[–]▶ No.915319>>915328 >>915430 >>915502 >>915846 >>916391 >>916656 >>917894 [Watch Thread][Show All Posts]
There is 30 million lines of software between the hardware and your programs. All these layers slow everything down and make it buggy and unreliable. The reason it's there is because hardware has gotten to complex and varied. We need to go back to the DOS days where the OS would just launch a program, or amiga where every single game was a bootable disk.
tldr: https://www.youtube.com/watch?v=kZRE7HIO3vk
▶ No.915324
>game is bootable disk
way ahead of ya
▶ No.915328>>915331 >>917538
>>915319 (OP)
>There is 30 million lines of software between the hardware and your programs.
That's a fallacious statement to say the least. Are you seriously implying all 30 million lines of code are going to be loaded into memory at once?
▶ No.915331>>915332 >>915337
>>915328
> Are you seriously implying all 30 million lines of code are going to be loaded into memory at once?
yes
▶ No.915332>>915334
>>915331
[citation needed]
▶ No.915334>>915393 >>917547
>>915332
30M is a very conservative lower bound, it's actually more like 55M.
▶ No.915336
Commodore 64 is where it's at.
▶ No.915337
>>915331
Say the average line of code is translated into 10 bytes worth of instructions. That's about 300 MB of instructions, which is not ridiculous for modern computers.
▶ No.915340
▶ No.915385>>915391
so can we actually discuss this?
▶ No.915391
>>915385
What specifically do you want to discuss? You can't just do a low effort post with one fact.
I think we're standardizing more which will help reduce bloat, manufacturers are settling with communications on board being PCIe or serial, NVMe will help reduce bloat, nvram even more.
I envision a computer which is nothing but NV-DIMM, CPU and PCIe lanes and it will be comfy.
▶ No.915393>>915398
>>915334
wordpress has 400k LOC? how?
▶ No.915398>>915407 >>915412
▶ No.915407
>>915398
All I see is bloat. Everywhere bloat.
▶ No.915412
>>915398
I watched the talk earlier. If he recall he said that he made the presentation in 2015 but has only just recorded it now. So that estimate has probably at least doubled by now.
▶ No.915420>>915430 >>915442
Lisp machines already solved this problem. Lisp has a GC, so there's one GC for everything. Hash tables, bignums, rationals, objects, classes, structures, arrays, strings, packages, and functions are all standard data types that can be shared between any program and created at runtime. Strings have all the properties you expect like being able to grow and shrink to fit the size of the text, and on the Symbolics machines they even had font attributes. Arrays support multiple dimensions with displacement like a rectangular part of a window. Everything on a Lisp machine also has built-in bounds checking and data is promoted to larger types if necessary. All of this is done by a combination of hardware, microcode, and OS, so programs don't need nearly as much code.
https://people.eecs.berkeley.edu/~kubitron/cs252/handouts/papers/symbolics.pdf
Maybe it was on someone's calendar to fix, but they
never see it because they can't run the program either.
Hmm. I used to think the strength of lisp machine tools
came from the fact that the developers actually used them
regularly in their work and depended on them in order to
develop everything they were going to need in the next
generation system. That is, I though that there was a
causal link between using your own tools and making them
better.
But maybe it's not whether you use your own tools that
makes them good, but rather that the goodness or badness of
your tools is just magnified over time by continuing to use
them. That would explain a lot of things about Unix...
▶ No.915424>>915426 >>915433 >>915593
What are good UNIX-Like FLOSS systems that aren't harmful bloat like this guy describes then? Ae we better off using a version of System V from 1998, or should we go off-UNIX to FreeDOS? I'm tired of UNIX-bloat personally, but Plan 9 isn't a daily-driver.
▶ No.915426>>915428 >>915491
>>915424
Modern Unix is terrible, and we're stuck with it. OpenBSD is probably your best bet for a daily driver.
▶ No.915428>>915433 >>915435
>>915426
Wew. Well... are there any new developments in Herd?
▶ No.915430>>915431 >>915821
>>915319 (OP)
>guy rants about OS being bloat and needing hardware access
>durr wha about redox.. we dun wroted an OS in rust.. ith dat gud now
lmfao rustfags
>>915420
Damn it sounds like this Lithp machine will blow us all away.
