Arantor

  • As powerful as possible, as complex as necessary.
  • Posts: 14,278
Fixed SMF bugs
« on February 29th, 2012, 02:16 PM »
As you may notice, there are quite a few(!) topics in this board now that represent SMF bugs that are suspected or known to exist in Wedge.

It's sort of a to-do list for me, and if no discussion results, I'll just merge the notice (with a note to confirm when it was fixed) into this thread to keep it tidy :)
SMF bug 4882 (eAccelerator support should be removed)
« Reply #1, on February 29th, 2012, 03:13 PM »
Well, it should. I seem to recall looking at this before, but it should definitely be removed now there is absolutely no use for it, without the key/value cache, it has no benefits in the caching subsystem.
Posted: February 29th, 2012, 01:37 PM

Fixed in Wedge r1407.
SMF bug 4959 (missing free_result calls)
« Reply #2, on February 29th, 2012, 03:14 PM »
While I think this is unnecessary fussing about a tiny amount, it doesn't hurt to put in such calls (and as my comment from ages ago in the related thread implies, it may have benefits that aren't immediately obvious)
Posted: February 29th, 2012, 01:48 PM

Fixed in Wedge r1407.
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

Dismal Shadow

  • Madman in a Box
  • Me: Who is Arantor? Cleverbot: It stands for time and relative dimensions in space.
  • Posts: 1,185
Re: Fixed SMF bugs
« Reply #3, on February 29th, 2012, 03:19 PM »
I was like WTF!? :lol:
“I will stand on my ground as an atheist until your god shows up...If my irreligious bothers you much, and if you think everything I do is heresy to your god I don't care. Heresy is for those who believe, I don't. So, it isn't heresy at all!


   Jack in, Wedge,
   EXECUTE!

Arantor

  • As powerful as possible, as complex as necessary.
  • Posts: 14,278
SMF bug 4794 (error warning in wrong place)
« Reply #4, on February 29th, 2012, 04:26 PM »
For consistency, warning boxes/success notifications should all be in a consistent place and consistently styled - above principle content in the default area (but under menu tabs/tab data) - which the stated places do not adhere to (search personal messages, search generall judging by this post)
Posted: February 29th, 2012, 02:33 PM

Fixed in Wedge r1408.

MultiformeIngegno

  • Posts: 1,337

Nao

  • Dadman with a boy
  • Posts: 16,078
Re: Fixed SMF bugs
« Reply #6, on February 29th, 2012, 06:33 PM »
Me too!

And no time to deal with it today (I'm having guests) and possibly tomorrow... UNTIL WE HAVE TO GO LIVE >_<

Very bad timing for guests, eh?

Arantor

  • As powerful as possible, as complex as necessary.
  • Posts: 14,278
Re: Fixed SMF bugs
« Reply #7, on February 29th, 2012, 07:46 PM »
Eh I wouldn't worry, nothing here is *that serious* because if it was, I'd be hard at work fixing it, as opposed to sitting on my butt. It's simply that there's a lot of bugs that slipped through the net and this was the quickest way to record them so I didn't forget them.
Re: Fixed SMF bugs
« Reply #8, on February 29th, 2012, 11:07 PM »
And, bah, I should have posted this thread *first* because now half the threads would be messed up if I merged them here. Stupid, very stupid >_<
SMF bug 4838 (Wrong HTTP-Header created)
« Reply #9, on February 29th, 2012, 11:11 PM »
Ref http://www.simplemachines.org/community/index.php?topic=411507.0

Regarding the use of reason phrases being translated, I have to say that the reporter is incorrect. Certainly as per HTTP 1.1, section 6.1.1 of the specification:
Quote
The individual values of the numeric status codes defined for HTTP/1.1, and an example set of corresponding Reason-Phrase's, are presented below. The reason phrases listed here are only recommendations -- they MAY be replaced by local equivalents without affecting the protocol.
The example quoted in the thread is 404 Not Found - if the language in use were Pig Latin (for sake of argument), 404 Otnay Oundfay would be entirely appropriate to send because the browser doesn't care what the text is, only the number.

Regarding of SERVER_PROTOCOL, that's a much trickier question, but HTTP/1.0 is a valid response even to an HTTP/1.1 request given the situation.

Ultimately, I don't think anything needs to be done, but I'm documenting the state of play anyway.
Posted: February 29th, 2012, 01:04 PM

Even in HTTP 1.0, as per 6.1.1 of that spec:
Quote
The individual values of the numeric status codes defined for HTTP/1.0, and an example set of corresponding Reason-Phrase's, are presented below. The reason phrases listed here are only recommended -- they may be replaced by local equivalents without affecting the protocol.
So yeah, no need to actually do anything.

Nao

  • Dadman with a boy
  • Posts: 16,078
Re: Fixed SMF bugs
« Reply #10, on February 29th, 2012, 11:24 PM »
Just edit the first post in the new topic ;)

Arantor

  • As powerful as possible, as complex as necessary.
  • Posts: 14,278
Re: Fixed SMF bugs
« Reply #11, on February 29th, 2012, 11:24 PM »
I suppose I could do that, yes. Surprisingly enough that didn't actually occur to me >_<
SMF bug 4865 (headers incorrect in manage-permissions simple mode)
« Reply #12, on March 1st, 2012, 01:27 AM »
While it's possible this will be pulled, in the meantime until the fate of permissions is decided, it probably should be fixed.
Posted: February 29th, 2012, 02:55 PM

OK, I know why this is happening in SMF and why I apparently couldn't reproduce it myself. It's hidden permissions.

The first 'gap' where it skips a windowbg, should be approve_posts permission - but it's disabled by default. And the row is inverted regardless of whether anything was displayed or whether an input type="hidden" was inserted in its place. Just need to quietly move the flipping of state.
Posted: March 1st, 2012, 12:31 AM

Fixed in Wedge r1410.
Because the form isn't validating properly, only relying on the session data being in $_POST to identify that it's supposed to receive content, it gets confused if the form immediately before happens to be the validate-password triggers from admin security (which also sends the session data in $_POST the same way)

Suggested solution is good (to name the submit and check for that instead)
Posted: February 29th, 2012, 02:46 PM

Fixed in Wedge r1411.