FreeBSD development seems lost

MrBSD said:
We need total moderator dictatorhip.
Click to expand...
I don't think so. You lot usually get this under control yourself, and even the rant threads usually get some good nuggets to the top. The same here, I think. Took only 6 pages to realize the sky is not falling. That is one thing I like about this place, to be honest. People care about FreeBSD, about each other and how to improve things. In the end, if I don't care about something I don't even criticise it. But I see the reason for many a rant is a lack of information (same as happiness, which also is a lack of information).

So, can we come to a good footing about this or is this one of the few threads that really need to be locked, trashed, put into wet peat and ... you get the idea, yes?
 
linimon@ said:
"someone" was wrong. perl was removed over a decade ago. It became too complicated to try to keep perl-in-base up to date w/rt our release schedules.
Click to expand...
Yeah, I can't remember where I read that, it was on these forums somewhere fairly recently. Anyway I can understand stripping it back to a posix-type core for the base and removing dependencies so those no longer need to be tracked. I didn't know it had gone 10 years ago 😂
 
As for rust in the base... well, since "they" have mandated it, because "they" have been sold the story that it's going to make everything "safe", it seems we are inevitably going to see more of it as time goes on.
www.tomshardware.com

White House urges developers to avoid C and C++, use 'memory-safe' programming languages

The languages may pose a security risk when used in critical systems.
www.tomshardware.com www.tomshardware.com
(If I didn't know better I might think it was part of a nefarious plan by a geopolitical competitor to slow down the pace of software development in the west; but nah, it couldn't be that, could it?)
 
blackbird9 said:
As for rust in the base... well, since "they" have mandated it, because "they" think it's going to make everything "safe", it seems we are inevitably going to see more of it as time goes on. If I didn't know better I might think it was part of a nefarious plan by a geopolitical competitor to slow down the pace of software development in the west; but nah, it couldn't be that, could it?
Click to expand...
That's a suggestion from they, the White House, not FreeBSD. FreeBSD decided not to use Rust in base. The White House has also stated that Rust isn't the only option. The only language in base that makes any sense, is anything that natively compiles C and C++, is close enough to them. and is safer than them. It may be a successor of Zig, as it may not be Zig. From reading, Zig seems better than safer C replacements/implementations C3 and Odin.

If anything, it will be a compatible safer language that will natively compile C and C++, as to not have two different major compilers. As having Rust in any base would have redundancy of compilers in that base. If it happened, it would be a language most would agree on, and as good as Rust is, that's not Rust, due to the excess weight of the overlap of having two separate compilers.
 
If introducing any new language(s) into FreeBSD base as the successor of C, it should generate C as intermediate language, like C can generate asm.

It would ease reviewing proposed patches by skilled and experienced C programmers by generating C codes.

It would be OK for partial inline asm to cope with added ELF sections, if any.
 
Crivens said:
I don't think so. You lot usually get this under control yourself, and even the rant threads usually get some good nuggets to the top.
Click to expand...
Perhaps there's a cultural thing here too. In the western tradition, certainly in the uk, we have a healthy culture of "combative robust discussion" 😁 . We slag each other off, raise all kinds of objections and outlandish points, call each other idiots, argue the hell out of things; it's a very effective strategy for arriving at a good solution. We do it in good humour, and don't take it personally. I guess it goes along with western individualism. Whereas some other cultures are different. I remember working with a team of engineers from somewhere in asia, I'd better not say which country; if you gave a talk and then asked them "is there anything you don't understand" no-one would dare to say so, instead they would struggle on regardless to avoid having to admit they didn't understand something; which meant that things didn't get resolved, or took a very long time.
 
blackbird9 said:
Perhaps there's a cultural thing here too.
Click to expand...
Oh, there totally is.
Once I figured that out, how the patterns behind the curtain work (so to speak), it became a hobby of me to look for them and sometimes exploit them. In Europe, you have the climate and seasons to deal with. You slack off in summer or autum, you go hungry in winter. Doing the planning and sticking to it, but not at all costs, is what made you survive. In Asia, where you maybe have 3-4 harvests a year, keeping union is much more important. Ta(l)king down the emporer/king/warlord/chief could split the tribe/nation/... and reduce the chances for survival. Up here, when your chief had the bright idea to go north in autum for better hunting, you better call him names and make him see the blizzard. Central europe was the pressure cooker where bad ideas from the leaders would make you go follow the dodo real fast. This also meant there was little continuity in technical and other developments, this was the thing the southern/asian empires were better placed for. What we got out of it is different, it's like (1,0) is different from (0,1), and when you combine them you go further.
 
