the japanese characters for the titles is a pain in the butt)

make all the tags tag1,tag2,tag3,… unnamespaced tags (not very important though, I can tag them manually coz I don't use much tags)

make the pixiv image:XXX tag where XXX is image_id I don't know if this is smart because I don't know if it is possible to hide tags from being displayed in the 'selection tags' tab on the left—if it is I would make a shortcut to hide/display it, if it is not this is not so important

If you know a good place that uses the same regex as hydrus I would also be happy. There was a link in hydrus but they used the * sign different there than how hydrus used it.

>>6022

The picture saved on the booru were incorrectly saved from pixiv and are one of the many pixiv samples.

this:

https://gelbooru.com/index.php?page=post&s=view&id=1379039

is not this:

https://www.pixiv.net/member_illust.php?mode=medium&illust_id=23002825

>>6025

found more detail on the cause:

https://github.com/r888888888/danbooru/issues/2261

It's a medium sized pixiv thumbnail always converted to jpg EVEN if the source is smaller.

I guess people were caught unaware during the change, and may stilll be if they don't use a script to download.

Is there a hotkey I'm missing to clear the list of tags in the current search?

If I search for "tagA + tagB" and then a bit later want to make a separate search for "just tagC", I first have to double-click on the first two tags in the search box to remove them. (Otherwise I'd be searching for A + B + C, which I don't want.)

I'm sure there's some way to do this with the keyboard instead of having to use the mouse, but I'm not sure how.

>>6061

There's no way to do it just with the keyboard, but if you click on a tag and then CTRL+A you'll select all of them, which you can quickly remove with the delete key.

>>6074

You can't right now, there's a future feature where the "filename:" namespace would replace the filename when you upload it somewhere, but I don't think that's even begun to be worked on. If you're paranoid (the only thing anyone would be able to do is connect posts with the same image to the same person, if someone were autistic enough to try to stalk you like that) you can drag and drop the image to your desktop and rename it there, then upload that.

How does booru parser work? If I create local booru like website e.g.with Django ,is it possible to use the parser?

>>6074

You can use 4chanX to generate random unix filenames if you're talking about posting there. That's what I do.

>>6074

There is no personally identifiable info in the file names. The hash is generated from the file contents only. Identical files will generate the same hash. If 100 people download the same file and imports it into Hydrus, it will be given the same file name for every person. The file name won't be unique for each person.

Enabled the Like/Dislike feature but it's non-responsive. Additionally the program crashes on whim frequently. Ver. 258 with 87k pictures.

Anyone know if there's a way to load images in blocs like in boorus?

I dont want hydrus to search and load 200000 images before giving them to me. Just the first 50 it finds, then the next 50, and so on.

how can I mass-rename a namespace? I have a namespace called noinseries that I made just when starting out, how can I rename that one to nis? I find the other one too long

Is there any plans on fixing the Sankaku downloader, or is the fix reliant on a change on their end?

>>6117

you can just make it a tag sibling, no?

>>6118

it's coming with the downloader overhaul

>>6119

yeah, I can indeed, was just reading up on those, rather new to this

so, say I have a namespace being foo: with all the properties of foo in it, and I also want a tag for bar:foo, to generally show that foo is in the image.

How can I add a parent relationship like this?

god damn do I feel stupid RN, how do i launch the slideshow? I 've looked around some, and even tried to make a hotkey for it, but I can't find it.

Hello, I'm late to the party. It would seem that the tutorial and the actual program get a few subtle differences, which I can't seem to understand myself. Help me out on how to find my access key. I'm trying to "conect" to the admin's public file repo (and possibly the tag repo too) and the tutorial really insist on the fact that "YOU HAVE AN ACCESS KEY", but never tells you where it actually is. Oh, and review services doesn't show a key of some sort either.

>6163

The acces key is below the pictures

>>6132

you need to view an image in fullscreen, then right-click > slideshow

>>6136

You can use "help > I don't know what I'm doing > just set up some repositories for me" to set up the public repos automatically

Not sure if this has been asked in the previous Q&A, but are there any scripts to help with updating tags for bulk-images downloaded from boorus?

Is there some way to search only for a namespaced or a non-namespaced tag without including the other (Without using - the other either?)

As an example, I have "series:final fantasy" and also a single picture mistakenly tagged "final fantasy" without the namespace. I want to find that picture and delete the tag from it, but searching for "final fantasy" returns all with both that tag AND the namespaced tag. And -series:final fantasy doesn't work, most likely because it is ALSO tagged with that.

I suppose it is possible to just tag sibling it, but that isn't ideal as this will pop up semi-often when mass tagging pictures and making a simple mistake, all with different tags that would get awkward to all tag sibling for a single example.

>>6181

Seconding this. This functionality does seem to exist as there are scripts when you look under 'services -> manage parsing scripts', but I don't know how to use it.

Unrelated: Is there a way to automatically import the images posted on a user's blogpost?

>>6184

Typo…

Meant "Is there a way to automatically import the images posted on a user's blogspot site?"

>>6184

>>6185

You can try to do 'pages > new download page > webpage of images' but it might not work.

Other than that I wouldn't know.

>>6183

Can't you just add both "final fantasy" and "series:final fantasy" to the search?

Alternatively: search for "final fantasy", select all, add "series:final fantasy" to all and remove "final fantasy" from all.

Was this well known before or is something recently discovered?

https://e621.net/forum/show/235250

https://e621.net/wiki/show/howto:sites_and_sources#tumblr

>>6189

shit I thought it was _1280… fuck

at least I have a list of artists whose tumblr I downloaded… but still

>>6188

Oh, that alternative option would work. How do you remove a tag from all images…

…nevermind. I played around with it and figured it out. Wow, I've never done mass tagging for multiple images before. This helps a TON, I can't believe I didn't think of it. I guess because I was so focused on tagging each picture. This is huge for cleaning tags though.>>6188

Would it be possible to use this together with the public booru databases to create a decentralized booru?

>>6257

>rider of black

>>6260

Some guy already did a personal rip of some boorus but "decentralized" unfortunately is not how a hydrus server would work.

>>6264

link pl0x

>>6265

Read it again. It's "personal", not public.

>>6264

Ah, okay. I just saw it linked on the IPFS git. But a decentralized booru would be real nice to have. Think of the possibilities. Not just for anime images, but think about porn webms for example.

>>6265

Search for ACGNHPCol on nyaa

https://nyaa.si/?f=0&c=0_0&q=ACGNHPCol

(horrible seeded probably, I think it's just one guy)

https://nyaa.si/view/887969

This is the one which should have the most seeders.

I want to use python to grab or add to a list of tags associated with a particular file hash, or get a list of files associated with a tag or who don't have any tags at all (in the main hydrus archive). HydrusTagArchive.py will only let me open one of the four databases (client.master.db) on account of the others not having the namespaces table, and that database doesn't have any tag mappings in it so it's useless for editing tags.

Is there any way to actually interact with the main tag archive with this module or am I better off accessing the sqlite databases directly to do what I want?

