My level of frustration with Microsoft is growing to a point that I'm unable to contain it.

They buy Github, it was a great tool for developers because is FUCKING WORKED! New features were never beta tested on users unless they requested it.

Why in the absolute fuck am I getting all these new experimental bullshit features that literally make it harder for me to do my god damned job?

They provide me NOTHING but grief and sleepless weekends while I fix the god damned pipeline that's worked perfectly fine for YEARS.

Your business model is bad and your products are SHIT.

Fuck you Microsoft, I cannot even stress it enough. If I had a time machine that could go back 5 years and my options were: Tell the world about Covid, make sure Trump never became president, or stop the Github purchse. I would hunt down and brutally attack the team of executives that decided to buy Microsoft.

Words cannot adequately describe how much I want Microsoft to fuck off. If the company was a person and they died in a house fire it wouldn't be enough.

I just want a VCS that does what it's supposed to do. I don't need pipelines, I don't need image repos, I don't need static code analysis.

I JUST NEED A FUCKING VCS THAT CONTROLS VERSIONS OF SOFTWARE YOU IGNORANT FUCKS.

I despise it when software developers remove features because "too few people use them".

Is this what those shady telemetry features are for? So they can pick which useful features to get rid of because some computer rookies whined that it is "feature creep" rather than just ignoring it?

Now I have to fear losing useful (or at least occasionally convenient) features each time I upgrade, such as Firefox ditching RSS, FTP, and the ability to view individual cookies. The third can be done with an extension, but compatibility for it might be broken at some point, so we have to wait for someone to come up with a replacement.
Also, the performance analysis tool in the developer tools has been moved to an online service ("Firefox profiler"). I hope I don't need to explain the problems with that.

But perhaps the biggest plunge in functionality in web browser history was Opera version 15. That was when they ditched their native "Presto" browsing engine for Chromium/Blink, and in the process removed many features including the integrated session manager and page element counter.

The same applies to products such as smartphones. In the early 2010s, it was a given that a new smartphone should cover all the capabilities of its predecessors in its series, so users can upgrade without worrying a second that anything will be missing. But that blissful image was completely destroyed with the Galaxy S6. (There have been some minor feature removals before that, such as the radio and the three-level video recording bitrate adjustment on the S4, but that's nothing compared to what was removed with the S6.).

Whenever I update software to a new version or upgrade my smartphone, I would like it to become MORE capable, not LESS (and to hell with that "less is more" nonsense).

Actually I have two stories

The first one, that one project I talked about with a big company when I was at school. It wasn't that much coding since it was mostly researching, but it was a big project that seems really interesting, with Image Analysis and Machine Learning.
The projects at school this year got drawn randomly for each group, so when I've been announced that I've been chosen for the biggest project, thinking about every side of the project, I was hyped. And even a year after we finished it, I'm still happy and excited about it.

The second is something a little more funny :
So we got some projects to do during December for school including cryptography. Again, those were randomly drawn (but some can really fuck you up) and I got to do a Password Manager, like KeyPass. We were 4, and we thought we had the time to do it.
But we misread the date. At the end of Christmas break, I got a call of a friend saying that the project is due in two days.
Thing is, one of my three co-workers weren't contactable. And we got nothing.
So I kinda took the lead : I said to one to do the UI, another to do the cryptograph helper, and I'll do the linking and all the behaviour of it.

In two days, I literally spent all the time available on it.
Then first meeting with the teacher for saying what is wrong, where bugs are if they exist, ect. so we can fix the issues and deliver a clean code. They were like only 4 big problems. More is, I fixed them all in like two hours while thinking fixing only one. And we got something like the 2nd or the 3rd best mark of the prom. And everyone congratulated me for that. I got so excited I was able to do that in few time.

But never that again lmao

[Long post]
My last big project at school.

There was some pretty interesting projects, some shitty one, but there was one big project that interested almost everyone : a project in collaboration with Siemens. The project implied Machine Learning and Image Analysis. There were like 11 applies, with a total of 13-14 groups.

The project was randomly chosen for each group. I've learned that my project was the big one with Siemens. I remember how excited and hyped I was in a quarter of second.

So the whole project was tutored by one teacher that know us pretty well (since we already did a pretty cool project last year tutored by him) and by a former student at my school who's now at Siemens. And to be honest, it was one of the coolest project I've been into, despite the difficulty, since the whole subject (not gonna tell it just in case) was pretty new. We had some troubles, but we and our tutors always had discussion every week that helped us quite a lot.

There was some development planned at first, but the more we went into the project, the more we all saw the complexity of it and didn't quite hope to do a single line of code, but mostly research.

