/dev/chat - 2017-01-26

« first day (2301 days earlier) ← previous day next day → last day (2636 days later) »

10:01 AM

Thoughts from John Goerzen

Not really on-topic. Though John is a long time Debian Eeveloper.

10:54 AM

@maulinglawns Yes, the new US Commander-in-Chief (shudder) seems to have inspired many to exceptional heights of eloquence. And John writes quite well normally. I'd also read Russ Allbery's post (link from John's post). Russ is always worth reading.

4 hours later…

2:26 PM

in Ask Ubuntu General Room, 4 mins ago, by Eliah Kagan

I feel like this question on U&L, though it is reasonably answered now, might also be possible to answer with (historical?) information about why some *nix OSes have system-provided root-owned executables owner-writable and others don't -- or, if not why, at least when the decisions were made and where, if anywhere, the choices have been documented. (I don't know enough about it to write such an answer, myself.)

in Ask Ubuntu General Room, 4 mins ago, by Eliah Kagan

4

Q: Why are executables in e.g. /usr/sbin writable by root?

Could you please explain why a binary compiled file (in, for example, /usr/sbin) has write permission for root user? For me, this is compiled. Meaning that direct write has no use and may expose file to some security issue somehow. A script (e.g. a bash file) may be writeable because it is a te...

linux files permissions root

@Kusalananda ^^ if you know and/or care to add a historical perspective.

Pretty sure that's just by default though.

@terdon Thanks terdon! I would need to do some research as I'm not currently aware of the history of these choices. I will update my answer if I find something.

I don't expect you will though.

in Ask Ubuntu General Room, 1 min ago, by terdon

@EliahKagan I doubt there was any decision. The decision was to make /usr/sbin writable by root, for obvious reasons. They just didn't think to explicitly go on and make binary files in there non-writable.

I don't think it was a choice to make binary files in that dir writable. They just get the permissions from their parent.

I don't see an "obvious reason" for having the executables writable (or not) by root other than aesthetics.

2:42 PM

@Kusalananda The only reason I can think of is "root can do anything" and/or "if it ain't broke, don't fix it"

Yes.

But @EliahKagan is curious and he points out that, as you say in your answer, other OSs took a different approach so there might be benefits/drawbacks to each.

I'm looking through the OpenBSD source tree and reading CVS logs right this minute ;-)

It's inherited from NetBSD in 1995. That's as far as I can see. There's a Makefile fragment bsd.own.mk that contains a BINMODE?=555 setting.

Interesting... thanks for looking into this!

This is the first revision of that file on NetBSD: cvsweb.netbsd.org/bsdweb.cgi/src/share/mk/…

So it's been there from the start. I can't see where NetBSD got it from, but it probably from the original Berkeley sources.

3:07 PM

@terdon The answer to that question now has my quick forensic work added.

Thanks!

Ping @EliahKagan too. Sorry I couldn't say more about Linux...

Nice!

I'd +1 but I already have

4:04 PM

Same here. :)

4:59 PM

Hi folks. I was wondering whether I should post a question about trying (and failing) to build the current unstable version of Okular on Debian stable. Or should I just talk to the project directly?

Talking to the project directly would enable you to get qualified help. Asking here would get you generic help. Also consider the audience and who would benefit from reading about the issue and the solution.

@Kusalananda True. It's probably not of general interest.

5:34 PM

@EliahKagan Tracked it down to a SCCS commit at the CSRG in 1990 by Keith Bostic. Still wondering about the Linux story :-)

@Kusalananda You mean someone added that at some point? It wasn't the default?

Well, the story really doesn't tell I'm afraid. The BINMODE?=555 was probably added to the Makefile fragment as a way of storing a default value somewhere. If that value was what was used previous to this or not, I don't know.

5:56 PM

Ok.

6:38 PM

@Kusalananda do not, I repeat not, engage with this user. Well known troll and wearer of tinfoil hats.

I hear you.

I guessed as much.

@Kusalananda It's this guy, if you're wondering: google.gr/…

2 hours later…

8:23 PM

@FaheemMitha That... sounds like a challenge. I'm guessing you'd need to backport a lot of KDE stuff.

@derobert Well, I satisfied the dependencies of the Debian package. But it failed to build anyway.

useful errors or not?

