Re: [WSG] Reduce Page Weight
The point of putting metatags on pages is mostly for search engines.
What if I create a conditional statement on my dynamically generated
pages that print these meta tags (especially ketword and description
ones) only if it is a bot.
Maybe just forget about them altogether? Use good titles, headings and
urls.
Google doesn't care about meta at all. Inktomi is the last one that takes
them
into account and it's authors say that meta has only marginal influence on
indexing.
Is there an easy way to recognize a bot?
There is a list of robot's User-Agent strings floating around, but that's
few thousands different strings...
For less important things I check for presence of "http://"; in User-Agent,
but is incorrect in many cases! ("ia_archiver", and browsers with
addons/proxies that advertise themselves).
Apart from targetting specific User-Agent strings you might create
bot-trap in /robots.txt - use Apache mod_rewrite to handle that file with
PHP for example.
Google tips mention that bot detects if different page is sent to the bot
and to the browsers and may threat it as a cheat.
--
pozdrawiam, Kornel Lesiński
**
The discussion list for http://webstandardsgroup.org/
See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list & getting help
**
Re: [WSG] question - follow, index meta tag
If I have the following on my index page, do I need to repeat it on every page at my site? Doesn't this tag appearing once send the robots forward to all the other pages? You don't need that at all. Index, follow is default behaviour and won't override robots.txt anyway. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] LOCAL XHTML TESTING - HOW TO DO IT?
For serious testing install web server locally. There is for example easyphp.org (and many others) that install and configure Apache, PHP, MySQL with almost one click. Then you can configure Apache to send .xhtml files as application/xhtml+xml or use PHP for that. httpd.conf: AddType application/xhtml+xml .xhtml or each file: Once you have proper content-type set browsers (except IE ofcourse) will do basic validation of document - requiring that your document is well-formed XML (XHTML). -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Changing Standard
Why using a:link ?
means that the word inside is a link
a { color:blue; text-decoration:underline; }
is the same as setting
a:link { color:blue; text-decoration:underline; }
Link is a redundant tag
No, it isn't. Think about these:
foo
bar
a {} matches both, and :link matches only the second one.
Additionally :link matches only unvisited, inavctive, nonfocused links,
but a{} sets all at once.
:link will become very important in XHTML2 where
every element can have href attribute.
--
regards, Kornel Lesiński
**
The discussion list for http://webstandardsgroup.org/
See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list & getting help
**
[WSG] iframe 100%
Do you know a good way to get iframe inside messy table-based layout to stretch to 100% available height? In IE it is as simple as , but Opera is very picky about heights :( I'm trying to make hotfix for someone else's site, so mess, tables and iframe has to stay :/ I know two methods using javascript: - compute and set fixed height - find *all* parent (parents of parents, etc.) elements of iframe and give them height: 100% I rememver I've read somewhere that Opera sticks to odd CSS rule: element can have height in % only if parent element has height set, but I can't find in in W3C TR. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Netscape 4 - let it die
Don't blame the end user for the lousy browser their employer sticks them with. Educate the employers instead. Sure, if it works - don't change it. By supporting NN4 you just keep it alive and give message to bureaucrats that it is OK browser. So make it stop working. I think that employee saying "Boss! This doesn't work!" is the best way to educate employers... -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
[WSG] Netscape 4 - let it die
Why do you let 8-year-old browser to stop you from making good pages? I bet that 90% of Netscape 4 users are bored webmasters ;) Whenever some good solution is mentioned hearing "but Netscape 4 doesn't support this" is unavoidable. I agree that webpages should be accessible to all - they should work without CSS and JavaScript. Personally I use @import for CSS and use object-detection to gracefully degrade pages. NN4 should be threated as a text browser. It is just too buggy to get anything better. These days web looks so bad in NN4 that one more page looking ugly in this dinosaur doesn't matter. Actually it is even better - it proves that finally user needs to upgrade. c'mon! 6 years isn't a short notice! -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] alt or title...
alt attribute is for some elements (images, buttons), that might not be fully rendered by browsers. should be visible only when element itself is not shown. title attribute is for almost every element in html. it is supposed to give additional information. optionally visible, usually as a tooltip or on status bar. WWW -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Firefox screen-reader emulator
Do any of these, or any others for that matter, support aural style sheets? from that list, emacSpeak only, unless things have changed recently... Opera 7.60 for Windows has support for voice xml and some aural styles. The software: http://snapshot.opera.com Docs and tutorials: http://my.opera.com/community/dev/voice/ -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] A quick breakdown of some code today
What browsers support CSS3? I'm guessing Firefox does/might. Are there others? Ofcourse there isn't a browser that supports more than little bits of CSS3. Gecko supports CSS3 selectors Opera supports CSS3 media queries Explorer supports some of the CSS3 text module (text-justify: newspaper; is quite nice) -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] No skipping to content needed?
