[ANN] Introducing Waves - Web App Framework

can anyone who is having this problem see if it is fixed now? i
updated the theme, but i since i'm not seeing this problem, i am not
sure if i fixed it or not? thanks again to everyone who helped me
resolve this. -dan

Your HTML states:

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd&quot;&gt;&lt;html&gt;
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type"/>
<title>Ruby Waves: Home</title>
<link type="text/css" rel="stylesheet" href="/css/site.css"/>

But according to
  http://validator.w3.org/check?verbose=1&uri=http%3A%2F%2Fwww.rubywaves.com%2F
you should avoid /> in HTML 4.01 strict.

HTH
s.

···

On Thu, 7 Feb 2008 15:11:09 -0500 Dan Yoder <dan@zeraweb.com> wrote:

On Feb 7, 2008, at 10:03 AM, Jesús Gabriel y Galán wrote:

> On Feb 7, 2008 6:49 PM, Dan Yoder <dan@zeraweb.com> wrote:
>> Are you using FF3? I've gotten a couple of reports on that ... not
>> sure what the problem is. :frowning:
>
> I see the same weird layout as Trans, I'm using FF 2.0.0.11 in Ubuntu.
>
> Jesus.
>

just out of curiosity, how do you test your framework?

···

On 6-Feb-08, at 3:46 PM, Dan Yoder wrote:

The walk-through looked promising, although I haven't had a chance to try it. Have you attempted Haml integration? Haml and Sass are such incredible productivity enhancers (honest, without them I feel like I have one hand tied behind my back). It would be great to know whether they are easy, sorta-easy, or really hard to integrate.

really easy. the only reason i didn't go with haml out of the gate is because markaby accomplishes a lot of the same things but is much more flexible since it is actually ruby. of course, if you want client-editable templates, than that becomes a drawback, one which i plan to address using sandbox. that said, it really isn't difficult to add haml support if it turns out to be a must-have. if you're interested at giving it a shot yourself, take a look at the Waves::Renderers stuff in the API docs.

dan

But according to
  http://validator.w3.org/check?verbose=1&uri=http%3A%2F%2Fwww.rubywaves.com%2F
you should avoid /> in HTML 4.01 strict.

are there any browsers that actually interpret /> per the spec? i've always just ignored that.

it appears the underlying problem is related to the way some versions of firefox calculate the offset dimensions of block elements. the way it is laying out the page suggests that it doesn't think it has enough room to put the main content and the sidebar side-by-side. although ie6, opera, safari, and mac / windows ff2 don't agree.

anyway. i will take this off-line since it has nothing to do with ruby.

thanks again for everyone's help and suggestions.

dan

just out of curiosity, how do you test your framework?

for functional testing, i have several apps running on it, and i test those. i would like to formalize that into actual test apps as contributors come on (knock on wood).

for performance testing, right now i am using ab, although performance will become more of a focus in future releases.

btw, thin support is coming in the next release. i've already had a couple of requests for it. :slight_smile:

···

On 6-Feb-08, at 3:46 PM, Dan Yoder wrote:

The walk-through looked promising, although I haven't had a chance to try it. Have you attempted Haml integration? Haml and Sass are such incredible productivity enhancers (honest, without them I feel like I have one hand tied behind my back). It would be great to know whether they are easy, sorta-easy, or really hard to integrate.

really easy. the only reason i didn't go with haml out of the gate is because markaby accomplishes a lot of the same things but is much more flexible since it is actually ruby. of course, if you want client-editable templates, than that becomes a drawback, one which i plan to address using sandbox. that said, it really isn't difficult to add haml support if it turns out to be a must-have. if you're interested at giving it a shot yourself, take a look at the Waves::Renderers stuff in the API docs.

dan

one suggestion... use tables and forget about it. all the drivel about
the superiority of divs mean squat in the face of practical realities.

T.

···

On Feb 8, 2:20 pm, Dan Yoder <d...@zeraweb.com> wrote:

> But according to
> Validation Results - W3C Markup Validator
> 2Fwww.rubywaves.com%2F
> you should avoid /> in HTML 4.01 strict.

are there any browsers that actually interpret /> per the spec? i've
always just ignored that.

it appears the underlying problem is related to the way some versions
of firefox calculate the offset dimensions of block elements. the way
it is laying out the page suggests that it doesn't think it has
enough room to put the main content and the sidebar side-by-side.
although ie6, opera, safari, and mac / windows ff2 don't agree.

anyway. i will take this off-line since it has nothing to do with ruby.

thanks again for everyone's help and suggestions.