▶ No.915431>>915437 >>915510
>>915430
you should fuck off
▶ No.915433
>>915424
The only reason plan 9 isn't a daily driver is because web rendering engines are bloat squared.
OpenBSD is probably the least harmful UNIX OS that will still satisfy as a daily driver
>>915428
>expecting anything out of the most bloated side of unix development
At this point if you want anything hurd like you should go with the Genode framework.
▶ No.915437
>>915431
Please post pictures of your lisp machine.
▶ No.915442>>915470
>>915420
>so there's one GC for everything
What's kind of interesting about it, is that is is like a generational GC in which you have a generation for commonly collected things and one for rarely collected things on a lisp machine you could create your own areas aka "generations." Another interesting things areas let you do was say if you wanted that area to be garbage collected at all (though if you set it if you manually invoked the GC it would still run).
▶ No.915470>>915479
>>915442
If LISP is so great, why is it not used productively? Why is there no usecase still prevalent today? Why didn't LISP penetrate into anything useful than just garbage glued together?
▶ No.915479>>915487 >>915533 >>916389
>>915470
>why is it not used productively
It is used productively. Not by a large amount of people though.
>Why is there no use case still prevalent today
There are still use cases for it, you just don't see it being used very much since it is not very popular.
>Why didn't LISP penetrate into anything
I don't know as I've only discovered the beauty of lisp about a month ago due to one of the unix haters on this board. If I had to guess it would be a combination between people thinking of list as a language for beginners, the AI winter in late 80s / early 90s, lisp machines becoming overpriced due to the companies behind them not lowering the price / not making progress fast enough, and early lisp interpreters being slow. Oh, and how could I forget u-weenies are a big reason it isn't as popular it could be.
▶ No.915487>>915489 >>915593
>>915479
>I've only discovered the beauty of lisp about a month ago
<Time to go tell everyone they're doing it wrong!
▶ No.915489>>915494
>>915487
>Time to go tell everyone they're doing it wrong!
I didn't do that though.
▶ No.915491>>915495 >>915499 >>915512 >>915593 >>915626
>>915338
>>915342
>>915426
OpenBSD confuzzles me. They claim to be all about security and stuff. But security has 3 parts when it comes to information. Confidentiality, Integrity, and Availability.
So when looking at the CIAnigger triangle, how does OpenBSD stack up?
well it has confidentiality down, with all its cryptography and encryption focus, and how few security holes are there (or at least have been found. Its not like its got marketshare and is a common target)
But Integrity? Why does it still use FFS, with no TRIM support or volumes or anything? Where's ZFS, or Btrfs, or that Hammer2 thing? They're on different OSes of course. OpenBSD has no bitrot protection, no CoW, no snapshots >_<
And availability? well u can't really access your data right if its integrity is bad..
▶ No.915494>>915496
>>915489
It just made me laugh that it's only been a month and you're already evangelizing. Seems to be a thing with lispers though, many such cases! Real men use Scheme, or Racket
▶ No.915495>>915499 >>915512 >>915593
>>915491
>and how few security holes
Well ones you know about. They try to downplay as much as possible vulnerabilities which are exposed.
▶ No.915496>>915500
>>915494
>and you're already evangelizing
Rather than evangelizing it, I get more of a feeling that everything else sucks more. It just gave me another perspective to see that things did not have to be the way that I'd always seen them.
▶ No.915499
>>915495
This is not true, they post their vulns whenever they do an audit, they're the only BSD to do proper security errata, this misconception is caused by them trying to explain to zealous people on the mailing list that an exploit that blew up FreeBSD does a lot less or is mitigated on OpenBSD.
>>915491
ZFS has a licence issue, Suns CDDL has managed to make ZFS radioactive for OpenBSD and Linux.
HAMMER2 might still be a thing but not for a while.
But yes, you're right, OpenBSD would be better off with a newer filesystem.
▶ No.915500
>>915496
Fair enough, but I tend to think that if it were to work, then it would have done so at the time.
▶ No.915502>>915682 >>915685
>>915319 (OP)
I'm only about half way through. I have a serious issue with some of his premises.
> Every bootable game was its own operating system.
