Jazzcafe problem

cg2112 · Post by **cg2112** » Fri Apr 14, 2006 4:50 pm

Hi. I'm currently using the jazzcafe theme, which I really like and have only had to do very slight modifications to. Kudos on the theme.

I'm having a weird problem, though (it existed before I made any modifications). I have some funny characters showing up, and I can't see an easy way to get rid of them. Anyone else see this?

My site is here: http://www.whiteboard.net.

You can see the characters at the top of the page under the header and above the sidebar.

Any help would be greatly appreciated.

carl_galloway · Post by **carl_galloway** » Fri Apr 14, 2006 7:42 pm

I believe the problem is the language encoding of the template files. Apparently my andreas00 does the same thing. With JazzCafe, the original design was done in Russian language encoding (UTF-8 I believe), then the serendipity author worked in turkish (ANSI or Turkish UTF-8), and then finally it found its way into the English world.

What we need is for one of the encoding experts to tell us how to strip these extra characters out of the file.

cg2112 · Post by **cg2112** » Fri Apr 14, 2006 8:03 pm

Interesting. I don't see the same characters on other sites (including your themes test site, Carl). Is there a way to just remove them from the templates? I don't see it (or anything that might cause it) anywhere.

carl_galloway wrote:I believe the problem is the language encoding of the template files. Apparently my andreas00 does the same thing. With JazzCafe, the original design was done in Russian language encoding (UTF-8 I believe), then the serendipity author worked in turkish (ANSI or Turkish UTF-8), and then finally it found its way into the English world.

What we need is for one of the encoding experts to tell us how to strip these extra characters out of the file.

Ianardo · Post by **Ianardo** » Sat Apr 15, 2006 12:47 pm

I had exactly the same problem.
It occurred when I edited the index.tpl in GoLive then uploaded it to the site.
I solved the problem by editing in TextWrangler instead.
It is indeed a UTF problem.

carl_galloway · Post by **carl_galloway** » Sun Apr 16, 2006 9:23 pm

are there any encoding experts in these forums, I have a problem with andreas00 where there seems to be an extra character appearing at the beginning of the left sidebar and which means that in IE the left column drops a few pixels. Is there anyway of viewing these sorts of characters in a text editor or some other program so they can be deleted. Sorry if I sound vague but I just don't know anything about encodings.

cg2112 · Post by **cg2112** » Sun Apr 16, 2006 9:44 pm

Hmm.
I'm using vi, I'm not sure if that would cause problems that GoLive might cause. I'll investigate further.

cg2112 · Post by **cg2112** » Sun Apr 16, 2006 9:49 pm

I took a look at the template files in another editor (I used pico) and sure enough, there were the characters I couldn't see in vi. I removed them (they were the first characters in index.tpl, trackbacks.tpl, entries.tpl, sidebar.tp) using pico, and everything is good.

Thanks for the help!

carl_galloway · Post by **carl_galloway** » Mon Apr 17, 2006 7:07 am

Awesome, glad we were able to fix your problem.

Could I ask anyone who uses Linux to take a look at my andreas00 files in pico and let me know if a similar problem exists. I'm stuck using Windows and haven't yet found a text editor that shows me anything different, yet I know there has to be an extra character in one of the files. Thanks everyone.

Post by **yellowled** » Mon Apr 17, 2006 10:50 am

carl_galloway wrote:Could I ask anyone who uses Linux to take a look at my andreas00 files in pico and let me know if a similar problem exists. I'm stuck using Windows and haven't yet found a text editor that shows me anything different, yet I know there has to be an extra character in one of the files. Thanks everyone.

I just tried. Doesn't work on my linux machine (Debian Sarge). Not in pico, not in emacs, not in vim, ... no obvious evidence of any character which wouldn't belong in there. Strange.

@cg2112: Which language environment do you use on your linux machine?

judebert · Post by **judebert** » Mon Apr 17, 2006 3:27 pm

I'd like to help, Carl; I've got Linux and Windows environments with multiple text editors (although Vim is my expertise). Unfortunately, I can't figure out where to download Andreas00.

carl_galloway · Post by **carl_galloway** » Mon Apr 17, 2006 7:05 pm

Jude, sorry, its here

However, yellowled has had a look at the files using linux text editors and tells me he couldn't find enything. In firefox/opera everything works as it should for me, but in IE I get

Code: Select all

<div id="LeftSideBar">ï»¿

The three strange characters at the end are the problem I think. Yellowled also tells me he ran the files through a filter known as 'recode' and it couldn't complete index.tpl due to being erroneous.

Thanks, Carl

Post by **yellowled** » Mon Apr 17, 2006 7:18 pm

carl_galloway wrote:Yellowled also tells me he ran the files through a filter known as 'recode' and it couldn't complete index.tpl due to being erroneous.

Erm, sorry ... I have to admit that I actually misinterpreted recode's output - it actually said my use of recode itself was erroneous, not the file

However, the (HTML) source code looks just fine in Firefox and Opera under Linux too, but the sidebars look asymmetric in both browsers.

Anybody with a Linux system should really try to reproduce this - maybe there's something wrong on my machine which makes it impossible to reproduce the character in pico ... you never know.

carl_galloway · Post by **carl_galloway** » Mon Apr 17, 2006 7:26 pm

damn, that gave me a glimmer of hope - ah well, maybe I'll have to print out the source code and retype the whole thing if I can't figure out where the problem lies - makes you realise we're still a long way from truly ubiquitous data sharing though.

Post by **garvinhicking** » Mon Apr 17, 2006 7:37 pm

Those first three characters are called "UTF-8 BOM" (Byte Order Message). They indicate a file is in UTF-8 format for usual files.

In web-environments they make trouble, and those characters need to be removed.

I just removed the characters (editors usually hide them) and committed updated versions to CVS.

Regards,
Garvin

carl_galloway · Post by **carl_galloway** » Mon Apr 17, 2006 7:41 pm

Garvin, that's great, but how can I find them myself? I'm still struggling with CVS and encodings because of problems like these, what software do you use to identify the characters. Every text editor I use looks fine.