Gaffer

Hi. Just thought I'd add to this topic. I have 3 main areas where spambots may go ahunting: 1) mailto: My mailto is not in my main html pages. It's part of a write function in an external JavaScript, so it's a link that gets dynamically created when the menu is. 2) I included a name@REMOVEdomainTHESEname.CAPITALScom.....so someone still has to manually edit the address....something which I think won't be too tiresome for people? but will definitely catch out the bots who will simply harvest the undeliverable address as is. 3) My guestbook is hosted/provided free by lycos. They include a captcha tool as well as me specifying that I have to approve posts beforehand. I still get spam in here sometimes though (from humans, not bots). I also use the number; style to try to "disguise" the links. Despite ALL of the above, I still got spam. So I did the most important method of using a throwaway email address. Now I get lots of spam, but in a "non-essential" account that has good spam filtering anyway. So my best method ended not being in my attempts at "devious" JavaScript.....it ended up purely being using a different free e-mail account. Finally, I have a link at the bottom of my main page which takes you to a separate "AntiSpam" page. It's main purpose is to use the following JS to dynamically create a long list of meaningless e-mail addresses that the bot can harvest at will: In the head: <script type="text/javascript" language="javascript"> <!-- function makeEmail() { var strValues="abcdefg12345"; var strEmail = ""; var strTmp; for (var i=0;i<10;i++) { strTmp = strValues.charAt(Math.round(strValues.length*Math.random())); strEmail = strEmail + strTmp; } strTmp = ""; strEmail = strEmail + "@"; for (var j=0;j<8;j++) { strTmp = strValues.charAt(Math.round(strValues.length*Math.random())); strEmail = strEmail + strTmp; } strEmail = strEmail + ".com" return strEmail; } --> </script> In the body: <script type="text/javascript" language="javascript"> <!-- Begin for (var rePeat=1; rePeat<10; rePeat++) { /* Set the second number above (after rePeat<) to one number higher than the number of e-mail addresses you want displayed on the page. */ document.write('<a href="mailto:' + makeEmail() + '">Email Us<br><\/a>') } // End --> </script> Then, at the bottom of the page are 2 links. One takes the bot to a separate page with a list of e-mail links, the other takes the bot to a page of "infinite" links (the bot gets to the bottom of the page, and then follows the final link back to the top of the page via a dynamic redirect.......so it will be endlessly trapped harvesting "useless" adresses). Then, when the spammer comes to check his/her server, it will be full of useless data and will hopefully crash Here's that "endless list" link (safe for browsers, harmful for bots) http://www.bfndevelopment.com/cgi-bin/home...orKGg/List.html Here's the other one (ditto): http://www1153356205784.blogpipe.com/ This last bit is purely "payback".....I'm not sure how effective it is (maybe spambots are "clever" enough nowadays to ignore such links, but everything I've read on the net seems to indicate it still works).

^^^ Thanx for that. What an elegent solution . I've done 2 things: 1) Use the defer="defer" attribute to "fix" the only browsers that have problems 2) I set window.setTimeout to 0 -----> apparently does nothing but is meant to allow asynchronous loading with all other browsers (which ignore the defer and so load the menu immediately)? ....Is that correct?.... IE on the other hand doesn't have a problem with this.....it's already been deferred in step one The only issue remaining is that now i.e.5 mac crashes completely . Aaaaaah sod it. It's not as if any other browser on Earth uses Tasman......using browsershots.org proves that the page loads fine in all windows versions of IE now, what do I care about i.e.5.mac? It's shortcomings are rapidly wearing thin (I had to laugh at that chaps "solution" to ie5 mac compatibility). I've now disabled JS in ie5mac and am using that to purely check the site with "I haven't got JavaScript enabled" functionality. Another example of the wonders of ie5mac Thanx for your help....that defer function is really cool (simple yet powerful). Cheers Gaffer.

It's actually quite slow (I think it extensively uses a Javascript based engine), but the built in validator is the reason I use it. Funnily enough, I hadn't heard of it until recently, but now that I do use it occasionally, I'm seeing more and more references to it on the net...

^^ I agree...... and this is great news. I like the way they appear to be implementing this... have your CSS point them to a remote http:// web address (like the example). This is intriguing... I can foresee a slew of websites springing up not unlike the current "get these dektops/icons for free" that simply store thousands upon thousands of free fonts.......you won't even have to have the font yourself, your own browser will get it from there too....brilliant. They could also specify that the fonts will have to only be useable by the browser...ie making it impossible/difficult to store it locally on your HD, thus solving piracy issues (which would be a great idea if true).