Is there any easy way to automatically add images from a specific folder to my inbox?

>>6334

set the folder up as an import folder

file > manage import folders

I'm using PixivUtil2 to download from Pixiv, but I'd like to add the downloaded images to Hydrus DB without copying/moving the images since new images always add, and I'd like to use PixivUtil2 to update them.

so Is there a way to import images to Hydrus but without copying/moving from the original location?

>>6342

http://hydrusnetwork.github.io/hydrus/help/faq.html#external_files

You could download the images to whatever folder and set that folder as an import folder and not have it delete files after importing. This would give you two copies of the images, one in your Pixiv folder and one in the Hydrus DB, but you can't have it manage the files remotely.

>>6343

Thanks, I know that I can keep two copies but I don't want to it'll take too much size.

is there a way to add different booru sites? or is that only possible on the devs end.

Is there a way to import the "rating" namespaced tags when importing from boorus? Like rating:safe or rating:explicit.

It doesn't seem like the PTR uses these sort of age-rating tags at all. u guys be crazy, how do you search for dem lewds?

>>6357

Just double-click the entry. A (-) will then appear. Hit ok and the relationship will be removed.

How can I rate my files?

>>6364

The menu is a little different, but it's still mostly the same functionally. In manage services: click add>local numerical rating service. The options from there should be self-explanatory.

>>6370

I don't know about all of them, but once you start typing, the tags you have in use that match what you typed should appear underneath. It's kind of slow. If you're typing out the whole word quick enough, you'd probably never notice. For some reason it's faster when browsing than when you're actually tagging.

is there's away to make sure that i'm not sharing/uploading my tags when using tag repository? i tag in a very bad way and I don't think it's a good idea to share it.

The subscription function no longer retrieves images. Is their a way to fix it? Ver. 265.

>>6378

I would love to be able to input a list of tags and the program would check the ptr to see if the image is tagged with those, and to automatically add those tags in my local tag repo if they are present. I'd get everything I want and nothing more

>>6397

If it's gelbooru that's giving you this problem it should be fixed in v266>>6397

>>6399

Okay thanks.

Is there a way to make parent tags automatically apply recursively via siblings?

e.g. If I have a sibling relationship "catgirl" -> "nekomimi", and a parent/child relationship "kemonomimi" -> "nekomimi", how do I make hydrus automatically consider "kemonomimi" to be a parent of "catgirl"?

why do so many files on the public tag repository have dumb tags like medium:png on them

doesn't the system:mime type tag already do a better job of this same thing automatically

New to Hydrus. How do I access the public tag repository? I try to add a hydrus tag repository service with the key given in the help but it says it couldn't be understood.

>>6427

go to

help > I don't know what I'm doing > just set up some repositories for me, please

I think this also connects to the dev's file repo, but you can just go to services > review services and remove it

Surprised i haven't found the answer to this with a bit of googling, but:

How do you guys post your archived images?

Say i want to post a file here. I'd select "add file" or whatever, and my browser would open windows explorer to go look for the file. Do you have to find the file in Hydrus, then copy the hash and look for that hash in windows explorer when posting? Seems like a lot of effort to post one file, it would make dumping really arduous.

What way do you guys do it?

>>6430

I'd just just highlight the files I'd want to share, then open context menu > share > copy > files and paste them on the desktop or some other folder

>>6432

>>6431

seconds after now, realizing that you can just drag and drop from hydrus onto the quick reply box…. that makes things easier

>>6433

>>6432

>>6431

>>6430

Ctrl+C on a thumbnail copies its path to the clipboard. You can just paste that in whatever file picker your OS uses.

Is there a way to have manual subscriptions that only play when prompted to, and at which point all new things are placed in a single tab for review or deletion?

Alright stupid question. When updating on Mac, which files should be replaced, and which ones should be left alone? I know if you replace everything, you pretty much delete all your stuff. Does anyone know exactly which files can't be replaced when trying to update? Thanks in advance.

I just found out about this and it looks great. I would just like to ask something, that this software may or may not help with.

For a while now, I've been looking for a way to be able to tag files within Windows' file browser, and to be able to use those tags with its search engine, or at least in the File Upload window. This would make uploading to imageboards so absurdly easier. But I'm yet to find software that can do this, and in fact I don't even know if it's possible without the OS' source code.

I figure that Hydrus' dev or the more knowledgeable folk here could answer this better than anyone. O please hear my plight ye taggers of Micronesian porcelain art.

Since firefox is breaking my "save image in folder" plugin, and I'm lazy, I just drop my images in my download folder.

But for some reason, hydrus detected the android sdk (exe) as an mp3.

It happened to a qemu installer as well.

Also had a film being detected as mp3 (think it was mp4 or avi).

Is there a reason why this happened? did hydrus not do a decent mime check? does it decompress exes and check if there's a file in there it can use and somehow applies that the to exe?

Is there any way you can randomly select an image with hydrus after sorting by tags I want? the majority of times I open and use hydrus I find I want a shuffle like feature for the images.

how do I filter pixiv to donwload only r-18 images?

https://www.pixiv.net/search.php?s_mode=s_tag&r18=1&word=ass

as an example is how it looks, I wanna do this but in hydrus because I think it's missing images since it's not in r-18

>>6496

nevermind the reason it can't find them is because they are a set, and hydrus doesn't do that (yet?)

>>6499

Use pixivutil2 then add them to hydrus, it is currently the best way to do it.

How do I disable the bandwidth limit?

Can someone help me install this on an .deb based system?

There's the AUR package on arch, but I had to migrate to debian and I need my library :C pls help

Can the duplicate queue be cleared? I've checked up to similar now and not duplicates are starting to show up. I don't want to go through hundreds of not duplicate pairs whenever I download something new to check if the latest batch had some real dupes.

>>6516

Just run from source.

Just downloaded this and currently importing files right now. Does this have a sort of account system so that I could get this on another computer and have the same files on it?

>>6524

The files stored in the database are all in the installation folder so just copy the whole folder to your other computer.

If I have 2 databases how do merge?

I was away from my HDD for a while and redownloaded hydrus and started some downloads, and now I have some files to merge now that I have my HDD again, what do?

>>6526

A hacky solution would be to export one database and then import all the files to the other one. Not sure if there's a better solution.

>>6527

export folder -> export everything -> import folder -> import everything?

>>6528

Yes

>>6529

feels like it missed some files but im probably imagining it, thanks

does/will hydrus support downloading entire boards of images?

i.e just entering http://8ch.net/hydrus/index.html or http://8ch.net/hydrus/catalog.html

Can all duplicate relationships be reset? I’m looking to start over with my pairing and can’t seem to find a way to reset the ‘discovery’ part of the duplicate filtering tab.

>>6416

Can I not do this on my local repo in the current version (v268)? I don't have the option under review services > local > tags.

@hydrus_dev, when do you plan to re-add HTA support for the local tag DB?

