/hydrus/ - Version 557

Name
Email
Subject
Comment
Password
Drawing	x size canvas
File(s)	Drag files to upload or click here to select them
Spoiler

Version 557 Anonymous Board owner 01/03/2024 (Wed) 22:42 Id: cffa1a [Preview] No. 1575

https://youtube.com/watch?v=xNKXyr6YN8Y [Embed]
windows
zip: https://github.com/hydrusnetwork/hydrus/releases/download/v557/Hydrus.Network.557.-.Windows.-.Extract.only.zip
exe: https://github.com/hydrusnetwork/hydrus/releases/download/v557/Hydrus.Network.557.-.Windows.-.Installer.exe
macOS
app: https://github.com/hydrusnetwork/hydrus/releases/download/v557/Hydrus.Network.557.-.macOS.-.App.dmg
linux
tar.zst: https://github.com/hydrusnetwork/hydrus/releases/download/v557/Hydrus.Network.557.-.Linux.-.Executable.tar.zst

I had a good week back after the holiday. There are some bug fixes and improvements to system:hash parsing.

Full changelog: https://hydrusnetwork.github.io/hydrus/changelog.html

highlights

Tag filters (which operate the various tag whitelists/blacklists across the program) now edit much much faster when they are full up with stuff. You can paste 5,000 items into an empty tag filter in less than a second now, and removing one or a handful of items from a filter with thousands of items already in is now instant. Previously, these things could take many seconds or even minutes due to inefficient overhead.

CBZs that have four or fewer pages should now be recognised correctly.

When you copy 'system:hash' and the 'system:similar to' predicates to clipboard, you now get a much longer string that includes all the hashes in the predicate. These copied strings are all parsable, meaning you can paste them into the same client elsewhere or a different client entirely and it should all just work in a complete loop. The human labels are updated to give you more information, too.

There may be other predicates that are currently parsable if you type carefully but do not themselves copy a parsable string, but I am not sure which they are, so if you run into one, let me know and I'll see what I can do.

If you run from source on Windows--or you'd like to--but you haven't installed the extremely convenient but bulky-to-install 'Git for Windows', I had to go through the installer again to set up my new dev machine, and I wrote out a guide for the full 12-page wizard here: https://hydrusnetwork.github.io/hydrus/running_from_source.html#core It is actually easy to do, and almost everything can be left as default and things will be fine. In any case, Git is great and lets you update in about three seconds, so if you run from source on Windows, give it a go!

next week

My old dev machine died just before Christmas, but I bought a new one and am back to normal work. For next week, I'd like to get some sort of system:num_urls figured out.

Anonymous 01/06/2024 (Sat) 00:33 Id: 15bd0d [Preview] No.1576 del

Hi there, is it possible for a client to use another client as a file repository? If not, what is the purpose of "Copy Service Key" under Review Services->Local->file domain?

Basically I want the answer on how to actually do this thing mentioned in the "You don't want the server" page
>Do you want to use multiple clients and have everything synced between them? You don't want the server.

Anonymous Board owner 01/06/2024 (Sat) 21:53 Id: 66a034 [Preview] No.1577 del

>>1576
Not yet, but in the future I expect to link clients together via the Client API this way. Fingers crossed, you'll be able to browse another client's files as simply as selecting it as the location in the normal file service selector.

The copy service key button is just there because I think I have it set as available for everything. It mostly gets used for weird debug purposes atm.

Anonymous 01/07/2024 (Sun) 07:40 Id: 15bd0d [Preview] No.1578 del

>>1577
That will be good to have. In the meantime, is it possible to use the same db for both a client and server?
e.g. The client is all set up, I shutdown the client, point a server to the same db folder. Would this cause any problems, especially if I then stopped using the server and booted the client?
If it matters, it's dockerized, I'll probably just make a backup to play with, but I wondered if it's expected for the two to play nice with the underlying database.

Also really like what you're doing with the system predicates.
If this is already possible let me know, but I'd like a way to set custom shorthand for specific system predicates... like system:landscape for system:ratio=landscape, for instance, or something even shorter - it didn't seem like siblings work for system predicates (technically they're not tags anyways, so that part makes sense to me)

Release Tomorrow! Anonymous Board owner 01/10/2024 (Wed) 02:40 Id: 9a416c [Preview] No.1579 del

I had an ok week. I fixed some bugs, improved some quality of life, and figured out a basic first version of 'system:number of urls'.

The release should be as normal tomorrow.

>>1578
Unfortunately the client and server databases are not compatible. Even if you point both exes at the same directory, the client makes client.db and server makes server.db. They are very different programs and databases and outside of the normal repository syncing, they cannot talk to each other or do each other's jobs, I'm afraid. This has come up before, so some of the discord guys wrote up this document:

https://hydrusnetwork.github.io/hydrus/youDontWantTheServer.html

I quite like the idea of 'system:landscape' style shorthands. I'm in the midst of fleshing out the system predicate parsing system right now, and there's a ton I'd like to do with it once things are a little less thrown together (I'm rearranging a bunch of system preds, like system:hash last week, so they parse easier). For now, you might like to play with the 'favourite searches' button (the star beside the normal search input text box), which let's you save full searches, or templates of them, for quick loading of quite complicated search setups in just a couple clicks.

Anonymous 01/10/2024 (Wed) 02:47 Id: 15bd0d [Preview] No.1580 del

>>1579
That page is exactly why I am using the client and not the server.

I believe that page is written from a high-and-mighty misleading perspective, not to mention incredibly lacking in information.

So, my other original question. That page explicitly states that you can run multiple clients without the server and sync files between them.
>Do you want to use multiple clients and have everything synced between them? You don't want the server.

How do you accomplish this without a complex export/import process between the clients?

Anonymous Board owner 01/10/2024 (Wed) 18:59 Id: e4bac8 [Preview] No.1581 del

>>1580
>How do you accomplish this without a complex export/import process between the clients?
Unfortunately, there isn't a good way right now. I plan to, over the next year or two, buff up the Client API and add a way for clients to browse each other. You'll be able to load up another client's file services in the normal search and it'll all just happen over the Client API instead. Once that tech works, I'll be recommending people have a master client that does all the subscriptions and stuff, and then other computers can dial into that one and read its content live.

Although it sucks to say, I know some users who simply set up VPN situations to browse their master client from another machine or tablet.

If you have a relatively slender client, let's say less than 50GB total 'db' directory, another option is simply to have a master client and then FreeFileSync its whole install dir to your other computers once a week, simply cloning it.

Posting mode: Reply