CodeStore.net comments on "Meeting Base Targets"

Reply from Sean Burgess

Sean Burgess — Sat, 8 Jan 2005 14:43:00 +0100

Well, as far as separating out internal and external links, the current script I have actually checks to see if the hostname is the same as my site. If you look at the source on my site, you will see the most up-to-date script. I am sure that if I took a couple of minutes, I could remove the hard coded host name and make the function be completely independent.

Sean---

Reply from Stan Rogers

Stan Rogers — Sun, 31 Jul 2005 23:07:00 +0100

...and if the target attribute is invalid in a particular doctype, it doesn't matter whether you set it declaratively (in the document source) or programmatically, except that there wouldn't be a validation error on load. Instead, the UA should either return an "invalid attiribute" error when the code runs or ignore the nonsense attribute once it is set. (Which of those should happen depends on whether you believe users should report concrete errors or "weird behaviour" when they file their trouble tickets.)

Reply from harkpabst_meliantrop

harkpabst_meliantrop — Sat, 30 Jul 2005 17:55:00 +0100

Tony, at least HTML 4.01 and XHTML 1.0 Transitional still contain the base element with target attribute. When writing for these standards, I couldn't care less about what's in the Strict specification.

And while the JavaScript approach probably works absolutely fine (especially, if the user environment is so well defined as in this case), I still wouldn't call it an ideal solution. In the first place, we should have an easy way in our CMS to seperate internal from external links. And I think, if all external links are entered by users using our own little app only, it could be done even with Domino. If Jake was paid enough for that ... ;-)

Reply from Richard Schwartz

Richard Schwartz — Thu, 28 Jul 2005 22:36:00 +0100

I used the multiple base tag trick in my very first commercially shipping Domino web template despite the fact that I knew it was invalid HTML. It worked, but whenever I thought about it I felt like I needed to take a shower to wash off the slime.

-rich

Reply from Sean Burgess

Sean Burgess — Thu, 28 Jul 2005 12:06:00 +0100

I can hear my dad saying "that's the way it's going to be because I said so!!!"

Sean---

Reply from Jake Howlett

Jake Howlett — Thu, 28 Jul 2005 11:36:00 +0100

Interesting that you can have multiple tag and toggle the target for links that way. Seems perverse but if it works...

However, I'm probably going to stick to my guns and make sure all links I create have target=self in them.

Reply from IanB

IanB — Thu, 28 Jul 2005 09:31:00 +0100

I still like the javascript idea as neatest.

I tried out the base tags (multi per page) and it does work, even if it isn't valid.

Reply from Sean Burgess

Sean Burgess — Thu, 28 Jul 2005 08:49:00 +0100

@IanB - That was my first choice, but there are 2 problems with that solution. The BASE tag does not have a closing tag and you are only allowed to have one BASE tag per page.

Ding, thank you for playing.

Sean---

Reply from Tony Summerville

Tony Summerville — Thu, 28 Jul 2005 08:10:00 +0100

I don't see why you think the use of Javascript is like "hacking" a page. I think this is a situation when javascript is BEST suited for the job, mainly since the target attribute has been phased out of (X)HTML standards, but IS still part of the DOM standards.

Reply from IanB

IanB — Wed, 27 Jul 2005 17:09:00 +0100

following on, I've had another idea.

have in the header.

then wrap whatever element contains links you [i]know[/i] are internal in .

Oh...is that what jorge meant 8-(

Reply from Tom

Tom — Wed, 27 Jul 2005 15:59:00 +0100

Hi Jake,

I think we could distinguish external links from internal ones by looking for "http:\\", as external links are absolute links, what do you think?

I know so we'd use Javascript...why not really.

Reply from Sean Burgess

Sean Burgess — Wed, 27 Jul 2005 15:53:00 +0100

I confess, the reason I wrote the code that is linked above is that I am a very lazy person. I got tired of putting target="_blank" on all my external links. The kewl thing about the code is that I no longer have to worry about mistakenly setting the wrong target. Plus, it actually decreases the amount of HTML that is downloaded. What more can a lazy person ask for????

Sean----

Reply from Jake Howlett

Jake Howlett — Wed, 27 Jul 2005 11:46:00 +0100

Good point Jorge. Yes I could. Hadn't thought of that. Cheers.

Reply from Jorge Coelho

Jorge Coelho — Wed, 27 Jul 2005 11:40:00 +0100

Could you just use a different base tag for your view templates? If your view will only contain internal links, why not just create a different base tag, which will require less work everytime you create a view?

Reply from Jake Howlett

Jake Howlett — Wed, 27 Jul 2005 10:38:00 +0100

You right YoGi, that would do the trick too.

I don't want to use JavaScript really as I always see this as hcaking the page. I'd much rather control it all from the server and the actual HTML. Far less to go wrong.

Reply from Chad S

Chad S — Wed, 27 Jul 2005 07:17:00 +0100

IanB is right, why not use javascript? There's code at

{Link}

that will accomplish it.

Reply from IanB

IanB — Wed, 27 Jul 2005 06:55:00 +0100

Surely it would be easier to treat external links to the special target treatment, as they would be the exception?

I know we shouldn't use javascript but could you not iterate through all tags in the page and check for the presence of external links and then add target-"_blank" to the innerhtml?

Reply from YoGi

YoGi — Wed, 27 Jul 2005 06:29:00 +0100

well you may use non-html domino views, just deactivate the "show values in column as links" and made your owns (ie : "["+document_title+"]", but i'm sure you've already thought about that. or didn't i understood your problem ?