I have a bunch of files downloaded from boorus using a different program, which have filenames something along the lines of "artist_name ##### tag_1 tag_2 tag_3 etc.jpg". I made a regex to extract the artist name from there when importing, but it includes the underscores still which is kind of crappy given that Hydrus supports spaces in tags and the PTR has all of its creator: tags with spaces in them.

Is there a quick, easy way to replace the underscores with spaces when importing? Or for that matter, for replacing underscores with spaces in tags that are already in the DB, since I imported some of them without realizing that it would fill my DB up with these kind of crappy tags.

Is there a way for the tumblr downloader to ignore reblogs?

When I download from tumblr, I usually use creator:username. However, it also downloads all their reblogs, which could be other artists work, which then gets tagged wrong.

How noticeable is the difference between splitting the DB and media between an SSD and HDD (respectively) compared to just throwing the entire thing onto an HDD? Having the split setup would be more work but is it worth it?

I've finished all PTR sync updates but there are still images without tags and I can find them on boorus so they should have tags. Am I missing something?

Any good way to batch download tags from public boorus for files already saved by md5?

>>6556

I see how I can add tags with the md5 parsing script, but what if I want to do this on batch? say if I have 100 files and I want to automatically fetch the gelbooru tags of each one?

Is there a way to download images from this kind of archive?

https://archived.moe/h/thread/4556412/#4563025

>>6562

If you click on a picture you'll see that it redirects you to https://archiveofsins.com/.

Use this link in the page of images downloader:

https://archiveofsins.com/h/thread/4556412/#4563025

archived.moe does not store pictures so you'll have to use the correct 4chan archive rather than these aggregating archives that only store thumbnails. For example desustorage or the already mentioned archiveofsins.

Is there a way to (mass) copy the tag from online repository to local?

>>6565

did not know that was an aggregate site, honestly thought it was a redirect to an image to dissuade mass downloading as much as possible, thanks, Ill keep in mind this for the future.

Yo anons, is there some way to rename a tag? What I am doing atm is searching for all images with a tag, selecting them and replacing the tag. But it gets really boring when you have too much to replace (I fucked up trying to use some regex), and when you press F3 there's a UI that looks like should be a way to rename them. But if I select images with different tags and double click a tag, it gets removed and if you try to add a new one, it's added to all the selected images.

And, on another note, is using the tag "page:" really the best way to order all the images from a pool? What I am doing is using "pool:" to add them into a collection and "page:" to sort them.

I gather that the "correct" way to organize doujin and manga and what-not is to use a single title: tag on each page of the manga, then a page: tag on each page to indicate what order the pages of that title go in.

I've seen that it's possible to have these collections display in the search results as a single thumbnail, but is there a way to filter the results so that ONLY things that have more than one page are shown? Just filtering things that have a title doesn't seem workable since a lot of single images have titles in the PTR (and things like the HentaiFoundry downloader apply images' titles to them as a tag).

a question for those who know how the db works:

how can i get the filename or path of a file starting form the hash? the file hashes are stored in client.master.local_hashes right?

i looked at the source and the only function that gets the path is _GenerateExpectedFilePath() in ClientCaches.py, i tried to emulate it but didnt work.

Thanks in advance

>>6587

nvm resolved

>>6568

There is no way to rename tags atm. I have suggested a right click -> replace option before but it doesn't seem like a priority for the dev right now.

>>6617

Fixed it. For when it breaks again, the firewall had been blocking upnpc_win32.exe

>>6627

You need to enable it under options/tags

>>6640

Select the files, right click manage -> file's tags, choose "public tag repository" on the left, then petition the bad title tags. It sounds like those tags are basically just bad and not just something you specifically don't want, so I think the best option here is to try to fix the PTR rather than to hide the effects for yourself personally.

Is there no way to edit a tag after you've pressed enter? It's kind of annoying having to retype the whole thing every time I make a typo.

I have a local numerical rating service that I use to record the number of characters in an image. It currently goes up to 10, but I would like to increase it to 20 because I found some images with a lot of characters.

Is there a way to increase the maximum of a numerical rating from 10 to 20 without doubling all of my previous ratings?

>>6647

Rename your old rating. Make a new numerical rating with the old name that goes to 20. Then search your database for (old rating = x) for x in 0 to 10, and just set the new rating to that x for each x.

what is the sql query to get the hashes from the tags?

should something like:

get tag_id from subtags where subtag in tags

get hash_id from mappings where tag_id in tag_ids

get hash from hashes

this works only with tags to include, how should it work if there's a tag to exclude?

The deviantart gallery downloader is stuck at 400-something and shows "waiting on bandwidth" for the last hour. What gives?

>>6652

You probably hit your download cap. Change it under services > review bandwidth usage.

Is it possible to group alternates images next to each other in general searches?

I'm trying to setup a shortcut that allows me to press numpad 1 to set a numerical rating service.

The command says "flip on/off ratings "1.0" for 'global rating', but pressing numpad 1 doesn't seem to do anything.

How exactly do I make this work? Thanks

>>6661

nope but it's a planned feature

sorting by resolution ratio kind of does it now, if the alts are the same resolution

Is full audio support something that's expected to come soon? It's the only thing preventing Hydrus from being an ideal porn viewer.

Is there a way to have search return only one copy of a same quality image pair in searches?

Does danbooru have ugoira support? https://danbooru.donmai.us/posts/2007388

If i try to save the video it says sample, but using the view original link shows basically the same thing.

I'm thinking about switching from my own image organization app to hydrus, is it possible to automatically mass-assign tags based on file md5 if I have local dump (xml and plaintext) of a website?

Sorry if this was asked before.

>>6713

when importing images there is an option to look for files with tags, I wanted to make a tag pre-processor, but have no idea what hydrus looks for in these files, please tell when you get it to work and how it works

>>6760

pages > new special page > duplicates processing

>>6762

Seems it doesn't work for videos, but it's still pretty nice.

Thanks btw.

>>6764

Gee i wonder why

>>6766

I thought it might have something simple like comparing if one video is exactly the same as the other since I have some duplicate ones, but I guess not.

>>6769

The thing is that it can only determine two files are exactly the same if they have the same hash, and if they have the same hash then the file will only be imported once in the first place. The duplicates finder finds files that are visually the same or similar, but have different hashes. This is really difficult for files with multiple frames though.

Is there a way to sync a tag repository faster or in larger chunks?

>>6785

How long have you been running Hydrus? It takes a while for it to sync with the tag repository. As far as I can tell, what it's doing is not querying the tag repository to find tags for the images in your DB, but instead it's downloading every set of hash -> tag mappings in the public tag repo and mirroring them on your system. This is better from a privacy standpoint but it does mean that it takes a long time before you can be sure that you're seeing all of the tags on the repo.

>>6794

>>6785

Oh, and to add on to that, you can check your progress syncing with the remote tag repository by doing:

services -> review services -> remote tab -> click "tag repositories"

and it'll show at the bottom of the panel on the right side.