The "operating system" of, for instance, DOS, was separated between the BIOS (through agreement with the hardware vendors) and DOS itself. The BIOS provided a library of interrupts to access the hardware in an abstract way: the BIOS was a library that exposed a hardware abstraction layer for much of the hardware. DOS was a library that exposed process control and filesystem functionality. Pretty much every bootable game used the BIOS as its OS.
> USB and video cards made for the explosion in OS complexity
The OS provides a consistent software interface to the hardware. A good portion of Linux's out of control LOC count is drivers: programs that provide a consistent interface to hardware, sometimes doing in software what is expected to be done by that hardware. And many drivers are for hardware that hasn't been sold in 15 years by companies that have gone under.
Another layer is keeping the software from doing anything to the hardware that would damage the hardware or system stability. The average person cannot count with their fingers and toes how many times their Windows 95 system bluescreened. I can count with my fingers alone the number of times my Windows 7 system has bluescreened. Actually, I can probably do so with a single hand. All of that protection is expensive: both in code and in performance. Microsoft's file explorer has enough trouble managing my porn collection, do I want every single game company to be able to fuck over my HDD?
A lot of the bloat is in the bells and whistles that the OS supports. Some of it is in giving the user a consistent experience. Some of it is in giving developers a baseline of functionality that they don't have to buy or implement or read standards documents for. He talks about video card manufacturers having a consistent interface to their hardware. They have a de facto one: DirectX.
I mean, my take away thus far has been this:
What you are describing is a console.
And even consoles have their own operating systems, in part to add libraries that the third party developers don't have to buy or write.
▶ No.915510
>>915431
Why though? I am truly impressed with what lishp has to offer.
▶ No.915512>>915513
>>915491
>with no TRIM support or volumes or anything? Where's ZFS, or Btrfs, or that Hammer2 thing
All of this is useless bloat. (Prove me wrong). I've never lost data with FFS nor is it slow. I've been using an ssd with it for 3-4 years and haven't had any issues. And as someone who follows the mailing lists I'm not the only one.
>>915495
I don't know why openbsd's security offends you so. They downplay the vulns that don't affect them, either because they already had mitigations in place or because they didn't implement retarded ideas and bloat (recent intel debug instruction CVE for example). They treat the ones that do affect them very seriously but it just so happens that every one except for 2 so far has been ineffective without physical access.
▶ No.915513
▶ No.915533>>915538
>>915479
>LISP still used and productively
Post em. What companies use LISP for their products? What products?
>Could be interpreters slow
They are still slow.
▶ No.915538>>915559
▶ No.915543>>915621
lol so wat computer are very fast!! it doesn't matter!!!
▶ No.915559>>915668
▶ No.915566
The talk put my feelings really well into words for me.
I've always wanted computers as "intimate" as possible, I've used the Raspberry Pi as a desktop for years because it's so simple and well supported and documented, yet modern enough to be usable. I've even toyed with Amigas and C64s.
Only recently I've given in and switched to doing everything on a thinkpad, because I do stil need to use the modern web, and I've got a 1440p display which the VC4 in the Pi can't output well enough to. And I also want to start programming with Vulkan, which will only be supported by the future VC5 IP in future Broadcom chips...
▶ No.915593>>915615 >>915625 >>915678 >>916258
>>915424
>What are good UNIX-Like FLOSS systems that aren't harmful bloat like this guy describes then?
They don't exist and will never exist. UNIX-like can never be good because making it good would no longer be UNIX-like. The closest thing to a good UNIX-like OS is Multics, but Multics is much better than UNIX. You probably want UNIX-like because you want to port bloatware like X, systemd, glibc, bash, and Chromium to the new OS. UNIX spreads by porting because that code all sucks and nobody would write it from scratch unless they had Stockholm syndrome and brain damage. If you eliminated bloat and used a better language than C, you won't need as much code, so porting wouldn't matter as much because the programs would be simpler to write from scratch and take less time to replace.
>>915487
><Time to go tell everyone they're doing it wrong!
UNIX weenies are doing it wrong. OOM killers, panic, null-terminated strings, C arrays, C signals, and the rest of that UNIX bullshit is wrong. UNIX shills don't want you distinguishing right from wrong because UNIX is objectively bad. This is why it's important to learn about a lot of operating systems.
