Re: metalink still unuseable the 2nd day ...

> On Nov 11, 4:05 pm, Noons <wizofo..._at_gmail.com> wrote:
>> On Nov 12, 1:38 am, MBPP <mpacheco_bra..._at_hotmail.com> wrote:
>>
>>
>>
>> > IE also worked fine (version 8). I also noted they call the SWF
>> > directly without a "wrapper" to check Flash version, etc. My Flash
>> > version here is 10.0.32, try to upgrade if you have version 9, maybe
>> > it can help? I read somewhere that they demand 9.0.115 or greater I
>> > think.
>>
>> Oh great!  So now, besides having to figure out all the db versions
>> and patches to make things work, we also are supposed to track the
>> "consistency" of browser version and flash version?
>>
>> Welcome to the brave new world of dba2.0: spend all your time trying
>> to get the "easy, time-saving UI" to work!
>>
>> I'm on 10.0.32 Flash and latest FF and still having minor nagging
>> problems with it.  It'll go away, I'm sure.  But did it really need to
>> be like this?
>
> Of course not, I too think the new interface could be simpler and
> faster. Disclaimer, I don't work for Oracle but I work with Oracle
> databases as a DBA and developer for +14 years. Note, the idea behind
> RIA interfaces is to make the applications more interactive and faster
> to navigate since you don't (re)load pages for every navigation and
> you don't need to mimic visual effects using tons of CSS/Javascript.
> The visual "appealing" in my opinion is a secondary consideration that
> must be touched only after the application meets all its functional
> requirements. For the past 2 years I am working with Adobe Flex
> (Flash) creating RIA applications for customers that use Oracle as
> database. Our architecture on the Oracle side is based on PL/SQL
> stored procedures that dynamically generate XML via EPG or MOD_PLSQL.
> This is in fact the same method that the old Metalink was using but
> generating dynamic HTML. But now it seems they are using a Java middle
> tier so most likely they recoded big portions of it. In my opinion if
> they would have continued to use MOD_PLSQL they would just need to
> adapt the output to XML (via a parameter) and things would be much
> simpler and faster. They would be able to run both versions in
> parallel without the need to migrate any data. Once users are happy
> and it is stable you turn the old off. On the interface side they
> could just keep more or less the same general interface as the old
> Metalink in version 1 and enhance it visually in the next versions.
> Apparently they did too much in one shot and the results are here for
> everyone to see. This is something I saw many times in the past, the
> development/management team is induced with "hype" technologies and
> make things much more complex than they should be.

Yep, over engineering seems to be one of the most common issues I run into. Agree 100% that the first objective should be good usable functionality - add/enhance the visual asthetics once you have nailed the functionality. People will put up with an ugly but usable site far more than a pretty unusable one.

Of course, getting the true necessary functional requirements in the first place can often be a challenge. There is also the 'first impressions' problem that can be a challenge, especially with some managers.

Tim

-- 
tcross (at) rapttech dot com dot au