/hydrus/ - Hydrus Network

Ok with being able to save page of pages as their own sessions, i'm splitting my current cluster fuck session up into other sessions that are ~40k images in total, this should eliminate most of the programs issues with massive amounts of files being open

Is there a way to change what a page of page is saved as though? right now it saves it as

[user] pop name

Personally I want to change [user] to [PoP]

While im talking about saving sessions, I have started to really only save 2 kinds of sessions prior to this version, 1 was the pre upgrade session and 2 was the date and time save.

Is there any way to lets say change the default session name to the current date and time?

>>7423

Is the unsetting debug option in yet? i'm not obviously seeing anything, but I will also admit I didn't comb through everything.

>The shortcuts system now allows you to set 'content' shortcuts (tag and rating mappings, typically for the 'media' shortcut domain) to explicit 'set' in addition to the old 'flip on/off' behaviour. Set will only ever set on.

I can't seem to understand how this works.

I have a shortcut inside media that reads:

>shortcut:f1

>command: set ratings "1.0" for x

However, upon press F1, nothing happens, even after refreshing. I tested to make sure the f1 key worked with another basic command, and it seems to work.

So I'm wondering what I'm doing wrong.

>>7498

Going through my session i'm noticing that top level page of pages don't have anything in front of them, but if you go down a level they get [user] im front of them.

I honestly like this, but wish I could change the [user] bit if only to make them easier to differentiate from actual sessions.

Ok, had an issue thats ultimately my fault however it lead to a suggestion

So, my session I found out when I consolidated everything was 460k images big, ok, thats fucking huge

I have my session saved, I went onto splitting everything into chunks and came out with 15 chunks, all under 50k but because of how some things were sorted some were 8000

so, now to get a new blank tab

-and I want to make a suggestion to have a 'restart as blank tab' as the last time I I did the test, loading up 60-100k images, then removing those tabs did not improve performance, this may have changed, but I would still prefer to start fresh

So getting the blank tab loaded, I go to load up the first image set… and 10 seconds later I realize I fucked up and saved over the session.

I know, my own fault, however I want to make the suggestion to have this

have save session have the first option 'save current session'

a second option 'save session (session name here) again'

This would be an option that is greyed out until the session is saved or loaded, this just makes it easier when going though the session menu seeing as I personally have 38 sessions saved, around 20 of which would still be relevant if I cleaned everything up.

Then another menu that would drop down with a list of sessions so you could save over a different session if you wanted.

so yea, back to loading the 460k session to resave the one I fucked up on…

Ok have a question, with more and more of a move to adding more websites to where images are found, is there a method, let's say if the hdd dies, and you don't have a backup, to full the images from where they were found?

I know a good deal of images I have wouldn't be able to be imported through a method like this unless you implement redirects for 4chan archives, but another good deal could be recovered this way.

Im asking because I don't think I will ever have the spare money to just have working backups, and instead I do have the money for a rollback backup (my current hdd is an 8tb, that has a backup of a 4tb drive, that has a backup of a 1.5 700 300 and 250gb drives, a backup like this) and Just got confirmation of a 3tb drive i'm pretty much going to exclusively use for hydrus archives, and seeing my current game drive that is acting as my hydrus archive is sitting at 700gb I want to get it off there and onto the 3tb as fast as possible, then i'm also thinking of cold storageing my hand picked saved images to blurays, and deleting that hdd backup too as its sitting in the 180-240gb range.

but i'm thinking that if the hdd fails within the first 3 months, when crib death usually happens, how recoverable would the archive be?

I am the anon who cannot download files. Here's log from this version:

2017/12/14 16:49:36: booting gui...
2017/12/14 16:49:36: The client has updated to version 286!
2017/12/14 16:51:09: Traceback (most recent call last):
  File "/opt/hydrus/include/ClientNetworking.py", line 1556, in Start
    response = self._SendRequestAndGetResponse()
  File "/opt/hydrus/include/ClientNetworking.py", line 1902, in _SendRequestAndGetResponse
    response = NetworkJob._SendRequestAndGetResponse( self )
  File "/opt/hydrus/include/ClientNetworking.py", line 1087, in _SendRequestAndGetResponse
    response = session.request( method, url, data = data, files = files, headers = headers, stream = True, timeout = ( connect_timeout, read_timeout ) )
  File "/home/pat36/.local/lib/python2.7/site-packages/requests/sessions.py", line 457, in request
    resp = self.send(prep, **send_kwargs)
  File "/home/pat36/.local/lib/python2.7/site-packages/requests/sessions.py", line 546, in send
    while request.url in self.redirect_cache:
TypeError: argument of type 'NoneType' is not iterable

