Arantor

  • As powerful as possible, as complex as necessary.
  • Posts: 14,278
Re: SMF 2.0 final THIS MONTH?
« Reply #120, on June 1st, 2011, 12:42 PM »
That's precisely my point. I could forgive this delay if there were any real evidence of any changes. The tracker doesn't agree, and the only line that the team are giving is that there are bugs being fixed.

I sync'd Wedge's SVN to SMF's on 21st May, that was rev 10542. They're now on rev 10560. I see at least two revs not related to 2.0 in those 18, I see a few credits updates, but I'm not really seeing anything that's setting my world on fire with bugfixes - I see more template bug fixes than other bug fixes in fact, hardly anything that would - in my mind - justify their delay, when the tone of some of the team members who are toeing the party line is that there was something big enough.

See, here's where I have a problem. On the one hand, they're not going to fix a number of minor bugs, but they're going to fix other minor bugs, that are equally aesthetic and minor as the ones they've actively said they're not going to fix this time around. On the other hand, they're saying about how there's something that required them to delay it to get it fixed, but no-one's saying what, and it's not like SVN indicates anything, so unless there's a security hole or major show-stopping bug that isn't documented anywhere outside the team area, I can't call it as anything other than 'shit, we called a deadline and now we failed on that again'.
When we unite against a common enemy that attacks our ethos, it nurtures group solidarity. Trolls are sensational, yes, but we keep everyone honest. | Game Memorial

Lex

  • Posts: 31
Re: SMF 2.0 final THIS MONTH?
« Reply #121, on June 1st, 2011, 12:48 PM »
I now have an idea of what this is about, after Sinan posted on the dev blog - if I'm right, it is understandable that they do not want to call it Final without fixing it to work properly...

Arantor

  • As powerful as possible, as complex as necessary.
  • Posts: 14,278
Re: SMF 2.0 final THIS MONTH?
« Reply #122, on June 1st, 2011, 02:04 PM »
Yeah, it's the OpenID bug, but as I just pointed out... that ain't new, nor even a surprise - it's been broken since at least RC3 (with the same reason)

TE

  • Posts: 286
Re: SMF 2.0 final THIS MONTH?
« Reply #123, on June 1st, 2011, 03:56 PM »
Quote from Arantor on June 1st, 2011, 02:04 PM
Yeah, it's the OpenID bug, but as I just pointed out... that ain't new, nor even a surprise - it's been broken since at least RC3 (with the same reason)
This. The OpenID bug is known since RC3... IMHO a "communication problem" within the team, seems nothing has changed since I left for good reasons.
From a tactical view it was a big mistake. I would have released it as promised on the end of may, then fix all remaining bugs and release a 2.0.1 short time afterwards. At least the promised release date could have been observed.

Now the release is delayed AND there will still be some bugs.
Thorsten "TE" Eurich - Former SMF Developer & Converters Guru

Arantor

  • As powerful as possible, as complex as necessary.
  • Posts: 14,278
Re: SMF 2.0 final THIS MONTH?
« Reply #124, on June 1st, 2011, 04:06 PM »
You know, it looks like they're enjoying playing at being all grown up and organised, but having no idea how to actually go about it.

billy2

  • Trying to earn brownie points for a lads trip to the Red Sea. Minus 1 already - just for asking!!
  • Posts: 350
Re: SMF 2.0 final THIS MONTH?
« Reply #125, on June 1st, 2011, 04:34 PM »
Quote from Arantor on June 1st, 2011, 04:06 PM
You know, it looks like they're enjoying playing at being all grown up and organised, but having no idea how to actually go about it.
Piss-up and brewery come to mind  :niark:
<br /><br />cough, cough.

TE

  • Posts: 286
Re: SMF 2.0 final THIS MONTH?
« Reply #126, on June 1st, 2011, 05:24 PM »
I'll bet there are lots of other bugs.. just found another one..
how to reproduce: add this to index.php
Code: [Select]
class smfException extends Exception
{
public static function error_handler_callback($code, $string, $file, $line, $context)
{
$e = new self($string, $code);
$e->line = $line;
$e->file = $file;

throw $e;
}
}
replace SMFs internal error handler with the above object oriented version:
Code: [Select]
set_error_handler('error_handler');
replace with:
Code: [Select]
set_error_handler(array("smfException", "error_handler_callback"), E_ALL);

