PLplot

Hi Alan,

> My recent commit messages (up to revision 8947) pretty much describe what I
> was doing for my liquid-layout style.  I have used the liquid layout style
> as the default with the single static style used before accessible as an
> alternative.  To switch between the two styles use the View ==> Page Style
> menu for firefox or the View ==> Use Stylesheet menu for konqueror. I

Would be more convenient to have a button for that, I'll have a look at 
a website which does this (http://www.miscdebris.net/blog).

> successfully tested a method to get the background image to scale for 
> liquid
> layout.  However, since that image is very similar to a solid colour, I
> replaced it with a solid colour for simplicity. I kept the logo at fixed
> width, left-justified on a black background. (this is similar to a
> suggestion Rafael made.) I think the resulting website looks good for a
> variety of resolutions with width ranging from ~100 px to 1024px (the most
> available with my old monitor).  This is consistent with advice I have read
> to optimize for a width of 1024px, but use liquid layout to address 
> concerns
> of users with smaller and larger widths.

Looks very nice! I found a problem though, between the main body and the 
right sidebar is a vertical dotted line, which moves not correctly if 
you resize the browser window. The line is hard to see, so watch 
closely. The solution of the problem would be not to resize the sidebar 
as well, so make it fixed size, or position the line also relative (if 
that's possible).
> 
> I have uploaded everything to our SF website. Please check it out with all
> the browsers accessible to you and let me know what you think.
> 
> I, of course, tested that the result HTML validates at
> http://validator.w3.org/, but I don't think that validates either our
> static or liquid styles.  Do you know how to validate CSS?

http://jigsaw.w3.org/css-validator/

1 error and over 150 warnings. It used to be 1 warning (many warnings 
are about the same issue though), so there should not be much to do to 
fix this.
> 
> Alan

Good work,
Werner

-- 
Dr. Werner Smekal
Institut fuer Allgemeine Physik
Technische Universitaet Wien
Wiedner Hauptstr 8-10
A-1040 Wien
Austria

email: smekal@...
web:   http://www.iap.tuwien.ac.at/~smekal
phone: +43-(0)1-58801-13463 (office)
        +43-(0)1-58801-13469 (laboratory)
fax:   +43-(0)1-58801-13499

On 2008-10-21 23:23+0200 Werner Smekal wrote:

> Looks very nice!

I am glad you like it.

I got rid of the CSS error with the illegal "float: center;" (see validation
below). I also found some additional style problems which I have now fixed
(as of revision 8951).

> I found a problem though, between the main body and the 
> right sidebar is a vertical dotted line, which moves not correctly if you 
> resize the browser window. The line is hard to see, so watch closely. The 
> solution of the problem would be not to resize the sidebar as well, so make 
> it fixed size, or position the line also relative (if that's possible).

At first I was going to say I couldn't see it at all, but the virtually
empty http://plplot.sourceforge.net/development.php does help make the line
more visible at all widths (although it still is pretty hard to see).  Is
that what you were referring to?  To me it looks like a slightly differently
coloured rectangle with a 1px edge (which makes the line that we see) with
the same fixed width as rightside for the static case (160px). However, I
have so far not found anything near that fixed width for the liquid style
sheet or the existing (fixed width) images.  I will look further at this,
but this may be a case of two sets of eyes being better than one so I would
appreciate it if you looked for the source of this as well.  N.B. look for
further e-mail tomorrow (Wednesday) when your read this in case I have
found the source of this line by then.

If you diff style_static.css and style_liquid.css you will see the
straightforward changes I did.  There is certainly no large number of pixels
left in style_liquid.css except for the fixed height of the logo, and fixed
height of the codebox.

One thing I did to try and find this strange fixed interloper in my liquid
style was to adopt "obvious" background-color values such as #FF0000 for
various bits of the style so I could colour code each bit of style and find
out how that affected our website.  However, that approach showed that every
component of rightside scaled nicely so I am a bit stuck finding the
remaining fixed component.

>> I, of course, tested that the result HTML validates at
>> http://validator.w3.org/, but I don't think that validates either our
>> static or liquid styles.  Do you know how to validate CSS?
>
> http://jigsaw.w3.org/css-validator/
>
> 1 error and over 150 warnings. It used to be 1 warning (many warnings are 
> about the same issue though), so there should not be much to do to fix this.

Thanks for that reference. It looks like both style files have the same
errors and same warnings (as is expected since I introduced few differences
between the two files other than some size options).  I have already gotten
rid of the same error for both (see above).

I will take responsibility for getting rid of the warnings as well for both
style_static.css and style_liquid.css.  All those warnings seem pretty
innocuous, but there may be something buried in there that will give a clue
to the fixed vertical line interloper that I am trying to chase down.

