Key Features Implemented

1. Authentication and Profile Setup: Early in August, foundational work was laid out with the implementation of user authentication. This was a critical step to ensure that users can securely access and manage their accounts. Following the authentication setup, basic profile management features were introduced, allowing users to personalize their experience within the app.
2. Chat Functionality: A major focus has been on building a robust chat system, which is central to the user experience in Nag.bot. The chat now supports functional communication, with a server-side prompt system that enables AI-driven responses. Additionally, the frontend has been optimized to handle real-time conversations, ensuring a smooth and responsive user experience.
3. Error Notifications and Markdown Support: To enhance usability, an error notification system was integrated into the chat. This helps users quickly identify and address issues during their interactions with the app. The chat also supports markdown, allowing for formatted text, which is particularly useful for displaying structured information and links within conversations.
4. UI/UX Enhancements: Significant attention has been given to refining the user interface. The chat now includes a list styling feature, making it easier for users to navigate through messages. Additionally, the text input area has been improved to dynamically adjust as users type, providing a more intuitive and user-friendly experience. A standard button component has been added to maintain consistency across different parts of the app.
5. Conversation Initiation from Home Screen: One of the more recent updates allows users to start a conversation directly from the home screen. This update is aimed at making it easier for users to dive into the chat with the AI, streamlining the process of getting the support they need right from the start. The starter tile on the home screen now forwards users to the chat with pre-seeded prompts, creating a seamless transition into the conversation.

Looking Ahead

With the core functionalities of authentication, chat, and error handling in place, the focus will now shift to refining these features based on user feedback. Additionally, there are plans to further enhance the AI’s capabilities, enabling it to provide more personalized and insightful responses to users’ queries and challenges.

Stay tuned for more updates as Nag.bot continues to evolve into a powerful tool for personal accountability and goal management.

Time is a precious resource, we should spend it as wisely as we spend our money. That is what this app will do. It will discuss your projects with you and help you break your goals or ideas down into actionable tasks. Then, it will check in from time to time and see if you’re working towards those goals. If not, Nag with NAG you to get to work!

Running into issues with it though. I was hoping I could depend on ALL CAPS and other identifiers to separate sections out. In a few cases, it’s difficult to decipher between section titles and body. In the PDF, I might be able to use italics and other variations in the font to figure out what’s what. This increases the complexity of sorting out the different sections of the document into something with a usable structure.

Goal 1: Read and parse the document

Goal 2: Extract sections.

Read and parse the document

What are some readily available open source projects that I can use to parse PDFs into text?

1. Poppler Utils: Poppler is a PDF rendering library that includes a variety of utilities for extracting text, images, and other content from PDFs. The pdftotext command is part of this suite and is widely used for converting PDF documents into plain text.
2. Xpdf: Similar to Poppler, Xpdf is a toolset that includes the pdftotext command. It’s an older project but still effective for extracting text from PDF files.
3. PDFMiner: PDFMiner is a tool written in Python specifically for extracting text, images, and metadata from PDF files. It’s more flexible than pdftotext and allows for more detailed analysis of the PDF structure, making it suitable for more complex extraction tasks.
4. mutool: Part of the MuPDF suite, mutool can extract text and images from PDF files. MuPDF is known for its speed and the quality of its rendering.
5. Apache PDFBox: Although primarily a Java library for working with PDF documents, PDFBox comes with a command-line utility that can be used to extract text from PDFs. It’s useful for those who prefer a Java-based solution.
6. Tesseract OCR: For PDFs that contain mostly images of text (like scanned documents), Tesseract OCR can be a powerful tool. It’s an optical character recognition (OCR) engine that can convert images into text, and with the right preprocessing, it can be used to extract text from image-based PDFs.

Let’s try some of these out and see how the result varies. I’m most interested in #1 poppler and #3 pdfminer

Using the poppler option I found provides a good starting point for text cleanup.

pdftotext -layout -enc UTF-8 WDI39597.pdf poppler.txt

PDFMiner has more options in term of formats (text | XML | HTML). The first thing I noticed though, it’s significantly slower to execute. Annnd. the output is far less usable. I was hopeful for the HTML or XML output. The most ridiculous output was XML. There was literally tags around every letter.

pdf2txt.py -o pdfminer.txt -t text -A WDI39597.pdf 
pdf2txt.py -o pdfminer.html -t html -A WDI39597.pdf 
pdf2txt.py -o pdfminer.xml -t xml -A WDI39597.pdf 

POPPLER WINS! It creates a usable output and is WAY faster in terms of execution, not that that is a huge factor.

Now we have something that looks like this: https://snovak.com/wp-content/uploads/2024/03/poppler.txt

Now, Let’s strip out some garbage and format this a bit more.

I’m using a Python script to do this part.
First it detects page number and formats that appropriately.
Then, it gets rid of leading whitespace.
Then, ditch the date, and the line under that, which has some crazy special characters….
Then, ditch any lines that end in ‘SEN. APPRO’

Now we have something that looks like this… https://snovak.com/wp-content/uploads/2024/03/WDI39597.txt

I’ve preserved the page numbers and line numbers for citation purposes. So, if I want to recall where appropriations were made in the bill, I can cite “Page 36 Line 22” for example.