>>915491
OpenBSD "security" is bullshit. If they had any concerns about security at all, they wouldn't be using C. It's like selling a "secure tent" that's made of a material anyone can tear with their bare hands. Actually it's worse, because at least the tent makes it obvious that someone broke in.
>>915495
>Well ones you know about. They try to downplay as much as possible vulnerabilities which are exposed.
That's because they're caused by using C and being UNIX. They can only be solved by replacing C and UNIX completely. An honest description of the vulnerabilities would make C and UNIX look bad, and of course these bugs can never be fixed, so the OpenBSD "security" myth would be dead. They'll probably say "OpenBSD was secure for its time" and blame the users for "configuring it wrong" because that's how these weenies are.
With respect to Emacs, may I remind you that the original
version ran on ITS on a PDP-10, whose address space was 1
moby, i.e. 256 thousand 36-bit words (that's a little over 1
Mbyte). It had plenty of space to contain many large files,
and the actual program was a not-too-large fraction of that
space.
There are many reasons why GNU Emacs is as big as it is
while its original ITS counterpart was much smaller:
- C is a horrible language in which to implement such things
as a Lisp interpreter and an interactive program. In
particular any program that wants to be careful not to crash
(and dump core) in the presence of errors has to become
bloated because it has to check everywhere. A reasonable
condition system would reduce the size of the code.
- Unix is a horrible operating system for which to write an
Emacs-like editor because it does not provide adequate
support for anything except trivial "Hello world" programs.
In particular, there is no standard good way (or even any in
many variants) to control your virtual memory sharing
properties.
- Unix presents such a poor interaction environment to users
(the various shells are pitiful) that GNU Emacs has had to
import a lot of the functionality that a minimally adequate
"shell" would provide. Many programmers at TLA never
directly interact with the shell, GNU Emacs IS their shell,
because it is the only adequate choice, and isolates them
from the various Unix (and even OS) variants.
Don't complain about TLA programs vs. Unix. The typical
workstation Unix requires 3 - 6 Mb just for the kernel, and
provides less functionality (at the OS level) than the OSs
of yesteryear. It is not surprising that programs that ran
on adequate amounts of memory under those OSs have to
reimplement some of the functionality that Unix has never
provided.
▶ No.915615>>915625 >>915685
>>915593
Hi Mister Lisp meanie! ^.^
>Unix does not provide adequate support for anything except trivial "Hello world" programs.
>OpenIB Basic requirements: A computer running a Unix or Unix-like OS (OpenIB has been specifically tested with and is known to work under FreeBSD 10.3), Apache, MySQL, and PHP
tfw 8chan is a Hello world program.
>This is why it's important to learn about a lot of operating systems.
do your lisp OSes use the superior microkernel design?
>You probably want UNIX-like because you want to port bloatware
like it or not, a modern web standards-compliant browser is kinda essential for basic usability these days.
m-mister lisp meanie? wheres your modern lisp OS? where can I try it?
▶ No.915621>>915623
>>915543
>no one needs a computer in their home
>no one needs 16 bit computing for home use
>no one needs more than 640k
>no one needs a multi user multi tasking operating system unless youre a big business
>why would anyone need to travel faster than 35 mph?
history is not on your side, anon
▶ No.915623>>915624
>>915621
I would argue that people don't need a multi-seat multi-tasking OS given that it's very feasible for each person to have a whole computer exclusively to themselves. Android OS is technically capable of being modified to provide a multi-seat configuration given it is built on top of Linux. In practice, Android is designed to service only one user profile at a time because its intended use case is for a computer in which only one exclusive user is operating the Android machine.
▶ No.915624
>>915623
>multitasking OS isn't necessary
Nobody would be able to listen to music while working. Running a SSH server and a web server at the same time would be impossible.
>multiuser OS isn't necessary
Everyone would have root privileges. I hope you know why that's a bad thing.
▶ No.915625>>915630
>>915593
>>915615
Finally the two fags are in one thread. I wanna see a battle between these two.
▶ No.915626>>915628
>>915491
>integrity
>snapshots
It's called RAID you fucking faggot
▶ No.915628>>915629 >>915631
>>915626
but raid is just one of the things ZFS/btrfs/hammer2 offers to keep data safe.
standards have gotten a lot higher since the 1980s.