Can I download all the tags and the image from a booru link? As far as I know my only options are to download just the image or go find it again using the browser in hydrus

>>6797

This script might help: https://8ch.net/hydrus/res/1914.html

Is there a way to embed the tags into downloaded images as metadata? I just got used to using tag repositories for downloading tags along with images, and I was thinking it would be cool to be able to use those tags cross-platform (specifically, with Android) and not just on my PC.

Is there a way to see all tags when im going to set parents and siblings? because FUCK having to type all that shit out since I FORGET WHAT TAGS IVE CREATED.

Might be the most stupid question in this thread, but is there a way or a workaround to search for tag1 OR tag2, displaying all images that have at least one tag out of tag1, tag2?

my subscription url count constantly stops at 40-42, despite me being sure that I checked "get everything" this happens both at old, and at newly created. What am I doing wrong?

Is there any way to automatically tag a group of images with page: tags, in order, for things that are already in the DB? I know you can do it while doing an import, but I downloaded a doujin from a chan thread using the thread watcher, and was wondering if there was a way to label the pages without doing each one manually.

I'm trying to direct hydrus to rhythmbox on linux to play some mp3's and the it gives me an error saying no such directory even though i put in the right path under the options menu.

can anyone help me with regex.

I've folder contains sub folders in this pattern

E:/Copyright/artist/character name/image tags separated by spaces.ext

i can use hydrus import option in misc to import #1, #2 and #3 directory names, but what if i've this folder is #5 directory already (like on desktop)

what regex should i use, or where i can find reference to help me on this?

thanks!

Im sorry if this sounds dumb, but i cant tell the difference between the thread watcher and the page downloader when i dump threads from 4/8chan, could someone clarify the differences. Also how would i access all my now downloaded images in the specific albums they were downloaded in without apllying an album specific tag. also can i dump a character page from lelhentoi.net with this.

>>6864

The thread watcher will keep checking the page for new files until the thread becomes inactive. The page downloader is a one-and-done sort of deal.

>>6814

maybe a bit late, but in the settings you can set the dropdown to appear sooner/with less typed, maybe you can set it to 0 characters typed

Is there a way to see the current to do list? I'm interested in what is coming to hydrus and would like to see it!

Trying to build from source, getting:

Traceback (most recent call last):
  File "C:\Hydrus Network\client.py", line 12, in <module>
    from include import HydrusConstants as HC
  File "C:\Hydrus Network\include\HydrusConstants.py", line 43, in <module>
    import sqlite3
  File "C:\Users\Aaron\Software\hydrus-win-1.6\Environment\Python2.7\lib\sqlite3\__init__.py", line 24, in <module>
    from dbapi2 import *
  File "C:\Users\Aaron\Software\hydrus-win-1.6\Environment\Python2.7\lib\sqlite3\dbapi2.py", line 28, in <module>
    from _sqlite3 import *
ImportError: DLL load failed: %1 is not a valid Win32 application

I'm using the hydrus-win environment.I can import sqlite3 in python on its own, but when trying to launch client.py I get this error. Answers on stack overflow say its caused by a 64bit/32bit mismatch, but sqlite3 is a default module.Help?

>>6889

Yep, if you go to file -> shortcuts and click the ? button, a help dialog will pop-up. The paragraphs after "There are two kinds–'reserved' and 'custom" explain how.

>>6888

dont know what your error is but these are the steps ive put together to get hydrus running from source. maybe it'll help you until hydrusdev or someone else can fix your specific problem

Running Hydrus from source on 32bit python without installing compile toolchain
tested on win7 64bit and win10 64bit

install python-2.7.14.msi (download from https://www.python.org/downloads/release/python-2714/)
tick option to put python on PATH

install wxPython3.0-win32-3.0.2.0-py27.exe (download from https://sourceforge.net/projects/wxpython/files/wxPython/3.0.2.0/)

extract sqlite3.dll from sqlite-dll-win32-x86-3200100.zip (download from https://www.sqlite.org/download.html)
and drop it in C:\Python27\DLLs or wherever you have python installed. overwrite the one thats already there

client:
pip install PyYAML beautifulsoup4 psutil Send2Trash requests PySocks win_inet_pton lxml pafy numpy opencv-python Pillow hsaudiotag PyPDF2 comtypes
pip install lz4-0.10.1-cp27-cp27m-win32.whl (download from http://www.lfd.uci.edu/~gohlke/pythonlibs/#python-lz4)
pip install Twisted-17.9.0-cp27-cp27m-win32.whl (download from http://www.lfd.uci.edu/~gohlke/pythonlibs/#twisted)

you need flash activex installed (installed by default on win10)

you need ffmpeg.exe. get it out the compiled hydrus release and place it in Hydrus\bin


server:
pip install pycrypto-2.6.1-cp27-none-win32.whl (download from http://www.voidspace.org.uk/python/pycrypto-2.6.1/)
pip install PyOpenSSL service_identity

if you install the compile toolchain you wouldn't need to download the wheels (.whl files), you could just pip install lz4 twisted pycrypto. but thats for another time

>>6897

>Is there any way to search for file type?

system:mime

>>6897

2. You would have to set a meta-series tag or smth like that and add data to that namespace, like for example:

series-meta:trap
series-meta:yuri
(series-meta:...)

3: You can set tag siblings that do just that, I have requested the possibillity to set whole namespaces as tag-siblings, dev might have seen that, maybe not.

[i]services -> manage tag siblings[/i]

>>6904

Tag siblings don't actually rename tags and just display them as another tag. Un-setting the tag sibling will revert them to whatever they were before.

>>6909

Tag repositories are online tag collections that are completely optional and Hydrus normally doesn't connect to anything. Just click file -> import files and add your own images, then tag them by right-clicking and choosing "manage -> file's tags".

>>6909

I'd ignore tag repositories for now, unless you're really lazy about tagging your own shit and just want it done, regardless of tag quality.

>>6910

>>6913

So I have to tag everything manually? I thought this program could fetch the tags directly from the boorus, kinda like Grabber and Happypanda do.

>>6914

Booru downloading is also possible. Its in pages -> new download page -> gallery -> booru.

So, is this only for your stuff or you can get galleries from anons somewhere?

How?

>>6916

What is your collection mostly of?

imageboard memes?

or anime pictures?

If its the former, you either tag manually or use the repo which is currently completely unmoderated. Unless there comes a day where someone starts up a purely meme based imageboard, these are your only two options.

If its anime pictures, yes you can grab tags from the boorus, but its a bit more complicated than "Here's my picture, please grab tags from danbooru".

Y'see, hydrus doesnt actually recognize any pictures, it converts images to a hash string and compares it to other pictures-turned-into-hash-strings. So what this means is unless your picture is a 100% match of the one found on a booru, you wont get a match. Even one pixel out of place will fail a match. So if your pictures are crops(im pretty sure most uploads on chan threads go through some sort of compression algorithm, but dont quote me on this one)/edits/demotivationals, you can forget about it.

