is to possible to sync the public tag repo while using the client?

>>1498
Yes--if you check the PTR's page under services->review services, it should have 'process now' at the bottom. The button is only clickable if there is pending work to do and everything seems good to go. Hit that, and the client will start grinding processing work. It makes a popup with progress and current speed, too. It won't work as hard as it would in idle time (in order to not blat your browsing too hard), but it should let you keep using the client. You'll get up to 500ms extra lag on any changes you make, and every 90 seconds or so you might get a few seconds more lag as it saves to disk.

I had a good week. I fixed some bugs--including the recent problem with editing favourite searches--improved some UI quality of life, and added a new services object to the Client API.

The release should be as normal tomorrow.

other (advanced) highlights

The v527 update went ok, no massive problems, but I wish I had done a bit more testing. Some Win 10 users are getting a two-second delay on opening any page due it seems to a Qt bug that got fixed in the next update, and putting these new libraries in front of more eyes would have caught this. Therefore, I have made a 'future' beta build that is the same code as the normal release, but it uses the newer library versions I am planning to next update to, for instance Python 3.10 instead of 3.9 and Qt 6.5.0 rather than 6.4.1. I am not sure how often I will put these future previews out--maybe once a month, maybe every week--but for now, if you had any weird issues with the recent update like the two-second bug, please check out the 'future' version of last week's 528 here: https://github.com/hydrusnetwork/hydrus/releases/tag/v528-future-1 . I'd also like to hear from anyone who has an older or unusual system and just wants to try it out. I particularly want to hear about problems--if you need to perform a clean install to update, if you have an older OS and it straight up won't boot, just any feedback on what is good and bad, and I can tweak it before I roll the updates into the master branch.

Thanks to a user, the HTML parsing formula now lets you search the previous or next siblings of a tag (instead of just descendants or ancestors). If you know you need the 2nd 'p' tag after the 'div' you have, it should be pretty easy now!

next week

I'll keep at small jobs and cleanup. I realised just now that my new similar files paste button should probably take direct file URLs too, so I'll give that a go. It'd be nice to hammer out some more Client API stuff too, and I really need to catch up on github issues, but we'll see. I am not firing on all cylinders right now, so I am keeping it simple.

I had a good week. I mostly worked on quality of life improvements--highlighting large downloaders is now smooth and cancellable, and the file search tag autocomplete can now take pastes of multiple tags and will also parse typed system predicates.

The release should be as normal tomorrow.

I had a good week. There's some misc small work, a new build for advanced users to test out, and neat similar-files 'get source' tech that looks up files quickly from your clipboard, without having to import them.

The release should be as normal tomorrow.

I had an ok week that unfortunately had less work time than I had hoped. There's some quality of life improvements with snappier search updating/cancelling and more flexible system predicate parsing. The macOS App is also fixed.

The release should be as normal tomorrow.

I had a good week. There are two things happening tomorrow, and nothing else: there's a prototype new media player that I would like some people to test, and there are long-delayed library and build updates that will require users to update in a special way. Users who extract will be making a ‘clean install’, and almost all users will be updating their shortcuts to point to the new executable, which is renamed from 'client' to 'hydrus_client'. I'll walk everyone through what they need to do in the post.

The release should be as normal tomorrow.

I had a mixed week. I was a bit short on work time, but there are some bug fixes, including some more logic/label fixes in my new date calculation code, and quality of life improvements, such as your recent petition 'reasons' now being remembered in the manage tags/siblings/parents dialogs and easier column-size reset for multi-column-lists.

The release should be as normal tomorrow.

I don't suppose there'd be any interest in making it possible to store media in a blob table in the database?

That requires some tuning to retain good performance (mostly related to the in-memory cache size), but it'd make backing up a lot easier and partial/incomplete files from crashes and other annoyances would be gone. SQLite (https://sqlite.org/c3ref/blob_open.html) can do incremental IO in blobs (so can others) and Python (https://docs.python.org/3/library/sqlite3.html#sqlite3.Connection.blobopen) actively supports it.

I had an ok week. I fixed several bugs (including a really bad one related to deleting inc/dec rating services), added a couple unusual features like jpeg subsampling detection and File URL redirects, and updated several core libraries (e.g. Qt) for all users.

The release should be as normal tomorrow.

I had a decent week working on a bunch of different stuff. There's some bug fixes, quality of life improvements, and more work on timestamps--now you can import/export them with sidecars, merge some in the duplicates system, and store timestamps before 1970.

The release should be as normal tomorrow.

Is there way for nitter importer to apply post date as file metadata date and if it is possible for multiple images in one tweet to have relationship between each other?

How could I set Hydrus to download the entirety of e621? I'm new to Hydrus but want to backup the site due to how much seems to get removed from it

Database is locked after my pc was force shutdown yesterday. I have a backup from yesterday, but ideally, I'd like to restore all the changes I made after creating the backup. Any way to fix this?

I changed to a new PC and am trying to move my hydrus database/files to the new computer from backup. I want the contents of the database on a secondary drive while the program is on my main drive. I followed the help page to migrate the database location to my secondary drive (https://hydrusnetwork.github.io/hydrus/database_migration.html#intro). But, now I can't import my backup because it says my "database is stored in multiple locations." But there's only 1 location in "migrate database" window - the new location on my secondary drive that I specified. How can I import my backup to the database?

Hello
I lost my db, and so my backup is a few weeks old. all my files were on another drive.
how do I get the db to search through its client_files and import or find files not in the db?

I've been head-down on a rewrite of the timestamp management system (for editing import, archived, modified etc.. times), and while the work has gone well, I am not done yet. I only have some simple UI for editing a handful of timestamps, but I want full editing of all of them, and ideally sidecar and API support too. Therefore, there will be no release tomorrow--I'll work instead.

v523 should be on the 12th of April. Thanks everyone!

I had a great couple of weeks overhauling how the program manages file timestamps. Most of the work was boring behind the scenes stuff, but the upshot is a new dialog now lets you edit every file timestamp--import time, deletion time, archive time, modified time, and last viewed time, and I've added 'system:archived time' for searching. There's also an important bug fix for file deletion reasons and some quality of life improvements.

The release should be as normal tomorrow.