MrBSD said:
And this is another thing. Why even pay attention to those things? Who cares about linux and political bs. Wolf does not care what the sheep think about him
Click to expand...
I pay attention to the casual bashing by users and rebuffing by the devs so that I can make an informed decision for myself. For example, the DIY nature of FreeBSD - there's pain points like wifi that have been very well resolved by Android/iOS/Windows, but are still not ironed out very well on FreeBSD - that is a point to be aware of for both users and devs, I'd think. What's been actually coded up, whether it works right, what expectations are for wifi - I think it's good to know and have that info out in the open.

Sometimes, one just needs to have a thick skin. For users, that often means to not be offended when they are told 'Sorry, you just don't know something important'. For devs, that often means realizing that users do have expectations. I don't condone personal attacks, but sometimes, one does have a vested interest in making the other party connect the dots. And sometimes, one does need to learn the lay of the land, and accept that this is how things work, and not stir up too much shit.
 
Crivens said:
Oh, there totally is.
Once I figured that out, how the patterns behind the curtain work (so to speak), it became a hobby of me to look for them and sometimes exploit them. In Europe, you have the climate and seasons to deal with. You slack off in summer or autum, you go hungry in winter. Doing the planning and sticking to it, but not at all costs, is what made you survive. In Asia, where you maybe have 3-4 harvests a year, keeping union is much more important. Ta(l)king down the emporer/king/warlord/chief could split the tribe/nation/... and reduce the chances for survival. Up here, when your chief had the bright idea to go north in autum for better hunting, you better call him names and make him see the blizzard. Central europe was the pressure cooker where bad ideas from the leaders would make you go follow the dodo real fast. This also meant there was little continuity in technical and other developments, this was the thing the southern/asian empires were better placed for. What we got out of it is different, it's like (1,0) is different from (0,1), and when you combine them you go further.
Click to expand...
So, cultural Endianness issue? From Swift to Cohen and back from Cohen to Swift 😁

__builtin_bswap{16,32,64} FTW! 😎
 
There are some other fun things. In western stock market charts, 'red' means a stock's price is falling, and 'green' means it's going up. Whereas in Shanghai, red means the price is RISING and green means it's FALLING! So you have to be careful if you're trading on the shanghai stock market!
 
T-Aoki said:
If introducing any new language(s) into FreeBSD base as the successor of C, it should generate C as intermediate language, like C can generate asm.
Click to expand...
That is an interesting idea.

One could introduce FORTRAN code in the kernel and then use f2c.

I would like to have something like a scripting language that compiles to C, but I never found such.
 
I don't think it's really feasible either, some things just have to be done in the compiler. For example C++ started out as a translator that wrote C code ('cfront'), however to support some features of C++ such as late binding (I think, to support polymorphism) that approach was found too limiting resulting in them rolling their own compiler (at least, that's my understanding).
 
Compiling to C is more tricky than it appears at first. C has limitations that get in the way, for example integer overflow checking and exceptions. GC will be crippled.

Scripting languages are usually too dynamic and would require using the compiler at runtime very often.

Optimization information is lost and the optimization cannpt be performed. That is why most fortran work is done in real compilers and not in f2c.

Also, compiling to C doesn't really help with the source language using C libraries.
 
linimon@ said:
I was reading this while watching TV and eating a sandwich, and even so, it was a complete waste of my time. (cracauer's and cy's as well.)

There is a reason why I do not often visit the forums, and this thread is an excellent example.
Click to expand...
I can understand that stupid critics are demoralizing, but not all critics can be despised.

In my case, I only want to work as I worked since decades, and the idea that the system becomes wonderful, but unrecognizable, is an horror.

I am not a kernel developer, OS developer, I just use BSD / UNIX for my work.