Well, I'd like to add to that. The argument against tables is often an irrational one, but not completely.
The idea, is that using semantic divs for layouts is more flexible and more accessible (for screen readers and such)
Tables are more brittle and make the code itself harder to read. Tables on the other hand, are certainly easier to do layouts with, but there are plenty of raw recipes out there using CSS to create div based layouts.
Tables are appropriate when the CDATA (character data) makes sense as a table.
All of that is just suggestions anyway. Ultimately you can do whatever you like that works.
You could use spans and make them display:block; in CSS.
You can avoid CSS altogether and just use HTML containing all of the presentational markup, but it's tougher to maintain.

The biggest problem with tables for layout is that people are often trying to approximate print document layouts that are fixed and controlled, but the reality with html is that you must be flexible and accept the fact that different user-agents (browsers) will render things differently and that users can resize text in most browsers and that they all have different possible screen sizes/window sizes.
The main philosophy of div based layouts is that you should give up trying to exercise complete control of the presentation of the document in html.

Tables will do hideous things when designed for one platform in pixel-based sizes...

In the end though, it all comes down to doing what suits you and your target audience.

The argument is similar to frameset problems. You can avoid framesets and just use the overflow CSS item to make scrollable sections.

···

On Feb 8, 2008, at 9:14 PM, Trans wrote:

On Feb 8, 2:20 pm, Dan Yoder <d...@zeraweb.com> wrote:

But according to
Validation Results - W3C Markup Validator
2Fwww.rubywaves.com%2F
you should avoid /> in HTML 4.01 strict.

are there any browsers that actually interpret /> per the spec? i've
always just ignored that.

it appears the underlying problem is related to the way some versions
of firefox calculate the offset dimensions of block elements. the way
it is laying out the page suggests that it doesn't think it has
enough room to put the main content and the sidebar side-by-side.
although ie6, opera, safari, and mac / windows ff2 don't agree.

anyway. i will take this off-line since it has nothing to do with ruby.

thanks again for everyone's help and suggestions.

one suggestion... use tables and forget about it. all the drivel about
the superiority of divs mean squat in the face of practical realities.

T.

>>> But according to
>>> Validation Results - W3C Markup Validator
>>> 2Fwww.rubywaves.com%2F
>>> you should avoid /> in HTML 4.01 strict.

>> are there any browsers that actually interpret /> per the spec? i've
>> always just ignored that.

>> it appears the underlying problem is related to the way some versions
>> of firefox calculate the offset dimensions of block elements. the way
>> it is laying out the page suggests that it doesn't think it has
>> enough room to put the main content and the sidebar side-by-side.
>> although ie6, opera, safari, and mac / windows ff2 don't agree.

>> anyway. i will take this off-line since it has nothing to do with
>> ruby.

>> thanks again for everyone's help and suggestions.

> one suggestion... use tables and forget about it. all the drivel about
> the superiority of divs mean squat in the face of practical realities.

> T.

Well, I'd like to add to that. The argument against tables is often
an irrational one, but not completely.
The idea, is that using semantic divs for layouts is more flexible
and more accessible (for screen readers and such)
Tables are more brittle and make the code itself harder to read.

No way are they are more brittle --it's divs that are more brittle.
I've seen many a site screwed-up because of messed up divs. To me divs
are beneficial in some of the ways you mention, but they were designed
rather poorly I think. It boggles my mind that it effectively pushes
us back to using absolute sizes all over the place --and then we end
up with things like Blueprint --one monstrosity on top of another. Do
developers like torture or what?

Tables on the other hand, are certainly easier to do layouts with,
but there are plenty of raw recipes out there using CSS to create div
based layouts.
Tables are appropriate when the CDATA (character data) makes sense as
a table.
All of that is just suggestions anyway. Ultimately you can do
whatever you like that works.
You could use spans and make them display:block; in CSS.

You have to measure ROI. You can spend a whole day getting divs
working the way you want. Just spend 10 minutes making it a table. And
it can be worse. Case in point, Waves' site has been broken for a week
now. Just use tables and get it fixed --you can always come back and
redo it with divs when you have the time.

You can avoid CSS altogether and just use HTML containing all of the
presentational markup, but it's tougher to maintain.

Well that's silly. Declarative stylization is hugely beneficial
regardless of whether you use divs or tables for layout.