A lot of people put an in-page anchor at the top to "skip navigation" or "skip to main content." Are there any hidden gotchas with simply putting the navigation last and positioning it first? E.g., The gotcha is that text browsers and screen readers don't understand this css. Does it negatively affect anyone's browsing experience? Yes, all those that needed 'skip' link in the first place. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] designing for the cell phone and PDA
Are there any successes or failures out there? I have no idea if my attempts works in the real world, since I do not even have one of those small devices. Have only been able to test in Opera SSR on PC. I have compared PC SSR to real device. It is very close. The only thing is that mobiles use smaller fonts, so 1em is fine and don't go below 0.9em. Remember that handhelds may not have any pointing device, so don't rely on :hover and onmouseover/onclick. I've adjusted css for handhelds on: http://pornel.ldreams.net http://osiolki.net http://browsehappy.pl I also tested OpenWave. It is very simple browser, lynx-alike. One problem that I've found is that it has limit to 4kb of text in a textarea. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] designing for the cell phone and PDA
On Thu, 2 Dec 2004 14:15:42 -0800, Ted Drake <[EMAIL PROTECTED]> wrote: Is there anyone out there that has had some success building a style sheet to make their web site look good on a pocket pc or cell phone? I'd like to add this feature to our site but I haven't had much luck. Are there any successes or failures out there? http://my.opera.com/community/dev/device/ Mobile Opera 6 uses stylesheet for media handheld and ignores styles for screen. Mobile Opera 7 fallbacks to stylesheet for screen if handheld stylesheet is not present, altough you don't have to worry about that because it only takes basic styles plus does smart reformatting so page will look good anyway. I cannot find information about Pocket IE, but AFAIR it attempts to read styles for media="all" and media="screen". If you want bare (X)HTML for all of them that should do it: You could also write stylesheet for handheld and test it on desktop Opera by pressing Shift+F11. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] What's "right"
hmm. ok, of course you are right. My next question is, what writing
a declaration like a#id, what is happening? Certinally don't wanna
throw out a "it works so it must be right", since it is obviously not
correct - but the question begs to be answered, why does it work? is
that in some way implying display:block?
Not at all. This is just a selector. It selects element "a" which has id
equal to "id". Only that.
Read about selectors:
http://www.w3.org/TR/REC-CSS2/selector.html#id-selectors
Various selectors may be joined together to select more specific element
in document.
#id and a#id basically are the same, because only one element can have
certain id, BUT each selector has its 'specificity' value which affects
cascade: http://www.w3.org/TR/REC-CSS2/cascade.html
a#id {color: green;}
#id {color: red;}
will be green. Not many webmasters actually know why Cascading
Style Sheets are cascading :)
--
regards, Kornel Lesiński
**
The discussion list for http://webstandardsgroup.org/
See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list & getting help
**
Re: [WSG] "Code" or "Markup"
Was CSS developed first for XSLT/XML and then applied to HTML? Or vice versa? CSS was developed for HTML. Hakon Lie, co-autor of CSS, doesn't like XSL-FO. http://people.opera.com/howcome/1999/foch.html SGML existed first and DSSSL was used to style Wow, compared to DSSSL CSS is s easy and clean. Back to code vs markup: For me "code" is reserved for languages that aspire to be turing-complete. These at least should have data storage (variables) and flow control instructions (if, loops or goto). Everything else, including XML and CSS, is data for me. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Why does firefox tooltip only "TITLE" attribute and not ALT?
It's much easier to check alt attributes with the tooltip than by reading source code after all. You can install Web Developer Toolbar extension or press G in Opera. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Flash menu (was: Remote control CSS - A quick little article about lists)
Article mentions some disadvantages of Flash. I'd like to add another one, IMO biggest one: * Flash links cannot be opened in new window (ie. user has no control over it) With growing popularity of tabbed browsing this starts to be huge disadvantage, especially if Flash is used for menu. Avoid! -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Why does firefox tooltip only "TITLE" attribute and not ALT?
I am curious as to why Firefox only understands the title attribute when creating tool tips, why cant we just use ALT? ALT is for alternate content only and should be visible only when image is not. It is like and other HTML fallbacks. TITLE was created especially for tooltips. All standards-compiliant browsers stick to it. IE is the only one that tries to be "smarter". See: http://my.opera.com/forums/showthread.php?s=&threadid=15784 -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Images in Nav, Splash Screens.
I try to explain to clients (and designers...) that user always knows where he got to - people just don't type random adresses :) Site design should already have logo and "corporate look'n'feel" so such splash screen basically has no information on meaningful content. Visitors seeking information will get annoyed and bored by splash screen, because it is not what they came for. There is also an explanation on useit.com: http://www.useit.com/alertbox/990530_comments.html -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Are conditional comments the way to do this?
First I'd check if that problem can be fixed so that you can keep one
stylesheet.
You might have some error in styles that some browsers ignore and other
don't.
If by "PC browsers" you mean IE, check if it isn't one of it's well-known
bugs:
http://positioniseverything.net/explorer.html
Some of them have easy workarounds.
CSS rule for IE only:
* html #foo {height: 1%;}
CSS rule for better browsers only:
*> #bar {position: fixed;}
If problems are caused by incompatible box model (do you use valid
doctype? no xml prolog?) in such case you might end up with separate
stylesheet for IE and conditional comments are right solution.
