IRC Logs

Network: freenode Channel: #php

	Enter your search terms Submit search form
	Web www.irclog.org

[20:26]<lsa-nzya>lol
[20:26]<wjlllrws>Sun... Certified.... I'm officially scared now
[20:26]<kdrg>dools: you have the choice in java too. you can choose to catch all and rethrow as a RuntimeException
[20:26]<mjjfs>consider the difference between mysql_query($sql) or die(); and mysql_query($sql) or throw new SqlException();
[20:26]<kdrg>but that is so clearly dumb that you tend not to do that
[20:26]<mjjfs>KDan: you don't have the choice
[20:26]<kdrg>whereas php basically does it for you :-)
[20:26]<mjjfs>it's enforced by the compiler
[20:26]<dgvgzdygysv>dools, not in safe mode...
[20:27]<byre>Ugh, I hate allergies.
[20:27]<mjjfs>Determinist: okay, and if you execute that command on the command line, what happens? oh, hold on you are redirect to out.txt
[20:27]<mjjfs>so what is shell_exec going to return?
[20:27]<lsa-nzya>BigE i hate rotator cups
[20:27]<dgvgzdygysv>it works fine on the command line mate but for some reason i'm not getting any output from the script or from the log file...
[20:28]<mjjfs>Determinist: you are doing > out.xtx
[20:28]<suv9w9>dools, what considerations did you have when you selected session_start? limitations, performance, RAM?
[20:28]<mjjfs>s/xtx/txt/
[20:28]<mjjfs>Determinist: so whatever output would be returned by shell_exec is presumably going into out.txt
[20:28]<dgvgzdygysv>dools, that's changed, i'll show you the updated thing
[20:29]<mjjfs>suv4x4: not really
[20:29]<suv9w9>dools, what then
[20:29]<dgvgzdygysv>http://hashphp.org/pastebin?pid=7554
[20:29]<mjjfs>suv4x4: what indeed?
[20:29]<rdvxzrwff>suv4x4, i missed it... wtf are you trying to do now?
[20:30]<mjjfs>Determinist: if i put wget --timeout=30 --tries=60 -o log.txt -O /tmp/buffer "http://myplanetside.station.sony.com/outfit.jsp?outfitId=6539&worldId=21" on the command line it gives absolutely no output
[20:31]<suzpmun>isnt this php :|
[20:31]<dgvgzdygysv>dools, bash, standard?
[20:31]<mjjfs>tcsh
[20:31]<dgvgzdygysv>dools, those quotes are meant for a bash shell mate
[20:31]<mjjfs>Determinist: and what shell do you think php user is running?
[20:31]<dgvgzdygysv>i'm not sure
[20:32]<suv9w9>cythrawll, I have classes that start session on demand (if it's not) and they are loaded on demand by autoload, which is sometimes in the middle of HTML code. So I'm faced with this: start session in advance for every page, in case a class needs it, OR ob_start every page, so the classes can start the sessions.
[20:32]<mjjfs>Determinist: try a simpler example, like $var = shell_exec('echo "im a pommy dork"'); or something and see if you get it bac
[20:32]<mjjfs>k
[20:32]<dgvgzdygysv>dools, alright, checking.
[20:33]<rdvxzrwff>are the classes you would start with ob_start be used on every page?
[20:34]<dgvgzdygysv>dools, the simpler example works. must be something wrong with the wget syntax then.
[20:34]<mjjfs>Determinist: i'd say that the php user is not using bash
[20:34]<dgvgzdygysv>dools, i'll try modifying the string then. brb
[20:35]<e0ee>why is that, dools ?
[20:35]<suv9w9>cythrawll, thing is, this starting to become a large network and I absolutely can't guess what classes may or may not be loaded on every page. Of course I can log what classes are loaded but an "if" might change everything (I use autoload, and some classes are loaded truly "on demand")
[20:35]<suv9w9>network = framework cythrawll sorry
[20:36]<dgvgzdygysv>dools, nada, still. there must be something wrong with wget. other commands work well.
[20:36]<mjjfs>Determinist: if i remove -o log.txt then i get output
[20:36]<rdvxzrwff>so session_start would be the easy lightweight way to do it
[20:36]<mjjfs>Determinist: i'm pretty sure wget is stable
[20:36]<suv9w9>cythrawll, I see. Is ob_start putting any strains on the server?
[20:37]<dgvgzdygysv>dools, it may be, but perhaps the hosting service removed it from the server?
[20:37]<dgvgzdygysv>dools, this exact script has been working for over a year and just stopped doing so a month or so ago. i just came around to fixing it now
[20:37]<rdvxzrwff>suv4x4, depends on RAM etc...
[20:38]<suv9w9>cythrawll, I see, and session_start I suppose then is pretty harmless for performance
[20:38]<mjjfs>Determinist: this is what i get: http://nanobin.nanosouffle.net/1484
