Natural sorting is not straight forward on computer system. When sorting the numbers 1 to 10, too many computer programs will start with 1 10 2 instead of the natural 1 2 3. As natural sorting is not supported by the database, I sought different options for sorting catalog codes which may be anything like P-11b, XPK-987-33, Km#12.11 and whichever combination you may consider. So what is currently done is that all numbers are padded with spaces to some set length and so the database knows how to sort them properly. Don't worry, it's not the original information just a copy of it used for quick searching and sorting.
Another issue now settled is that of searching by 'Name'. This was my mistake to begin with as 'Name' on coin countries list, for example, would search a country name and on the main coins page would search a coin name. Now proper pages would have both "Coin name" and "Country name" search so prevent confusion.
A few duplicate catalogs and duplicate series have been merged using a newly duplicates processing script which would make it easy for editors to report such duplicates and process them.
On the multiple-language front an important change has been made. A user that has never visited Colnect will now have his default language according to his country (conferred by the IP address). That is, of course, if it's one of the 35 languages supported by Colnect. Once a user sets a different language, it's remembered and all subsequent URLs would contain the language letters. So even if a German-speaking collector send a /de/ URL to his Spanish-speaking friend, it'll be converted to /es/ for the Spanish-speaker.
And finally, a new improved welcome page is supposed to make more sense than the previous one. Yes, usability and design have to be re-thought and re-made on Colnect but there's nothing quite as bad as a horrible first impression, so here we go...