--
regards, Kornel Lesiński
**
The discussion list for http://webstandardsgroup.org/
See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list & getting help
**
Re: Focus highlighting, was Re: [WSG] Some links for light reading (30/11/04)
The problem with declaring all three in one is that IE 5 (possibly 5.5 also, can't remember which right now) for PC chokes on any declaration that contains :focus. Combining your :active and :focus rules will effectively cancel that entire declaration in dodgy old IE. oh, dodgy old IE :/ remember Opera in smallscreen mode choking on that combined rule as well, but I think that's a seperate discussion. :) Probably old version. I've tested latest and seems to parse all CSS2.1. Ofcourse styles unsuitable for SSR are ignored or lost in reformatting. Without SSR latest mobile Opera handles CSS Edge demo pages*, many literarymoose experiments and even displays CSS lines hack - http://www.infimum.dk/HTML/rotatingStar.html *) can't test pure css menus, as :hover is impossible to archieve on keyboard-only browser. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] charater sets
In UTF-8 files you can use extra characters in their "natural" form instead of HTML entities - like nbsp, shy, mdash, ndash. You may also use quotes, elipsis, etc. They take less space and are safer for string manipulations on server-side. You don't have to worry about copying and pasting from other sources (MS Word creates quotes and dashes that (formally) are incompatible with ISO-8859-1). Foreign names are preserved. There are problems, though. Many editors that claim to support UTF-8, but internally operate on strings translated to codepage, so they may lose characters not present in current system codepage. As I've mentioned in other post, Notepad, ASP Web Matrix and most likely other Microsoft text editors insert invisible BOM character to mark file as UTF-8. This character prevents DOCTYPE or XML Prolog from being recognized and makes output buffering useless in PHP4. If you heavily use UTF-8 (most notably soft hyphen) you need to check if browser can handle it (check Accept-Charset header plus serve UTF to IE anyway, because it sends meaningless headers) - if browser (bot?) can't handle UTF-8 you need to make conversion. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] Foreign Translations
On Tue, 30 Nov 2004 14:30:11 -0500, Ketan Vakil <[EMAIL PROTECTED]> wrote: Is it an okay plan to take an existing English website programmed in ColdFusion and basically duplicate it across a few other languages (French, German, Italian, Spanish) either in their own directories like /english and /espanol, or at localized domains like .com, .com.es, etc... ? What workflow issues have people run into going with this method? Updates of website will be difficult. Every change will need many times more work to update and synchronize all versions. When you forget to update one language it may stop working and you won't notice that, unless you do x times more testing. You should think about keeping only text strings in an external file and having one language-independent version of pages. I can't help with CF, but in PHP it is usually done by including a file that defines associative array (key=>value pairs being some_id=>"translated string") Do you use the ip-to-country database to do the auto redirects to localized URLs? That won't work, people may use proxies, etc. Just parse Accept-Language header (but really parse it; don't assume it to start with suitable language code). It is good to have domains or subdomains with translated versions. If you have host-based localized versions you don't need to change paths on your pages plus search engines will index all languages (they won't if you use accept-language detection alone). -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: Focus highlighting, was Re: [WSG] Some links for light reading (30/11/04)
So, please, please, if you want to make your sites more accessible to
keyboard users, add :focus and :active rules to match your :hover rule.
ok, I will :)
a:focus {color: #346095; background-color:#fff;}
a:hover {color: #346095; background-color:#fff;}
a:active {color: #346095; background-color:#fff;}
BTW: It would be better to write:
a:focus, a:hover, a:active {color: #346095; background: #fff;}
It is simply shorter, more compact form. Some perfectionists prefer
shorthand background property, because there is slight chance that
background-color alone might not work if user stylesheet defines
background-image...
--
regards, Kornel Lesiński
**
The discussion list for http://webstandardsgroup.org/
See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list & getting help
**
Re: Focus highlighting, was Re: [WSG] Some links for light reading (30/11/04)
Does that really matter? In Firefox and IE there is a focus border anyway. IE doesn't support :focus or outlines, so there isn't much you can help. In Firefox Cursor-Mode (F7) uses small text-cursor that isn't good for bad sighted people anyway. Opera with spatial navigation always adds background on focused links. I don't know how mac browsers deal with this though. -- regards, Kornel Lesiński ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
Re: [WSG] UTF-8 ,charset and Standard
UTF-8, a flavour of unicode, is an universal character set. You don't define any codepage/language for it. You just simply use whatever characters you like. This creates "Content-Type" header being "http equivalent". Content is text/html with charset UTF-8. It would be even better to send real http header with charset. In PHP it's: Note: Don't use Notepad or other Microsoft tools for UTF-8, because they tend to add unvisible "BOM marker" character at the beginning of every file. This helps them recognize UTF-8 from other files, but confuses many browsers. I use freeware "Notepad2" for UTF-8. -- regards, Kornel Lesiński osiolki.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help **
