Handling partial recv()s
in [Unix Programming]
|
Prev: Proxy site open your favoraite sites facebook myspace ...etc
Next: And, following a series of proxy sites blocked open any page or download from RapidShare download sites have all accelerated the internet for free
From: Rainer Weikusat on 1 Jul 2010 07:13 John Gordon <gordon(a)panix.com> writes: > In <4c2bbb28$0$10436$426a74cc(a)news.free.fr> Nicolas George <nicolas$george(a)salle-s.org> writes: > >> Rainer Weikusat wrote in message <87bpasfcqd.fsf(a)fever.mssgmbh.com>: >> > The original topic happened to be 'network protocols'. Network >> > protocols usually don't employ 'arbitrarily large' records > >> Like HTTP? > > HTTP is an application-layer protocol, not a network-layer protocol. HTTP is actually a counterexample because the protocol definition does not specify a maximum header length. It just states that a server should be capable of handling any URL which could possibly fall within its responsibility and defines an error code for 'line too long'. Practically, HTTP headers are 'reasonably short'. But this is a ghost discussion, anyway. People who don't understand pointer arithmetic tend to want explicit lengths because they fear that their already needlessly complicated index operations would become too complicated to deal with otherwise. Similar to the Java-cult, where working code must not be used because it is taboo unless written in Java, this is probably a weird (and rather primitive) religion. One can have some hope that its aherents will (at some point in time in the future) have gained enough attributes of actual human beings to get past 'stone the heretic!' and gain some discussion abilities. Until then, it's best to treat them like all other kinds of foaming public preachers.
From: Nicolas George on 1 Jul 2010 07:43 Rainer Weikusat wrote in message <874ogjihfi.fsf(a)fever.mssgmbh.com>: > People who don't understand pointer > arithmetic tend to want explicit lengths because they fear that their > already needlessly complicated index operations would become too > complicated to deal with otherwise. And people who understand perfectly pointer arithmetic want explicit lengths because it makes the code simpler, and simpler code is better. You should derive no pride in writing complex code when a better design could have made it simpler.
From: John Kelly on 1 Jul 2010 08:02 On 01 Jul 2010 11:43:33 GMT, Nicolas George <nicolas$george(a)salle-s.org> wrote: >Rainer Weikusat wrote in message <874ogjihfi.fsf(a)fever.mssgmbh.com>: >> People who don't understand pointer >> arithmetic tend to want explicit lengths because they fear that their >> already needlessly complicated index operations would become too >> complicated to deal with otherwise. > >And people who understand perfectly pointer arithmetic want explicit lengths >because it makes the code simpler, and simpler code is better. > >You should derive no pride in writing complex code when a better design >could have made it simpler. Yes simple is better. Einstein said something like that too. And by now, this conversation's horse is dead. Why keep beating it. -- Web mail, POP3, and SMTP http://www.beewyz.com/freeaccounts.php
From: Rainer Weikusat on 1 Jul 2010 13:21 Rick Jones <rick.jones2(a)hp.com> writes: > Nicolas George <nicolas$george(a)salle-s.org> wrote: >> Fortunately, the days where network protocols were directly >> connected to a terminal ended a good decade ago. > > Speaking as a member of the Luddite Lunatic Fringe I will speak-out in > defense of protocols that a human can (attempt to) emulate via a dumb > terminal (emulator) and telnet :) The statement 'network protocols directly connected to a terminal' doesn't mean anything in itself, especially not when 'terminal' (as was originally claimed) is supposed to be something which 'randomly generates carriage returns'. It is already unclear what this 'terminal which randomly generates \r' is supposed to be but given that quite a few internet protocols where designed in the late 1970s, it ought to refer to something which was then in common use as a terminal. Such a thing cannot run the software necessary to interact with a TCP-based service even if a RS-232 line could coneivably used to transport IP. Confronted with a statement which obviously doesn't make any sense, there are generally two options: a) request that the person who made the statement explains what he is actually referring to. b) try to guess a sensible meaning based on one's own knowledge of the topic. b) is usually a bad idea because it is conceivable that the person who wrote the original statement was just throwing around terms he got from somewhere. My attempt at a) resulted in the affirmation that Mr George isn't willing to express himself comprehensibly because he considers a being incapable of reading his mind too lowly a lifeform to tolerate it on HIS earth (or something close to that). Assuming that your assumption regarding a possible technical meaning of the term combination was correct, George's statement would roughly parse as "I [he] haven't dealt with anything related to internet protocols either as developer or even as mere 'user support guy' for more than a decade" (the notion to do either of both without 'connecting protocols to terminals' in this way is absurd). Which kind-of settles the discussion: People tend to have opinions on lots of topics they know very little about and insofar their opinions differ from the way things are usually done, they are usually wrong.
From: Nicolas George on 1 Jul 2010 13:30
Rainer Weikusat wrote in message <871vbnm83d.fsf(a)fever.mssgmbh.com>: > People tend to have opinions on lots of topics they > know very little about Fortunately, not everyone here act like you do. |