@derobert Um. Not sure. Do you want more information?

@FaheemMitha I'll see if I can spot quickly why it failed to build

@derobert Here's the build log -> gist.github.com/172e1716e7e47d6559e71cd560c72e3e

8:38 PM

@FaheemMitha check those two "see also" files

you may also need to backport cmake

Here is one curious error:

> /usr/bin/ld.bfd.real: cannot find -lpthreads

I thought it was called -lpthread

@FaheemMitha well, I'm posting a question about successfully (?) building the kernel using deb-pkg recipe, talking about that....

@FaheemMitha that's a major undertaking

@FaheemMitha yes, it's -lpthread

> Need to get 536 MB of archives.

@StephenKitt Bummer.

8:43 PM

@FaheemMitha hmmm, maybe try with a new cmake and see what happens

I have downloaded like 3GB of updates in the last week!

@Braiam debdelta?

@derobert testing/sid

ah wait

that project took off?

@StephenKitt Is okular seriously tied to the most recent KDE libaries, then?

@Braiam not sure, it used to work... but I've been on fast unmetered connections so long I haven't used it in years

8:46 PM

@FaheemMitha yes, the KDE bump between Jessie and Stretch is huge

@StephenKitt I wonder if we should start using "yuge" instead... that seems to be popular nowadays

Yeah, I figured it'd take backporting a fair bit of KDE.

and I think Okular is closely tied to the rest of KDE

@Braiam urgh

@FaheemMitha have you considered a testing chroot? Would probably cost a GB of disk space, but... would work.

@derobert yes that's probably the easiest solution

there might be some weird IPC issues running new KDE in a chroot from an old KDE desktop

but it's worth a shot

8:48 PM

(Well, at least work mostly. Not sure what'll happen when different versions of KDE libs try to talk to each other)

Or just upgrade to testing... Though of course that's an undertaking.

@derobert a useful one though, help get the bugs ironed out before the release :-)

Yep. And the freeze has started, so it'll be more stable than usual.

we reached debfreeze, so upgrading to Stretch could be fine

@derobert yes, updates have slowed way down, although there's likely to be a rush over the next few days before the real freeze on the 5th

@StephenKitt don't you say :D

8:51 PM

Running testing/unstable on my desktops for years, I'd say it's normally right after the release that everything breaks. Not right before.

btw, for the ones that haven't seen it thenib.com/this-is-not-fine

@derobert yes, when the breaking changes go in early in the cycle

I don't see why trying to build programs that are independent of specific versions is so hard. And people wonder why everyone uses Windows.

sadly, one of my system (cheap baytrail laptop) has to be kept on jessie due a bug in the nobody-knows-what's-wrong-in-the-kernel

@FaheemMitha because they're not independent

8:55 PM

@FaheemMitha on Windows you would have 5-6 copies of gtk depending on what you install

but Flatpak & co will improve things

@Braiam I'm not holding up Windows as a model to be emulated.

@StephenKitt Couldn't they build in abstraction layers?

@FaheemMitha it would cost a lot, and I doubt upstream would consider it worth their while

@derobert firefox 51 is on unstable

@FaheemMitha Well, KDE went through some pretty major changes for version 5 (a pre-KDE5 version would probably work). And its hardly surprising the new versions of programs might need new features added in new library versions.

9:04 PM

@StephenKitt Sigh. Well, it would be nice. Other programs manage without being tied to very specific versions of stuff.

@Braiam yep, running it now

@derobert not fair :(

how is it?

So far, about the same as normal. It's currently fast, but of course I just restarted it...

my stress testing is opening web.whatsapp.com and try to send a document. Currently it slows to a crawl, and I suspect is because animations

Ah. I don't have whatsapp.

9:20 PM

libwinpr-dsparse0.1:
Installed: 1.1.0~git20140921.1.440916e+dfsg1-12
Candidate: 1.1.0~git20140921.1.440916e+dfsg1-13

wut?

