Jarkko Sakkinen
jarkko
Edited 1 month ago
I have now also syntax highlighted source code viewer in ccd. I stole that out from tree-sitter code of a small text editor project that never really got going.
My stash of local repos over the year really is like a super powerful LLM for me at least or could not be replaced with any of those :-)
Preloader will mean +1 thread, which again only uses non-blocking I/O.
That's really as much one should use at most. Context switches are expensive and IO is mostly just waiting from CPU's standpoint.
I think user space programs advertising the size of their thread pool are made by developers who really don't understand what they are doing TBH 🤷
My stash of local repos over the year really is like a super powerful LLM for me at least or could not be replaced with any of those :-)
Preloader will mean +1 thread, which again only uses non-blocking I/O.
That's really as much one should use at most. Context switches are expensive and IO is mostly just waiting from CPU's standpoint.
I think user space programs advertising the size of their thread pool are made by developers who really don't understand what they are doing TBH 🤷
0
0
0
Jarkko Sakkinen
repeated
repeated
nixCraft 🐧
nixCraft@mastodon.socialLinkedIn Is Illegally Searching Your Computer. Microsoft is running one of the largest corporate espionage operations in modern history https://browsergate.eu/
Is there anything Microsoft is not doing these days?
6
11
0
Jarkko Sakkinen
jarkko
Edited 1 month ago
The DAW project I found a path to reduce the current 100 KSLOC and it has about halved.
1. Reduce radically custom rendering by levering GSK of GTK4 and base on GTK4 for window creation etc.
2. Drop plugins from the main project and migrated deleted UI that they need to that project.
After week or two of repivoting now I have a lively 50 KSLOC base :-) It has strong non-portable Linux focus like Logic has macOS focus so this is not an issue. It will be totally designed-for-pipewire DAW.
I really don't start projects anymore that I don't eventually finish. This might take year or two but I just occasionally decide +1 step and move it forward.
Given AI copying trend you really don't want to push in-dev repos out there. I only publish stuff that I consider to be a usable version of something.
1. Reduce radically custom rendering by levering GSK of GTK4 and base on GTK4 for window creation etc.
2. Drop plugins from the main project and migrated deleted UI that they need to that project.
After week or two of repivoting now I have a lively 50 KSLOC base :-) It has strong non-portable Linux focus like Logic has macOS focus so this is not an issue. It will be totally designed-for-pipewire DAW.
I really don't start projects anymore that I don't eventually finish. This might take year or two but I just occasionally decide +1 step and move it forward.
Given AI copying trend you really don't want to push in-dev repos out there. I only publish stuff that I consider to be a usable version of something.
1
1
1
Jarkko Sakkinen
jarkko
Edited 1 month ago
For caffeine command the only included themes are Catpuccin variations, as it fits to the brand :-)
It uses https://codeberg.org/jarkko/nippu to bundle the default themes. I hate dangling files at ../share.
Started to post about ccd because it suddenly starts to look like an app. Has been totally unintentional and off-topic project :-)
Powered by C.
It uses https://codeberg.org/jarkko/nippu to bundle the default themes. I hate dangling files at ../share.
Started to post about ccd because it suddenly starts to look like an app. Has been totally unintentional and off-topic project :-)
Powered by C.
0
0
0
Jarkko Sakkinen
jarkko
Edited 1 month ago
I've created over last few months or so a single pane filemanger called Caffeine Commander aka ccd:
1.It has a VFS layer.
2. It does not invoke external programs as subconsequence of its predefined actions.
3. It does not have plugins.
4. While not always unavoidable, minimizing and reducing the build configuration is a continuous goal.
1.It has a VFS layer.
2. It does not invoke external programs as subconsequence of its predefined actions.
3. It does not have plugins.
4. While not always unavoidable, minimizing and reducing the build configuration is a continuous goal.
1
1
2
Jarkko Sakkinen
jarkko
I have quite pretty big Yocto project to be published in near weeks/months including its own meta layer and complicated stuff such as OTA updates.
More on that later but https://github.com/siemens/kas was a great finding.
Unfortunately after working with Yocto for 2-3 weeks I found Kas only yesterday but was definitely worth of rewrite... (and i sleepless night)
Previously when working with Himmelblau Siemens surprised me with with linux-entra-sso browser extension. I guess they know how to deliver :-)
More on that later but https://github.com/siemens/kas was a great finding.
Unfortunately after working with Yocto for 2-3 weeks I found Kas only yesterday but was definitely worth of rewrite... (and i sleepless night)
Previously when working with Himmelblau Siemens surprised me with with linux-entra-sso browser extension. I guess they know how to deliver :-)
1
0
0
Jarkko Sakkinen
jarkko
Thank God I have one M2 Mini. It's life saver in aarch64 builds despite having to docker wrap them. And orbStack is great!