I think developers tend to be unempathic with users, everywhere. They think every user has a powerful computer with a lot of RAM and fast like theirs, they think all users are also kernel developers, etc, or the opposite, every user is stupid and needs windows like systems.

I do not use Linux because their distributions are not BSD like, not even UNIX like, and always less UNIX like. They are forgetting what that is and making just 'wonderful software'.

On the mean time there is almost no email client for the command line that satisfy modern requirements, there is no really lightweight file and applications manager for X11, there seems to be no interest on user-land software in the
style of UNIX, but just wonderful software and innovations.

On the other side, where is a wonderful free OS from scratch? Plan 9 is one, but are there others, modern ones?
 
I've reread your messages. Very informative. But there's a bit of acid nihilism in them.
So far, I don't see any reason to deviate from the chosen path of development.
I'm not overly concerned about every little detail, so it will be easier for me to survive the "leap" in FreeBSD's evolution.
That is, I'm not overly picky about licensing, outdated protocols, and editors.
I think we'll get through this stage and move on.
 
cracauer@ said:
Scripting languages are usually too dynamic and would require using the compiler at runtime very often.
Click to expand...
Yes, I know it, I see the problem, the idea is perhaps only reachable with limitations, but perhaps just to have
some syntax sugar to make some programs less verbose would be a great thing ...

And regarding f2c, you know where it comes from: netlib. It is sure used for serious things.
 
hruodr said:
I would like to have something like a scripting language that compiles to C, but I never found such.
Click to expand...
V lang?

hruodr said:
Perhaps that is the reason. I have 'only' 4GB in my Desktop. Perhaps I upgrade to 8 GB.

The other solution is to install FreeBSD with UFS if I decide to work with FreeBSD 15 and when I do it.

For A NAS I will need ZFS, or perhaps Hammer 2 ?
Click to expand...
Hammer 2 = DragonflyBSD, dont know about it.
UFS and FFS in NetBSD have shapshots that good for backup..
 
hruodr said:
I think developers tend to be unempathic with users, everywhere. They think every user has a powerful computer with a lot of RAM and fast like theirs,
Click to expand...

I'm not sure. Myself I don't have any DDR5 based computer, so I am quite behind what many normal users have. I know many developers that use some random old Thinkpad (not because they can't afford a new one, because of the better keyboard) or outright potato PCs.

But one thing is fact: if you want to build heavy ports from source you should get some decent horsepower. That's unavoidable, otherwise you wait a long time and then you are screwed if there is a build error. A fast builder still costs less than $1000 for mainboard+CPU+RAM.

ETA: also, if you get many cores you need to back them up with enough RAM.
 
CeXP1917 said:
UFS and FFS in NetBSD have shapshots that good for backup..
Click to expand...
The reason I wanted ZFS and moved to FreeBSD was to have self healing using redundance and check sums.

It seems Hammer 2 has it.

I do not need till now other features of ZFS. I was always happy with UFS.

I will investigate V lang, thanks. There is also a LISP compiler, LISP is in some way like scripting language, but verbose.
 
One of these with a couple of hefty 8-core or 16-core xeons would do the trick. You might need a bit more than $1000 though, at least, over here. A nice Antec P180 type case and power supply, and buy a load of used server pull ram from ebay. Set it up as a headless build server. Then it doesn't matter what the front end you edit the code on is. I use old thinkpads for the same reason, I love the keyboards. Though I've been wondering about trying an X1 carbon recently, since I got my T490, although the T490 gets a bit warm on the underside when its on my lap; it's keyboard is better than I was expecting, although still not in X220 territory. Just bring back the old thinkpad keyboards, enough of the thinness already! Oh well, that's not going to happen... :'‑(

www.scan.co.uk

Asus Intel Z13PE-D16 Dual 5th Gen Xeon EATX Sever Motherboard

Buy from Scan - Asus Z13PE-D16 Motherboard, Intel C741, S LGA 4677, DDR5, SATA3, PCIe 5.0, 2x M.2, 2x 10GbE, USB 3.2 Gen1,E-ATX (12”x13)
www.scan.co.uk www.scan.co.uk
 
You must log in or register to reply here.
Back
Top