@Braiam you're a package revision behind. By one single revision

:P

so many numbers...

i'm trained to read package rev. numbers :P

most of that's git data

date, time, version string, commit reference, debian changes, package rev number.

interesting, marking packages as automatic/manual doesn't block /var/lib/dpkg

wait, what's the .1.?

the one between the date and the commit

good question, that I don't know, but I do know how to recognize a git commit hash right after that

9:25 PM

@Braiam probably in case they need a second version from the same day

you could always ask the maintainers :)

@derobert oh, that sounds... sensible

1 hour later…

10:48 PM

@derobert I don't know how much you are following this, but what are your thoughts about the orange terror tearing up the free trade agreements? It's a bit hard to imagine him actually doing anything helpful, but I suppose I could be wrong.

@FaheemMitha Tearing up free trade agreements is generally harmful. So no surprise its an activity he enjoys.

@derobert Even if the free trade agreement is a bad idea?

@FaheemMitha Free trade is a good idea.

The TPP one, at least, seems to have been quite controversial.

Yeah. It was also not really a free trade agreement.

10:50 PM

@derobert What, always? I've heard plenty, saying (plausibly) that it can often lead to bad things.

Protectionism has historically worked quite well for economies that have practiced it.

@FaheemMitha it can lead to bad things for certain groups

but as a whole, most of free trade is benefits

There's a book about this:

Bad Samaritans: The Myth of Free Trade and the Secret History of Capitalism

Bad Samaritans is a book about economy written by Ha-Joon Chang, a South Korean institutional economist specialising in development economics. The book criticizes economic mainstream and neo-liberalism. Chang mentioned developed countries require developing countries to change and open their markets. Rich and powerful governments and institutions are actually being 'Bad Samaritans': their intentions are worthy but their simplistic free-market ideology and poor understanding of history leads them to inflict policy errors on others. == Summary == Chang argued uncomfortable truth of capitalism and...

@FaheemMitha There are a few things almost all economists (i.e., the relevant experts) agree on. Free trade is one of them.

It makes good points, imo. And I've heard similar arguments elsewhere.

the trick is in trying to don't leave behind those that don't benefit the most

10:53 PM

@derobert See above. Alexander Hamilton didn't think so.

ie. you know that the free trade is going to open up your rice markets, you need to make sure that your producers are up to the competency or look for an alternative to move them onto

@FaheemMitha Alexander Hamilton? The guy from two centuries ago?

@derobert Yes, him.

I think I actually have a copy of "Bad Samaritans" somewhere. I should take another look at it.

Naturalists of his time didn't believe in evolution, either. Knowledge advances. Citing the beliefs of people from centuries past is not convincing or even relevant...

@derobert Er, what? His policies were adopted by the US, and were very successful.

You're presumably aware the US was very protectionist until the 20th century, right?

10:56 PM

Yes. Much of the world was. But that doesn't mean it was a good policy. Just not a policy so destructive it could counter the technological progress (and it wasn't a new policy, either).

There are certain policies that work in certain context, so whatever you read, you need to take into account that chaveat

Sadly, economist isn't a science of absolute truths, but smaller truths that need to be adjusted

And also, policies that are overall good (like free trade) can be bad for some people—often are. Trade definitely is. There are just (at least, to economists) better solutions to that than rejecting free trade.

@derobert Well, the policy had a reason. To protect the US's developing industries.

@FaheemMitha do the US need that anymore?

afaik, their industries are as mature as they can be

@Braiam No, of course they don't. But when they needed protectionism, they did practice it.

10:59 PM

@FaheemMitha I'm not sure what the consensus of economic historians is on whether it was actually helpful in that or a hindrance. But the consensus of economists about today's situation is that free trade is a positive.

That's the point.

@derobert It's a big part of why the US has been so successful. The US economy grew very fast during that period.

@FaheemMitha exactly the argument I'm trying to make: there are certain policies that works under certain circumstances. Right now, protectionism isn't needed.

@Braiam Well, other developing economies might need it. I'm not saying the US does.

Free trade tends to hurt the weaker participants. Though I'm not saying it's a simple picture, of course.

that's why I said, there's no absolute truths in economy

@Braiam I didn't say there was.

11:01 PM

@FaheemMitha But that's merely a correlation. I'm not sure it's a causation, and more importantly I'm not sure if the experts feel it is.

I recommend reading economics.mit.edu/files/10403

@derobert Shrug. Fair point - you can't prove anything in economics.

I wonder if there are really any experts in economics.

is not loaded with math or theory, its most likely a history lesson

@Braiam Ok.