then login to the forum and type a wrong password, the result is:
Code: [Select]
Fatal error: Uncaught exception 'smfException' with message 'Undefined offset: 1' in C:\Program Files\Apache Software Foundation\Apache2.2\htdocs\svn\Sources\LogInOut.php:701 Stack trace: #0 C:\Program Files\Apache Software Foundation\Apache2.2\htdocs\svn\Sources\LogInOut.php(701): smfException::error_handler_callback(8, 'Undefined offse...', 'C:\Program File...', 701, Array) #1 C:\Program Files\Apache Software Foundation\Apache2.2\htdocs\svn\Sources\LogInOut.php(282): validatePasswordFlood('5', '') #2 [internal function]: Login2() #3 C:\Program Files\Apache Software Foundation\Apache2.2\htdocs\svn\index.php(163): call_user_func('Login2') #4 {main} thrown in C:\Program Files\Apache Software Foundation\Apache2.2\htdocs\svn\Sources\LogInOut.php on line 701

wizzle

  • I'm not allowed to have em. But I do sneak one in once in awhile
  • Posts: 6
Re: SMF 2.0 final THIS MONTH?
« Reply #127, on June 1st, 2011, 05:41 PM »
Quote from Arantor on June 1st, 2011, 02:04 PM
Yeah, it's the OpenID bug, but as I just pointed out... that ain't new, nor even a surprise - it's been broken since at least RC3 (with the same reason)
Well... I beg to differ!  <insert sarcastic smiley>
Quote
I considered it an "unpleasant surprise" because it was found out just days before the planned release.
Which also explains the slow timeline. 3 days SMF time = 7 months real time.

Dragooon

  • I can code! Really!
  • polygon.com has to be one of the best sites I've seen recently.
  • Posts: 1,841
Re: SMF 2.0 final THIS MONTH?
« Reply #128, on June 1st, 2011, 05:45 PM »
That's normal? You're throwing an exception from an error handler, there is no exception handler and the error handler is treated as a normal function. So...I see no reason for it to be a bug.
The way it's meant to be

TE

  • Posts: 286
Re: SMF 2.0 final THIS MONTH?
« Reply #129, on June 1st, 2011, 06:10 PM »
the exception is thrown because of this line:
Code: [Select]
$_SESSION['failed_login'] = @$_SESSION['failed_login'] + 1;
$_SESSION['failed_login'] isn't set at this time, thus proper (cleaner) code should be:
Code: [Select]
$_SESSION['failed_login'] = isset($_SESSION['failed_login']) ? ($_SESSION['failed_login'] + 1) : 1;

Dragooon

  • I can code! Really!
  • polygon.com has to be one of the best sites I've seen recently.
  • Posts: 1,841
Re: SMF 2.0 final THIS MONTH?
« Reply #130, on June 1st, 2011, 06:24 PM »
Ah you were catching the error and intentionally throwing the exception, I thought you were bugging about the exception being uncaught.

Nao

  • Dadman with a boy
  • Posts: 16,082
Re: SMF 2.0 final THIS MONTH?
« Reply #131, on June 1st, 2011, 07:36 PM »
Quote from TE on June 1st, 2011, 06:10 PM
the exception is thrown because of this line:
I have no idea what this does, but rest assured that I've applied your fix to the two occurrences in Login2.php ;)

TE

  • Posts: 286
Re: SMF 2.0 final THIS MONTH?
« Reply #132, on June 1st, 2011, 08:24 PM »
np, it's not very complicated. php has it's own build-in exception class for catching errors, which is IMO very useful, especially when using object oriented code. The class smfException extends this internal class with a simple function (error_handler_callback). This function creates an object every time an error is reached and throws the related error to the screen. Instead of using SMFs internal error handler I just replaced it with the new function.

PHPs internal exception class is somewhat more nitpicking than the normal error handler, thus the script aborted.

Example: the normal error_handler would allow this, simply because of the @ before $test. (int) would force an undefined variable (without the @ you'd get an undefined index error) to become 0. the exception class doesn't allow such "dirty" coding.
Code: [Select]
<?php
$test 
= (int) @$test;
echo 
$test;
?>

Xarcell

  • Posts: 30
Re: SMF 2.0 final THIS MONTH?
« Reply #133, on June 2nd, 2011, 12:04 AM »
GOD, how many times is SMF going they say the release is near and it never is?

Arantor

  • As powerful as possible, as complex as necessary.
  • Posts: 14,278
Re: SMF 2.0 final THIS MONTH?
« Reply #134, on June 2nd, 2011, 12:09 AM »
Quote from Xarcell on June 2nd, 2011, 12:04 AM
GOD, how many times is SMF going they say the release is near and it never is?
And you wonder why they get so much ridicule. Point of interest, it was said to certain people there last summer that if Nao and I were left alone to get on with the dev work, 2.0 would have been finished by the end of 2010. (Which, barring external circumstances, it would have been. The only reason I'm not full pelt on it is SimpleDesk, as the team blog over there will testify with results of the last 3 months' work on it)