Now, lets say your picture doesnt have any of these problems, there's still one more hurdle to go.

Since there is no current infrastructure in place in hydrus to retrieve tags from other sources besides the repo, you will have no choice but to repurpose the image downloader to do so. Like the previous anon said, its under pages->new download page->gallery->booru. This is going to be a manual affair unless you have some scripting knowledge, since it relies on searching the hash string for each image.

>>6917

>Since there is no current infrastructure in place in hydrus to retrieve tags from other sources besides the repo, you will have no choice but to repurpose the image downloader to do so. Like the previous anon said, its under pages->new download page->gallery->booru.

Can a kind anon make a mini tutorial for this? So instead of using grabber to download any booru tag that interests me and only having the downloaded filenames as rudimentary tags, it would be better if they are downloaded and tagged by Hydrus instead.

>>6918

It's not hard, most boorus have a search by search by hash feature. So instead of entering tags, you enter the hash of the image you want tagged instead.

So here are the steps-

1. import image into hydrus.

2.right click the imported image, go share->copy->hash->md5.

3. open up a new instance of downloader, choose a booru as gallery (i suggest gelbooru)

4. under pending queries, type md5:<ctrl-v>

where <ctrl-v> is where you paste the md5 you saved earlier.

Now this will work for one image. if you have hundreds of images, its probably faster to grab the md5 hashes OUTSIDE of hydrus, but this requires experience with scripting.

>>6919

Please note, if the gallery does NOT have a search by hash feature, this will not work.

>>6919

Oh, I forgot step 5, which is to enable all the checkboxes underneath import options-tags.

If you do not enable them, hydrus does nothing.

>>6919

Fuuuug, I guess it's not viable to do this for entire folders. Hopefully an automated way will be implemented at some point.

>>6922

Had some spare time so I banged up a bash script that does the job.

copy whats between the double quotes into notepad and rename it to hash.sh, and drop it in hydrus network/db/client files

"#!/bin/sh

for dir in $(find . -maxdepth 1 -type d|sed 's/\.\/[tr].*//'|sed 's/\.\///');

do

for FILE in $dir/*;

do

hash=`md5sum $FILE|sed 's/\s.*//'`

echo -e "md5:$hash\r\n"

done

done > o.txt"

open up command prompt and move the directory hydrus network/db/client files. Run the script using

sh hash.sh

Should produce a file o.txt, where there should be the output all of the files converted into queries ready to send into hydrus for processing. Dont be alarmed if you see a lot of whitespace in the textfile, I cut some corners when writing the script so ignore all the lines that are missing things after the colon.

Good luck.

>>6923

Is there one like this for windows?

I tried making one using CertUtil but I don't know shit about batch.

>>6924

This is for windows.

>>6924

woops, I see what you mean. This is what happens when you dont test your own scripts properly. Gimme a sec

>>6926

Okay, I give up. I have absolutely no idea how to translate this into batch. Lets hope someone more knowledgeable can fix my mistakes

Is there any way I can add it so I can add tag to an image from a context menu?

>>6928

Sorry, it appears I had a stroke.

Can I set it up so that I can quickly add a tag via the context menu?

I just went through 150 false positives on the dupes system, is there a way to delete the found couples, the next time that I search too wide?

>>6927

>>6924

Fuck batch. I made one in python, it's slow as shit but that's what I get for not using JIT lua.

import glob, os, subprocess, sys

os.chdir("./")

if(len(sys.argv)>2): #for retards without python in path
	open("links.txt", 'w').close()

outFile = open("links.txt", 'a')

for file in glob.glob("f*/*"):
		hashProcess = subprocess.Popen(["CertUtil", "-hashfile", file, "MD5"], stdout=subprocess.PIPE)
		hash = str(hashProcess.communicate()[0]).split("\\r\\n")[1]
		hash = hash.replace(" ","")
		
		link = "https://gelbooru.com/index.php?page=post&s=list&tags=md5%3a"+hash+"\n"
		outFile.write(link)
	
outFile.close()

You just run it like normals and it creates hashes.txt with one hash per line, run it with 2 args(eg. hashes.py a a) to clear the file before writing.

Also, it only works if you put it in the /client_files/ folder and if every file you want to search is under a folder that starts with "/f "(which is the case for me and I guess for everyone unless dev changes it).

>>6938

Just noticed it can really fuck you RAM if you have a lot of files, here's V2.

import glob, os, subprocess, sys, datetime

os.chdir("./")

print(datetime.datetime.now())

if(len(sys.argv)>2): #for retards without python in path
	open("hashes.txt", 'w').close()

outFile = open("hashes.txt", 'a')

for file in glob.glob("f*/*"):
		hashProcess = subprocess.Popen(["CertUtil", "-hashfile", file, "MD5"], stdout=subprocess.PIPE)
		hash = str(hashProcess.communicate()[0]).split("\\r\\n")[1]
		hash = hash.replace(" ","")
		
		link = "md5:"+hash+"\n"
		outFile.write(link)
		outFile.flush()
	
outFile.close()


print(datetime.datetime.now())

It now prints the start and finish time for research purposes, you can just delete the prints and datetime if you don't want that.

Went over my 1650 files, took about 9 minutes. Good luck on bigger DBs.

>>6939

I just implemented some multi threading for a program that I made a while back based on a thread on /g/, I use this as dummy code to see how it works. Since you are waiting on the net most of the time you might be able to create like 50 threads or so for a significant speed up, I'm quite busy rn, but after friday I should have time to implement it for you if you can't get it to work:

import queue    # For Python 2.x use 'import Queue as queue'
import threading, time, random


def func(id, result_queue):
    print("Thread", id)
    slptime = random.random() * 5
    time.sleep(slptime)
    print("Thread ", id, " slept", slptime, " time" )
    result_queue.put((id, 'done', slptime))

def main():
    q = queue.Queue()
    threads = [threading.Thread(target=func, args=(0, q)), threading.Thread(target=func, args=(1,q))]
    for th in threads:
        th.daemon = True
        th.start()

    results = [0, 0]
    results[0] = q.get()
    results[1] = q.get()


    resultssorted = [0, 0]
    resultssorted[results[0][0]] = (results[0][1], results[0][2]) # we sort the results by making use of the fact that the first [0] slot in the table is the threadID, this ID we then use to determine
    resultssorted[results[1][0]] = (results[1][1], results[1][2]) # the spot it gets in the sorted table, thus making sure that the first [0] slot in that table is the one from the first [0] thread

    print(resultssorted)


if __name__=='__main__':
    main()

>>6929

Shortcut is F3 my good man.

Does the new downloader engine support logins for individual sites yet?

>>6945

if(podraceing):

kek

You should be able to run like 20 threads at a time, since most of the time you are still waiting on your internet, but it might be smart take a look at a task manager yourself to see how it is going on there

Any way to upgrade already existing images with better quality ones without having to tag the new ones manually? I can't find a way to import them and just remove the old ones. The images are identical except for resolution and file size

>>6957

pages -> new page -> duplicates processing sounds like what you want. If it finds a pair and you say one is better in the duplicate filter, it will transfer any tags on the worse image to the better image.

