Re: Xaraya Aruba 1.2 roadmap open for comments

> 
> MySQL's default charset settings are one of the most ludicrous things 
> I've ever seen.
> 
> ADOdb has a SetCharSet method in both the mysqli, oci8 and postgre7 
> drivers.
> 
> While we're at it, we're running 4.93, should upgrade to 4.990.
Btw according to http://phplens.com/adodb/change.log.html it's version 4.990 from 11 Jul 2008, but on http://sourceforge.net/project/showfiles.php?group_id=42718&package_id=34890 is a version 4.991 as a last version supporting php 4 (http://sourceforge.net/project/shownotes.php?release_id=636415&group_id=42718)

> 
>> I'd skip collation settings as it seems to be more difficult 
>> (according to Dracos adodb does NOT support collation).
>> This would be at least for mysql and postgresql. I have no idea how is 
>> encoding used in sqlite.
> 
> SET NAMES takes care of several things, details in this thread in the 
> adoDB forums: http://phplens.com/lens/lensforum/msgs.php?id=12467&x=1
> 
> Collation has to do (mostly, I think) with how MySQL orders records 
> during sort/order by operations involving varchar and text fields.  If 
> the charset settings are in sync, I think collation takes care of itself 
> for the most part (which may be why adoDB doesn't support setting it).
> 
> Speaking of the installer, I've been thinking about expanded db-specific 
> settings for a while.  I doubt it would be hard to do.
Great :-).

> 
> I'm also considering creating an installer theme so that Xaraya_Classic 
> doesn't need to be kept around all the time.  One page template, one 
> stylesheet, and images necessary to dress up the output.  I already have 
> a hack in Xaraya_Classic/pages/installer.xt that can detect if the 
> current script is install.php or upgrade.php.
I'd agree to dismiss dependency of Xaraya_Classic during upgrade (and install).

> 
>>
>>
>>
>> 2) only ASCII characters in Short-URLs
>> In talks about ASCII characters in Short-URLs in articles I've heard 
>> that is has to be defined generally for xaraya.
>>
>> So, it would be good to have only ASCII characters in Short-URLs.
>> E.g. In Czech "žlutě" would be "zlute", but "zlute" would be also 
>> "zlute".
>> So is that possible? I mean decoding from short url wouldn't be one 
>> way :-( ("zlute" could be "zlute" or "žlutě".) If not, it should be 
>> somehow achieved avoiding NON-ASCII characters in Short-URLs.
>>
>> There is a bug about this issue:
>> http://bugs.xaraya.com/show_bug.cgi?id=3138
>>
> 
> The url encoding process would need to differentiate the "zlute" and 
> "žlutě" urls somehow when they are both converted to "zlute": "zlute" 
> and "zlute2" or something.
I have no idea how :-(. Is there any possible way?
I case we wouldn't be able to differentiate them I'd be for adjusting just articles as I know about this use just in particular case of articles with short urls enabled. In this case "žlutě" and "zlute" would be considered as a same titles (there is a setting howto handle the same titles).
There could be on way map for languages, e.g.
cs_CZ.UTF-8:
'ž' => 'z'
'š' => 's'
'ě' => 'e'
...

> 
>>
>>
>> 3) test more with PostgreSQL (several problems with MySQL put me off 
>> from ideas to switch from MySQL to my favourite PostgreSQL).
>>
>>
> 
> Few of us know/use Postgres, so whatever testing it gets can probably go 
> a long way.
> 
>> Pev