Recent Forum Posts
From categories:
page »

You remind me that code quality overlaps with, but is not the same as technical brilliance.

A machine learning scientist may be technically brilliant but poor at coding due to lack of enough programming experiences.

Software engineering or code quality is not machine learning nor network protocols like zeromq or nanomsg.

Pieter hintjens and C4 didn't directly focus on code quality. In his talks, C4 could improve code quality indirectly by making a community work better and stick to certain rules of API evolution.

C4 is like a set of documented habits for an organization. Even a project without a process has implicit habits that maintainers follow. The feudal model also has implicit habits.

Actually, I made a comment about C4 as it seems to be an alternative to the "feudalization" model, but I've finally left it out not to distort the message. (I delete ~50% of what I write.)

It still would be interesting the investigate the outcomes for the 0mq experiment from both sociological and code quality perspectives.

by martin_sustrikmartin_sustrik, 12 Sep 2017 13:23

I know that you had a bad past with pieter hintjens.
The book he wrote before he died, social architecture, explains the economics of open source projects in depth. It spends a lot of pages to explain how C4 works well.

His strategy is to keep projects small by splitting big projects into smaller ones.
You used the same strategy in nanomsg.
Reputation economy works in small projects.

That's why small modular open source softwares are the way forward.

However, the difference C4 makes is that it doesn't treat outsiders poorly.
There's no clear border between core contributors and outsiders.

Whether or not ZeroMQ is technologically brilliant, I have to admit that pieter hintjen's community strategy reduces friction significantly in open source communities.(except when the owner tries to legislate the process in an already established project)

I conjecture that C4 is compatible with technological brilliance in the form of remake. nanomsg is a brilliant remake of ZeroMQ although it is in need of more manpower.

yoshi120 (guest) 02 Sep 2017 11:48
in discussion Hidden / Per page discussions » On Tolerance

Tolerance is one of the contemporary false values, inculcated with a great deal of civic education at a school, the aim of which is to enable people to cohabit, beyond their differences (sex, origin, class , religion, etc.)

In truth, and as the American philosopher Christopher Lasch, in his ultimate book, The Revolt of the Elites and the Betrayal of Democracy, this tolerance, masked only the failure of the union of society, by confusing respect and tolerance.

"Democracy also demands a more stimulating ethic than tolerance. Tolerance is very nice, but it is not the beginning of democracy, not its destination. Today, democracy is more seriously threatened by indifference than by intolerance or superstition … We are determined to respect everyone, but we have forgotten that respect must be won. Respect is not synonymous with tolerance or price for "different lifestyles or communities". This is a tourism-based approach to morale. Respect is what we experience in the presence of admirable successes, admirably trained characters, natural gifts put to good use. It involves the exercise of discriminatory judgment and not of indiscriminate acceptance."

by yoshi120 (guest), 02 Sep 2017 11:48

I am amazed at the ignorance reading the comments as if martin_sustrik has no understanding.

"You obviously do not know C++"
"You do know C++ is a SUPERSET of C?"
"You do know that C++ is not just an OOP?"

Blimey. The guy mentioned he has been programming in C++ for 15 years (at the time of writing) and is talking about SYSTEM-LEVEL programming, rather than Application Development.

Why is that so hard to digest? Seriously, if C++ programmers want to continue being butt-hurt that the "inferior" C language can still be preferred over C++ then I recommend you watch -
Mike Acton Data Oriented Design CppConn 2014

by Martyn S (guest), 29 Aug 2017 14:49
Auguste (guest) 17 Aug 2017 08:54
in discussion Hidden / Per page discussions » On Tolerance

It's not about tolerate someone but tolerate its act it's a huge difference.

by Auguste (guest), 17 Aug 2017 08:54

I think so.

Here's a thought experiment: Wipe out all antiterrorist memes so that nobody has any idea of what antiterrorism is or any memories of how it used to be. After the next terrorist attack you'll obviously get some kind of reaction. Will it be the same as the previous round of antiterrorism? Unlikely. It will probably have somehow different dynamics, different techniques, different rethorics, etc.

That seems to suggest that antiterrorist memes are not contained in the terrorist memeplex. Rather, presence of terrorism creates a favourable habitat for all kinds of antiterrorist memes. That sounds very much like mutualism.