[20:38]<rdvxzrwff>if your doing a framework that you plan to redistribute, output buffering may not be wanted by every user
[20:38]<kdrg>19:27 < cythrawll> if your doing a framework that you plan to redistribute, output buffering may not be wanted by every user
[20:38]<kdrg>^^ the users are evil
[20:39]<kdrg>they must be punished
[20:39]<rdvxzrwff>suv4x4, yeah, the only thing i see bad about session_start is it will set session headers on every page, needed or not, but alot of pages do that these days
[20:39]<mjjfs>Determinist: is it your own host? it's possible that the host changed, for some reason, what shell apache user uses
[20:39]<dgvgzdygysv>dools, it's a shared hosting service i'm using, cybexhost.com
[20:39]<mjjfs>Determinist: something to consider
[20:39]<suv9w9>cythrawll, ok thanks. I'll do that and hope make my framework a bit smarter on the next interation
[20:39]<dgvgzdygysv>dools, indeed. they've been giving me alot of grief lately.
[20:43]<suv9w9>cythrawll, one question: doesn't session_start make a page to load and desearialize the session file too?
[20:45]<wjlllrws>KDan: rathger than users, I'd punish the developer's stupidity
[20:45]<rdvxzrwff>suv4x4, don't know that exactly, but i assume that it deserializes the session file
[20:46]<kdrg>Wolfpaws: you are kinder than I
[20:46]<rdvxzrwff>but suv4x4, something to think about, if your making a framework, wouldn't it be better to take away the dependency on sessions?
[20:46]<rdvxzrwff>sounds like your designing towards implementation more than abstraction, frameworks should do the later
[20:48]<wjlllrws>KDan: I am?
[20:48]<lgvggvgzsjy>türkçe bilen arkadaslar varmi
[20:48]<ddr9ww>hey guys i cant find the find "change SQL root passworrd"
[20:48]<ddr9ww>where do i find it ?
[20:48]<suv9w9>cythrawll, I use session in a very limited way, but still I use it. Here's what I use it so far (only two classes out of probably 40): keeping login info, userid, username, user privileges. And the other is to create unique transaction id-s for actions that need to happen once. When the form arrives, I check if the hash exists in the session and unset that hash. If there's no such hash I deny action. How could I replace session in these uses?
[20:48]<wjlllrws>ha?
[20:48]<wjlllrws>!tell LeventersoY about english
[20:48]<mjjfs>Mma411: you need to be root user on the machine
[20:49]<ddr9ww>im in the php admin where do i go now ?
[20:49]<kdrg>Wolfpaws: I'm Chaotic Evil. :-)
[20:49]<mjjfs>Wolfpaws: what's we don't support x script gl?
[20:49]<wjlllrws>!+g10
[20:49]<lxl-2jv>Guideline #10) We don't support script(s). We help you *write* PHP, not recommend or download and install/hack/modify/adapt/use pre-written scripts. Contact the authors of the script for support options they provide.
[20:49]<vznd2d>Mma411: if you mean phpmyadmin, i don't think you can do it from there
[20:49]<wjlllrws>KDan: My paws are itching... So...
[20:50]<rdvxzrwff>suv4x4, I wouldn't replace them, just take them out of the framework and use them in implementation of the framework classes instead
[20:50]<kdrg>Wolfpaws: tried anti-histamines?
[20:50]<ddr9ww>is it the php my admin or do i do it through ftp ?
[20:50]<mjjfs>suv4x4: i use session allll the time
[20:50]<suv9w9>cythrawll, the problem in fact is the transaction id - when the form "realizes" it's a POST form, it appends a hidden field with the transaction ID. This happens in the middle of the file, so the form fails to set the hash in the session.
[20:50]<wjlllrws>KDan: I have different cures for that
[20:50]<mjjfs>Mma411: this is either a question for the phpmyadmin mailing list or #mysql
[20:51]<mjjfs>suv4x4: so don't code spaghetti
[20:51]<kdrg>suv4x4: that does sound awfully like spaghetti
[20:52]<suv9w9>Kdan, dools, how come it sounds like spaghetti to you
[20:52]<mjjfs>suv4x4: because something is happening halfway through html output that affects the session
[20:52]<suv9w9>Kdan, dools, the transaction id / check is done by a class Transaction with two methods: Transaction::getUniqueId and Transation::isUnique($id)
[20:52]<kdrg>^^ what he said
[20:52]<dgvgzdygysv>i'm gonna really get pissed with my hosting service soon.
[20:53]<mjjfs>Determinist: do you have php running locally?
[20:53]<suv9w9>dools, how would you avoid it and achieve this functionality I need
[20:53]<mjjfs>suv4x4: yeah, set a variable in your php template

Back to #php
Or go to some related logs:
#suse
#debian
#gentoo
#php
#blender