The biggest problem with tables for layout is that people are often
trying to approximate print document layouts that are fixed and
controlled, but the reality with html is that you must be flexible
and accept the fact that different user-agents (browsers) will render
things differently and that users can resize text in most browsers
and that they all have different possible screen sizes/window sizes.
The main philosophy of div based layouts is that you should give up
trying to exercise complete control of the presentation of the
document in html.

Tables will do hideous things when designed for one platform in pixel-
based sizes...

Why would you do that? You can size your table in % not px.

T.

···

On Feb 10, 11:30 am, John Joyce <dangerwillrobinsondan...@gmail.com> wrote:

On Feb 8, 2008, at 9:14 PM, Trans wrote:
> On Feb 8, 2:20 pm, Dan Yoder <d...@zeraweb.com> wrote:

Tables are more brittle and make the code itself harder to read.

This is incorrect if anyone asks me.
I am generating tables programmatically from ruby, and on the layout
question, it is simply a lot easier to have a table arranged nicely with
columns rather than messing with multiple divs. You can use CSS in both
cases anyway.

The "harder to read" argument is really a no argument as far as I see
it, and I will even go as far as to suggest to compare a specific table
with 8 rows and 18 columns with the equivalent div solution.
I would be curious to see the solution so I can compare but until I have
seen it I will claim, based on my own experience, that the table
solution is the better one.

···

--
Posted via http://www.ruby-forum.com/\.

From a designer's perspective, tables are not accessible since they
don't retain the semantic structure of the document. I don't really
know a whole lot about accessibility and what have you, but I would
imagine that a screen reader treats tabular structured markup
differently than layout structured markup (and would, presumably,
treat a page laid out with tables incorrectly).

--Jeremy

···

On Feb 10, 2008 6:33 PM, Marc Heiler <shevegen@linuxmail.org> wrote:

> Tables are more brittle and make the code itself harder to read.

This is incorrect if anyone asks me.
I am generating tables programmatically from ruby, and on the layout
question, it is simply a lot easier to have a table arranged nicely with
columns rather than messing with multiple divs. You can use CSS in both
cases anyway.

The "harder to read" argument is really a no argument as far as I see
it, and I will even go as far as to suggest to compare a specific table
with 8 rows and 18 columns with the equivalent div solution.
I would be curious to see the solution so I can compare but until I have
seen it I will claim, based on my own experience, that the table
solution is the better one.
--
Posted via http://www.ruby-forum.com/\.

--
http://www.jeremymcanally.com/

My books:
Ruby in Practice

My free Ruby e-book

My blogs:

http://www.rubyinpractice.com/

Jeremy McAnally wrote:

From a designer's perspective, tables are not accessible since they

don't retain the semantic structure of the document. I don't really
know a whole lot about accessibility and what have you, but I would
imagine that a screen reader treats tabular structured markup
differently than layout structured markup (and would, presumably,
treat a page laid out with tables incorrectly).

Theoretically tables are less accessible, but the fact is that a large portion of the web is made of table layouts, so screen readers have adapted and learned how to make sense of table layouts. So in practice there is no accessibility problem. Of course it might become problematic if you have 25 levels of nested tables, but as in everything moderation is key.

IMHO making a column layout with a table is a small HTML hack; making a column layout with divs is a monstrous CSS hack. I think that, as a community, web/graphical designers have not yet realized that monstrous hacks are a bad thing rather than a sign of technical prowess.

···

--
Daniel

I'm sorry, but I keep looking at this thread expecting to hear more
about waves and have forgotten it was hijacked. I believe this
discussion has very little to do with Ruby and very little to do with
Waves. Would you folks (I don't mean to target Daniel specifically,
here) who want to argue this point (which is often argued and never
settled all over the internet) at least please change the subject
line? I would consider it a personal favor. Thank you.

Ben

···

On Thu, Feb 14, 2008 at 4:53 PM, Daniel DeLorme <dan-ml@dan42.com> wrote:

... continued XHTML/CSS debate

AAgreed. :let's hear more about Waves.

···

On Thu, Feb 14, 2008 at 5:44 PM, Day <iamday@gmail.com> wrote:

On Thu, Feb 14, 2008 at 4:53 PM, Daniel DeLorme <dan-ml@dan42.com> wrote:
> ... continued XHTML/CSS debate

I'm sorry, but I keep looking at this thread expecting to hear more
about waves and have forgotten it was hijacked. I

--
Collins Richey
     If you fill your heart with regrets of yesterday and the worries
     of tomorrow, you have no today to be thankful for.