2017/12/14 16:51:09: Failed to refresh account for public tag repository:
2017/12/14 16:51:09: Traceback (most recent call last):
  File "/opt/hydrus/include/ClientServices.py", line 876, in SyncAccount
    response = self.Request( HC.GET, 'account' )
  File "/opt/hydrus/include/ClientServices.py", line 773, in Request
    network_job.WaitUntilDone()
  File "/opt/hydrus/include/ClientNetworking.py", line 1696, in WaitUntilDone
    raise self._error_exception
TypeError: argument of type 'NoneType' is not iterable

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.

http://www.mediafire.com/file/jodg6u33idbjep6/hydrus-286.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.

Hi there.

I just saw this change in the 285 changelog:

"like the recent subscription query randomisation update, subscriptions themselves are now synced in random order (this stops a subscription named 'aadvark' always getting first bite into available bandwidth)"

Is there a way around this?

I named most of my subs in a way that they are done in a certain, deterministic order. And I've grown rather fond of it working that way.

>>7498

Thank you for this report. The [USER] stuff is supposed to be hidden–it is a quick hack to test if a page name is user-created. I will check this week.

You can only set the the default session name under options->gui at the moment, but I am open to new thoughts. Can you explain your date/time session names a bit more?

For the set/flip stuff, I skipped the debug step entirely and just wrote set/flip support into the shortcut system. Please check your shortcut settings and change the new set/flip dropdown in the edit shortcut action panel. The ui is a bit ugly as I just threw it in, but I think it works ok! Let me know otherwise.

>>7499

I think the 'content' shortcuts only work in the media viewer at the moment. I need to put a bit more work in to get them to work at the thumbnail level, which I assume is what you are seeing. If you open a file up in the media viewer, does your F1 work then?

>>7501

Thanks. I think moving to the client being aware of the current session is one way to go. At the moment, it doesn't know that stuff, and only keeps track of 'last session'.

I am also thinking about just moving to the new page of pages sessions stuff, to treat sessions less as full 'replacing' sessions and instead just as handfuls of pages.

>>7502

There isn't a recovery like this, but it might be possible. I can't really promise anything when it comes to recovery operations, as by definition things are broken, and fixing from other data is often more complicated (and hence more work) than one expects. My main advice remains to set up a decent backup for all your documents and everything, not just hydrus. A WD Passport is usually a good cheap solution.

There will be more url actions as I develop the domain manager. Maybe a mass-export of known urls will serve what you want?

>>7503

Thank you. I will look into this this week.

>>7505

Thank you for this link. I will check this out!

>>7509

I will add an option to revert back to the old behaviour. Let me know if you have any further problems!

>>7513

with the date and time, really when I save a session it's either me upgrading the program, or me going though a session and wanting to save a progress report as up till the saving page of pages, I was always riding the line of stability.

I, at least in my case, find just a date and time being the most useful session name, as if something fucks itself, I can restart blank from a time when it didn't fuck itself.

So with that in mind, a setting that you could check that when you went to save a session, it defaulted to the date and time, like 12-18 10:53 as an example, if you want to append more data to that, you could, but if checked it would just add current day and time, possibly a few different ways like instead of 12 it would say december, or dec, possibly a 4chan x style where the user puts the format they want in there for it to pull from.

On the recover side, im thinking more like this

hydrus knows if an image is not there, or can check

Hydrus also knows the url it was acquired from if it was a url

If the file is gone, and it sees it was acquired from a url, could it try to recover the image from the url, or potentially urls.

At least the way i'm thinking, I have the thumbs and client in a separate area then the full images, if the full image drive goes kaput, a way to recover from that.

thinking more on this, would it be possible when a full image is gone/not findable to move the thumbnail of said image to the dup finder/archive? that way it would at least be in the archive, and if the same file was found again in the future, the dup detector sould show them as exact or close to exact matches.

Forgot, I was here for a reason.

So I have an issue coming up and i'm wondering if there is anything that can be done.

