/hydrus/ - Hydrus Network

…

One-click start page selector.

A full-fledged search bar with back-forth buttons and tabbed autocomplete. I expect that with time Hydrus will poach more and more ideas from online browsers to increase usability in general.

Tags are interactive labels colored after corresponding namespaces and can be grouped into expandable groups which can also be pinned to list top.

Add tags on the go with add field at the bottom which also supports autocomplete. Delete them by hitting x on the right. Select several images to remove the same tag simultaneosly. To avoid accidental deletions hit the lock icon at the bottom of panel.

Pages are named after search queries to be more informative.

Tags can also be grouped by parents.

Change file's descripton at the special panel. Select several files to assign same description.

Hit the second button with circles at the bottom to see only the tags that all selected files share. Hit the third to see list without common tags.

Panel on the right contains all tags available for current tag service. Type something at the bottom field to find relevant tags and drag some images on one of them to add this tag to all selected images.

Tags are draggable themselves. Select several images and add ten tags in one drag'n'drop!

Adding tag by copypasting. What can be handier? Copy some tags and paste it all around with ctrl+v. Also submit them between available tag services

Try different type of deletions. Simply remove tag or be raducal and get rid of all images tagged by this tag. You can always bring them back from trash bin.

Assign common namespace to several tags at once or merge several tags into one.

Switch beteen siblings in one click from submenu. Assign tag relations with quick&simple copypaste.

Start looking up tags before hitting F3.

Logiacl operators. Self-explanatory pretty much.

Pre-assign some page parameters without having to hit some checkboxes. Pretyped missing part is also handy for quick editions.

Open Hydrus with page you need already there.

Change thumb pararmeters as you see them.

Remove gaps.

Set horisontal tiling.

Or both.

Vertical tiling is also neat.

Cycle tab to hide side panels first.

And all other next. Pure thumb browsing experience. Hit tab to back.

As an option for single tag galleies - header with tag description. Handy for unfamiliar tags from PTR.

Oh yes, favorites panel. Bookmark tags and queries across system of folders. Bookmarked tags will appear in "favorites" autocomplete tab. Folder can also host a list arbitrary sorted by user.

Folder also remembers thumb settings assigned for it.

Set to display icons and other service information in slug.

Assign ratings manually outside media viewer.

You can go pretty crazy and assign a ton of parameters for slug as well as status bar.

Query itself may contain lots of tags in different relations.

Paste tags copied from images, paste images themselves, queries from search bar or paste other folders. Paste your heart out!

Browse and edit URL sources in specified panel.

If slug wasn't enough for you, browse file info in pop-up tips.

Tip can partially replace 'current tags' panel presenting clickable tags.

If you put so much effort maintaing giant folder structure on your HDD, just store it in favorites. Also choose wether to preserve orignal date created/modified, if it is important for you.

Or you can parse folder names and set them as parents for their subfolders. Select any N-th directory and assign a namespace you want.

Import pages named after date of import instance. View different instances and manage import folders on the same page with special panel.

Fetch files with aplpha blending via system tag and control how transparency is rendered.

Previews sometimes can be less then super-effective. Help yourself with fancy mime-dependent overlay.

And/or set custom selection overlay.

Go further and use images as covers for other files. Undoable, duh.

Assign the list of programs of choice, instead of relying on OS defaults.

Copy-paste images between pages to add / remove tags according to keikaku query.

Be a nice sempai to your page-chan. Lock page to aviod changing its query, protect it from closure and pin to fix it at the start of page list. Or add it to favorites in any form.

Merge queries and revers order of multiple pages.

Open 'deleted tags' section of trash bin to browse your tag deletion instances. Clicking on them will display images you deleted the tag from.

Browse 'deleted hases' and deleted some latest if you delete those file by accident and want to re-import them.

Dupes panel. 'Preparation' section has never made much sense to me. Maintenance is a must anyway, so progress bar and pause button is all I need. Search results are grouped by class pairs.

Clicking on group displays images in it. Some options are available on panel and in context menu. For closer inspection open selected files in filter.

