More OPML Browser corrections
September 26th, 2007 1:19:04 pm pst by 
Sterling Camden
Shack Dougall spotted several HTML coding errors in the latest verion of my OPML Browser plugin. Get the version 1.3 below.
Posted in OPML, PHP, Web, WordPress | 
25 Comments » RSS 2.0 | Sphere it!
Subscribe by Email
[...] watch: Lune uses my OPML browser widget. Better get the latest version, [...]
[...] More corrections here. Tags: feeddemon, markbarnes, OPML, PHP, phpopmlbrowse, phpopmlbrowse12, plugin, tdavid, [...]
I love this plugin! It keeps my blogroll from going on forever. One question, though: is there any way to have categories collapsed by default?
Sure! In the widget options panel, check the box beside “Start with folders closed?”
Thanks for the compliment!
Ah, guess it helps if I have the right version.
Thanks for the reply, though.
[...] is OPML Browser, running off a local copy of the OPML file because my hosting provider doesn’t have [...]
Great plugin. Thanks for putting together such a nice little tool – I’ve been looking for something like this for years.
I installed it, started making a few tweaks, and ended up getting a little carried away. I pulled out a lot of the embedded styles, added a number of configuration options (including being able to specify the CSS from the widget control panel), and updated the javascript to use jQuery. I am now using it to power a single-category list of blogs on my front page, and a full list on a separate page with a custom template that just calls the widget function.
I was planning to write a quick post about this, but wanted to run it by you first. You can see what I’ve done here.
Wow, Justin! Let me take a good look at this on Monday — maybe with your permission I could incorporate your features into the next version?
By all means, do with it what you wish. There was no license on it that I could see (?), and I was hesitant to slap one on since it wasn’t my work.
I was checking through my error logs last night and discovered a whole bunch of 404 errors – due to a missing atom.png called by your plugin. I have no idea what an atom feed icon is supposed to look like, so my “solution” was simply to copy rss.png and rename it as atom.png. You might like to update the code in the next release however. Thanks for a great plugin.
Thanks for the heads-up, Mark. Will do.
[...] More OPML Browser corrections — Chip’s Tips for Developers (tags: opml browser wordpress php) [...]
I like the sound of Justin’s changes,
but when I edited the plugin, replacing all the original contents of the php file with Justin’s version, I got “Plugin deactivated because changes caused a fatal error.”
Jonathon
Don’t know what could be wrong… it activated for me. But I did get errors in the admin panel because content was emitted before WP wanted to output headers. I haven’t finished merging in and testing Justin’s changes with what else I want to do to this plugin.
Can I create an OPML document which (combined with xslt) creates a table where users can CHECKMARK which feeds they want to subscribe to, and then subscribe to all of them at once?
That’s my reason for exploring OPML, but I have not found anyone doing that, or anywhere talking about it.
I expect I need to combine it with an “xform” (forms for inputing data into an xml file).
I’ve spent time at w3schools.com/xsl, w3.org/Style/XSL/, opml.org and others.
Do you know?
After fiddling with someone else’s sample opml files, and finally reading w3schools guide, I was able to produce a semi-basic opml display (imagiscape.ca/opml.xml), so I have some idea of somethings we’re talking about.
Thanks,
Jonathon
That’s a great idea, Jonathon!
So then, you’d build a new OPML file containing just the desired subscriptions? How would you import that into their feed reader? Or would you leave that up to the user to do?
The syntax of your XML file is more like how I wish OPML were spec’d. OPML puts all this data into attributes of the “outline” element.
Hmph. I thought I had triumphed over my OPML nightmare, but I guess I just evaded OPML, which doesn’t help me to use it.
When I used proper OPML format – all the attributes in the “outline” element – the result was useless.
Using Firefox, everything was invisible. I get this problem on most or all opml pages, when using Firefox.
Using Internet Explorer, on most or all opml pages I see an unstyled opml-format xml file. What good is that? What is anyone supposed to do with that?
I don’t think my Firefox add-ons conflict. (They are: coComment, delicious, download statusbar, google gears, google notebook, noscript (I approved the sites with opml files), one-click installer for wordpress, video downloader, web developer, mozilla’s talkback, and others which were not active at times i have had this problem.)
Is there an easy explanation for the invisible output?
I have this problem, for example, on Justin’s OPML page which uses newsgator – http://services.newsgator.com/ngws/svc/opml.aspx?uid=106209&mid=1, and a page produced for me by Feedblitz.
I emailed Justin last Friday asking “why is your OPML invisible to me …?”
Previously, I asked the question to Feedblitz. In December, Feedblitz replied that the [data] is there but not rendered (not visible), and they would look into it, but they haven’t, despite 2 reminders from me.
Do you happen to know the answer?
(If you don’t, I know there are OPML groups to which I can copy this question.)
Thanks, Chip.
P.S. I’ve seen documentaries and movies about autism (such as “After Thomas”). Coincidental to your subscribing to my blog, yesterday’s news had what to me is a remarkable story about autism – I deemed it blog-worthy; I blogged the article.
I’ll create an alternate “properly”-formatted OPML file.
OPML is an XML grammar. It’s not supposed to be rendered by browsers. To render it, you need a widget like this one, or you could roll your own using XSLT or some other transform to produce HTML from the OPML.
That solves that problem – thanks.
But:
What is someone supposed to do with an invisible OPML file?
Are we supposed to save the file to our computer and then add it to our local feed reader and/or upload it to an online feed reader? That’s rather indirect.
When we view rss/atom feeds, we see a link to subscribe to the feed. Surely browsers should prompt us with a link to import the opml file into a feed reader, no?
I have installed the OPML Support Firefox extension. It doesn’t do anything if you go to an opml page. I eventually found that if I open Bookmark Manager (“Organize Bookmarks”, and click File – Import, it offers OPML as a choice, AND I can enter a url into the Open: dialog box (which appears as if it will only process local files). It works, for importing an OPML file somewhat-directly into a local feed reader.
But then I still have to upload it if I want to get it into an online feed reader.
Why doesn’t it work like when we open a feed [page] – prompt us to import it into our local feed reader or an online feed reader?
Does IE do any better than Firefox? In IE the xml is visible – in [computer-friendly format] – but still there is no prompt-link to import it to a feed reader.
Grazr has, next to its opml links: “Use this URL to open this reading list in other tools, such as a feed reader”, but still it just opens the semi-useless plain opml file.
Is there still something I’m missing?
Thanks.
I think you’re just seeing the state of the art at this time, Jonathon. I agree that browsers should be able to do something with an OPML file, like offer to import it, or present a series of links to the RSS. That would be a cool Firefox extension, actually.
We have to remember that it was only starting with Firefox 2.0 that RSS autodiscovery could be used to add a feed to your favorite feed reader (other than Firefox bookmarks).
Chip,
I sent you an email about this, but possibly to the wrong address.
The site providing my OPML file went down last week. It brough my entire blog with it, because your plug-in couldn’t gracefully handle the site being down. Would you be able to update the plugin so that it could set a sensible timeout for checking for an OPML feed?
Thanks,
Mark
Ooh, sorry about that, Mark. I’m sure I could catch a failure on the file_get_contents statement to avoid this, or at least check for its existence first. I’ll be sure to include that in my next version, whenever I can get around to pulling that together.
I’m about to release the next version, and I don’t seem to be able to reproduce a problem here. I get a fail out right away that only prints a couple of warnings and then proceeds along its merry way. There may be some site-specific setting in PHP that governs this.
I just posted the new version here.