just the other day I noticed that I am no longer able to use the web based e-banking of my bank without manually setting a bogous user agent in my browser...

how did that make things better, exactly?

"Oh, you're not using Windows, Linux or MacOS with chromium? we clearly have to protect you from yourself"

the situation gets worse once I tell you that I'm a swiss citizen located within Switzerland - the land of banks.

That last is probably a FreeBSD User-Agent. We are suffering the pain of being an insignificant minority. I have to do that too. I set it to Linux because it would pain me more to set it to Windows. I highly recommend the User-Agent-Switcher plugin which can automatically switch due to URL names. I need it for chase.com or they refuse to authenticate me.

you understand my point - right? this is outright ridiculous. User-Agent fields were never meant for this.

The developer obviously did not think things through. It's user provided data, which could be ANYTHING, so using it for an authentication filter is just wrong.

we have technology from the 70s and 80s which still performs exceptionally well in 2024 and meanwhile we have crap like node.js, go (and potentially rust) which added absolutely no value - and everybody claims that it solves problems - while introducing about two times as many new problems.

rwp, if I may ask - and this is entirely irrelevant - are you located in the EU?

But those solutions are introducing DIFFERENT PROBLEMs not the original problems. People can claim that the original problem is now solved. We are left to deal with the new problems now.

I live in Colorado in the middle of the USA snuggled up against the foot of the Rocky Mountains.

but the new problems are often outright ridiculous (in my subjective, personal opinion)

I know I might sound like a 80 year old dude, the said truth is that I'm literally 30 years old at this point)

Colorado is nice.

Saying that you liked C++ because it had so many nooks and crannies (my paraphrasing) does not sound like an 80yo honestly. So no I did not expect you to tell me to get off your lawn. :-)

I can't argue against that - being a 90s kid.

again: really just enjoying the discussion. I have no problems accepting that other people might have different - equally valid - opinions. Something that seems increasingly rare today.

Yes. Having a civil discussion about differences of opinion is increasingly rare today.

I like that with C that at one time I pretty knew the entire feature set and the stdio library in its entirety.

I dislike that with C++ I need a large bookshelf of wisdom that I need to refer to and still will never be able to hold the entirety of it in my head.

Rust I dislike many parts of the syntax of the language. There are some clunky things in there. But I understand the goal is to avoid memory overrun errors. Errors which I don't have a problem with in C. Errors which I see being made daily by other people. And therefore I value the addition of a language to make that impossible for newcomers to the language. It's a great experiment to see if that helps in the long run. I hate the

bootstrapping that is needed to make a Rust compiler.

zig nim and Go also have memory stuff for that right. what about them?

Go got the "branded mark" for me when it was initially only statically linking to consuming libraries back in the day

I know nothing about zig nim. Go-lang is a memory safe language that uses garbage collection.

nim has gc zig does not

"Oh, you wrote a totally safe library in Go? Great. Oh wait, there is a vulnerability in OpenSSL? Better rebuild & distribute the entire thing because OpenSSL is statically linked"

Go-lang was created by, among others, Ken Thompson of C fame, after being hired on at Google and becoming frustrated with the hugely long compile times of C++ which was the main language there. Therefore Go-lang has many features to make compile times faster.

jbo, That's the main argument against container deployment these days. Even for dynamic library compilation after everything is bundled into a container then it is the same problem.

Remember, as you stated, every solution creates new problems.

rwp, C++ compile times are notably shorter since the introduction of modules (albeit "barely supported" as of today)

and that is what I love about C++: It is evolving in a way that allows you to keep using the crap you wrote down 20 years ago

Those C++ modules were created to solve a problem so must be creating two new ones. And it only took three decades for that solution to arrive.

do I need to remind you of the debacle of python 2 -> python 3 migration?

www/chromium...

No. I am still suffering the effects of that. And I don't even use python myself.

exactly.

my point is that C++ is not as bad as people like to seem to make it.

it is extremely balanced. It's not perfect. but it achieved something that I have not seen in any other language - even remotely.

How can it be bad? It is driving people like me to look seriously at other new languages like Go-lang, Rust, Clojure, and others. :-)

clojure?!

we seem to be digressing >:D

Lisp. I am trying to improve my functional language programming skills. And we have been digressing for a while. We should really talk about FreeBSD issues here.

(again, don't forget that I clearly have a subjective preference for C++ - so I am obviously biased)

lol <3

I like you man.

makes me feel like it's 2003 again

I think I asked this before but is 16GB of free space enough to build world

or do I need more like 20-24?

at least 20 i think

for a standard build

(including source)

i see

can you get updates as binaries also, or do you have to build world for the latest security/stability improvements?

SponiX: yes, you can get updates as binaries too; man `freebsd-update`

rtprio: thanks for the reply. I need to get FreeBSD in a vm to start tinkering with all these things

no time like the present

i keep one on my windows system in hyperv which is rather handy

johnjaye: I've built world on a VPS (30GB).

Usually RAM has been more of an issue. But I don't know if that's a freebsd or hardenedbsd thing. It can use close to all of 30GB RAM though.

rwp: you can give Haskell a shot to boost your FP skills, though its strict lazy-evaluation is an acquired taste, something I can't get used to.

It's one of the reason I like Python is the ability to "control" where I use lazy evaluation with generators and generator expressions, but that's my niche preference.

Hi there

is there a better way to upgrade php version ? Now I'm just collecting information about existing modules and then remove existing, install new.

What's up?