>>6958

>new page

*new special page, sorry

More just plain curiosity rather than a serious question, but does anyone have multiple hydruses running on their PC?

I like decentralizing stuff as much as possible, in this case it helps with load times(less files) and search(just "2girls" as opposed to "porn 2girls"). Currently, I have a porn one I'm slowly tagging, a semi-dead picture one(reaction pics and the like) I'll tag one day and now I'm tagging a third drawfag one with tutorial pics, infographs and the like.

>>6958

>it will transfer any tags on the worse image to the better image

O-oh. I didn't know that, I kept copying and pasting manually.

>>6958

Thanks! I didn't know a duplicates filter existed

Any way to create a parent/child relationship using wildcards? For example, I want all tags ending in "_bikini" to also have the "swimsuit" tag.

Creating the relationship like this isn't working:

Child: *_bikini

Parent: swimsuit

I've been getting 404 errors from all of my tumblr subscriptions lately. When I go and look at the blogs manually I see the images there, though.

Has anyone else been experiencing this?

>>7004

It's probably this >>6968

>>7005

Oh, I guess that'd do it. Hopefully that gets fixed soon.

Is it possible to download user bookmarks from pixiv?

>>7016

Dev is currently reworking the network engine, so if you ask nicely it might become possible soon, currently I don't think so

how can i start a Hydrus server over SSH and use the Client on my main Computer, executing the server.py or server give me error messages

So what guarantee do we have that the developer(or for that matter a virus) couldn't turn Hydrus into a ransomware attack?

>>7081

Couldnt you say the same about every other piece of software?

>>7081

Hydrus is open-source, so you are free to go audit the code and then build it yourself if you like, but there's of course no guarantee that the binaries don't have extra code clandestinely added in. But that goes for all of the software you run on your computer, unless you are considerably more paranoid than 99.999% of users.

>>7089

…put a minus before it in search?

sexy_ladies -animal_dicks

How does one properly edit pictures namely rotation.

>>7101

Hydrus doesn't have picture rotation though it would be cool as a view option

If you want to edit your Hydrus images you should probably export them (right-click: "share->export" or drag the images into another window) and open them in Paint or something.

Is there a way to transform tags as I parse them? Most boorus have the artist name as "creator:" but when I started tagging my collection I set the artist as "artist:". Is it possible to change "creator:" to "artist:" when parsing tags?

I have just now started trying out the client, but I can't figure out how the search system works within the GUI.

Let's say I have 4 images, two of Superman and two of Batman. All have tags of "series:DC" for example, but the Superman images have the tag of "character:superman", and obviously Batman has "character:batman".

Using the search bar, searching for both "character:superman" > ENTER, "character:batman" > ENTER, does not bring up all 4 images. In a nutshell, I'm trying to search for Superman OR Batman among all tags, not both Superman AND Batman in a single image.

(I'm sick, so I'm probably just missing something extremely basic - but, how does the searching system work?)

>>7106

Or, to put it another way: What is the proper syntax for searching for "[tag1] OR [tag2]" ?

Can hyrdus search like that?

>>7106

can't do or searches yet fampai

>>7110

Ah well, thanks for the reply. You said 'yet' - does that mean it's coming later on?

>>7111

I assume so, it's a common request. Keep in mind the Current to-do list has: 1,147 items so "yet" is relative

>>7105

Tag parents/siblings forget which one, should do the trick

Newfriend here, how do I download directly from boorus?

So far I only have figured out how to import folders, but it doesn't really work folders within folders either.

>>7114

I can't get it to work, if I have tried setting siblings between the "artist" and "creator" namespaces but it isn't working. Is there a specific way to set it up? I've tried setting "artist:*" as a sibling of "creator:*" but it doesn't change the namespace when parsing tags

>>7115

pages -> new download page -> gallery -> booru

Select whichever booru you want to download from and set the search tags.

>>7116

You have to do it with each individual tag I believe. If not, then you'd need to use * for wildcards but I doubt that's gonna work

On the 'page of images downloader' can I paste multiple urls at once or do they have to be done one at a time?

What do people here like to do with duplicate images where neither one is obviously better or worse than the other? There's a "same quality" button in the duplicates processor, but it seems like kind of a waste to keep both copies of the image around.

>>7156

Arrows move pending queries up and down in priority. The X deletes it.

>>7160

moreso what happened between version 272 and now that there is no support for https://rule34hentai.net/

Is there a way for hydrus to remember the sort and collections individually? So that when I restart the client it autmatically collects my comics tab by title? Same with the preview panel. Would be nice if it remembered its size after rebooting.

btw the default sort/collect option in settings doesnt work. You have to manually uncheck and check again to sort/collect.

Is there a way to enter my danbooru account information (like with pixiv) so that I can search and download restricted tags?

>>6795

Something I just discovered in relation to this is the network usage limits. If your client is refusing to sync any more, it might be because you have run out of network usage for the day. You can change those limits by going to:

services -> review bandwidth usage -> edit default bandwidth rules

When Hydrus is "processing" the synced data from the repository, what is it actually doing? It's currently sitting at 3/3993, my CPU usage is minimal, and it hasn't made any progress in the last half hour at least.

As far as I can tell it has just decided not to bother doing any work today.

is there a way to sort/search for files only downloaded from certain domains in hydrus? ie, pixiv danbooru etc

>>7209

Never mind, I just needed to redownload the tag after closing the tab with all of the tags selected in import options. It doesn't redownload any of the images, but gets all the tags.

>>7206

Nevermind, I've figured it out. There's a setting (on by default) to make Hydrus only do that kind of stuff when the computer is idle.

Does the local booru support a password or some sort of authentication?

I'm just getting started and still learning my way around everything before I get into the more advanced stuff, so I'm sorry if this is a crazy basic one. What's the easiest way to display the list of tags that I've used before so I can throw it on a second monitor and go down the list making sure I don't miss any of those tags for my new imports? Right now I have to have a tab open for files->system:everything, check the first tag in that, switch back to the tab with my imports, select all the ones with that tag and tag it as such, then go back to the first tab again to see what the next tag I used was, etc. Lotta tab switching that I assume I could cut down on.

sorry , but , uhm , whats , like , the , uh , point ? i can do te same thing with an usb stick and i get to carry files around

is there a way to download derpibooru yet?

>>7218

file -> options

Click the tab 'tags'

Under the "suggested tags" heading, there is a tab called "favorites". Add some common tags to this and they will appear to the left of the image's tags when you are editing it. Double clicking them will quickly add them to the current image.

There is another tab to the left of the image's tags called "recent", that just displays the last X tags you've used. The number of recent tags to show can be set on the same options page, under the "recent" menu instead of "favorites"

>>7144

I just pick one and say that it's better. "Same quality" just clutters things for me.