I’ll have to get to extracting the sections tomorrow…

This week another Uniparty Omnibus spending bill was passed without much a fuss. I was thinking Speaker Johnson was going to be a force to stand up to the machine and reduce spending. I thought he was going change things. I may have been mistaken. We need to get inflation under control, its like a brush fire that could consume the country. Meanwhile the money printing machine is in overdrive. Instead of whining about it on X, why not do something that’ll bring some visibility and comprehensibility to these massive bills?

Many years back, I’ve registered a domain politipal.com, which I had grandiose plans for. Naturally, I’ve done nothing with it. It’s time to change that too.

If you haven’t seen one before, these bills are published in the most unusable format possible. A super lengthy document, that no one can easily read and/or understand. Example

No way to compare to previous years, no way to visualize using common graph paradigms. Hopefully, this project will fix that.

How does a project like this make money? I have not f’ing clue, but I’m tired of doing nothing and watching the shit show carry on uninterrupted.

The first step is a POC. Can I parse this bill text into usable data with readily available open source scripts, programs, etc?

Automated Workflow:

1. Read and parse the document, extracting sections.
2. For each section, extract relevant details.
3. Format those details into a JSON object.
4. Insert the JSON object into Database.

I recently watched “Leave the World Behind“. It’s a message, a clear and terrifying message from our adversary. It’s a message about what happens when we resist they/them. When I refer to “they/them”, I’m not referring to the confused millennial non-binary they/them sorts. I’m talking about “The Machine”…. you know, one that “Rage Against the Machine” raged about before the band by that name was corrupted, consumed, and assimilated into the very machine they raged against. I’m talking about The Machine that has largely had a monopoly on influence and power for the last century or so. I’m supposing the 1913 creation of the Federal Reserve is a good marker for that level of influence and power, and the global elite class that wields it.

I internalized the message they wanted to deliver in the movie. They want people to duck and cover, to hide in the basement, with a cache of food, and a box set of “Friends” DVDs to keep our little minds occupied while the world tears itself apart. They want us out of the way, while their carefully choreographed chaos unravels the fabric of society.

After watching the video below, I’m thinking THAT, would be taking their bait. THAT is playing into their hands. THAT is exactly what they want. Instead, in the video below, Bret Weinstein, an extremely brilliant man/scientist/educator advocates for forming coalitions. Getting people together to share ideas and combining the power of the multitudes who stand against The Machine. He suggests that Goliath (as Weinstein refers to The Machine), has lost the first onslaught in their war for power. He speculates that the heros, who have emerged through the first wave largely fit the description of lone wolves. The Machine is learning and leveling up. So we, “The Resistance”, need to learn and level up as well. Those lone wolves need a pack.

With the first wave behind us, the confusion persists. Goliath is looking for a rematch. Some of they/them have been exposed. Presidents of the last decades are all on the list, mix in a bit of Jeffrey Epstein and Hunter Biden, and you have a hot bowl corruption soup. It seems, their exposure only reveals more questions than answers. These revelations now float on the surface, but this bowl runs deep.

I listened to Weinstein’s Dark Horse Podcast throughout the COVID crisis. He and Dr Robert Malone were a bastion of common sense and inquisitive curiosity about the confusion that didn’t fit then. His insight earned my respect and trust. During the below interview, I can see he has a sort of existential alarm about him. He has left the rage behind, and gone to “war with the machine”. I will pray for him, as I will pray for us too.

This is a first run at the UI, basically I need it to add, edit, and remove a nested hierarchy of topics. Only admin users will see this, so it doesn’t have to be pretty, just functional.

Each topic has a color, which I suppose should trickle down to it’s child topics. This way there will be some visual separation among the different lessons.

Here I have some initial topics, generated by ChatGPT, these will certainly change. I may use KhanAcademy as a template. They have certainly put careful consideration into their taxonomy.

Also, pay no attention to the branding. I’m still undecided between, “Homeschool Link (homeschool.ink)” or “Learnalot.net”.

I would love your opinon @ x.com

I’m continually impressed by ChatGPT. This morning I thought it would be really nice to be able to track my health statistics on Nextcloud, my private cloud that I have running just behind me in my closet. What a cool little project to give to ChatGPT and see how quickly we can get something up and running. It’s 8am on a Tuesday morning, I’m back to work on my day job, but I have about an hour to fiddle around with it. Let’s see how quickly ChatGPT can get this started….

A little background, I’ve been tracking some health parameters for a while with iHealth, mainly because they’ve made it easy to do so. I have a bluetooth bloodpressure cuff, every time I take my BP, it’s logged to the cloud. It has a nice UI. But, I’m not very happy with giving my health information away anymore. So I’ve been looking for a new home for my health data. Lately, I’ve been using “Waistline”, an open source app found on F-Droid. It works, but not nearly as nicely as iHealth. The data is siloed, and I’m not really sure how to get it out of the app. So, passively I’m still looking. That’s were we pick up the story for this idea.