Ah

FreeBSD has Plasma 6 pkgs already?

SponiX: apparently yes, freshports.org/x11/kde6

Title: FreshPorts -- x11/kde6: KDE Plasma Desktop and Applications (current)

Nice

Why the following messages happen in /var/log/messages from time to time.

pid 57935 (sockstat), jid 0, uid 1001: exited on signal 11 (no core dump - other error)

what could be the reason ?

signal 11 is SIGSEGV

(program crash)

I don't understand the reason of the crashing of sockstat with signal 11

sockstat just read info from kernel, why it crash?

nerozero: best way is to run a test environment with your new PHP version including all new modules and test it

jmnbtslsQE, the question was not about testing but upgrading

yeh, i'm talking about upgrading

unless you don't care about whether it works

not sure what you mean by module though - if it's a PHP extension, those need to all be built and upgraded on the new PHP version

ports lang/php83 and lang/php83-extensions for example

I'm attempting to generate a patch for the Nginx port, which includes a Makefile and a Makefile.extmod. However, when executing the command make makesum, I notice that the distinfo file only lists checksums for a select few of the additional modules. This issue arises even when no modifications have been made to the port, simply by running make makesum. Does anyone have insights into why this might be happening?

Anyone using Plasma 6 on FreeBSD 14?

hello all, I have a Intel HD Graphics 520 (Skylake GT2). What would be the command to list it's max resolution?

Hello

Can i reduce ZFS?

well, talking about size, probably shrink is a better word

AFAIK, No. You can only expand it. To reduce one would need (again AFAIK) to create a smaller array and then data can be poured onto the smaller array. (using send|recv most likely)

oh no. freebsd-update install hung during my upgrade to 14.0, I tried a freebsd-update rollback, and now I'm booting to read-only :(.

Which was the earlier version?

would any of you folks have any suggestions for jumping back - or is this a better time for a fresh install/rebuild because I'm probably in an inconsistent state

13.2-RELEASE

evas: no chance you created a ZFS boot environment before running the upgrade?

I was not very dilligent about keeping on top of updates, I have to confess

I think that freebsd-update did, but I did not independently

try 'bectl list' from the single-user environment then 'bectl activate <name>' if one looks relevant - i don't think you need to remount root rw to do that

lw: thank you so much!! that worked :D

evas, When freebsd-update install hung, what it was doing?

I mean what were the progress messages? Did you see anything in /var/log/*?

parv: I expect it is my issue -- I had some errors about /usr/include/c++/v1/__string existing but not being a directory and I interrupted it (and other subdirs). that was probably dumb. I'll check /var/logthough, one sec.

evas: that error is expected upgrading to 14.0, it's harmless *iff* you update to the latest 13.2 patch release before upgrading to 14.0

evas, Ha! I saw the same & thought installation was screwed up. Later found that the message can be ignored. See reviews.freebsd.org/D41945 & bugs.freebsd.org/bugzilla/show_bug.cgi?id=273661

Title: ⚙ D41945 freebsd-update: handle directories changing to files, too

... also what lw had mentioned

you can only ignore it if you updated to the latest 13.x patch before upgrading, otherwise you will get missing files after the upgrade is done

is there any tool to disable bios fan control?

(if you're upgrading from 12.x, i think that got the patch as well, but check the EN to be sure)

I saw that too lol. My impulse was OH NO! interrupt quick before anything bad happens lol

I had wasted my time upgrading 13.2 release to 14.0 only to throw that away that I could have safely ignored (if that would have been mentioned by "freebsd-update") to install 14.0 from scratch.

I'm not seeing anything suspect in /var/log/messages. I'll tail it while running it again

evas: I'd open a bug if none that have been mentioned so far cover it.

there's no need to open a bug for this issue, it was fixed and an EN issues a couple of months ago

Embarassing question. While using `freebsd-update upgrade --currently-running $(freebsd-version) -r 14.0-RELEASE` -- should I be fetching/applying all updates first? I hesitate to open a bug, I think I may have done this incorrectly.

evas: usually it probably doesn't matter, but in the specific case of upgrading to 14 from 12 or 13, yes, you have to apply all patches for your current release first

... which is not a bad idea in general anyway

thanks lw! lol, well, I won't forget this next time (I say as I add this hastily to my notes)

this was a pretty unfortunate issue, it's probably the single most frequently asked question about upgrading to 14.0

not sure how it didn't show up in testing

I think I saw an issue where it did. It looks like it came up in the BETA release IIRC. Hang on, I'll see if I can find it again

I really do think this is my own mishap

johnjaye: FYI I'm building world on a VPS with 26G free. 15G was not enough. It'd run out of disk space.

sorry, I was looking at the same link that parv linked earlier: bugs.freebsd.org/bugzilla/show_bug.cgi?id=273661 . this was my goof

Title: 273661 – freebsd-update install: ///usr/include/c++/v1/__string exists but is not a directory

jbo: bugzilla is sending me email about the sublime-music bug (+ 4 Feb 18 bugzilla-noreply⊙Fo ( 36) Problem reports for lexi.freebsd⊙lo that need special attention) so i set it to maintainer-feedback+, this one is still waiting on you though right? there's nothing i need to do?

jbo: (the last patch is ready to commit as far as i'm aware)

am I the only person who's ever seen panic: tcp_hostcache: bucket length too high at 236: 30 (bear in mind I dont think this is actually a FreeBSD issue, but a downstream issue)

can't say i've seen that one