The last few weeks have seen a complete overhaul of both the UI and previously-disclosed multi-map system. The new multi-map/docking/panel system, as well as the new UI, require somewhat of a rewrite of my initial next-gen rendering engine from the fall, because we now (advantageously) render everything within one full-size GPU swapchain output window, including the GUI, as opposed to the old design where different maps were different Windows OS-level panels having different swapchains and using native Windows UI controls (CPU rendered).

I will have more details later about this incredible refactoring which has extreme promises for:

  1. Speed/efficiency - especially minimizing CPU rendering overhead in multi-map scenarios
  2. Rapidity of UI development as I build concrete features
  3. Visual appeal of the UI
  4. Ability to have rich on-map graphical editors/UI components directly related to map content, minimizing user navigation
  5. Ability to customize UI for space-constrained/mobile applications where WSV3 Tactical Mesoanalyst might not be given full monitor space by the user
  6. Ability to add rich GPU-rendered chart plots/graphs for numerical timeseries point data
  7. Detethering from Windows to facilitate eventual native versions for different platforms, if ever deemed desirable
  8. Massive further minimization of EXE file size for fast, efficient updates
  9. Launch time reduction



As a result of the major beneficial refactoring (which adds 1 - 2 months of work before I can resume work on the core raster and vector content rendering engines), I am likely abandoning the idea of intra-map quad/dual parameter display. Instead I think there should be one high-level display mode selection wherein quad panel is actually Map 1-4 (separate "maps") and a flexible multimap mode allows the user to dock/tab/orient/resize/drag map window panels at liberty.

This reduces more complexity than it adds, but it does add the complexity of allowing the user to customize synchronization between dual/quad map panels.

Back in the original design where one single map could have intra-map same-view parameter display, the camera view/probe tool/timeline state/style would have all automatically synced across the 2/4 param display within that particular map.

Now since I am abandoning the "intra-map multi-parameter" concept and simplifying to a high-level selection of "single map, dual map, quad map, custom windowed multimap" (or something similar), this means there should be flexible controls for when the user wants to synchronize camera/style/timeline state.

I think it can be as simple as checkboxes in dual or quad mode for synchronizing each category. If you have a quad panel display, you probably want either all 4 to synchronize or none to synchronize, so you can take advantage of the unique ability for multiple maps with distinct camera views, timeline states, and even styles. This is incredibly powerful.

If there is a need for, e.g., keeping the top two panels in quad display automatically synced but the bottom two non-synced and independent, then the necessary extra UI controls allow these complex options can be safely hidden under the multi-level collapsible "COMMAND" side menu you see in the screenshots above, which is super easy and fast to navigate on a laptop using two-finger up/down scrolling. But the common case is that the user either wants them all synced (4 maps only differing in data parameter but not view/timeline/style), or complete independence between maps.

Ideas/comments/discussion/user requirements welcome.

This looks polished, professional, modern, slick and streamlined. I cannot wait to test drive this! Great, exceptional work, sir!

    Kirk It is motivating for me to hear that even at such an early, unpolished stage. It was 2 weeks ago today that I ripped-out the old VCL-based UI and effectively restarted a large portion of the rendering engine as a lightweight raw Win32 app with no VCL UI framework, but everything solely in graphics, including UI.

    I plan to implement octo-map display as well:

      Octo?! That’ll look amazing on a 4K display!

      Just curious. Is it possible to make it so one can have different viewpoints in the panels? With as many as eight panels, being able to have different viewpoints could be useful for viewing micro scale, mesoscale and synoptic scale features simultaneously.

        Kirk Absolutely. Diverse-view multimap concurrent display is one of the signature features of WSV3 Tactical 2025 and the new hardware GPU map rendering engine. See my discussion of this concept above. The approach I am going with for now is a simple checkbox "Interlocked" which will lock all display panels in camera view/timeline state (with the intention that content engines, such as raster data, will have support for submitting diverse content across map slots 1 - 8 for concurrent multiparameter display). If this is unchecked, all of the display panels are fully separate maps with independent style, camera, timeline, and data. I will consider later whether additional logic for synchronizing only certain display panels becomes necessary.

          A "mapset" will contain 8 maps, as this maps well to hardware, visual processing, and maximum perception of diverse data representation across different panels of the same map view. The user can change on the fly to view one, some, or all maps within the mapset.
          A "Multimap" screen (possibly to be renamed) allows the user to create arbitrary 2D templates/arrangements of maps, with a "view" mode contemplated, to which 2D vector graphics such as text, PNGs, and scrolling marquee bars could be added.


          This looks amazing! If it looks so good at that early stage, how good will it look when finished? 😃 You are blowing me away with every screenshot post.

          That looks absolutely insane. Very impressive! I’m really loving where things are going for WSV3 Tactical. This is practically studio-quality, and I honestly think everyone, from smaller stations with smaller budgets, to online streamers, chasers, the casual weather hobbyist, etc. will all flock to this product once it’s fully released. Amazing job Paul!!

          I like WSV3 with all of the features it has now. It gives you all the tools just like your local meterologist has access to. I look forward to having WSV3 just like it is now except it would be nice to being able to have a linux version that will run natively on any linux distro but offers all of the data and features that you can access now on the windows version. Also a feature I would like to see in future versions would be being able to zoom down to street level and show street names just like your local meterologist can do on tv. This would be useful during severe weather situations where you see Future plans for WSV3 is promising. WSV3 is the only program that I need to have a native linux version for so I can dump Windows altogether for good. Other than that the future of this program is promising for those of us looking to dump windows permanently.

          I am incredibly excited about this. It's going to be such a massive upgrade from the current version of WSV3.

          This looks stunning to say the least. This is stuff that used to only be available to the very fanciest of TV stations... and then it was a matter of whether or not the meteorologist would show it on TV.

          18 days later

          I have been offline for a while. Has 'plugin' support been talked about? If not, what language would you support? From my ongoing research into Delphi apps, it appears you can pear the C family of languages with it.

            TheDopplerKid I can dynamically link with any 64-bit Win32 DLLs compiled with any language. I think better than a traditional plug-in system will be the development of a new optimized general content format, better than Placefiles, allowing for arbitrary raster/vector/text ingest through the web system. It would be binary based and likely use Sqlite as an efficient container format with unbeatable language support and existing documentation. These are post-initial-release considerations as I've taken the time from now through Decemeber 2025 to optimize the core product value offering as much as possible, but not too early to hear ideas.