[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: KDE is now broken (Fwd: Heads-up: KDE4 hitting testing tonight (UTC) )



In <[🔎] 20090527205951.GP5158@cat.rubenette.is-a-geek.com>, lee wrote:
>On Wed, May 27, 2009 at 12:10:54PM -0500, Boyd Stephen Smith Jr. wrote:
>> >It's
>> >something hidden from the user instead of telling them about
>> >it. That's never a good idea.
>>
>> Actually, according to actual HCI studies it is often better to hide
>> things from a user instead of telling them about it.
>
>What means "better" in that case?

They are able to complete a series of tasks in less time or with less 
prompting/guidance.

>> Now, once the user starts
>> looking for that setting, it should be available, but too many settings
>> and too much explanatory text confuses rather than enlightens.
>
>Whatever you do, there's nothing that helps against stupidity.

And being confused by an over-abundance of options is not a sign of 
stupidity.  While the number of facts one can track shows some 
correspondence to financial success, academic success, and the one-size-
fits-all monstrosity that is IQ, it is much lower than the number of options 
that will fit on a 800x600 panel, even for the very 
successful/"intelligent".

>> >> >> These same issues can be hidden when using RDBMS backed, but the
>> >> >> translations are usually much faster.
>> >If you have to make so many "translations" of a configuration file
>> >that nowadays' computers run into performance problems when doing so,
>> >I don't consider the file as a human readable configuration file
>> >anymore.
>> I never said the translations were causing performance issues.  I
>> said they'd be faster with an SQL backend.  That is definitely *not*
>> the reason Akonadi wants an SQL backend.
>Then it's irrelevant that translations can be "usually much faster"
>when an RDBMS is used.

Okay, but it is not irrelevant that the translations can still be hidden 
from the user so there's no reason to worry about minor migration issues 
with it anymore that with plain-text files.

>> They are quite readable.  Usually a translation is just changing the
>> name/spelling of a key.  But, it might be converting a value that is a
>> list into multiple stanzas or something like that.  Generally, they
>> leave the values alone, but they are the migration path from 'old'
>> configuration files to 'new' configuration files.
>
>So these translations need to be done only once?

Yes.  They are preformed on the file and then a value is updated with the 
name of the translation to prevent it from being run against the file again 
(and for dependency tracking).

>> >> >Try to read
>> >> >your current kde configuration in 35 years, or try to read your data
>> >> >from the the RDBMS you're currently using in 35 years. You'll find
>> >> >that it won't be easy.
>> >> I hope so.  I plan on using different, hopefully better, software by
>> >> then.
>> >And what if you need the information stored in it?
>> I won't.  I'll export the data as I abandon that software.  Actually,
>> I'll export the data before I abandon the software so I can import it
>> into the new software and test it.
>That is a possibility, but it requires to plan for it and to work on
>it, and mistakes can be made. It would be easier if that wasn't
>required.

Change happens.  Trying to prevent it is futile.  Instead, focus on shaping 
the future into something you like better than the present.
-- 
Boyd Stephen Smith Jr.           	 ,= ,-_-. =.
bss@iguanasuicide.net            	((_/)o o(\_))
ICQ: 514984 YM/AIM: DaTwinkDaddy 	 `-'(. .)`-'
http://iguanasuicide.net/        	     \_/

Attachment: signature.asc
Description: This is a digitally signed message part.


Reply to: