/Δ/ - diy/projects

As Stallman and other oldfag hackers said, "hacking" is a process of finding non-trivial solutions to ordinary problems. This is not how software should be developed. Expanding entropy by contributing to non-general Linux-based operating systems or compiling Gentoo on Powermacs is utterly pointless.
>POSIX
>soykaf shat lead us to "cloud" computing and invasion of user's privacy, not suitable for home computing, has redundant bloatware features that are not needed on home computers
>a bunch of BSDs incompatible between each other, as usual. Proprietary drivers and FBI backdoors included + cuckold license
>Plan9, a Chrome OS for hipsters who still think it's 1968 outside
>Minix
>A joke on Intel ME security

>>670
Hacking is often used to mean tinkering on or playing around with software. That's what Stallman means when he end his letters with "happy hacking."

>>671
In June 2000, while visiting Korea, I did a fun hack that clearly illustrates the original and true meaning of the word "hacker".

I went to lunch with some GNU fans, and was sitting down to eat some tteokpaekki, when a waitress set down six chopsticks right in front of me. It occurred to me that perhaps these were meant for three people, but it was more amusing to imagine that I was supposed to use all six. I did not know any way to do that, so I realized that if I could come up with a way, it would be a hack. I started thinking. After a few seconds I had an idea.

First I used my left hand to put three chopsticks into my right hand. That was not so hard, though I had to figure out where to put them so that I could control them individually. Then I used my right hand to put the other three chopsticks into my left hand. That was hard, since I had to keep the three chopsticks already in my right hand from falling out. After a couple of tries I got it done.

Then I had to figure out how to use the six chopsticks. That was harder. I did not manage well with the left hand, but I succeeded in manipulating all three in the right hand. After a couple of minutes of practice and adjustment, I managed to pick up a piece of food using three sticks converging on it from three different directions, and put it in my mouth.

It didn't become easy—for practical purposes, using two chopsticks is completely superior. But precisely because using three in one hand is hard and ordinarily never thought of, it has "hack value", as my lunch companions immediately recognized. Playfully doing something difficult, whether useful or not, that is hacking.

>>670
It's always good to hear other people's opinions.
The intention of this thread, however, was for people who are currently working on some part on a Unix system, to share and ask questions and stuff. Maybe I should've made that explicit in the OP. Oh well.
Anyway, that's why I put it in the Projects board.

>Plan9, originally developed by the creators of Unix to aim for a modern not-quite-Unix
From what I remember, they though Unix was not Unix enough, so they tried to make something that really did the ideas they had in mind while making Unix.
I really wish it took off, god knows what wonderful things we could have had.
Also,
>9front
posting obligatory (9)

I recently did an in place upgrade of an existing RHEL 4 system to RHEL 6. 6 was the most recent supported version since it was installed as a 32-bit system. This is a machine that handles a number of financial transactions and other important data via sftp and Samba. Yes, Samba. I am also a part of a project to transfer this functionality to a new system that doesn't use CIFS, but it's a multi-year project and who knows what will happen.

Red Hat would like you to believe that you can't do in place upgrades between major versions, but it is possible with certain modifications to the underlying OS and taking care to upgrade the versions of RPM manually during the OS installation process. Regardless, it's still not considered supported and they aren't going to help you going forward.

It was decided that we needed to upgrade based on the vulnerabilities in SMB v1, so I took the whole project on as a challenge at work; since the bosses assumed there was nothing to be done. You need to step up one version at a time, from 4 to 5, then from 5 to 6. I was fortunate in that it had been converted from physical hardware to a VM, so I had snapshots to rely on.

I've also recently converted a number of HPC machines from one Hyper-V cluster to another, resolving a number of network issues and putting all machines behind a firewall. Don't ask me why we are using Hyper-V for this - my guess is that it's "free" (we have a Microsoft site license) and familiar. The money comes from grants, so the less you spend the better. These machines are various flavors of Linux - Ubuntu and RHEL. We ran in to one problem where a certain outside company could no longer access a machine and there was a stiff SLA penalty for downtime which no one knew about before the project started, so we needed to fail a single machine back. Now we have two Hyper-V clusters, neither of which can run all the machines on them. The forecast is grim.

I can barely work up the care to maintain my own laptop. I told management if the RHEL 6 machine isn't gone by the end of support, I will get a new job. 2020 here we come.

>>674
Plan9 still has some potential, or at least we could learn a thing or two from those guys. There seems to be a growing trend for distributed systems, I think it's a wise move since there is little use for a computer that's not connected to a network these days.
I wonder if linux could accomodate to such a model. I don't know much about the inner structure of the kernel, but it's development model is already quite distributed at several levels. On the other hand, there is also a growing trend for using linux as a mere firmware (another wise move imo).
Long have I fantasized of linux-based OSes with a completely custom userspace, kind of like what android does. The same thing could be done with a great deal of VMs of all sorts. I remember reading someone says linux is "a library emacs uses to interact with intel processors".
I would like to see linux used more in this sense and quite frankly, less of a component of a Unix-style user OS.
I am just thinking out loud here, yet, I think that the potential of linux is wasted on Unix-like or Windows-like (looking at you GNOME) environments.

>>669
I would to have a second PC with Unix on it for doing various hacking/experiments, but I only have some soykafty Pentium 4 PCs or worse. What should I do?

>>697
Not OP, but are you new to *nix? You may just want to learn the basics first.

Also, don't worry about your hardware. Old hardware isn't that bad, it's kind of a marketing meme that you need to upgrade your soykaf all the time. I have a bunch of core 2 duo pc's, and they do eerything I need them to.

I would just like to mention Bedrock Linux.

>>672
A hack isn't making something more difficult, but easy and quick. Your description was more like an anti-hack.
Also cringe level from the other people was likely over 9000.

>>743
Um. That's a story on Richard Stallman's site that they copy pasted. Not sure if you missed it, or I'm missing your joke about Stallman.
https://stallman.org/articles/on-hacking.html

>>679
>growing trend for distributed systems
i agree but why would you fork the entire os and kernel just to do something that already has many implementations in gnu/linux?

I really would love to try Plan 9, but it seems like there is a huge entry barrier… must learn more….

Despite what a lot of people say the #cat-v community on FreeNode is very helpful in setting up 9front on your hardware as long as you do not act like an idiot and have patience. The OS itself is actually fairly simple and the best way to learn is to simply use it and read the man pages(which are of the highest quality).

>>670
>cuckold license
It's better than "you are free to use my free code, BUT DON'T TRY TO USE IT IN A WAY THAT I DOESN'T LIKE".