Chart of ReactOS BootCD size

[karlexceed]

I got curious, so I used Python to parse the huge HTML file generated when you visit https://iso.reactos.org/bootcd/. I then shoved the data into a CSV for easy plotting.

Here's the chart, thanks to Excel:
[ external image ]
Full size here: https://i.imgur.com/dfJ2G2i.png

[hbelusca]

What are these strange spikes?
Also, have you separated the bootcds from GCC and from MSVC?

[ThFabba]

Hey, I did that too...
https://home.tfaber.net/~reactos/cdsize.html
https://gist.github.com/ThFabba/592205b ... b61f3610e1

Unfortunately I never updated it to support git... or Python 3.

[karlexceed]

What are these strange spikes?
Also, have you separated the bootcds from GCC and from MSVC?

The spike down to 0 is at 2010-6-20, which is a parsing error due to the fact that that entry doesn't have an "M" following the size value for some reason. Thanks for making me take a second look!

The spike up to 69 happens at 2011-10-3 and 2011-10-4, starting at r53981 and ending at r53990 inclusive. Looking at the commit, I don't see anything obvious as to why this happened: https://github.com/reactos/reactos/comm ... 975b437d70

And no, I didn't separate GCC from MSVC - the noise at the end of the graph starts at 2017-10-03, when the first MSVC builds show up. The lower bounds of that noise are the MSVC builds, the upper bounds are the GCC builds.

Hey, I did that too...
https://home.tfaber.net/~reactos/cdsize.html
https://gist.github.com/ThFabba/592205b ... b61f3610e1

Unfortunately I never updated it to support git... or Python 3.

Nice! My approach was much more of a 5 minute hack to make a quick chart. As you can see from my script: https://pastebin.com/wfvy8gD5 I used a local copy of the ISO page's HTML.