Day: 8 June 2003

Ok I looked at the opml spec and what a simple standard. Its almost too simple and I feel goes slightly against the idea of xml. I mean it structures data but using the outline tag is like me using the paragraph tag in html to structure all data.

Anyway seeing how people and applications are using this simple xml schema, I've been thinking maybe I should use it for my bookmarks. Because I've been looking for a standard way of storing those in xml for a long time now. Also been thinking I could import xlink and use that on top of opml rather than html href's? I swear there has to be better way of storing bookmarks.

Also kind of related in a round-about way. The newsreader on my ipaq pocketfeed is working but seems to have problems connecting to certain types of rss streams. I got a feeling it has problems with rdf based ones. But also has problems with my own stream which is weird because it works with other blojsom streams. Maybe the port 81 thing is causing problems or something? Also my newsreader on the pc doesnt support opml, so i keep having to copy and paste things between my tabletpc, workstation and ipaq. I really need to sort out the file sync issue on windows xp, because this is becoming a bit of a pain now.
Oh also why the hell does feedreader put its config and xml files in the user directory? ANd why cant you change it? Its a bloody nightmare. Hope they fix this issue in the new version. Tempting to use something else. Specially with the webserver not having any template file I can modify to change the code it generates. Non standard html code at that too…
Sort it out!

So I finally got a copy and installed it on my workstation.
I'm hoping to see what Microsoft have done when it comes to xml, because if the hype is right (very unlikely) it should bring xml to the desktop market.

First things first…
Not much has happened to Office when it comes to installation. Still a real pain. God help anyone who does typical install of office 2003.

Ok so first things first, I loaded a plain xml document into word 2003. Inpressive, it actually knows its xml and will let me author it in the same way as xmetal. Kinda of tag based if you want or non tagged if you like. Seems good enough so far.

But what if I load a xml document with a DTD attached? Ah it does it but doesnt like the DTD, kind of what I expected really. Doing the same with a XML schema fairs better, but it doesnt seem to restrict me from putting in non valid content. Didnt try saving, maybe thats when it does the checking? Also found if you open a xml doc with a xsl also attached it will show the html tags instead of the xml ones.
Could be useful, but I actually thought it would show the final result, so you could almost edit the webpage in word. Because editing the html isnt a good solution for non technical people.

So I started thinking lets have a look at infopath 2003.
It seems very much like xmlspy's stylesheet form editor. But with a little help you can drop a xml schema in. But lo-and-behold it doesnt like my schemas…Why?
My schemas have been checked against many xsd checkers and there not actually that complex. It seems Infopath doesnt like includes and imports at all. I had to revert to a schemas I wrote over a year ago to test it out, and even some of those didnt work. Maybe its the namespaces?
Once I finally got a schema in it a matter of dragging and droping elements and attributes around on the plain page and editing from there. You can add validation rules to the forms and areas but why does it not just add default rules from the datatypes in the xml schema?
If I make the datatype xs:AnyURI, that usually means no numbers to start off with, etc.

Once you preview the final thing it seems pretty good. No idea where the xml goes but it does the validation and nothing much else it seems. Anyway I didnt really test it that much.
Going back to word 2003, seems even when you save it doesnt do the checking against the schema. I even added the schemas to the schema library and it correctly realised there were imported and included schemas attached to the schema I was adding. So why does Infopath have a problem with the same schema?
I really need to read some docs on xml in word to see how it all works rather than fumbling around blindly. Missed the infopath online streaming demo on may 29th but I'm sure they have archived it, would look now but Reservoir Dogs is on tonight and I'm watching that till its finished.

Ah useful Office 2003 Assistance Home.

Ah it seems word 2003 works in the same way as open office. There is one xml file which ties the schemas, xsl, css and xml together. Manifest.xml. Seems you need the Smart Document Software Development Kit to make the whole package? Isnt this what infopath was ment to do?

Its so true they say about learning while in a project. I have been helping david with his osdavid project. Problem was the same as I had with adding search functianlity to courselets. He is using many xml files pulled into flash for dynamic presenation. Anyway when he searches for something or compares a node against another one, flash chokes and takes along time to return any results.
Bad news eh? yes…I'm not the number one fan of flash but if people want to use it thats up to them in the same way if you want to snort coke, I cant stop you.
Anyhow putting davids xml in exist and doing the queries in xpath on exist will stop flash choking because all it needs to do is format a http get request to the local server and exist will come back with the subsets of xml it needs to do its job. Easy.

This requires david to learn a little xpath because he wants to compare elements and attributes in different documents. And do content searches. It sholdnt be all that hard for him to learn.
I think the hardest thing for him will be getting the file the search results returned. Simply because flash uses a dom type of method. See the power of using standards! xpath to query the database and xpath to write the xslt! Anyhow for his osdavid project this should be good news.

After working with david yesterday, I've been using exist a bit more and finally understand it. Man there is so many ways of getting to the xml database. xml-rpc, soap, http, xmldb, etc, etc. Its one hell of a personal project I have to say.
Links to blojsom? maybe, because i did read you can put not well formed xml into somehow. But by default it checks when you put it in if its well formed, if not it rejects it.
Ah david, that could be the reason why some of your xml files couldnt be moved across to exist even on my own webdav client. Yes I remember had the same problem putting courselets in. ha i love xml.