From ZFS wikipedia:
>Designed for long term storage of data, and indefinitely scaled datastore sizes with zero data loss, and high configurability.
>Hierarchical checksumming of all data and metadata, ensuring that the entire storage system can be verified on use, and confirmed to be correctly stored, or remedied if corrupt. >Checksums are stored with a block's parent block, rather than with the block itself. This contrasts with many file systems where checksums (if held) are stored with the data so that if the data is lost or corrupt, the checksum is also likely to be lost or incorrect.
>Can store a user-specified number of copies of data or metadata, or selected types of data, to improve the ability to recover from data corruption of important files and structures.
>Automatic rollback of recent changes to the file system and data, in some circumstances, in the event of an error or inconsistency.
>Automated and (usually) silent self-healing of data inconsistencies and write failure when detected, for all errors where the data is capable of reconstruction. Data can be reconstructed using all of the following: error detection and correction checksums stored in each block's parent block; multiple copies of data (including checksums) held on the disk; write intentions logged on the SLOG (ZIL) for writes that should have occurred but did not occur (after a power failure); parity data from RAID/RAIDZ disks and volumes; copies of data from mirrored disks and volumes.
>Native handling of standard RAID levels and additional ZFS RAID layouts ("RAIDZ"). The RAIDZ levels stripe data across only the disks required, for efficiency (many RAID systems stripe indiscriminately across all devices), and checksumming allows rebuilding of inconsistent or corrupted data to be minimised to those blocks with defects;
>Native handling of tiered storage and caching devices, which is usually a volume related task. Because it also understands the file system, it can use file-related knowledge to inform, integrate and optimize its tiered storage handling which a separate device cannot;
>Native handling of snapshots and backup/replication which can be made efficient by integrating the volume and file handling. ZFS can routinely take snapshots several times an hour of the data system, efficiently and quickly. (Relevant tools are provided at a low level and require external scripts and software for utilization).
>Native data compression and deduplication, although the latter is largely handled in RAM and is memory hungry.
>Efficient rebuilding of RAID arrays --- a RAID controller often has to rebuild an entire disk, but ZFS can combine disk and file knowledge to limit any rebuilding to data which is actually missing or corrupt, greatly speeding up rebuilding;
Ability to identify data that would have been found in a cache but has been discarded recently instead; this allows ZFS to reassess its caching decisions in light of later use and facilitates very high cache hit levels (ZFS cache hit rates are typically over 80%);
>Alternative caching strategies can be used for data that would otherwise cause delays in data handling. For example, synchronous writes which are capable of slowing down the storage system can be converted to asynchronous writes by being written to a fast separate caching device, known as the SLOG (sometimes called the ZIL -- ZFS Intent Log).
>Highly tunable -- many internal parameters can be configured for optimal functionality.
>Can be used for high availability clusters and computing, although not fully designed for this use.
▶ No.915629>>915630
>>915628
>long ass post
>most of which is copy and paste from wikipedia
Nobody Fucking Cares
▶ No.915630
>>915629
maybe take a minute to read it though. In terms of Integrity, OpenBSD's security is kinda lacking.
>>915625
OwO this sounds fun!
▶ No.915631>>915635
>>915628
>but raid is just one of the things ZFS/btrfs/hammer2 offers to keep data safe
t. someone who has no clue what a RAID is
▶ No.915635>>915638
>>915631
Redundant Array of Independent Disks. theres different ways to do it, all of them other than RAID0 provide redundancy in the event of a disk failure. Like RAID1, which mirrors the data, or RAID5/6, which offer parity at the cost of one or two of your disks respectively.
again, thats just one of the many things ZFS has.
Read the list again, baka!
▶ No.915640>>915646
>>915638
>Native handling of standard RAID levels and additional ZFS RAID layouts
its like you're sticking ur fingers in your ears and trying not to understand me! >_<
▶ No.915646>>915653
>>915640
>thinking this fucking matters
God I swear you faggots have low eye kew
▶ No.915653
>>915646
>>915638
You seem like the moron here, do you know what soft-RAID is?
▶ No.915668
>>915559
Were you expecting an exhaustive list?