The project took around 3-4 months, we had a room that we can use with a GTX 1070 for training the neural network, and me and my friend knew how to work perfectly and efficiently.
At the end of the project, as expected we didn't do some coding, but we did a presentation of the project, with the big help of our tutor at Siemens that told us to redo from scratch our part in a more scientific way; the presentation was a real success, we got all the jury saying they actually wanted those kind of presentation and were really pleased. And we provided everything needed so a new fresh group with no knowledge of the topic could do some coding on it.

We got one of the highest notes of the promotion (not sure if the highest or not). Even tho it kinda disgusted me in researching, that actually was one of the best project I got to do that was that successful.

FUCK. YOU. AYLIEN.

- For your shitty hashtag generator, that generated #FCBarcelona for a game review
- For your shitty classification endpoint
- For your shitty sentiment analysis that only works in the demo
- For your shitty image tagging that clarifai is way better at
- And for your "semantic labeling" that doesn't work

FUCK YOUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUU

Sharing progress on a project I started two days ago.

Elevator pitch: a personal (open-source) photography website where you can search my repository of photos by color, brightness, orientation, resolution, and subject.

I got a new camera and it actually performs very well and it made me realize that there's a lot of interesting data in images, so I've decided to convert my personal blog into a photography-centered personal blog.

My first idea, the one that actually drew me into starting the project, was to pick colors from my images (colorpicker/ eyedropper style) and then on the home page, those colors would appear in a grid and you could click them to see related photos.

Cool idea, but clunky and not very useful in practice.

So I implemented median-cut algorithm to generate a palette from an input image. It has its weaknesses, but it's consistent and does a job that makes you go "yeah that's about right" in 99% of cases. It generates 8 colors and then a second algo removes any colors that are within 5% distance from one another.

Before the color theory nerds ask, YES - I am using plain-old RGB and you can suck my balls. I don't care about human perception enough - not for this project at least.

Then, these colors (and other basic image analysis like brightness) are stored in the DB and related to the image entity from the upload.

Then the user (not yet implemented...) can do the reverse. Either they can choose colors from a colorpicker, or they can upload their own photo to get a palette and then the colors are looked up in the database and the related images are shown in the search results.

This will be combined with a somewhat complex tree-style system for the subject of the photo. For example, a subject like "crow" can be related to the subject "bird" which itself can be related to the parent subject "animal". So if you search by crow, you'll only get crows, but if you search "animal", you'll get birds, crows, etc.

While it's not exactly a reverse image search like google images has, I think it's a somewhat refined take on a way to explore a photography repository - especially a personal one.

I am developing the management portal and the public facing portals as separate projects due to separation of concerns - and so I can avoid implementing (and upkeeping) authentication for ONE user (me) so it just runs on my local network.

Screenshot is the current iteration of the upload page for my management portal, which will eventually take care of watermarks etc etc.

ComPDFKit Solutions

For text extraction technology, ComPDFKit offers the following two solutions that effectively address text extraction for all types of PDF files. For documents containing only text information, our non-intelligent solution can suffice. But for more complex documents and image-based ones, ComPDFKit Document AI offers higher accuracy in text extraction. To learn about the accuracy of ComPDFKit's information extraction, see this article.

1. Algorithm: X-Y Cut Recursion Projection Method

The X-Y Cut Recursion Projection Method is a top-down page segmentation technique that decomposes a document image into rectangular blocks. It employs a recursive approach by projecting along both the X and Y axes to segment a PDF into independent rectangles, facilitating the extraction of textual components. ComPDFKit utilizes this method for efficient text separation and structural organization, including rows, paragraphs, and columns, to retrieve characters, words, lines, and paragraphs from the document.

The advantage of the X-Y Cut Recursion Projection Method is its speed, making it suitable for simple, structured, non-image-based PDF documents. However, for complex, unstructured PDFs, there might be recognition errors or omissions.

2. ComPDFKit Document AI

Document AI is an intelligent text extraction solution supporting all types of PDF files, including image-based. It uses artificial intelligence-based methods for document recognition and analysis to extract textual information from PDF documents (as well as images, tables, etc.).

- PDF Recognition and Analysis: This involves using deep learning models to recognize and analyze PDF documents, extracting elements like text, images, and tables while retaining their position, size, style, etc. ComPDFKit owns well-trained AI models to accomplish this process.

- Image Pre-processing: This process involves improving the quality and clarity of low-quality images in PDF documents, enhancing subsequent recognition and analysis. ComPDFKit employs multiple image processing techniques, such as image sharpening enhancement, noise reduction, document trimming and straightening, and stamp detection.

- OCR (Optical Character Recognition): OCR technology has a wide range of application scenarios such as license plate recognition, bank card information extraction, identity document (ID card) information recognition, train ticket information detection, etc. ComPDFKit supports recognition in dozens of languages. With extensively trained model zoo, it can accurately detect and recognize text in documents and analyze document structure.