When debugging media streaming applications you likely need to test a source. ChatGPT and Google leave you stranded with outdated URLs. Here is a good alternative solution.
--- Lately, I am working with some "AI" (booooo!) enabled cameras (nice!). They allow uploading Python or C++ code. Usually, this will be a combination of [OpenCV](https://github.com/opencv/opencv) code and some SDK-calling to move the camera or such. Hence, a good DX should be a local development of the CV-specific bits and only then uploading. I am generating parametrised test data locally for this and am streaming this into OpenCV, as this is a nice, non-"works-on-my-machine"-approach. This way the developer can decide where they get the stream from and swap out the source easily. Annoyingly though, you need to connect some pieces together for this that are: - the streaming source - possibly [ffmpeg](https://github.com/FFmpeg/FFmpeg) - and [RTSP](https://en.wikipedia.org/wiki/Real-Time_Streaming_Protocol)-Server such as [mediamtx](https://github.com/bluenviron/mediamtx) - your own OS (this part I am developing on Windows 11; meeeh!) - your own code Since you want to reduce as many points of failures in a debugging session for this, I wanted to replace the following: - the source, for example a video for something static, that ideally I would not need to check into git. Also, avoiding [codec hell](https://blog.codinghorror.com/video-codecs-are-the-next-dll-hell/) - the receiving end for [VLC](https://www.videolan.org/vlc/) Googling did not yield nice results, since a lot of people are not writing about stuff like this anymore (hence this blog ;-)). Also, no-one seems to publish nice RTSP public links anymore. Prompting ChatGPT for `Give me a public rtsp stream url` was as pointless as often when do something niche-y - why the hell does only simple work work with this thing: I leaded to invalid links and outdated `mediamtx` examples. Great! Luckily, diving deep led me to this [great guy](https://superuser.com/users/635712/rotem)'s comment, [here](https://superuser.com/q/1823654/341454). His snippet would read as: "use `ffmpeg`s internal test stream and forward it" ```bash ffmpeg -report -v verbose -re -f lavfi -i testsrc=size=720x404:rate=59.94 -f lavfi -i sine=frequency=400 -vcodec libx264 -pix_fmt yuv420p -acodec aac -ar 24000 -rtsp_transport tcp -f rtsp rtsp://localhost:8554/mystream ``` One should care about codec stuff here, but that is out scope. This will start streaming to the to `mediamtx` with its default config. For verbosity and if this gets out of date, here is the current config:
So, thank you [cabreratypo on Dafont](https://www.dafont.com/alejandro-cabrera-avila.d2473)!!!!
## Some Styling
Okay, now let's make this 3d. Chatgpt won't help...or not give great options.
But the humans (I suppose) over at [Avada](https://avada.com/) do on their [blog](https://blog.avada.io/css/3d-text-effects). Great compilation and I really wonder what the motivation behind this here was...lol:
> 19+ Best CSS 3D Text Effects Examples from hundreds of the CSS 3D Text Effects reviews in the market (Codepen.io) as derived from Avada Commerce Ranking which is using Avada Commerce scores, rating reviews, search results, social metrics. The bellow reviews were picked manually by Avada Commerce experts, if your CSS 3D Text Effects does not include in the list, feel free to contact us. The best CSS 3D Text Effects css collection is ranked and result in July 3, 2024. You can find free CSS 3D Text Effects examples or alternatives to CSS 3D Text Effects also.
Also, to you: thank you!!!
---
## [wip] Essential Reads: Systems and Requirements Engineering
URL: https://eljefedelrodeo.humans.airdress.co/blog/02-essential-reads-system-and-requirements-engineering/
Date: 2024-07-01
Author: Robert Jefe Lindstaedt
Categories: Essential Reads
Tags: systems-engineering, requirements-engineering, roadmapping, wip
---
Planning systems to become successful needs more attention. Let's learn from the best.
--- When setting out to create something new may it be: - a new venture - a new project - a new product ...we tend to ask ourselves: _How can we plan better than last time?_ Defining the tactics for such an endeavour obviously depends on on many factors. Since you have read the word _depends_ and probably rolled your eyes, we will jump ahead and assume some conditions - we are a small or medium or no organisation at all - we feel inspired by successful organisations, but will not blindly follow what they - we will read what very large or very large organisations are doing and trim it to our needs ## Why Looking Towards Others? We could also just invent something - and indeed this is something we need to do to get ahead eventually. However, other organisations have gone through very many maturity stages, by failing (and succeeding) forward. Those large organisation are following principles already, that we are striving to achieve, because we want to emulate their success. Dismissing their oftentimes public record of historic learning, would be unwise or just very uneconomic. Also, big organisations define the playing field, with things such as **compliance**, because they are relevant enough to be attacked. Rules are not only there to be followed, as they represent in the worst cases a kind of physical barrier, they can though also be used as guidance for us, what to do. If need be, we could of course take guidelines of such organisations and blindly emulate them, carrying or own innovation forward all the while, or just start innovating, once we have reached their level of maturity. ## What to Read? In recent years US government organisations (no affiliation or opinion about them) have shown to produce a lot of documents and statements that are actually very useful. And they need to be public to a degree. Compare your own googling of mediocre blog posts - it's not you, it's the internet - to this treasure trove of what I dub the [`DoD CTO`](https://www.cto.mil/) (not a person but _DoD Research & Engineering, OUSD(R&E)_ -> Office of [Under Secretary of Defense for Research and Engineering](https://en.wikipedia.org/wiki/Under_Secretary_of_Defense_for_Research_and_Engineering)). Super cool. They run the... ### DoD Systems Engineering Guidebook Available under this [link](https://www.cto.mil/wp-content/uploads/2024/05/SE-Guidebook-Feb2022.pdf) and their [library](https://www.cto.mil/sea/pg/) you will find a guide for free, that is complete, with no bells and whistles that you won't need. Well, ignore the military bits if you want. Let alone looking at their table of contents should give of list of things we need cover. Things that catch my eye: - implementation is tiny activity - requirement engineering is central We will dive deeper into those topics, but it is a great starter, is it not?