11 Feb
2004
11 Feb
'04
3:52 p.m.
Hi Sander,
I was trying to look at the RSS feed of the SmallWiki
at kilana and at
first BottomFeeder (my favourite Smalltalk-based newsreader) was not
able to read it. James Robertson looked at it and found problems with
the RSS you generate (look at his comments below). He made
BottomFeeder more forgiving, but it would be good if you could fix the
feed.
I think BottomFeeder should ignore any unknown namespaces and just
display the <description>...</description> thing, that is also
available and where I currently render the modification date only. In a
previous version I tried to render the first paragraph of the page or
the first 500 characters as plain text, but this didn't work out well.
This is the reason why I decided to render the whole document body as
HTML into its own namespace. My reader displays the page in-lined in
the browser-window and I think it is very convenient. Readers that
cannot handle HTML should ignore that section, but obviously this can
be a problem because the code inside
<content:encoded><![CDATA[ ... ]]></content>
is not guaranteed to be valid XML as SmallWiki users can put (wrong)
HTML code manually inside any page. Anyway, I don't know what to do on
this issue? I thought that XML-Parsers will ignore anything inside
<![CDATA[ ... ]]>, but obviously this is wrong or I am doing something
wrong with the encoding of the CDATA?
And a feature request (in case you have too much free
time ;). I
subscribed to the Changes page of the Moose wiki, to track all the
changes that occur to that wiki. It would be nice if I could click on
the entry and it would open the page for me in my browser (like I can
do on my Slashdot feed and others). I guess the entries in the XML
should contain some URL information.
[...]
I was wrong about my 'feature request'. I
looked a bit at the XML
SmallWiki generates and I saw the you actually _do_ generate item
links. So it should work. It must be a problem with my setup or with
BottomFeeder.
Yes, this is what I intended to write ;)
Cheers,
Lukas
--
Lukas Renggli
http://renggli.freezope.org
11 Feb
11 Feb
4:09 p.m.
New subject: [Fwd: Re: possible bug]
At 09:52 AM 2/11/2004, Lukas Renggli wrote:
<Talk Small and Carry a Big Class Library>
James Robertson, Product Manager, Cincom Smalltalk
http://www.cincomsmalltalk.com/blog/blogView
jarober(a)gosmalltalk.com
Hi Sander,
It does that in the latest dev stream. Even so, this is bad xml, and most
other readers are going to bail on it.
I was trying to look at the RSS feed of the
SmallWiki at kilana and at
first BottomFeeder (my favourite Smalltalk-based newsreader) was not able
to read it. James Robertson looked at it and found problems with the RSS
you generate (look at his comments below). He made BottomFeeder more
forgiving, but it would be good if you could fix the feed.
I think BottomFeeder should ignore any unknown namespaces and just display
the <description>...</description> thing, that is also available and where
I currently render the modification date only. In a previous version I
tried to render the first paragraph of the page or the first 500
characters as plain text, but this didn't work out well. This is the reason why I decided to render the whole
document body as HTML
into its own namespace. My reader displays the page in-lined in the
browser-window and I think it is very convenient. Readers that cannot
handle HTML should ignore that section, but obviously this can be a
problem because the code inside
<content:encoded><![CDATA[ ... ]]></content>
That's not the issue. The issues are:
1) You need to declare the <content> namespace at the top of your doc -
like this:
http://dotnetguy.techieswithcats.com/rss/allposts.xml
also, the content:encoded section should really be on it's own, not inside
<description>
is not guaranteed to be valid XML as SmallWiki users
can put (wrong) HTML
code manually inside any page. Anyway, I don't know what to do on this
issue? I thought that XML-Parsers will ignore anything inside <![CDATA[
... ]]>, but obviously this is wrong or I am doing something wrong with
the encoding of the CDATA?
No, just the wrong wrapping. Check it out at http://www.feedvalidator.org
And a feature
request (in case you have too much free time ;). I
subscribed to the Changes page of the Moose wiki, to track all the
changes that occur to that wiki. It would be nice if I could click on the
entry and it would open the page for me in my browser (like I can do on
my Slashdot feed and others). I guess the entries in the XML should
contain some URL information.
[...]
I was wrong about my 'feature request'. I
looked a bit at the XML
SmallWiki generates and I saw the you actually _do_ generate item links.
So it should work. It must be a problem with my setup or with BottomFeeder.
Yes, this is what I intended to write ;)
Cheers,
Lukas
--
Lukas Renggli
http://renggli.freezope.org
6:08 p.m.
New subject: possible bug
Hi,
thanks for the hint. I'm not very much into XML and didn't know about
this, I just copied some code I found on the web and tested it with a
few clients. The current SmallWiki fixes the issue and I updated the
kilana-server.
Cool, it validates now without problems:
http://www.feedvalidator.org/check.cgi?
url=http%3A%2F%2Fkilana.unibe.ch%3A9090%2F%3Faction%3DRSSChangesFeed
Cheers,
Lukas
--
Lukas Renggli
http://renggli.freezope.org
This is the
reason why I decided to render the whole document body as
HTML into its own namespace. My reader displays the page in-lined in
the browser-window and I think it is very convenient. Readers that
cannot handle HTML should ignore that section, but obviously this can
be a problem because the code inside
<content:encoded><![CDATA[ ... ]]></content>
That's not the issue. The issues are:
1) You need to declare the <content> namespace at the top of your doc
- like this:
http://dotnetguy.techieswithcats.com/rss/allposts.xml also, the content:encoded section should really be on
it's own, not
inside <description>
As a matter of fact this had never been the case.
is not
guaranteed to be valid XML as SmallWiki users can put (wrong)
HTML code manually inside any page. Anyway, I don't know what to do
on this issue? I thought that XML-Parsers will ignore anything inside
<![CDATA[ ... ]]>, but obviously this is wrong or I am doing
something wrong with the encoding of the CDATA?
No, just the wrong wrapping. Check it out at
http://www.feedvalidator.org
7391
days inactive
7391
days old
2 comments
2 participants
participants (2)
-
James Robertson -
Lukas Renggli