@Braiam Math doesn't work. I doubt anyone really understands economic systems.

most math doesn't work you mean

some math does

11:04 PM

@ThomasWard I was using shorthand. I meant that mathematical economic theories don't model economic behavior very well. Unlike, say, physics.

@ThomasWard 1+1=pi

:P

@Braiam 1+1 = 10.

@ThomasWard Only in base 2.

what's to say that's not the base we're using

@FaheemMitha you are comparing a psysical science with a social science

11:05 PM

and who the hell is buzzing my phone

@Braiam I think you mean physical.

and despite humans being predictable with uncanny accuracy, making them all fit on our models is daunting task

heck I think the only model we believe is correct is the curve of demand-supply

we cannot agree on what money is!

@Braiam I forgot - you're an economics student. Unless I've got it wrong as usual.

@Braiam money is worthless

@Braiam and unfortunately California wasn't convinced to run the experiment back when it was out of money and printing IOUs...

11:10 PM

yeah, that's the problem with economy: experimenting can't be scary, because you are messing with people lives

(where they could have taken those IOUs as payment for taxes, to see if the hypothesis that's a sufficient condition for the IOUs to become currency is correct)

@derobert My memory is that the left, at least, has been very sceptical of free trade agreements in the past. I don't understand economics myself enough to weigh in, but presumably some on the left could be considered expert.

run in your systems how-can-i-help --show help... the list grew!

user image

The general thinking seems to be that the dismantling of controls generally acts to make rich people richer, and poor people poorer.

11:12 PM

... from capitolstartup.com/…

@FaheemMitha I'm not sure which political party is now against free trade. Might be the Republicans now.

@derobert I don't know those people. But you do realise that many economists are trained in the same few places, right? It's not that they all arrived at their opinions independently.

igmchicago.org/igm-economic-experts-panel has some explanation of who's on the panel

@derobert ok

If you mean this group, it's the new people.

Abhijit Banerjee (MIT)
Markus K. Brunnermeier (Princeton)
Liran Einav (Stanford)
Amy Finkelstein (MIT)
Oliver Hart (Harvard)
Hilary Hoynes (Berkeley)
Steven N. Kaplan (Chicago)
Larry Samuelson (Yale)
Carl Shapiro (Berkeley)
Robert Shimer (Chicago)

igmchicago.org/igm-economic-experts-panel/participants

@derobert Ok. But those people are all from the same 5 places.

11:18 PM

Well, more like 7, but yeah. That's the top economics schools in the country, pretty much.

Try and get an international group, at least.

@derobert Ok, 7.

They have a Europe panel as well

@FaheemMitha and to finalize my argument earlier, trade can be bad for certain sectors, but that depends on the sector and whenever or not they can compete

@Braiam True. Like I said, it's bad for the weaker groups/people.

igmchicago.org/surveys/trade-within-europe

11:20 PM

on the gripping hand truck drivers is the most common job in the US npr.org/sections/money/2015/02/05/382664837/…

@derobert The EU is a bit of a different situation, though.

The EU question included free movement of people, not just goods. That's even less popular among the general public.

I guess my overall feeling is that professional economists at fancy universities work for the elites. And I realise that makes me sound like a voter for you-know-who.

@derobert I thought people in Europe are generally pro-EU. Even if not by much.

After all, they voted to join.

Unlike with trade agreements.

@FaheemMitha Actually, while that may be in the appearances, the heart of the science is to make sure everyone is satisfied

@Braiam I don't follow.

11:23 PM

@FaheemMitha Donno what the current status is. But I'm sure you'll find opposition well above the 0% they found of the economist panel.

Anyway, I have to run...

@derobert Oh, yes. That's for sure.

@derobert Sorry about the distraction. Have a good evening.

I mean, Brexit itself is evidence that no Europeans are starry-eyed about the EU. Though the British aren't typical Europeans.

@Braiam Interesting article.

I'm not british, but it sounds to me that they were looking for a scapegoat for the discomfort they had, and they used migrants to take the blame/allow people to vent out

@Braiam That's certainly possible. Though I don't know anything about it.

« first day (2301 days earlier) ← previous day next day → last day (2636 days later) »

Transcript for

Jan '1726

/dev/chat

General discussion for unix.stackexchange.com. If you have a q...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile