If you insist on using that abominable google groups for posting to
usenet, learn how to use it and start following the accepted posting
conventions of this group!
<URL:http://www.safalra.com/special/googlegroupsreply/>

If you won't do this much, it shows your inability to learn or adapt,
hence you'll just end up being ignored by those people who are most
likely to give you sound advice.

expertware@libero.it wrote:

> As already said, tables were first choice. But
> then we realized they didn't allow for all the flexibility
> and precision of formatting we need. That why the control is at cell
> level.
> Try to replicate exacly (I mean at pixel level)
> http://cam70.sta.uniroma1.it/Techni...atatimedemo.htm
> or
> http://cam70.sta.uniroma1.it/Techni...ransparency.htm
> with a table and you will understand what i mean.

One thing you apparently don't realize is that there is no need for
pixel precision here at all. So what if the overlaying graphic covers 3
pixels of the word "Argentina" or not? Nobody cares.

You apparently also haven't considered using multiple tables. That first
page should have 2 tables, one 6-column table for all the data on the
left side of the page and one 2-column table for the data on the right.
Position those 2 tables if you must, but not the individual cells. The
second page should have one 5-column table for each country. No
positioning required at all, except for the overlaying graphic(s).

As is, you have just created an accessibility problem with all that
positioning, as well as a layout that does not adapt well to different
browsing environments and has a huge amount of bloat, which negatively
affects download time. This is what happens when people don't understand
the technology they are trying to use and just hack away rather than
take the time to actually learn it. You just create a problem where none
should exist. :(

--
Reply email address is a bottomless spam bucket.
Please reply to the group so everyone can share.

expertware@libero.it schreef:
> Rincewind ha scritto:
>
> 
>
>
> sorry I don't get the meaning of all your expressions.
> What you mean?
>
> I also didn't get this one from Niek:
>
> 
>
>
> what is "legit" and "geek" ?

It means something like:
"Quello è un esempio 100% del legit di geek-comunica!"
translated by http://babelfish.altavista.com/

A geek is someone who says he/she spends his.her life with a compiler?
Legit means something like 'legittimo'in Italian.


>
> You guys talking too difficult!!

Spice up your English, this is a good way to do that, and learning the corre
ct way to build website alongside
(using tables for tabular dat for instance).


BTW I posted an answer (well.. not quite an entire answer) to one of your qu
estions in the other, more recent,
thread.


--
Niek

expertware@libero.it schreef: 
>
>
> it's just missing to translate those 2 words.
> Perhaps it is considered slang if the dictionary does not have the
> words
>
> 
>
> you mean out of real world? Sound like gig robot (cartoon)

Oh boy, this is going to be difficult...
What does Google say:
Geek è un termine di origine anglosassone di etimo oscuro (forse storpiatura
 del termine inglese - in italiano
geco - come antonomasia per la particolare abilità dell'animale di incollars
i agli oggetti).Indica una persona
solitaria che è affascinata dalla tecnologia e dalla fantasia. Il significat
o di geek non coincide con quello
di nerd.
it.wikipedia.org/wiki/Geek

~snip~
 
>
>
> Well there are good reason for me to maintais this approach.
> While I understand your reasons (you are kind of artists programming by
> hand)
> I have other reason which are not easy to explain to nonprogrammers.
> All these reason can be summarized in one word: code maintainability.

I understand, lets just say you have multiple blocks which are based on one 
'source-block'.

>
> However... if one of you is capable to do the following make a page
> which
>  - Obtain exacly the same results as in datatimedemo with tables only
>  - Has less bytes in the file
> I shall be happy to reconsider recoding the export routine

You can export one 'source-block' to something like this: <td style="some st
yle-elements here">Argentian</td>
If you know which block starts in a row of five blocks, you can put a <tr> i
n front of it an if you kwnw where
which block is the last one of the row, you can put a </tr> after it.
Same goes with the start and end of the table.

~snip~

> I wanto do add quotes: so
> height=200px should be changed in height="200px" or height="200"px or
> height="200" or what?
> last time I remove px the whole table disappeared.

It must be: height="200"

>
> pamela the gig r steel heart ... (I compensate with other virtues)

I hope you do ;-)

>
> PS
> A question I noticed that my code almost passed the 4.01 except
> the validator was complaining with the "WIDTH" tag in tables.
> Why the validator complains and how do I replace WIDTH.

Basicly putting width in a tble-cell or div using html is depricated, it is 
better to use CSS to define
style-elements (and so you seperate the design from the content).



--
Niek

expertware@libero.it schreef:

> note that cells in a row can have different styoe with different
> borders, cellpadding, alignments and so on... I don't see it as easy to
> do
> or maintainable

that will be no problem it is completely legite to do something like:
<td style="width:95px;height:47px;border: 2px solid #ff9900; vertical-align:
top">Argentian</td>
It just makes your file big.
Note that you should only use inline-style (such as the example above) when 
the style in that <td>-tag must be
different from the default-style.

 
>
>
> Ok will try. If report disappears I come to beat you.

I'm not so into the beating-game to be honoust...
 
>
> Yes but here the width is necessary for the td to align correctly. It
> cannot align based on
> the container DIV block

That's why you have to use the DTD mentioned earlier, it tells the browser i
t is "Transitional".


--
Niek

'sNiek wrote:
>
> it is completely legite to do something like:
> <td style="width:95px;height:47px;border: 2px solid #ff9900; vertical-align:top">A
rgentian</td>

Much better would be to assign a class selector and put all the style
rules in an external stylesheet, using class names that identify the
content rather than the styling. For example
<td class="country">Argentina</td>

> It just makes your file big.

Indeed.

--
Reply email address is a bottomless spam bucket.
Please reply to the group so everyone can share.

expertware@libero.it wrote:
>
> Well there are good reason for me to maintais this approach.
> I have other reason which are not easy to explain to nonprogrammers.

Don't be condescending. Read through a database, do a little data
manipulation and print out the results. It isn't anything new.

> All these reason can be summarized in one word: code maintainability.

Considering that you don't seem to understand either proper HTML or CSS,
I think you only imagine that using all those absolutely positioned divs
is somehow easier to maintain. It isn't. More like you just haven't had
enough experience to know a better, more efficient way to do it.

> However... if one of you is capable to do the following make a page
> which
>  - Obtain exacly the same results as in datatimedemo with tables only

Why does it have to be either all tables or none?
- Use proper multi-column, multi-row tables for the tabular data.
- Use divs (or whatever is most appropriate) for the other elements.
- Position (or float) the tables and divs as you see fit.

Contrary to popular belief, a pixel-perfect layout is not required.

>  - Has less bytes in the file
> I shall be happy to reconsider recoding the export routine

It's not that hard to do, you just seem to have some preconceived
notions that are getting in the way.

> Another one. Why that stupid validator complains about transparent
> background
> color. What is the change to remove that annoying warning?

Remove that annoying non-standard CSS. ;)
"behavior" is proprietary to IE and is not recognized by the validator.

--
Reply email address is a bottomless spam bucket.
Please reply to the group so everyone can share.

kchayka schreef:
> 'sNiek wrote:
> 
>
>
> Much better would be to assign a class selector and put all the style
> rules in an external stylesheet, using class names that identify the
> content rather than the styling. For example
> <td class="country">Argentina</td>

I told the OP that, but she wants to be able to chenge the style of any
cell on the page, so I mentioned that she dhould do it the way you said
and then overrule the style if necessary by using inline style.

> 
>
>
> Indeed.
>

very big!

--
Niek

expertware@libero.it schreef:
> I think one should first prove something and then speak. Otherwise
> anyone can say anything and everyone is right.

We allready told you to use tables for tabular data, your project is a
schoolaxample for tabular data. Proof enough I would say.

I do hope you take this advice at heart: study it, implement it and come
back with a new demo.
We certainly are curious: this is about the longest thred I've seen lately.


Ciao

--
Niek

expertware@libero.it wrote:
>
> My believe is that is not possible to achieve exactly the same result.
> Just think about the 3d effect. On the same row we can have inset and
> ouset. Also border and padding top bottom can be all of different sized
> for cells in a given row. But I may be wrong.

This is where you are going wrong, methinks. It sounds like classic
"form over function". That sample page you put up looks just awful from
a usability standpoint. I can't imagine why you would want to do such a
thing in the first place. The whim of some marketing guy?

> There are 3 criteria I see so far:
>
> 1- Code must attain exact rendering of layout report

HTML was never meant to acheive exactness. You already discovered that
CSS might do it, but in your case, it looks like a case of abuse.

> 2- code must validate according current specs

Validation only checks for syntax errors. Your sample page was loaded
with logic errors, insofar as using correct markup for the specific
content. As a programmer, you should understand the difference.
Validating according to specs should also include conforming to some
level of published accessibility guidelines. Your sample page was a
total disaster in that regard.

> You show WITH CODE you can fith the above criteria, and I will
> certainly
> implement gratefully your solution. All the rest is just bla bla...

A couple people have already shown you the path to better code. You
don't seem interested in using correct markup.

> So far what I can see is that, while I have rendered a complex layout,

Seems to me it is unnecessarily complex, which is where much of the
trouble lies. If acheiving a particular pixel-perfect layout is so
bloody important, forget HTML altogether. Try PDF.

--
Reply email address is a bottomless spam bucket.
Please reply to the group so everyone can share.

expertware@libero.it wrote:

> kchayka ha scritto:
> 
>
> Then do it and, then, you can talk with reason ...

Let go of this crazy notion that you must maintain some pixel-perfect
layout. If you can't do that, then there is little hope.

--
Reply email address is a bottomless spam bucket.
Please reply to the group so everyone can share.