Simplify the Setup?

[thestumbler]

I really wanted to make this work a couple of years back on a project with UDP datagrams. I got like 90% there, but eventually did a custom Python / Matplotlib solution. I wasn't 100% sure I understood how Serial Studio worked, because still had small issues from time to time, and also because doing it in Python myself (for a single custom project) turned out to not be that difficult.

Fast forward, I'd like to give Serial Studio try again, this time just using serial data. I find that I'm facing some of the same parsing challenges as before. It seems that is is still difficult (impossible?) to specify the delimiters such that a single line of white-space separated fields is recognized correctly. Looks like the custom JSON parser is the way to handle this, but I couldn't figure out how to specify using the custom parser or not. My first test was very simple, it turns out if you don't specify ANY parameter to the split() function, it parses white-space separated fields perfectly. When you put a space as the argument, you get a whole bunch of empty nonsense fields. I made this trivial change to the parser, but either it wasn't using this parser or I had some other problem.

Anyway, skipping that problem temporarily, I added the default delimiters to my line of text and now Serial Studio parses them correctly. Now I'm struggling with the JSON format definitions. I can 100% see that Serial Studio will display and plot my data as I need -- I think all the features I want exist and work. But setting up the JSON file is confusing and the documentation of the format seems to be mixed between the old and new styles.

Ignoring the display aspect, just the field definitions themselves are painful to setup. My telemetry has 20+ fields, and because the project is in development, these fields will change around, be expanded or reduced, redefined, etc. If I insert a field in the beginning of the line, suddenly all the remaining field numbers get shifted. Then I have to manually edit the JSON file for a dozen or more fields. I'm also having a tough time wrapping my head around groups, group datasets, datasets, widgets, data plots... Clearly there are good definitions of these, but I haven't grasped it yet. As well, the automatic project generation wizard doesn't make this any clearer, at least to my thick head. And furthermore, the automatically generated file has a lot of fields which (I think) don't apply for the majority of cases. So when you go to edit it, it looks like 90% default values and only 10% important information (in my case).

I wonder if there could be an even dumber wizard that would help the community? It seems to me that it should be possible to start plotting data almost immediately with very little effort... I'm thinking the moronic Arduino serial plotter, for example. My telemetry sends on startup, and can optionally send on command, a header line containing the column titles (field names). It would be nice if Serial Studio's "dumb wizard" could pick up on that and generate a minimalistic JSON file as a starting point for the user to edit. Or not even edit, if the project just needed something quick and simple. Maybe the thing I should do is write a simple Python preprocessor that reads my telemetry for a few seconds and generates a Serial Studio JSON file -- once I figure out how to make the file, of course.

Let me end by saying I really LOVE this project, and my remarks are meant as friendly suggestions -- not criticism. Serial Studio clicks so many boxes for general purpose telemetry plotting. If it could be made easier for the casual user to get started, I think it could help a lot more people.

p.s. I would like to see a comment feature added as well, like in a bash or python script. Such a character could be user-definable, but say the hash character.

[alex-spataru]

Hi @thestumbler,

Thanks for your feedback, and sorry for the late response. I’m currently redesigning Serial Studio to make the project setup easier, even for large projects. I recently had a breakthrough, inspired by regedit and code editors, which led to a better way to display and manage complex data structures. Here’s a preview of the new project editor (WIP):

I’m also planning to add a tree-like or table view for editing group & dataset properties. This will save screen space and make it easier to see all available options and features. Using this approach, I can also add an additional "help" column next to each option to help new users.

I also agree that the documentation needs improvement. I’ll be rewriting the wiki and might create some tutorial videos to help users get started. Thanks again for your support and suggestions—they’re helping make Serial Studio better for everyone.

By the way, here is a screenshot of the new main window, let me know if you like the changes so far :)

Best,
Alex

[alex-spataru]

By the way, I haven’t committed or made these changes public yet. I’ll let you know as soon as the continuous build with these features is available.

[alex-spataru]

Hi!

I’ve just added a new visualization mode called Quick Plot Mode, in addition to the existing Device Sends JSON and Project File modes.

Similar to Arduino's plotting functionality, this mode automatically plots comma-separated values received from the device without requiring any project setup. The only requirement is that the user must ensure an EOL character (\n, \r, \n\r, or \r\n) is included to allow Serial Studio to correctly separate frames.

For more details, please refer to the attached screenshots:

[alex-spataru]

Closing this discussion to keep things clean. If you feel the issue needs to be re-opened, just let me know.

I’ve been working extensively on the wiki and added some examples to help new users get started with the software.

As always, there’s room for improvement, but I believe the project—across features, stability, UI-friendliness, and documentation—has made a lot of progress in the past few months. I’d like to keep the project moving in that direction. As always, any support from users—whether through donations, pull requests, examples, or updates to the wiki—is always welcome and greatly appreciated.