The site is being given a slightly improved design, both visually and with regard to bandwidth usage. It will probably be a day or two before it’s done and in the meantime the website might not be looking its best. Will also try to use more CSS this time around.
Progress so far …
- Reduced the logo size by 50%, omitting the PNG 24 alpha data but ultimately threw it away in favor of a TEXT logo. For the time being at least. And I kind of like the simplicity.
- No more total style switcher. Might add a simpler switch option to just change font size.
- Split CSS using @import, weeded out old styles, duplicates, cleaned up all whitespace, line breaks. And could obviously slash CSS in half to begin with since the style switcher wont be active. In the future I can use @import if I wanted to load different styles, like with the switcher, to just control font size for instance.
- Updated the Creative Commons licence. And a new spiffy button to go with that. Not that I really believe in IP and rights but I have become convinced that we have to strike at corporations where they are vulnerable. I.e. out-licence the licensers. And to prevent commercial derivations by any means necessary. And yes, I agree with the critics, CC is not good enough and it might be saving paradise by destroying it. But as I have said in the past, it’s not there to protect any work as such, more like protecting against clever legal attacks. And supporting non-commercial principles. It’s the wired equivalent of ‘trespassers will be shot’, ’stay off the lawn’, ‘beware of the dog’, whatever …
- Added a number of Nucleus plugins that I had previously overlooked. Like XML Parser Support Library, Nucleus XBEL Reader, Nucleus RSS/Atom Aggregator, Views, Most Viewed, Referrer, QueryLog, ExtraSkin, Table … but most of that is background work.
The idea is of course that I will no longer have external pages that I have to update via ftp. And since I’m planning to shrink this site to its bare blog essentials I don’t foresee any problems with that. ExtraSkin goes a long way towards that.
One thing that I’m still having issues with is Referrer. For some reason I can’t get it to work properly. Not the original code, not the updated version and not Referer2. - The design is up and working. Most of the major things seem to be working with MSIE6, Firefox 1.0 and Opera 7.
XHTML quirks and validation issues (no validation):
- Strict vs transitional - Had to go with transitional. If nothing else, for the href targets.
- Target _blank - Don’t validate under strict. What a joke. So I understand the workaround includes javascript hacks. What a joke.
- Ampersand - there are always links with & for some reason w.bloggar doesn’t translate them well.
- Border, hspace, vspace - Img properties transferred to CSS.
- “An “id” is a unique identifier” - Easy enough to fix but why?
- UL inside P - I’m sure there are quite a few disallowed combinations but this is one of the more stupid. Also easy to fix, but what is the point.
- CaPs IsSuEs - XHTML takes case very seriously.
- Ending / - Big problem with some Nucleus plugins, scripts etc. Had to edit quite a few to make them suitable for inclusion.
- <!– –> conflicts, mostly in IE - seems to break certain width and margin properties in certain places.
- Display : block for images, again in IE - or you get a nasty little bottom margin or padding that you just can’t do away with.
Also … The server has had a great deal of downtime these last couple of days. But that was a temporary thing. I hope.
Contact
Lifestream





I have seen quite a few places that translate the ampersand to %26 inside of the Link and then (I assume) alter it through some CGI script.
Also, I think that the uniqueness of "id" was implemented so that a designer could have something to reference one unique tag which exists on multiple pages (such as a signture, home link button, etc…) and it’s properties can be manipulated by an external CSS file. The uniqueness is supposed to be explicit. For example: avioding oversights that could be made by other designers that may be working on the same code in the future.
Well, I don’t know about that ampersand coding. Maybe it’s an IE thing? The solution is to use &amp;, which I guess most browsers can handle as well. And then I can be sure the code is valid and that it can be passed in WML/WAP and so on.
Yeah, I changed all the ids. It was stupid mistake on my part, class is fine too, gets the job done just as well. Never used ids until this latest site, except for using javascript effects and such. I really need to read up on this, and how the use of id and class is supposed to be in current XHTML versions. It validates now, sure, but as you point out, proofing the code (and myself) for the future is always nice.