History panel is similar to the one in internet browser—it displays latest queries. Click to open or drag to favorites.

Switch between sessions with a manager. It it displays date saved and number of pages (and subpages).

Arrange order of namespaces when they are pinned to top and their appearance.

Another way to assign tag relations is by drag'n'drop.

Just drop tag in desired group. Namespaces can be assigned in the same manner.

When two types of grouping are on, a single tag may belong to both groups. Here is how you can handle this problem.

Censor and privatise tags.

Switch between databases on the go with a list.

This message appears when client fails to find database on startup.

Transfer files between databases.

Allow pagination and set number of items per page to quickly navigate across huge queries.

Power Notch mode displays the first item's thumb of page and lets you jump to page even if its number is omitted in main gui.

Media viewer contains simplified 'current tags' tab, but you can still add/deled tags and even browse tag history. The title displays index and page's query, while the rest of info can be found at the botom and on the right.

At the center of navigation keys is slideshow timer/starter. +/- 10 buttons appear when list contains 10+ items, +/- 100 buttons appear when list contains 100+ items and so on.

Context menu mostly duplicates bottom panel's functionality.

Drag tags on image to create notes.

Edit or delete notes by dragging outside image's margins.

Use Power Notch the same way as with pages aross vast lists. If item list is bigger than grid can handle, he Notch will freeze at center utnil the right / left end of grid appears on screen. It might take high sensitivity / big mouse pad to navigate acrose such lists in one move.

That's it for now I guess.

This is some pretty incredible stuff. It must have taken a long time–thank you for putting it all together.

I like a lot of these ideas and have saved the thread and the images themselves. Due to time, software, and skill constraints I don't think I can do them all complete justice in any reasonable time, but the workflows here make a lot more sense than what we currently have. I will try to push in this direction.

I especially like >>8187 as something I could do pretty quickly. We certainly have the screen estate to show more buttons, and if I am careful, I can keep easy keyboard navigation as well.

How did you mock these up, btw? Is it some UI builder or straight into photoshop? Are you a time traveller? Are you me from the future after I learn proper ui design?

Off-topic, on-topic post.

I recently had a dream with Hydrus Network, in my dream it was v350 or something like that, the icon wasn't a square anymore, it was a transparent version of the logo, and it had a dark UI by default, and so many new features, I was very confuse in my dream.

What a dream.

Sorry for the late reply lads, this week is killing me.

>>8264

I will see if I can hack together something simple for favourite tags. A tab in the autocomplete dropdown seems like a good solution. I will see if I can get it to play well with the floating window and all the focus stuff.

I am glad you like my software! I don't feel particularly stand-out myself–I am just trying to get some interesting code out there. THE IMAGEBOARD COMMUNITY means a lot to me, so I ultimately enjoy working on it. I like that my efforts have helped the memes to flow.

I guess .ai files use a similar header to .pdf, since they are both Adobe? My importer basically sends files off to a bunch of libraries in order saying:

You think this is an image?

no

You think this is video?

no

You think this is audio?

no

You think this is flash?

no

You think this is pdf?

yes

So I am guessing my pdf library, PyPDF2, is false-positive identifying it as that. I would be interested to know how many 'words' your ai files appear to have–I can't get an accurate count with PyPDF2, so I go num_pages, which it does provide, multiplied by 350. I wonder if num_pages here is being pulled from a crazy value or instead like the number of layers in the .ai.

I don't have any experience with translating software. I understand my gui library, wx, has support for it but I will have to rewrite the way I generate all the text in the program. I would like to do this eventually, and I would love yours and anyone else's support in this, but I am still recovering from the recent wx update and have to try out and get used to several other new features first.

>>8304

>>8309

I dream about the network or database code going catastrophically wrong, sometimes.

I can't promise anything, but I hope to continue working on hydrus as long as there is stuff to do. And since my todo list keeps growing faster than I can clear it out, I assume the ride will only end when I do. My personal situation is mostly stable and my schedule is mostly healthy. Even on weeks/months where I get blatted by IRL shit–like just recently–I am still enthused about hydrus.