▶ No.915678>>915809
>>915593
>this god-damn Unix-Hater retard
The reason I want a UNIX-Like is because I happen to like the UNIX Philosophy. We've already estabilished you're a high schooler, and are only readong other's comments, the others being sore losers in a culture-war. Don't you have anything better to do?
Giving the Eastern-Bloc internet was a mistake.
▶ No.915681
It's time to reinvent the computer. Not a new architecture, but a whole new concept of computer.
Yes, fellas. I think it's finally time for base-57 computing to take the world over by storm.
▶ No.915682
>>915502
>What you are describing is a console.
No, he's describing the classic home computers he grew up with. And thing were not much different in the professional world either. It might be hard to imagine now, but not to long ago hardware and peripheral makers still delivered products that worked correctly, were easy to work with and followed sensible standards.
▶ No.915685>>916258
>>915502
>The OS provides a consistent software interface to the hardware. A good portion of Linux's out of control LOC count is drivers: programs that provide a consistent interface to hardware, sometimes doing in software what is expected to be done by that hardware. And many drivers are for hardware that hasn't been sold in 15 years by companies that have gone under.
funny how i remove all the drivers i don't in my kernel and the loc is still in the millions
>after win9x, OSes were stable
no, fuck you. a few memes like memory protection changed nothing. and if you're seriously going to claim OSes are now secure against misuse of their API/syscalls, etc, i'm going to have tell you to fuck off
>Microsoft's file explorer has enough trouble managing my porn collection, do I want every single game company to be able to fuck over my HDD?
anything you run on a "desktop" system such as linux, mac, or windows already effectively has full privileges. running a game on a security sensitive device (even consumer banking) is retarded
>b-b-b but muh root and privileges n sheeeit
you have no idea what you're talking about. the bottom line is that if software is malicious, it can easily get full privileges on your desktop. if you're just talking about software inadverdently fucking up your system, then yes, that's a little more rare than win9x
>What you are describing is a console.
consoles are just a program embedded in a web browser on windows 10 in a proprietary plastic box now
(only a slight exaggeration)
>>915615
>tfw 8chan is a Hello world program.
posting messages to a website is the hello world of websites
▶ No.915809>>918003
>>915678
>I happen to like the UNIX Philosophy
Imagine if your CPU followed the UNIX philosophy. Who cares if it only works correctly for 90% of the instructions you throw at it. So what if it halts everyday. Just go walk up to your computer and restart it. Wouldn't you prefer a CPU that actually worked to specification rather than one that followed the Unix philosophy?
▶ No.915821
>>915430
>Lithp
For the uninitiated to this language:
<This otherwise unremarkable language is distinguished by the absence of an "S" in its character set; users must substitute "TH". LITHP is said to be useful in protheththing lithtth.
▶ No.915846
>>915319 (OP)
>posts 1 short paragraph text
>in tldr links to 2 hour lecture
▶ No.916258>>916757
>>915593
>OpenBSD "security" is bullshit.
No it's not. Most of the time, the weakest links is the user. The "secure by default" configuration is infinitely better than allowing for example "sudo su" on a user account. Even if you find kernel 0days, bugs happen, OpenBSD rather restarts than trying to magically fix unknown errors. This kind of "security above everything else" behavior is what you won't find in any other operating system.
>>915685
>anything you run on a "desktop" system such as linux, mac, or windows already effectively has full privileges
That statement is false on Windows Vista+. By default the UAC prompts happen on a different "secure desktop", completely independent of the main desktop.
▶ No.916389
>>915479
Moore's law killed Lisp machines back in the 80s. Symbolics could not keep up. If a new company tried to produce a Lisp/Smalltalk machine today they would have better luck.
▶ No.916391
>>915319 (OP)
Good talk about the evils of modern computing. Intel could fix this problem single handily if they follow Casey's advice. Intel has the CPU and a good enough GPU.
▶ No.916460>>916504 >>916579 >>916620 >>916757
>this whole thread
Do you guys not understand the meaning of abstraction? CPUs already have all the hardware to interface with USB devices, I2C devices, SPI devices and everything. A couple of hundred instructions is probably all you need to communicate with a piece of hardware connected to a bus because there hardware is all there, you just need to flip a couple of registers and the hardware will work as its supposed to. The additional LoC are to make the interfacing process easier to understand in the high level context.
