Re: Does xml:base apply to type="html" content?

["M. David Peterson" <xmlhacker@xxxxxxxxx>]

Hey Aristotle, 

Firstly, thanks for your recent effort to let the foks at O'Reilly know that some serious issues have been introduced into the system since the "upgrade" to Movable type.  I can tell you how aggravated I was to get an email from Lawrence Lessig last Saturday, which was in response to the email I sent letting him now I had made announcement regarding the stream and download... His paraphrased response was:

> What Post?  All I get is a "Page Not Found" Error.

Of course the reason he got this was the fact that MovableType refuses to pull their head out and realize "if I update the post, that doesn't mean I want you to change the name of the html file as well!", and because i addes a few extra links and such thats exactly what took place, breaking the link I had sent him.  

I originally thought "well maybe there simply building in a versioning system of sorts, incrementing the file name by one number with each update, but in fact thats not what they're doing at all, as its quite a random process as to what name it chooses to rename it to, often times reverting back to the original name it had it set to 
e.g. title_of_post_1.html will be changed to title_of_post_2.html and then back to title_of_post_1.html, or sometimes it will drop the number off entirely -- title_of_post.html -- and its all COMPLETELY random as you can make 15 updates to a post and nothing will change, and then suddenly it will decides to rename itself and break any and all link that either you sent out, or being stored on 
del.icio.us, or whatever else.

Anyway, the point is that the reason everything is currently broken is because they moved things to a MovableType system 2 weeks ago, and in doing so its been a bit of a mess as of late.  To Justin's credit though he's been fixing each and every problem as soon as that problem is made known, so if nothing else, at least theres someone both compitent and willing to do what needs to be done to keep the system running.  And in fact, one of the problems (although fortunately there have been MANY, so its not as bad as it could have been) came from yours truly, as I copied over the link for the popup Free Culture reader/player, and like an idiot didnt check to see if the place that provided the "copy & paste" code had quotes all of their attributes, and as such came to discover yesterday tha, In fact, this was the cause of breaking the feed.

How can a blogging engine written in the supposed Text Processing Wonder Language, and even further, been around the game as long as anyone has, not even check for simple things like unquoted attibutes, and instead output broken XML?

Okay, rant now over... :)

On 3/31/06, A. Pagaltzis < pagaltzis@xxxxxx> wrote:

No. That is only a problem if you just mash markup together
without taking care to preserve base URIs by adding xml:base
at the junction points as necessary.

Well yeah... but again, pick a planet, any planet.  ;)

I guess thats kind of my point... Even those of us who make code both our life and our livelyhood have continued to run into too many issues, and have just given up, or simply decided "Lets just wait til' Atom releases and then build our systems from the Atom feeds provodide by the same folks.  In fact, it seems to me that this might be the exact thing taking place, as I seem to be noticing that the overall quality has been getting better as of late.

That said, I'm not sure if my concerns are founded on much of anything beyond trying to ensure that the little things that might get in the way can be cleared up and simply not be an issue any longer.

As such, if all thats been proven is that my comments and base:concerns are in fact comptetely off base... ;)

SWEET!!! :D

Copying an atom:entry from one feed to another correctly requires
that you query the base URI which is in effect in the scope of
the atom:entry in the source feed, and add an xml:base attribute
to that effect to the copied atom:entry in the destination feed.
If you do this, any xml:base attributes within the copy of the
atom:entry will continue to resolve correctly.

Yeah, I can totally see that...  In fact I am trying to pull together some XSLT 2.0 transform code that can be used to, in essence, canonicalize the usage of xml:base and the associated URI's, and if I can get a solid hour this morning to finish it off, then I will post the location to this thread so you all can tear it to pieces, such that we quicly develop a simple utlity that can be pushed out to the masses to build a local cache, and keep the base and relative URI pristine clean.

This all relates to both the ChannelXML and AtomicRSS code that are my current areas of focus, so
 

It's much easier to get right than copying markup without
violating namespace-wellformedness, even.

I agree.  As long as all of the proper URI's are in place, is just a matter of determining the proper base value in regards to the URI's that each will relate to, and for the most part be able to call it good.

You up for PowerHack ExtremeXSLT session sometime in the next few hours? :D

Regards,
--
Aristotle Pagaltzis // <http://plasmasturm.org/>

-- 
<M:D/>

M. David Peterson
http://www.xsltblog.com/