Alan
__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

On 2008-10-21 18:00-0700 Alan W. Irwin wrote:

> N.B. look for
> further e-mail tomorrow (Wednesday) when your read this in case I have
> found the source of this (vertical fuzzy) line (which doesn't scale) by then.

Hi Werner:

I found the source of the problem (an unnecessary image used to make a white
background rather than the white color that was there already).  From my
reading and recent experience for some simple test cases, there is no way to
scale a background image in CSS 2 so the imposed width was being ignored for
the image.  However, I don't quite understand where the line came from and
its bad scaling properties. Anyhow, it disappears (both for the static and
liquid style) when you drop the URL to the white background image (revision
8952).  I have just now uploaded my latest work to our website as well so
please look carefully at our website with all browsers accessible to you for
any remaining issues.  I request other developers here to do the same.

The only issue left that I am aware of at the moment is the remaining CSS
validator warnings for both the static and liquid cases.  I plan to deal
with all those warnings tomorrow (Wednesday).

Alan
__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

On 2008-10-21 19:37-0700 Alan W. Irwin wrote:

> The only issue left that I am aware of at the moment is the remaining CSS
> validator warnings for both the static and liquid cases.  I plan to deal
> with all those warnings tomorrow (Wednesday).

Hi Werner:

I looked over the current validator warnings, and they all appear to be
about coincidences in the colour theme between some background colours and
others.  So this is a deficiency in the particular designed colour scheme
you adopted (or adapted) from the web.

I really would like to get rid of these color warnings if at all possible
since they will tend to hide more important CSS validation warnings if/when
those exist. Also, to my mind one drawback of the present colour scheme is
there is no way to distinguish between the various (visited, etc.) kinds of
links.  So if we could find a pre-designed colour scheme that has similar
fundamental colours to the present colour scheme but which has more
varieties of colours in it to get rid of the warnings and also distinguish
the various kinds of links, that would be great. I hope you have some advice
about finding such a colour scheme since I don't think we would do very well
designing our own.

BTW, I could not find our present colour scheme at the DesignsByDarren.com
site.  Did you grab it from another site?

Alan

__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

Hi Alan,

> Hi Werner:
> 
> I looked over the current validator warnings, and they all appear to be
> about coincidences in the colour theme between some background colours and
> others.  So this is a deficiency in the particular designed colour scheme
> you adopted (or adapted) from the web.
> 
> I really would like to get rid of these color warnings if at all possible
> since they will tend to hide more important CSS validation warnings if/when
> those exist. Also, to my mind one drawback of the present colour scheme is
> there is no way to distinguish between the various (visited, etc.) kinds of
> links.  So if we could find a pre-designed colour scheme that has similar
> fundamental colours to the present colour scheme but which has more
> varieties of colours in it to get rid of the warnings and also distinguish
> the various kinds of links, that would be great. I hope you have some advice
> about finding such a colour scheme since I don't think we would do very well
> designing our own.

The easiest would be to change the colors only a bit, let's say e.g. 
#menubar ul background color is the same as A hover/link/visited/... 
(#ff6600). So let's set the background color for #menubar ul to #ff6601 
and they are not the same anymore, but the users eye won't see any 
difference and we don't have to look for new colors which fit well. 
Warnings are also gone (presumably).

We could also change the color for visited links, e.g. to a darker 
orange or something like that if you wish.

> 
> BTW, I could not find our present colour scheme at the DesignsByDarren.com
> site.  Did you grab it from another site?

It's actually from http://www.oswd.org and the theme is called 
fruitopia: http://www.oswd.org/design/preview/id/3181 or 
http://www.oswd.org/files/designs/3181/Fruitopia/ .

Regards,
Werner

> 
> Alan
> 
> __________________________
> Alan W. Irwin
> 
> Astronomical research affiliation with Department of Physics and Astronomy,
> University of Victoria (astrowww.phys.uvic.ca).
> 
> Programming affiliations with the FreeEOS equation-of-state implementation
> for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
> package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
> Linux Links project (loll.sf.net); and the Linux Brochure Project
> (lbproject.sf.net).
> __________________________
> 
> Linux-powered Science
> __________________________
> 
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Plplot-devel mailing list
> Plplot-devel@...
> https://lists.sourceforge.net/lists/listinfo/plplot-devel


-- 
Dr. Werner Smekal
Institut fuer Allgemeine Physik
Technische Universitaet Wien
Wiedner Hauptstr 8-10
A-1040 Wien
Austria

email: smekal@...
web:   http://www.iap.tuwien.ac.at/~smekal
phone: +43-(0)1-58801-13463 (office)
        +43-(0)1-58801-13469 (laboratory)
fax:   +43-(0)1-58801-13499

