[dev] Re: [cvs] commit: horde login.php imp login.php redirect.php gollem login.php redirect.php horde/templates common-header.inc horde/templates/login login.inc framework/Browser Browser.php imp/templates common-header.inc imp/templates/login login.inc ...

[Vilius Šumskas]

Quoting Michael M Slusarz <slusarz at mail.curecanti.org>:

> Quoting Chuck Hagenbuch <chuck at horde.org>:
>
>> Quoting Michael M Slusarz <slusarz at curecanti.org>:
>>
>>>   Log:
>>>   Insert code in the login screen to auto-detect the exact version 
>>> of IE (version
>>>   5.0 +).  This allows us to only turn off page compression for the exact
>>>   versions of IE that are affected.  I am still not sure if I am 
>>> embarassed at
>>>   how hackish this is or excited by how hackish this is :)
>>
>> Is there a critical reason for this to be added while we're in a
>> feature freeze?
>
> Because, at last report, 91% of the world uses IE.  and a large portion
> of those
> IE users are using IE 5.5+.  Since the bug only affects 6.0 without at least
> SP1, and I am (hoping) that many people have upgraded IE to at least 6.0 SP1
> (or 5.5 SP1), that we are unnecessarily turning of page compression to
> a *huge*
> amount of users (75% or more I would guess) when it is not necessary.
>
>
> I think catching this before we release Horde 3 makes a large amount of sense
> (since a savings of 50% or more bandwith per page adds up pretty quickly when
> talking about thousands of page accesses).

Michael is apsolutely right. I just love my horde flying high with
alphaImagaLoader.htc and even higher with compression fix. My bandwith 
today is
more thank twice times lower than before.

>  And I personally don't think this
> is an added feature; i looked at it as a bug fix for an overbroad browser
> check.  Of course, comments/criticism on my argument are welcome.

-- 
   Best Regards,

   Vilius