CSS sledgehammer... #4

...bugs & extreme cases...

digging deeper…

I'm not hunting for bugs, but I hate bad browsers...
In order to figure out which browsers are in support of what on our homepages, the following is a normal set of questions from this web carpenter.

  • How much (x)html and CSS can each browser handle?
    - This is essential.
  • What's right and what's wrong?
    - Doesn't really matter, but it's good to know.
  • How to get around browser-bugs and weaknesses?
    - There's always a way if we look for it...
  • How much "overload" can each browser handle?
    - No need to go easy on them. I want some progress...

Note: it's all about how well the browsers supports my code — not the other way around. There are some gray areas in the standards I'm using, and the validator isn't flawless either. Besides, I use standards as tools — not as a set of defined limitations.

how are browsers handling this..?

Exploring browser-differences and looking for some predictability across browser-land. I'm using "minimal" test pages for this, in order to avoid code-confusion.

Some of my test-pages are valid[valid], some are non-valid[not valid], and some are not determined and/or finished yetvalidity unclear.

  1. nested rules and parse errors validity unclear
    - separating browsers.
  2. recovery from parse errors [not valid]
    - separating browsers.
  3. floats: removed and layered [valid]
    - Opera 8 can't handle this.
  4. floats: partly removed, nested and layered [valid]
    - Opera, Safari and IE can't handle this.
  5. floats: layering — sort of [valid]
    - included fix for Opera 8 weakness.
  6. faulty layering of covered links [valid]
    - IE6 and IE/Mac fails.
  7. calculations: off screen extremes, using absolute positioning [valid]
    - pretty wide limits, but not endless.
  8. nested rules and parse errors #2 [not valid]
    - separating browsers.
  9. nested rules and parse errors #3 [not valid]
    - separating browsers.
  10. corrupted rules and parse error recovery [not valid]
    - targeting browsers.
  11. corrupted rules and parse error recovery #2 [not valid]
    - targeting browsers orderly.
  12. corrupted rules and parse error recovery #3 [not valid]
    - targeting browsers orderly.
  13. overflow: hidden and negative margins on floats [valid]
    - Opera 8.0 handles this badly.
  14. Image Replacement nonsense - UGH [valid]
    - a joke, but it's working...
  15. Loony line-up of floats [valid]
    - survives in some browsers...
  16. exposing float-margins and float-dimensions [valid]
    - Gecko lost without dimension?
  17. centering of right-aligned float [valid]
    - with overflow to the left.
  18. basic image-control #1 [valid]
    - controlling overflow in a fluid-width environment, using floats and positioning.
  19. basic image-control #2 [valid]
    - controlling overflow in a fluid-width environment, using text-align.
  20. hacking @import [not valid]
    - testing valid and non-valid hacks to see if the relevant stylesheet is applied.
  21. stacking multiple nested floats [valid]
    - the basics for simple drop-shadows and similar effects.
  22. stacking multiple nested & positioned elements [valid]
    - the basics for simple drop-shadows and similar effects.
  23. stacking multiple nested in-flow elements [valid]
    - the basics for simple drop-shadows and similar effects.
  24. horizontal ruler styled vertically [valid]
    - in the absense of a vertical ruler. No good in IE.
  25. fluid “decoration” [valid]
    - positioning outside an element.
  26. date-stamp [valid]
    - Repositioning an article-date, or any element for that matter.
  27. counting stylesheet links [valid]
    and for fun: counting stylesheet links to the extreme [valid]
    - IE/win can only handle a limited number of links.
  28. counting @imports [valid]
    and for fun: counting @imports to the extreme [valid]
    - IE/win can only handle a limited number of @imports.
  29. inline and block styling and handling [valid]
    - basic testing across browser-land.
  30. horizontal ruler as separator [valid]
    - overstyling IE/win's defaults.
  31. branching extends the number of @imports IE/win can handle [valid]
  32. reducing link repetitions [valid]

be observant…

This is a warning to less experienced web designers, who happens to pass by in their search for answers and solutions.

The fact that something works in a certain way in a certain combination or in complete isolation, only hints at how something may work if we add or change some of the conditions and/or combinations and place it in different surroundings.

Everything is “conditional” and quite often “illogical” as long as we're dealing with software, specs, software bugs and designer bugs. Perfect answers can not be found by creating simplified test cases, but they sure are of great help when we're in the process of clearing things up a bit.

will be continued…

Test-pages will be added as I discover new problems related to my own web carpentry, and find time to explore them.

I'm also revisiting test-cases with irregular intervals, and update them with new finds as new browser-versions arrive.

Note: if you're looking for some ordinary bugs, solutions and workarounds, follow my example and look at my off site resources (in the right column on this page). I've shortened the list to the most relevant and useful pages/sites as of my latest revision.

sincerely  georg; sign

Hageland 19.mar.2005
last rev/addition: 04.oct.2008

CSS sledgehammer...


…not even W3C can handle this much junk…

Well, that's too bad…
…better test the validator.

Section

  • introduction
  • Table of Content

About

the usual
  • the author
  • Copyright
the unusual
  • Molly speaks up
the additional
  • Examples
  • Demo pages

Off site resources:


about…
…2005 - 2008