On 2008-10-23 20:57+0200 Werner Smekal wrote:

> The easiest would be to change the colors only a bit, let's say e.g. #menubar 
> ul background color is the same as A hover/link/visited/... (#ff6600). So 
> let's set the background color for #menubar ul to #ff6601 and they are not 
> the same anymore, but the users eye won't see any difference and we don't 
> have to look for new colors which fit well. Warnings are also gone 
> (presumably).
>
> We could also change the color for visited links, e.g. to a darker orange or 
> something like that if you wish.
>
>> 
>> BTW, I could not find our present colour scheme at the DesignsByDarren.com
>> site.  Did you grab it from another site?
>
> It's actually from http://www.oswd.org and the theme is called fruitopia: 
> http://www.oswd.org/design/preview/id/3181 or 
> http://www.oswd.org/files/designs/3181/Fruitopia/ .

Hi Werner:

Thanks for those comments/ideas.

Here is the current status:

I have now changed over to configuring style_liquid.css and style_static.css
from one common file, style.css.in.  So far, I have only done the
dimensional configuration, and not the colour one, but the idea works well.
Configuring the colours should work well also; relatively few CMake
variables are needed to describe our website colours since they are done
with a consistent colour scheme that repeats colour data often for various
CSS contexts. Also, I will use separate variables when the consistency is
too much (as in the background colour warnings we are getting).  When I am
done with this it should be possible to easily adjust our colour scheme by
changing a relatively small number of CMake variables.

I am also in the process of replacing the simple www/Makefile and
www/announce/Makefile with a CMake-based approach because the latter handles
access to the build tree (where all the configured files now reside) simply
and automatically.

The current status of www/announce is the CMake-based build system now
validates the DocBook source for our historical release announcements and
also builds html and text versions that agree with what the old
www/announce/Makefile produced.  I am working on putting together CMake
custom rules for uploading our base website similar to what www/Makefile
does now.  That upload contains all our historical formal release
announcements now, and I am assuming in the future it will include future
formal release announcements as well (see <aside> below).

<aside> I must say, it should be incredibly easy to do formal announcements
of future releases the way they have been done historically.  Currently our
website only points to detailed release notes (the README.release file that
Hazen uploads to SF as part of the release).  Those notes certainly have
their place, but I also think there is a place for a short formal release
announcements as well on our website like we used to do until the process
stopped when Tom Duck took over from Rafael as the release manager. I have
put some notes into www/announce/README about what would need to happen if
we decide to make a formal release announcement for the next release.
Anyhow, Hazen this is something to think about.</aside>

I am keeping scripts/generate_website.sh up to date as I change things
around.  Thus, once I am completely finished with my changes, the release
manager's task of updating the website will remain the same which is simply
to run the script.

In sum, there are still some fairly trivial bits of updating our website
that still need changes and testing, but by the end of this weekend I hope
to be done with everything affecting our website update procedure.

More later.

Alan
__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

On 2008-10-25 00:30-0700 Alan W. Irwin wrote:

> In sum, there are still some fairly trivial bits of updating our website
> that still need changes and testing, but by the end of this weekend I hope
> to be done with everything affecting our website update procedure.

The website upload procedure has been substantially modified to use a new
target www-install-base configured by our CMake-based build system rather
than the old method of cding to www and running make there using a
hand-crafted Makefile. My tests (which including generating and uploading
the whole website following the directions in
README.Release_Manager_Cookbook) indicate all is well with the changes.

Tomorrow (Sunday) I plan to finish off my current round of website work by
implementing the CMake configuration of our colour style which I discussed
previously.

Alan

__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

On 2008-10-25 21:48-0700 Alan W. Irwin wrote:

> Tomorrow (Sunday) I plan to finish off my current round of website work by
> implementing the CMake configuration of our colour style which I discussed
> previously.

That didn't pan out.  The colour scheme in the menus is essentially frozen
and non-configurable because it depends mostly on the colour of background
images which would potentially clash with any attempt to configure other
colours.

I also tried the idea of changing the colour of visited links, but I didn't
like the patchwork look of the results.

So I confined myself to slight tweaking of the background colours in the
menu to stop the validation warning messages about those colours being
potentially equal to foreground colours.  I also improved the footer by
respacing text within it and inserting the appropriate validation icons
(which currently show perfect validation for both the xhtml and css).

Thus, as of revison 8977 I think I am done with the website for a while.

Alan
__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________