If you want to do anything with a USB device, you don't have to go through a 1000 page manual about the particular CPU's USB subsystem registers and interrupts, you just have to pick up maybe the API guide and start.
This thread reaffirms my doubts that /tech/ doesn't know much about the hardware part of technologies they use everyday
▶ No.916504
>>916460
I agree with you. People are far too coarse with the words they use and they end up having a coarse worldview. In this specific case, their understanding of the word "bloat" and its relationship to modern systems is just like a buzzword. I agree in the principle that it's theoretically possible to provide software implementations that is objectively smaller in resource requirements. I just don't agree with their bitching and their lack of investment into writing such a system. The "less bloated" software platforms exist today, they simply don't care to use it at all.
▶ No.916579>>916739 >>917650
>>916460
you didn't watch the video
▶ No.916620>>916763
>>916460
>you just need to flip a couple of registers and the hardware will work as its supposed to.
yeah? what hardware have you implemented drivers for?
▶ No.916656
>>915319 (OP)
I recall need for speed could boot by itself to save memory usage of running DOS in the 'background', man that's fucking cool
▶ No.916658>>916660 >>916721 >>916757
My CPU can crunch 4 billion instructions a second. 30 million lines is a fucking joke.
▶ No.916660>>916662
▶ No.916662>>916665 >>916757
>>916660
>the reason I use MSN/AOL is because some clients or customers or programmers are using this service that I have to speak to at times. Freeware Gnu Garbageware alternatives such as GAIM and Trillian are even slower bloated crap
FOSS BTFO
▶ No.916665>>916667 >>918676
>>916662
Well, he is the person behind this website: http://gng.z505.com/
▶ No.916667>>916739
>>916665
I don't really get his arguments. He hates the GPL for software being free of charge but then advocates BSD which is also free of charge. Then he says you're not compensated for your work with GPL, but with BSD licenses, people can also freely take your work and sell it, but you receive nothing in return.
▶ No.916721
>>916658
>I have no idea how computer programs work
▶ No.916739>>916760 >>917521
>>916579
It's not worth it. The guy clearly doesn't understand how computers and the industry work.
>>916667
Don't try. These people just formed a cult around hating GNU and made justifications later.
▶ No.916757
>>916258
UAC doesn't and never has stopped anything. i haven't even heard the "UAC works" meme since 2006
>>916658
and you can audit about 1 line of code per day
>>916460
not all the code is for interfacing. there's tons of useless crap in there. and USB might as well be an HTTP server that your driver sends API calls to. it's complete shit and bloat
>>916662
pretty much every IM software ever made was shit really. there were a few that seemed alright but they were written in C which is retarded for an IM client. just putting in vulns for no performance gain
▶ No.916760>>917650
>>916739
>didn't watch the video
Too busy doing your homework?
▶ No.916763
>>916620
SPI, SMBus, I2C, USART, a couple of proprietary protocols in the late 90s. Nothing too extreme
▶ No.917521>>917650
>>916739
>The guy clearly doesn't understand how computers and the industry work.
And you can say that without watching the video? Just fuck off and come back with an argument after watching the guy.
▶ No.917538
>>915328
Electron.
A Browser + Javashit + Javashit frameworks .. all to draw a shitty, half-functioning "web page GUI". This is the state of things.
▶ No.917540>>917547 >>917550 >>917901
There's not much code between you and the hardware on Linux, especially in networking. If you use a full system debugger or even attach to qemu you can step through the lifecycle of a UDP packet from when the kernel is notified to when it's delivered to the application and it's not a lot of code. I think a lot of the whinging is from the kind of people who don't belong in tech and are overwhelmed by the complexity of the surrounding system.
▶ No.917547>>917921
>>917540
nice shill friendo :^), see >>915334
>linux: 17million
▶ No.917550>>917557 >>917921
>>917540
>let's take one process out of the hundred in the example
>let's ignore firmware, drivers and all the other software that does things with that UDP packet
>NOW IT'S EASY GUYS! JUST IGNORE EVERYTHING!
▶ No.917557
>>917550
You can step through the kernel and watch how the driver handles packets. As for the firmware, get a probe and watch it.
