It may have been overkill, but I took the datasheet of the Philips tube used in the Commodore 1084 monitor and used it to recalculate the angles of the CRT shader in Copperline. This is almost ready to be included in the next release.
#Amiga
Remote
Andrew (LinuxJedi) Hutchings
@LinuxJedi@fosstodon.org
mastodon 4.7.3Software Engineer at wolfSSL. OSS advocate. Restores and designs Acorn & Amiga hardware.
0 Followers
0 Following
16 Posts
Joined November 04, 2022
Blog:
Open post
I've had several people ask me why we use C instead of Rust or other languages at wolfSSL. One of my colleagues wrote an explanation:
https://www.wolfssl.com/why-c-remains-the-gold-standard-for-cryptographic-software/
156
61
128
0
Open post
An Amiga in your browser tab: Copperline 0.12.0 released today! Now with a WebAssembly build, big-box A3000/A4000 support, SCSI CD-ROM drives, and writable host folders. Boot one now at https://copperline.dev/try #Amiga
Big thanks to Nicolas Ramz for optimising the AROS ROM, this optimised version is now bundled in 0.12.0 (and the online build) and is significantly faster booting the Amiga.
13
0
8
0
Open post
Copperline #Amiga emulator 0.8.0 has been released, and this is one for those who prefer GUIs to command line. There is now a launcher that appears if you do not use command line options.
https://copperline.dev/
7
6
7
0
Open post
Replying to
@raymaccarthy a translation step adds extra complexity and reduces portability. wolfSSL’s predecessor (yaSSL) was written in C++. That was around 20 years ago. C++ is not very mature in yet embedded world anyway, and sometimes we are dealing with compilers from the 90s.
It is also a lot easier to hire good C engineers than languages such as Ada.
4
17
2
0
Open post
Replying to
@codewiz@mstdn.io not so much forgot, more like deliberately didn't shout out about it. There were so many demo regressions in it that I was expecting a lot of bug reports. I'm focusing on fixing it up for 0.11.0 and making a lot of noise about that with the features of both releases.
1
2
0
0
Open post
Replying to
@raymaccarthy we have been doing this a very long time, we know what we are doing.
Unfortunately, what we do may not align with everyone’s core beliefs about how things work.
Many of our engineers are like me, have decades of experience developing core internet infrastructure projects.
We also hire where talent is, not where is geographically convenient.
2
2
0
0
Open post
Replying to
@RealGene @raymaccarthy also, I’ve seen some embedded platforms with some really junk C++ compilers.
2
5
0
0
Open post
Replying to
@goedelchen I’ve seen a lot of people say PHP is dead too. But it still used in a large percentage of the web.
0
0
0
0
Open post
Replying to
@thejpster for us the wolfCrypt core is the most important higher risk part for this argument. This is our core cryptographic engine. X509 parsing is in the next layer up (wolfSSL itself), but that makes calls back to the wolfCrypt engine to do the parsing.
So, you could take wolfCrypt and implement X509 in Rust on top if you wanted (we have wrappers to help with that)
0
0
0
0
Open post
Replying to
@codewiz@mstdn.io I’m on several of them and I have my own channel on https://discord.gg/tsAbuAfqf9
0
0
0
0