>>7222

see

>>4475

Either I got the script there or I modified it and added derpibooru

Is it possible to set a custom order for the favorites tags? I would like to use it as sort of a checklist but it automatically sorts them alphabetically.

I feel like I'm being really stupid, but how do I empty my trash? there are a lot of files in there, and they keep showing up in searches. I feel like this is really obvious and I'm not getting it

>>7226

Thanks, I couldn't find it and I am not skilled enough to do it myself, can you please post it here?

My question may be similar too >>6713

I want to start using a mobile gallery app that allows to save tags on a XML file using XMP metadata. Is this type of data readable when importing files?

Sorry if the question is too stupid.

FYI the app I'm talking about is this http://www.fstopapp.com/metadata/

>>7267

Traps or futa go under male for me.

You could just do something like gender:futa, gender:girl and then have both of those under parent tag gender:female. That way you could search gender:female for both, or just gender:girl for girls only.

Or you could have an extra nametag for females like female:futa, so you could either search by female:futa for futas, gender:female -gender:futa for females only or just gender:female for both.

>>7268

>gender:female -gender:futa for females only

*gender:female -female:futa for females only

>>7268

>>7269

I have now tagged them as gender:female and gender:futa, including a dick:futa and a dick:male tag for the times that there is overlap, thanks for the advice.

Are there tagging structures available that people can use? I created my own highly namespaced (that's now a word) structure for all my files, but that does mean that I have to change over all the tags that get from boorus for my anime type files. Are there published guidelines for tagging?

Related to that, when I eventually attach my system to the main tag repo for hydrus, will my tags get accepted? I mostly do this for myself. So not a big problem if not. When testing the repo I didn't find any tags for the images that I have. So I just made my own system, but would that be a problem in the end?

I'm having a problem with parent tags.

In the public tag repo, someone has decided that "cum in mouth" implies "cum" and "tongue"

While I agree with the former, I disagree with "tongue", yet I can't find this relationship in the list of parents. Why is tongue being suggested?

>>7273

"Tongue" is a parent to "cum on tongue" and "cum on tongue" is aliased to "cum in mouth." "Cum in mouth" should probably be changed to be a parent of "cum on tongue" instead of an alias.

When I search "sakimichan" on Sankaku, it turns up 1336 results. However when I put it into Hydrus' gallery downloader or a subscription, it only finds 493 images. What gives?

>>7280

>>7270

A tag redirector would be nice, can I connect to the main repo just by pressing "help → 'I don't know what I'm doing' → 'Just set up some (…)'?

I'd love to be of use for the repo at least a bit.

>>7281

I went on to the help files and added the tag repo to my services. I had an IO error during here: >>7282

but I'm not sure if that is related.

>>7280

>> neg ratio

m8 do you understand how ratios work?

Sorry if this question has been asked a million times, but how to deal with sequenced images in hydrus?

If i have a pack of images that go together (could be 2, could be 10, but i'm not referring to vncg packs even thought it could apply), how to make sure they stay in order when i change up the sorting method?

Should i just make up tags for the name of the sequence, and then the order and use it as a sorting method?

Are there better ways now? I don't experiment too much

>>7285

Yes, but it's simpler to have a default value instead of gbr:0.33 and gbr:0.25 in different files. Besides, 'neg' was already in use, so I just used it to keep things as simple as possible.

>>7291

I just use a series:, group: and NIS: namespaces,

series: is the main overhanging namespace for everything that is to be placed together.

group is for series inside a series think like series:name, group:sketches (or group:finished).

NIS: is the number beneath that, that last sorting attempt, for all the groups and series without a name, I use USID (unknown series id) and GID(group ID), I place those just below their 'normal' counterparts in the sorting line.

>>7292

>not being degenerate and loving penises 'girls'

>>7299

It seems like it's different versions of the same file having their tags merged together inappropriately. As long as you're only deleting the ones that aren't actually true for that version of the file I can't see why anyone could be upset about it.

I wanted to test the local booru hosting function, it seems to be disabled, is it?

>>7277

Answering my own question. It's because Sankaku limits non-logged-in users to viewing up to 25 pages of results, which means just under 500 image urls. Logging in fixes this issue.

The next question then is, is it possible to have Hydrus log into a booru account?

>>7305

It's in the works lad. In the meantime, you can skirt around the page restriction by using some fancy query-fu to break up a tag into 500 image increments. I suggest something like:

sakimichan filesize:1kb..150kb

sakimichan filesize:150kb..300kb

sakimichan filesize:>300kb

Feel free to play around with the numbers

>>7307

Did you really connect completely? when you have added the connection you have only done a percent of the work, go to: services → review services → remote

and take a look at the PTR, I guess you still need to download the updates and apply them, good luck, as this will take a long time.

Is downloading all images from one artist tag the same in booru downloader and via first subscription check, do the bandwith rules apply the same way? I have a list of artist I want to both download all images from and subscribe to and I'm wondering if I have to first download via booru downloader and then add subscription or can I just add subscription and select to download everything on first check?

Also is there any difference between having multiple queries under one subscription and each query under each own subscription or is it just for organization purposes?

>>7312

it's now downloading them part by part, it's about 1 gig or 1.5.

Then it will apply all those updates to your local repo, this is the lengthy proces, by keeping that page open you can see a status report. Also the downloading might stop due to bandwidth restrictions. You'll have to edit those in the settings (I think) when it becomes a problem.

Depending on your hardware the integrating of the PTR will take days to weeks, on my I3 and hard disk it has taken about 5 days already and I think that in another night it might be done.

Some tips: Do not interact with hydrus except for the stop button when turning your pc off, any other interaction might freeze the frames, it's still working, but you won't be able to see the status bars.

If the program does freeze and you need to close it just kill client.exe in task manager. it will delete the progress on the current update, but since there are about 4000 you need to go through you won't loose a lot of progress.

If you have a laptop you don't need to stop the program when putting your laptop to sleep, I have not yet tried to turn it off whilst the client was doing it's work, but I imagine it will not run smoothly.

>>7313

how would i manage the bandwidth usage? the "manage network rules" option is (UNDER CONSTRUCTION)

>>7314

as found by ding ctrl+F in this thread:

> services > review bandwidth usage

>>6658

>>7316

yeah, you can double click on that to edit it's specific bandwidth rules, or edit the default bandwidth rules with the dedicated button, there you can set up how much bandwidth every piece of hydrus can use, this is mostly useful when you live in a third world country and have bandwidth limits on your internet, and for making sure that servers aren't spammed too much.

>>7317

i live in a third world shithole but i have no caps on data transfering, feels good mang.

so i modified a few of those "caps", then i had to close the client and open it again, and now its downloading stuff again.

slooowly getting the hang of it i guess. my media folders are fucking unsortable behemoths at this point.

>>7319

Im pretty sure that particular limit is for the dev's server end to make sure no one user overwhelms his bandwidth. Nothing you can do about it except wait

>>7320

oh well, that explains it at least, thanks anon.