In general, sex is a pretty specific mechanism for mixing genes. I don't think using the same name for stuff that's structurally different is very helpful.

by martin_sustrikmartin_sustrik, 14 Aug 2017 14:48
MugaSofer (guest) 14 Aug 2017 12:14
in discussion Hidden / Per page discussions » Infinite Mirrors and Sexual Selection

If you wiped out all antiterrorist memes, they would grow back pretty quickly after the next terrorist attack. Similarly, antiterrorism spreads terrorist memes.

Are they really separate memeplexes?

by MugaSofer (guest), 14 Aug 2017 12:14

While I am agnostic about existence or male/female memes, the example above doesn't seem right. Male and female are just two phenotypes of the same genotype, whereas left-wing and right-wing are two different memplexes. I think it would be more accurate to call them, using ecological terminology, "mutualist memes" or maybe "commensalist memes".

by martin_sustrikmartin_sustrik, 13 Aug 2017 12:33
tailcalled (guest) 13 Aug 2017 12:18
in discussion Hidden / Per page discussions » Infinite Mirrors and Sexual Selection

Sometimes, you do have """male""" and """female""" memes; it seems plausible, for example, that "terrorism" and "war on terror" memes reproduce with the help of each other. Some "culture war" seems to have a similar pattern with "left-wing" and "right-wing".

by tailcalled (guest), 13 Aug 2017 12:18

from Gift vs. reputation in hacker culture:

if I download a piece of open source, and it’s useful to me, and I find a bug in it, I do indeed feel a reciprocal obligation to the project owner (not just an attenuated feeling about the culture in general) to gin up a fix patch if it is at all within my capability to do so – an obligation that rises in proportion to the value of his/her gift.

while it's probably true about the feeling of proportioned obligation to fix that bug, your patch still comes as a surprise to the owner, esp. if the bug was not known.

If there's anything reciprocal there, it is in the owner's obligation to merge the patch, but only in proportion to the value of the fix. At that point it starts to look more gift-culture-ish but there's still no obligation and the appearance of reciprocity is incidental to a more central, motivating force: software value…

Yes, that sounds weird. I wonder what a linguist would have to say about that.

by martin_sustrikmartin_sustrik, 19 Jul 2017 06:20
DAB (guest) 18 Jul 2017 20:59
in discussion Hidden / Per page discussions » Linguistics and Programming Languages

One of my favorite examples of odd-seeming semantics comes from COBOL, where conditional expressions can be abbreviated.

For example " if foo = 'A' or foo = 'X' " can be abbreviated as " if foo = 'A' or 'X' "

That seems OK until you realize that " if foo not = 'A' and foo not = 'X' " must be abbreviated
as " if foo not = 'A' and 'X' ", whereas English usage would call for "or" in place of "and".

But in COBOL, " if foo not = 'A' or 'X' " always evaluates as TRUE, because what that abbreviates
is " if foo not = 'A' or foo not = 'X' ", which is always TRUE.

by DAB (guest), 18 Jul 2017 20:59
Arran Cudbard-Bell (guest) 07 Jul 2017 23:43
in discussion Hidden / Per page discussions » Using likely() and unlikely()

Unless of course the debug level can be altered at runtime, in which case marking all the debug statements as unlikely would be a very bad idea indeed.

by Arran Cudbard-Bell (guest), 07 Jul 2017 23:43
Eric S. Raymond (guest) 05 Jul 2017 18:46
in discussion Hidden / Per page discussions » Note on Homesteading the Noosphere

A minor error or two but some sharp thinking. I have blogged a response:

Gift vs. reputation in hacker culture:

by Eric S. Raymond (guest), 05 Jul 2017 18:46

Fixed. Thanks!

by martin_sustrikmartin_sustrik, 04 Jul 2017 05:23
Allan Wind (guest) 03 Jul 2017 19:24
in discussion Hidden / Per page discussions » Reputation Engineering, Part I


by Allan Wind (guest), 03 Jul 2017 19:24

Yes, it crossed my mind as well. In theory, one could use their PGP key as a way to accumulate reputation (signatures by different people). However, at least as far as I know, nobody uses it as a "store of value".

by martin_sustrikmartin_sustrik, 02 Jul 2017 19:52
mcz (guest) 02 Jul 2017 16:38
in discussion Hidden / Per page discussions » Reputation Engineering, Part I

Sounds a bit similar to the web of trust.

by mcz (guest), 02 Jul 2017 16:38
page »
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License