I upgraded to Intrepid (Ubuntu 8.10) today. What a long day. It took ages, but was really painless to do the upgrade. WELL DONE UBUNTU TEAM !!!

The only hitches I have had so far:

On a fresh install (another PC), Synaptic wouldn't search properly. I found I had to set the repositories to the Main Server first, reload, then switch back to my local.

On my PC, which does the REMOTE Support, there was a problem now. No go! After much diagnosis, I found that the xtightvncviewer was not responding as expected. It got the connection fine, then nothing. No remote window. Reinstall it. Nope. In the end, after much web-smurfing, I came across tightvncviewer as part of the package of SSVNC. This works great if you use the pre-built binaries, and even lets you work with the UltraVNC file transfer and scaling. Chat didn't work, but no matter.

I tried to compile ssvnc from source, but it kept failing for some reason, complaining about files not there or lack of perms, even though I was running under sudo. Oh well, not worth the hassle. Save the Unix only binaries in an easily accessible place in your path, then alter the startup script for your remote listener to got to /path/to/tightvncviewer -listen.

The latest Ubuntu, Intrepid Ibex, version 8.10, is hitting the streets right now, overloading servers, etc.

Go ahead and get it if you want. No, no links just yet. Don't be so impatient!

I'll stick with Hardy for now. Why? It works, and the update for my system has suddenly become 1.6GB worth of download (got every thing loaded)!! KDE4 + Gnome + + +

Anyhoo... have fun ubunteroos.

In working with Bitweaver, I have come across a nasty side affect of using MySQL- it's handling of encoding is sucky.

I have posted into the Bitweaver Wiki about this, and am sure that it applies to many similar softwares ( I did a quick search and found Moodle and EZ-Publish forum postings with people fighting this, along with numerous other blogging softwares).

In a nutshell, the default MySQL encoding settings are Latin1. Most modern software needs UTF8. This stuffs up the saving of characters,into the database, that are unusual.

UTF8 is huge, literally hundreds of unique characters. If you need to convert the Latin1 characters in your MySQL DB into UTF8, then you will need to understand what is going on and how to fix it.

To understand the character conversion craziness, I recommend you create a new page/blog post/ whatever in your software, leaving it in draft mode, and inserting all the weird characters that your audience is likely to use, in a SINGLE COLUMN.At the same time, do it within a spreadsheet column. Use a character map for this, or if you are unsure, check this page.

Now save the page. And check it in your PHPMyAdmin(I cannot tell you how to do this- it varies hugely). You should have a single column of garbage characters. Congratulations! Copy/paste this into your next spreadsheet column.

Now you have a spreadsheet table detailing the conversions you need to do - first column= character in UTF8 that you want, second= Latin1/UTF8 hybrid mangle that fills your database and needs replacement.

Read the wiki link above, apply those settings to your software if you can. Fix the cause.

Now to clean your database. Use the Export feature of PHPMyAdmin, export with DROP Tables. The resulting SQL file is the code required to rebuild your database.

Examine the following shell script (unix/macos natively; windows users need to use their web host's command line, unless somebody wants to write a BATch file):

#!/bin/bash
echo I\'ll work on "$@" ...
for i in "$@" ; do
echo -n $i...
sed \
-e 's/=latin1/=utf8/g' \
-e 's/é/é/g' -e 's/è/è/g' -e 's/ê/ê/g' -e 's/ë/ë/g' \
-e 's/ / /g' -e 's/«/«/g' -e 's/»/»/g' -e 's/°/°/g' \
-e 's/ä/ä/g' -e 's/â/â/g' -e 's/î/î/g' -e 's/Â/’/g' \
-e 's/ï/ï/g' -e 's/ì/ì/g' -e 's/ò/ò/g' -e 's/ô/ô/g' \
-e 's/ö/ö/g' -e 's/ÿ/ÿ/g' -e 's/ù/ù/g' -e 's/ü/ü/g' \
-e 's/û/û/g' -e 's/ç/ç/g' -e 's/É/É/g' -e 's/È/È/g' \
-e 's/Ê/Ê/g' -e 's/Ë/Ë/g' -e 's/À/À/g' -e 's/Ä/Ä/g' \
-e 's/Â/Â/g' -e 's/Þ/Î/g' -e 's/Ï/Ï/g' -e 's/ÃŒ/Ì/g' \
-e 's/Ã’/Ò/g' -e 's/Ô/Ô/g' -e 's/Ö/Ö/g' -e 's/Ÿ/Ÿ/g' \
-e 's/…/…/g' -e 's/’/’/g' -e 's/àƒâ‚¬/ä/g' -e 's/‘//g' \
-e 's/“/“/g' -e 's/”/”/g' -e 's/‹/‹/g' -e 's/›/›/g' \
-e 's/—/—/g' -e 's/–/—/g' -e 's/â€?/”/g' -e 's/àƒ’/à/g' \
-e 's/â’€’™/’/g' -e 's/â’€’œ/“/g' -e 's/â’€?/”/g' -e 's/àƒ’©/é/g' \
-e 's/â’€’¦//g' -e 's/–/\•/g' -e 's/€/€/g' \
-e 's/Ù/Ù/g' -e 's/Ãœ/Ü/g' -e 's/Û/Û/g' -e 's/Ç/Ç/g' $i|
sed -e 's/Ã/à/g' > $i.clean

echo done
done


This shell script uses SED to open your exported(and thrice backed up) SQL file, fix it, then save it with .clean at the end of the file name. The code is highly customizable, and will need to be, if you have another character set giving you grief. That above is mostly Western European and abbreviated (the Cleanup script below is "Bigger, Better, Stronger!").

Now to use it, save that code as "cleanup.sh", then do chmod +x cleanup.sh, which makes it executable. Then: sh cleanup.sh olddb.sql (where olddb.sql is the saved name of your exported SQL file from PHPMyAdmin).

Check the file in your text editor (needs to be UTF8 aware), and search for weird characters like  and Ã. If they are still there, then there is some more characters to add to your script.

When you are finished, create the new database in PHPMyadmin, with Collation set to UTF8_general_ci. Then use the Import feature to lift your new clean SQL file upstream.

Don't forget to change your program's settings if you've renamed your database.

For PHP/MySQL apps you will want to send the query "SET NAMES UTF8" to the database JUST ONCE, immediately after first connection, so you will need to find the right place in your code. HINT: find mysql_connect.

eg: $dbconnect=@mysql_connect($host, $user ,$password);
$db=mysql_select_db($dbname);
$setnamesutf8 = mysql_query("SET NAMES UTF8");
...etc.

========================
Resources: (requires Unix/Linux/BSD and OpenOffice Calc)

Cleanup script
UTF8 Charset
Replacement Table
Sed for Windows
========================
Thanks to schwarzvogel.de and Tommie-Lie.