OMG that's FANTASTIC... thankyou thankyou thankyou thankyou (I think I love you) I like the way you don't even specify a element in the css....it's just 'dot'class Thank you again Gaffer

I'm not sure why W3C has changed the concept of <font>. What we had: <font="">blah</font> Nice and simply. Easy to quick scan through html. What we changed to: <font class="plop">blah</font> along with a css element: font.plop {color} added bloat but still manageable. Still easy to scan through html. What we're being told to change to: <div class="plop">blah</div> along with a css element: div.plop {attributes here, font, size, colour, align...etc etc} JEEZ....forget bloat, it's turning into something Herman Melville would be proud of . No longer is it easy to scan through long tracts of html because when EVERYTHING has become <div>, how easy is it to pick out the lone font element you want? Yes, you can alter all your pages with the css, but what if you only want this font element on ONE page....it's easier to navigate to and alter that particular page isn't it? </rant > So I attempted to change some <font> tags to <div class="colour">. The text is silver, with comments on the right being darkgoldenrod(-ish). The problem I have is that whereas using font I get what you currently see: text is here. maybe a comment here. more text here. more text here. another comment. yet more text. yet another comment. what I'm getting is text is here. maybe a comment here. more text here. more text here. another comment. yet more text. yet another comment. I haven't added any extra spacing, so can someone please help (it looks like the <div> element is adding them)? I was planning on using the following: Silver text is div class="sil" CSS div.sil {color:#898989;} small silver text is div class="silsml" CSS div.silsml {color:#898989; font-size:70%;} small orange text is div class="colsml" CSS div.colsml {color:#c81; font-size:70%;} ^^ works fine re size and colours, but then adds those stupid extra line spaces. Help ... P.S. page is here (@ the bottom). Thanx for any help

Thanx .....only been here a couple of days and already learned so much . Re: Wii.....GREAT isn't it? can't wait for mario galaxy and metroid 3 to appear.

Hi, thanx for the info. Actually, I.E5 isn't too bad (well....OK....it is but at least it doesn't crash as much as I remember it would on win '98 and 2000). Re: problem, it turns out it's actually an inherent issue with the way ie (all of them) are handling pages.....as mentioned all other browsers are fine. M$ themselves have admitted the problem here (it occurs when the script inside a div element uses .innerHTML to try and modify the body before that has finished loading). This would also explain why, when navigating via the 'back' arrow, everything is fine. In this case, the <body> has already fully loaded and been cached, so when the menu loads, it's not trying to load before <body> has finished. I can just imagine the M$ bods who created the Trident engine saying :doh: right now I found another fix here, but hopefully, seeing as M$ has acknowledged the problem, the next updates for those people on win will solve the issue. (.......aaaaaah how I miss the "stability" of my old windows days ) Thanx for your help.... I'm off to try one of those fixes EDIT: confusingly, I already have one of those fixes......a window.setTimeout function at the bottom of the script (the script creates the floating toolbar, it was set to 1000.....I think the reason it didn't work at my sis' but works on other ie's is because she has a slower internet connection?????? ---> so <body> takes longer to load and therefore menu tries to load before it has finished in her case.....maybe?)......problem is, I since set it down to 250 without realising hoping for a "instant" rendering of the menu (rather than the 1 sec delay), which may worsen the problem.......time to set it back Cheers

You might want to try this free icon collection as an alternative to those Fam Fam ones: http://www.ndesign-studio.com/resources/mini-pixel-icons/ there are about 320 of them in the zip file.....

Seeing as they're different OS's, just in case there are inconsistencies, it's worth having one browser per layout engine . For instance, that means it's not worth using Firefox if you use Camino (same engine).......so I have: Gecko Mozilla - choose between Firefox/Camino/Flock/SeaMonkey/Netscape WebCore - choose between Safari/OmniWeb/Shiira Tasman - Internet Explorer 5 mac iCab - iCab (<-- worth it because it's free, has a unique engine, but most importantly has a GREAT built in validator) Presto - Opera That's 5 browsers on my mac . This leaves the most popular engine, Trident (I.E 6 and I.E 7 win) left exclusively for your windows machine.

OK, being a mac guy I have tried to use a broad range of browsers (that use different layout engines) to check compliance of my site.... WebCore - Safari Mozilla Gecko - Firefox Presto - Opera Tasman - i.e 5 for mac iCab - iCab The only engine I can't get on the mac is Trident (as used in i.e. 6 & 7)......and up till now I assumed that ie5mac that I have is good enough (seeing as Tasman is meant to have stricter protocols than Trident.....so if it displays fine in ie5mac, it should theoretically display fine in ie6/7win). Anyway.....long story short, went to my sis' today and her pc win based ie7 would not load my site. It kept saying: "Internet Explorer cannot open this site: aborting operation" and then redirecting me to a oage that said "server unavailable" or something. Now the odd thing is that pressing back on the browser would then take me back to my homepage, that not only would load and display fine, but navigate fine with nil "operation aborted" problems thereafter. This behaviour makes me think it's an issue with i.e.7 as opposed to any coding on my part (as otherwise it would never display/function), but I just wanted to check with others who use 6/7 on win whether it's happening to them too (click my siggie to go there). Hopefully it is a problem with the prog as I don't relish having to recode my entire site just for I.E. when every single other browser I have ever tried seems to work fine with it. Thanx for anyones help on this Gaffer

Thank you for the welcome . Sorry for the longwindedness (<--- is that a word? ) of the post, it's just that I've never come across a validator that compiles your code prior to validation before... .....just wanted to see 1) if others had come across it and 2) what you all thought of it. To tell you the truth, html validation is VERY confusing. I currently have 3 that I use: iCab - which PROMISES that it uses W3C Total Validator - which again PROMISES that it uses W3C Tidy - yes you guessed it......GUARANTEED to use W3C ....however, putting the same pages through them results in each one pointing out different things it deems are errors . No wonder the net is full of non-compliant code in web pages when the validators themselves can't agree on common ground .