I'm moving from my 1920x1200 to a tcl p 605 55 inch 4k tv for a monitor, my every metric it should kick my current displays ass handily, however, thumbnails are currently 200x200, I don't think that should change, however, you are looking into thumbnail stretching of sorts to fit an area

with this would it be possible to define the area to be larger than 200x200, like 300x300, 400x400, 800x800 and have the thumbnail stretch to be bigger? hell in the case of 400x400 or 800x800 it may actually be better to use a thumb as a placeholder and bring in the full size image

For me because i'm going to be using a 55 inch 4k tv, and if that fails me utterly, a 43inch 4k monitor, the size of the image will be roughly the same as it is now on my 1200p, but on an actual 4k monitor that is 24-30 inches, the size of the thumbnails must be pretty painful.

>>7513

>I need to put a bit more work in to get them to work at the thumbnail level, which I assume is what you are seeing

Yes that's what I meant. Can't I just like, pay you some money to prioritize my feature first lol

>If you open a file up in the media viewer, does your F1 work then?

I'm sure it does, but I don't use the media viewer since I'd rather view the images in irfanview

>>7523

>I looked at this today. It has I think 12 frames, but my ffmpeg reports fps of either 2.5 or 50, wew. Since it has a duration of about 10.9s, I am not sure how to combine 2.5 and 50 to make the correct result. I appreciate this example, and I will hang on to it for the next time I do a proper pass over the video parsing. Perhaps a new version of ffmpeg will suddenly fix this magically.

>I noticed it is a danbooru webm of an ugoira–I guess their webm generating script gives some funny framerates or something? I don't know much about that stuff, or how widespread this problem is.

I wouldn't know either, nor do I need the file itself as I have the ugoira zip; I just got it from the download, saw it was broke, and thought you'd be interested. The actual ugoira is here https://www.pixiv.net/member_illust.php?mode=medium&illust_id=63819160 if that would somehow help your testing.

Speaking of ugoiras: would it be fine to import the actual ugoira zip files for whenever the eventual ugoira->apng thing gets worked on, or would it be best to keep them out of Hydrus altogether until that time comes? Just curious as to what would be the best approach.

>>7524

Thanks.

For whether to import, if you would like to use hydrus stuff like tags to find those files again, for easy upload somewhere else or associating zips with an external ugoira viewer program, then please do. But make sure to keep track of them by adding an 'ugoira' tag or similar, so that when I add ugoira support, you have a way of finding them all again so you can help hydrus reclassify them or convert them to apng or whatever. I am not decided on how ugoiras will be handled in the client yet, but figuring out whether an arbitrary zip file is an ugoira is probably a bit difficult to automate, so make sure you have a quick way to find them all again. Although, having said that, if you don't have any other zips in your client, then the whole issue is probably moot.

If you don't have immediate need for hydrus searching for your ugoiras, I recommend you not import them yet. Just keep them in a 'import later' folder somewhere.

>>7523

I don't know what happened. My package manager thought I've got requests 2.18.4-1, but I had got 2.4.3. I updated it, but still got the error.

>>7533

2017/12/20 23:04:13: Traceback (most recent call last):
  File "/opt/hydrus/include/ClientNetworking.py", line 1556, in Start
    response = self._SendRequestAndGetResponse()
  File "/opt/hydrus/include/ClientNetworking.py", line 1087, in _SendRequestAndGetResponse
    response = session.request( method, url, data = data, files = files, headers = headers, stream = True, timeout = ( connect_timeout, read_timeout ) )
  File "/home/pat36/.local/lib/python2.7/site-packages/requests/sessions.py", line 457, in request
    resp = self.send(prep, **send_kwargs)
  File "/home/pat36/.local/lib/python2.7/site-packages/requests/sessions.py", line 546, in send
    while request.url in self.redirect_cache:
TypeError: argument of type 'NoneType' is not iterable

2017/12/20 23:04:13: 
2017/12/20 23:04:13: Exception:
2017/12/20 23:04:13: TypeError: argument of type 'NoneType' is not iterable
Traceback (most recent call last):
  File "/opt/hydrus/include/ClientImporting.py", line 5032, in _CheckThread
    network_job.WaitUntilDone()
  File "/opt/hydrus/include/ClientNetworking.py", line 1696, in WaitUntilDone
    raise self._error_exception
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 242, in run
    callable( *args, **kwargs )
  File "/opt/hydrus/include/ClientImporting.py", line 5541, in _THREADWorkOnThread
    self._CheckThread( page_key )
  File "/opt/hydrus/include/ClientImporting.py", line 5117, in _CheckThread
    HydrusData.PrintException( e )
  File "/opt/hydrus/include/HydrusData.py", line 936, in PrintException
    stack_list = traceback.format_stack()