0
0
0
Jarkko Sakkinen
jarkko
leveled up my localhost git hosting from ssh-to-NAS to separate git-account and forgejo :-)
life quality++
life quality++
1
2
4
Jarkko Sakkinen
jarkko
Siemens is doing BTW amazing job in open source IMHO. Bumping a lof of to cool projects they have spinned off all the time.
0
0
0
Jarkko Sakkinen
repeated
repeated
Servo
servo@floss.socialServo is now passing 1.9M subtests at wpt.fyi: 1,903,187 / 2,131,264 (89,3%) 🎉
See https://wpt.fyi/results/?product=servo
0
3
0
Jarkko Sakkinen
jarkko
I think I also publish at some point my personal backup tool of which I've been using for ages in various forms. I don't know how backup tools work, never used them, and it has been implemented over that basis :-) This is actualy only backup tool I've ever used.
/* KOPIO on-disk format structures */
static const uint8_t KOPIO_MAGIC[8] =
{ 'K', 'O', 'P', 'I', 'O', '7', '7', '7' };
/* KOPIO on-disk format structures */
static const uint8_t KOPIO_MAGIC[8] =
{ 'K', 'O', 'P', 'I', 'O', '7', '7', '7' };
0
0
0
Jarkko Sakkinen
jarkko
Edited 2 months ago
“nippu - does one thing right”
https://codeberg.org/jarkko/nippu
I just need something simple and stupid as it is quite often :-)
I.e.,
#include <stdio.h>
#include <nippu/nippu.h>
int main(void)
{
const struct nippu_node *node;
const unsigned char *data;
if (nippu_lookup(&assets, "/hello.txt", &node) != NIPPU_SUCCESS)
return 1;
if (!nippu_is_file(node))
return 1;
data = nippu_data(&assets, node);
fwrite(data, 1, (size_t)nippu_size(node), stdout);
return 0;
}
Jarkko Sakkinen
jarkko
I sometimes wonder how many of the people who think compiler injected safety checks are great have heard about icache :-)
0
0
0
Jarkko Sakkinen
jarkko
AI competition and deregulation in the United States have caused "pop-up" data centers to be created like mushrooms in the rain, and energy for them is largely produced with fossil fuels. Other resources, such as water, are also consumed in shocking amounts.
So remember that if you rely dominantly on agent infrastructure, it will come at the expense of basic human resources, and you will destroy the planet for future generations. Appealing to the economy is irrational, because the economy is always ultimately a virtual issue and a matter of contract.
Nvidia's hardware is quite bad in terms of energy use in terms of environmental specs, because it's half a graphics card, and not just for inference etc. designed. Only Google (maybe someone else) has energy efficiency in order, because they produce an ASIC designed for the purpose, the Tensor Processing Unit (TPU).
Tech is not problem, humans are once again. E.g., my use of inference in my sample browser pakki is probably example of such case as it runs smart sample search just fine on my X390 ThinkPad :-)
So remember that if you rely dominantly on agent infrastructure, it will come at the expense of basic human resources, and you will destroy the planet for future generations. Appealing to the economy is irrational, because the economy is always ultimately a virtual issue and a matter of contract.
Nvidia's hardware is quite bad in terms of energy use in terms of environmental specs, because it's half a graphics card, and not just for inference etc. designed. Only Google (maybe someone else) has energy efficiency in order, because they produce an ASIC designed for the purpose, the Tensor Processing Unit (TPU).
Tech is not problem, humans are once again. E.g., my use of inference in my sample browser pakki is probably example of such case as it runs smart sample search just fine on my X390 ThinkPad :-)
1
0
2
Jarkko Sakkinen
jarkko
Edited 2 months ago
Pakki - a sample browser i'm putting together :-) it has also AI and interference run with CPU/GPU (works fine on my X390 Thinkpad).
The only AI app I've made so far...
If there's a magic button that generates all the code in the world, I can only assume that everyone is pushing that button. How I adjust my focus on topics such as AI is to look into smaller scale not as crowded areas such as local AI, which has the similar optimization challenge as all optimization challenges I love :-)
1. pfffft for FFT
2. libonnxruntime for inference
3. https://github.com/LAION-AI/CLAP/ for epoch models.
The only AI app I've made so far...
If there's a magic button that generates all the code in the world, I can only assume that everyone is pushing that button. How I adjust my focus on topics such as AI is to look into smaller scale not as crowded areas such as local AI, which has the similar optimization challenge as all optimization challenges I love :-)
1. pfffft for FFT
2. libonnxruntime for inference
3. https://github.com/LAION-AI/CLAP/ for epoch models.
0
0
1