Thanx for the welcome folks Sage wisdom?......naaah....it'll probably be me asking all the Q's from you more knowledgeable guys... .....oh.....apart from "PLEASE PLEASE PLEASE stop using I.E as a browser" (yes, you've guessed it, I'm a Safari man myself.....although recently using Opera 9 has left me Mightily impressed)..... most recent IE for Windoze transgression: continual "internet explorer cannot open site: aborting aborting.....DANGER Will Robinson.." (OK, I made that last bit up ). P.S. As it happens, just ordered a new 20", 2.4GHz iMac for my sys (3Gb RAM.....mmmmm). Leopard should ship with it. I myself am still using my 2 and half year old PBG4 (PPC), but it's still going strong on Tiger. P.P.S. That's me in my Avatar

Hi everyone, newbie here. Have been maintaining my own personal page since about '96 now. What started as a place to keep all my bookmarks across all Uni computers has mutated into a bit of a hydra over time. Have helped create a few simple designs for colleagues and friends too ......but I only do this as a minor hobby - my actual job involves working in the "No Hope of Survival" . I'm also resolutely stuck in my ways....I loath things like iWeb/Dreamweaver/Frontpage (I prefer coding "raw" using Text Edit/Notepad (now BBEdit) and recently CSSEdit. I at least attempt to make most of my own graphics in Photoshop rather than that free rubbish you get online ). Apart from that I'm O.K at HTML,currently learning CSS, and utterly utterly abysmal at JavaScript. What do I hate MOST about the 'net? Has to be flash really . I've only ever seen about 3 well-designed flash based sites in my life / sites that extensively use it (http://www.thedesignersrepublic.com/ is one of them). Apart from that, I usually get SOOOO p'd off with them, I give them approx 1.2sec before navigating away What do I like? What's not to like? For instance, little gems like http://www.webpagesthatsuck.com/ are always good for a laugh/lesson. Have recently (well...Jan 07) discovered the Wii, so have spent this week getting it to display properly in the Wii Opera browser. This seemed like an uphill task.....but I FINALLY managed to get rid of frames (YAAAAAY ). It's now 22:30 and I've just debugged my floating menu so was going to turn in.....when I found this excellent forum. I'm a bit of a forum junkie at times (belong to lots of Hi-Fi ones as listening to music's my main way of relaxing) so was glad to come across this community. Looks great here and can't wait to get posting Gaffer P.S. Apple mac and OS X all the way........

Hi everyone......first post (please be gentle ) I noticed that there are two types of HTML validator out there. The most common kind: For instance, that used in Tidy (Safari)/ W3C / Total Validator (Firefox) ...etc.... What these do is simply get the raw data and (without rendering any page), simply validate the source as is. Example source: <h1>hello</h1> <script type="text/javascript" language="javascript"> <!-- Begin document.write('flippy' + ' & ' + 'floppy') // End --> </script> <p>were 2 wittle bunny wabbits What the validator "sees": <h1>hello</h1> <script type="text/javascript" language="javascript"> <!-- Begin document.write('flippy' + ' & ' + 'floppy') // End --> </script> <p>were 2 wittle bunny wabbits .....and you then get an error report. Compared to...... The less common kind: That which is found in the browser iCab Same source, but this time the validator "sees": <h1>hello</h1> <script type="text/javascript" language="javascript"> <!-- Begin document.write('flippy' + ' & ' + 'floppy') // End --> </script> flippy & floppy <p>were 2 wittle bunny wabbits Notice the penultimate line. The iCab validator actually runs the code first and then sends it to validation. The problem with this is that all the little tricks used to "get around" validation no longer work......because iCab W3C validation actually runs all the CSS and Javascript code before analysing. Let's say I have an external code to run that requires <embed src .....blah blah>. It won't pass validation unless you code it as: <script type="text/javascript" language="javascript"> <!-- Begin document.write('emb' + 'ed' + 'src=' + 'blah blah') // End --> </script> ....but iCab will "compile"(?) the script first and then run it....... As a result, it will fail iCab validation, but pass every other kind (iCab can "see" the fact that you're trying to run the tag embed.........sneaky ) In fact, that's another bugbear of mine, even within the first category of validator, differnet ones I try give me different reports when looking at the same code.....but that's another story Overall I tend to use both categories because even though the former catch most obvious issues, iCab tends to "compile" the inline CSS (which others might ignore) first, so I get to catch more errors this way (I have very little inline CSS, but that which I have can be checked). Anyway, that's enough of me rambling , so which type do you all prefer and why? P.S. I know i shouldn't use <embed..... but it's from an external site that uses it to encode my Radio Station (last.fm).......not my fault...honestly