Before you go crazy removing bandwidth limitations, keep in mind that all of them are there less to keep from fucking up your data cap if you have one, and more to be polite to the people who own the remote servers and also to avoid getting banned for hammering them.

so now that I'm connected to the public tag repo, how do I go about uploading my tags to it?

Do it use: "services → review services → local → tags → local tags → advances service wide update"? it gives some warnings about not being undoable, so I want to be sure before doing it.

>>7308

pleas resbond

>>7325

I want to do it for my reaction images, I don't know if there are a lot of tags for these in the PTR, and I don't know if the dev wants these in the PTR, but I imagine it'll work.

>>7323

so i should lower my data caps then?

alright so i started tagging a few images now, when you "commit" the pending tags, it shares them across the entire PTR right? if so, i'll follow the tagging guidelines in the tutorial.

is it possible to host local booru now?

whats the difference between "pend" and "petition" when talking about tags?

im tagging groups of images with some general tags (ie: series name/franchise name), but some of them are already tagged, so it gives me the option to either pend or petition. whats the difference?

>>7332

Pend = Add tags, Petition = Remove tags

I forgot how to update withouth breaking anything

I was still on v242, how do i get to the current version safely?

I tried to install it on top of it but it encounters an error at v246

how can i take the tags out of an e621 image?

>>7339

I got it working, althought i'm not really sure how, it seems i just need to do a bit of maintenance in the old version before recovering the dabatase in the newer one

Is there a shortcut in media viewer to toggle between 100% zoom and fit to canvas? I'm used to irfanview having it

If there isn't can i create it?

i left my client open overnight, now i cant close it. I cant even kill it through task manager. its just stuck and is fucking invincible. help?

nevermind, process finally died, but now when i open it again its frozen on "booting db"

>>7341

you should be able to, open the media viewer:

move mouse to top of screen, press the keyboard button and select edit keybinds to see them, you have to select eh media viewer in the top box and press edit in the middle to see the specific keybinds, there you can see and set the key, it's z for me

>>7324

anyone? I want my sense of pride and accomplishment for adding all my tags

Is it normal that hydrus' image viewer "blinks" every time you change image?

It gets really annoying when you're fast forwarding throught your images

>>7346

Hit F3 on an image. and then change to the Public Repository in the column on the right. There's also the advanced operation button in the lower right if you wanna upload all your tags.

>>7349

Yes, but that's probably unavoidable since it has to load the image.

How can update the gelbooru tag database?

I only have the gelbooru2016.7z but new pictures obviously aren't tagged.

I haven't updated Hydrus since the last 8 months and I want to start with a fresh install with the newest version.

>>7246

https://pastebin.com/XtpZAp5D

is there any way to put the "page:##" tag on all the images of a comic faster than manually adding the tag to each individual image with the corresponding number?

im tagging up some small comics with around 20-30 pages each, and its getting jarring to go pic by pic putting the page number on each one. i'd loathe to do this on some of those massive japanese doujins that have 100+ pages.

>>7355

the only way I can think of is with filenames, try and download the set somewhere else with the number somewhere in the filename and parse those with regex

>>7356

you mean doing the "add tags based on filename" thing right?

>>7355

I just use add tags based on filename, and add "order" or "page" or whatever

what does namespacing mean?

>>7361

Have you even been to a booru?

>>7362

yes? i just dont pay that much attention

I have few subscriptions but they don't activate automatically when the time for a check comes. If I just open services/manage subscriptions, change nothing and just close the window they update immediately. Doesn't hydrus try to check the queries as soon as it launches? what is happening?

>>7361

a namespace is the part that comes before the ':' part, it's there for sorting and logic purposes, say you want to add some tags about the body of a person in a photo, would you want all those tags just hanging loosely, or nicely in one space, in the namespace body:

Another thing is sorting: you can add page:1 to a file, page:2 to another, add series:dicksrighthere to both, and then sort based on series-page to get them in the correct order in the file viewer. Thus allowing for easier reading and such

>>7365

thanks, i just wanted to make sure. there are a lot of unnecessary tags already in the shit im uploading, for example

charactername

character:charactername

i try to remove the "charactername" one and put in a proper tag whenever i see that, specially because most of those have very few hits. I mean i knew what it was, i just didn't refer to it as namespace until now. you can imagine how much of a mess my media folders are

is there a query that will display images with character:cirno excluding all images in which cirno is not the only character?

is it possible to download pixiv albums?

>>7370

Not yet. I recommend using pixivutil and then importing the folders

>>7373

F9 > Download > Gallery > Booru

>>7375

Welcome to the darkside.

>>7374

You just ruined somebody's life

>enter 2 characters to search for a specific tag

>program freezes as it loads the MASSIVE fucking list of tags that contain that particular string of 2 characters i just typed

this is a problem i'd like to go away in future versions

when i search for a set of pictures with a particular tag, why does it only show me a fraction of them?

for example, i have 500 pictures with the tag "character:anon", so i put that same tag in a files tab for a search and it comes up with a number next to the tag (ie: "character:anon (150-500)" ), then i enter the tag and it only brings up 150 of the 500 images i have archived with that tag. why is that?

i've tried with several tags of which i have numerous files and it does the same, if i have 1000 files, it brings up 300-something, if i have 150, it brings up 60 something, and so on. whats up with this?

>>7380

Do you have any tag siblings? Those fuck with numbers for some reason

New to Hydrus Network and can't figure out how to do this. Any help would be greatly appreciated. I partially followed some instructions in an old thread, putting the Danbooru tag archive's .db file in a freshly created client_archives folder. But after that when I go into manage services and start to add a hydrus tag repository I have no idea what to do, as the format of the UI is very different from the instructions I was using. Don't know what needs to be put in the address and/or access key fields for example.

My goal is to automatedly tag the pictures in my personal collection with all their danbooru/gelbooru/etc tags.

>>7382

i've tagged a lot of pictures with a few parent/sibling tags because they were the correct tags

from what i gather, "66" is the actual number of pictures i have, i don't know what 145 is in that example.

Is there anything like Hydrus and its tumblr downloader for reddit? Ideally something where you input a username and it rips all their submissions.

>>7387

…as text? Isn't reddit just text posts with links to other websites? I'd like to have .txt files in Hydrus, but for other purposes

>>7389

They added direct image hosting recently

>>7389

Sure, anything will do. I want to toss a username in there and it either download all submissions or give me links to them rather than clicking each link manually. JDownloader won't do this, so I'm stuck using the gonewild archives torrents. Which is fine, for any content prior to October. Just, if they're scrapping it, then surely there's a way I could do that myself for specific users.

>>7391

since that isn't working in hydrus currently I use ripme,jar, java sadly, but amazing program updated 1-3 times a month with new service supper.

is there any way to import all tags from my booru of choice instead of having to manually open the tag import window and ticking every box?

i want all those boxes to be ticked by default dammit.

>>7403

options -> default tag import options -> add -> [your booru of choice here]

>>7407

thanks famalam, i didnt see that