▶ No.917650
>>916579
>>916760
>>917521
Stop shilling so hard, it's embarrassing.
▶ No.917894
>>915319 (OP)
This makes too much sense for /tech/. They want to force everyone to stay on Windows but mostly Linux for their ideological/autismo reasons. Unix is an opinion and a bad one at that.
▶ No.917902>>917976
When he shows lines of code in linux-kernel, how much of that are just code for drivers and not actual kernel-code in that sense?
▶ No.917919
>Every amiga game had a custom OS!
Does that fucker have no idea what amiga kickstart was?
▶ No.917921>>918000
>>917901
>>917550
>>917547
You might not be cut out for software development.
▶ No.917947>>917955
Not entirely related as the project is about a virtual machine, but I really like their philosophy with regards to a complete unchanging codebase.
https://urbit.org/blog/2017.5-frozen/
▶ No.917955>>918001
>>917947
Please create a seperate thread for Urbit because Urbit leads to nothing but shitposting between /pol/ and antifa.
▶ No.917976>>917978
>>917902
This is a question he should have asked or at least noted.
▶ No.917978>>917996 >>918010
>>917976
He kind of did. He mentions an experiment where some group strips linux down to the bare essentials and it ran much faster than stock linux.
▶ No.917996
>>917978
This is why compiling your own kernel is by far the best option. You can disable anything you want pretty much. It's why linux is capable of providing embedded systems with such a robust OS without stress.
▶ No.918000
>>917921
>hurr durr i think what i do is hard
shut the fuck up
▶ No.918001
>>917955
its CIAniggers trying to sabotage a good alternative to current trash OSs
▶ No.918003
>>915809
dude it pretty much does only work 90% of the time, check out AMD64 processor errata
I would prefer an alternative if it has the same features, but I am too comfy at the moment to downgrade to older tech.
▶ No.918004
RAID doesn't keep data safe, it keeps it available
backups keep data safe
please rember mr cutie
▶ No.918010
>>917978
Apologies. Commented after only getting 30min in. Watching the rest though.
▶ No.918017>>918224
>amiga where every single game was a bootable disk.
The Amiga had a 256k Kickstart ROM (growing to 512K, 1MB, 2MB in later versions), on a machine which typically had 1MB or so of memory.
Even the C-64 had an 8K Kernal ROM, 1/8 of its total memory.
In comparison, 30 million LOC is relatively more efficient than these "we need to go back" systems when you take into account typical modern memory sizes, hard disk sizes, CPU speed, disk access speed.
▶ No.918027
This might seem like too much but couldn't someone take a reasonable affordable and decently powerful FPGA, create a brand new ISA for it, and create an OS with accompanying software and achieve a similar effect? I don't mean this as a substitute but rather as an intermediary step.
I think it's fair to synthesize the issue down to too much abstraction and any hardware, as well as software, that minimalizes this would be an improvement.
▶ No.918224>>918708
>>918017
Percentage of ROM vs. RAM is irrelevant. It's the total size of code that matters, because that's the amount of stuff you have to wade through if you want to fully understand the system. And as Terry Davis says, it doesn't scale linearly.
▶ No.918676>>918770 >>918797
>>916665
>Once you release software under GPL, the GNU and FSF (and its members) own you and your code - you have given it up to the cult.
Holy shit this guy's a fucking moron
▶ No.918694
amiga did multitasking on 1MB disk
there is no excuse
▶ No.918708>>918782
>>918224
Total code size is a poor measure of what it takes to understand a system. Like in Linux's case, almost all of the code is just uninteresting drivershit that lives in the kernel tree because Linux never got a proper driver API.
▶ No.918770
>>918676
He's an imageboard level troll. I rate him a level below Uriel from catv.
▶ No.918782
>>918708
In that case total code just means everything except the drivers you don't use. But that's going to change wildly depending on which hardware you have, and what the system is used for.
With the Amiga (and even moreso the C64 or other 8-bit computer), the code is essentially always the same, since they use identical hardware in most cases. And there's a whole lot less of this code, and it doesn't change much over time (unlike modern OS that have new releases constantly).
▶ No.918797
>>918676
I know right its like he thinks we're the bad cult. Like go read Stallmans blog he explains everything to us.