Here is the chat in it’s entirety. I basically walk the bot through the process of coding the entire plugin for me.

At this point, I have yet to test it out, but as you can see, it’s an amazing start. I’ve got a plugin templated out, an API, and directions to get the frontend started as well. It’s 9am now, so I need to get to my day job. But, wow. Just WOW.

More to come as time permits.

I’ve been using a very crude Vosk keyboard on Android to fill the gap. I’d love to try to improve upon this project, but for today I’m interested in getting this functionality in Gnome, my Desktop of choice on Ubuntu Linux. This is not meant to be a tutorial, but more of a journal entry.

Documentation for gnome extension are scant. Here is what I could find:

• https://gjs.guide/extensions/
• https://github.com/GNOME/gnome-shell-extensions
• https://wiki.gnome.org/Projects/GnomeShell/Extensions
• https://gjs.guide/extensions/development/creating.html#gnome-extensions-tool

Here is a great playlist on YouTube to get more familiar with creating gnome extensions.

Damn, don’t you hate when you don’t save your work? I just lost a bunch of work. DOH!

Let’s see, I was astonished to see that, in general, the Gnome extensions area is not super active.

Development is a little rough, I have to switch from Wayland to X11, which makes reloading extensions a little easier. In wayland, you have to log out and back in for extensions to refresh. Yikes.

Here’s a directory of existing extensions: https://extensions.gnome.org/

I like to learn from other code. So I installed this extension, which allows you to manage your system clipboard: https://github.com/Tudmotu/gnome-shell-extension-clipboard-indicator

I haven’t found anything preinstalled to manage extensions. Seems like something that would be readily available in “Settings”.

Anyways, I started this at 9am, I hope to have something working by noon, but time is dwindling. I just spent some time on creating an icon in figma. No matter what I do, it’s still hard to see the “TXT” in the icon. I may just ditch it and just use the mic, but I’ll leave it in for now. Anyway, I hope Gnome supports SVG, which might render a little nicer. Let’s move on. We have some functionality to create.

Golly, documentation is THIN for gnome extensions.

I’m simply trying to get a button in the tray, when clicked it will change color. Also, reloading extensions is still a CHORE. I have to log out, then log back into gnome each time. Tedious.

I found a solution to that here: https://www.reddit.com/r/gnome/comments/eb4pn9/how_do_i_reload_a_gnome_shell_extension_during/

I’m using a reload.sh script to load up another session of gnome, which naturally reloads all the extensions.

dbus-run-session -- gnome-shell --nested --wayland

My SVG isn’t looking great in there though. I may have to use a ready-made system icon.

I’ve got the icon working now, but there still is styling issue, where the icon seems a little small.

I’ve messed with getting Vosk working appropriately. I’ve tried a few of the suggested methods, but I’m having a lot of issues making my microphone accessible in nodejs with the ‘mic’ library.

I’m currently leaning towards running vosk as a docker service with the following docker-compose.yml

version: '3'

services:
  vosk:
    image: alphacep/kaldi-en
    ports:
      - "2700:2700"

So far, only one test script that I’ve tried actually worked.

#!/usr/bin/env python3

import asyncio
import websockets
import sys
import wave

async def run_test(uri):
    async with websockets.connect(uri) as websocket:

        wf = wave.open(sys.argv[1], "rb")
        await websocket.send('{ "config" : { "sample_rate" : %d } }' % (wf.getframerate()))
        buffer_size = int(wf.getframerate() * 0.2) # 0.2 seconds of audio
        while True:
            data = wf.readframes(buffer_size)

            if len(data) == 0:
                break

            await websocket.send(data)
            print (await websocket.recv())

        await websocket.send('{"eof" : 1}')
        print (await websocket.recv())

asyncio.run(run_test('ws://localhost:2700'))

The problem here is that it’s sending a .wav file, not opening the microphone and transcribing the output.

That’s enough for today. I’ll pick this project back up at some point.

There are quite a few ways that we might use ChatGPT to plan for homeschool. It’s GREAT at generating lists. Now, our primary task is simply to ask good questions.

For instance:

I feel like some of the suggestions here are great, but definitely needs expanding on. So, let’s get an idea of what that looks like. Let’s drill down on combining reading and vocabulary.

Well, that’s plenty to choose from. I love the “Chronicles of Narnia”, it’s Christian based, great story, a good selection. Also, I think I could sell that to Eva. Let’s keep going. How about a list of vocab words from the first chapter of the first book?

Yes, there was an error in the process. I wasn’t completely happy with all of the words in the first go-round. So, I asked for more, and ChatGPT picked up from where it left off.

I think I can select 10 good vocab words from this list of 20 words 19 words.

What’s more important is that I could basically do this same sequence with the all of the books that my daughter selects to read. She could literally select any book and we could turn that reading into a decent reading and vocab list.