h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

There are two bugs in Pliny that affect the display.

The first bug means that the entry is not displayed. The title is there, but there's no content.

The second bug means that the page returns an error. Pliny says it can't find the page.

In both these bugs, the page is visible normally in any of the Ripley skins.

I've checked about 650 entries so far, and I've only come across 7 examples of the first bug and 4 examples of the second.

I intend to check every one of the 10,000 approx entries for these two bugs.

Oo... Not seen any that do that (yet), I'll let you know if I spot any, as that sounds like something that would affect the screen-reader, as its the actual page not coming up (I'm using Ripley, genrally on hootoo, as the conversation use is easier, but using Pliny for my looking at the entrys, as that is easier for finding the created and modified dates, and just easier for lookign at entrys as the headings on the pages work properly...)

Here's an example of the first bug:

A706277

This displays some text in Ripley but blank in Pliny.

I can't give you an example of the second bug because I've fixed all the ones I came across. The particular error in the GuideML that causes it can't be created in Pliny. It could only be created in the old h2g2, and only by the BBC Editors. So the good news is that when I fix it, it is fixed for ever.

oo... how odd! I can see it, as you say, in ripley, and its just empty in pliny

That one is caused by the trivial matter of having the GUIDE and BODY tags in the GuideML in lowercase when they should be uppercase.

Wouldn't it be easier to make Pliny respect lower-case tags?

There's a case for changing the editor on Pliny to convert tags into uppercase, but I believe Pliny should enforce strict coding standards.

I find it easier to read Guide-ML when the tags are uppercase, don't you?

It's more complicated than that. GuideML needs to be in uppercase because Pliny insists on it, but HTML tags can be in either, and the HTML tags are passed on to the world, so Pastey insists that they should be in lowercase which is the world standard. He says writing tags in upper case gives the site an old fashioned look and will lower its search rating on search engines because of this.

So when I was the Guide Editor in charge of such things, I set the standard to lowercase for HTML and uppercase for GuideML. That's clearly laid out in the GuideML documents on the site. But I don't think anybody else has noticed.

So _must_ be in uppercase, FOOTNOTE, LINK and HEADER should be in uppercase, and p, i, table, b, ul and so on should be in lowercase.

I'd prefer to see the standard changed so that everything was in lowercase and some sort of a robot to go in an rewrite all the existing quarter of a million entries to make them compliant.

I got used to writing all the HTML I do in upper case, from having gotten used to using all the Guide ML in upper case so long as its all one or the other; and there is some automated way to change all* the code on the site to be complient, with whatever is chosen, via some automated process
Its the big advantage of pliny, as far as I'm concerned; in ripley, things like 'headers' on entrys, weren't really* passed properly to HTML, in the browser, so didn't* really appear as actual header markup, which means, for me, using the screen-reader, soemthing like 'heading navigation' never worked on guide entrys, in ripley, wheras, it does in pliny; actually, the headings on the page in pliny makes it easier to use, for me, from that perspective, ; say, on an entry, I can start by hitting the link to the entry, the page loads, and I'm at the 'top' of the page (remembering top is top of the left hand half of the screen, and teh right hand screen, for me, is basically the bottom/lower half of the screen, as its all made liniar by the screen reader), From the top of the page, I can hit the letter 'h', and be jumped down the page, past all the links, the search box, etc., and jump to the first heading, which from memory, is the title of the entry, when reading the entry; now, in pliney at least, then, once I'm reading it, I can hit 'h' again, and jump to any subsequent headings in the entry if I want to skip any sections as I'm reading through it... (and, of course, I can do shift key plus 'h' to jump backwards through the headings on the page)

Its just the conversations that keep me using ripley for now, no unsubscribe by the conversation thread list, on my PS, and no easy way to get to the last unred/first new post
And, actually, now in the conversation thread, the headings on there I think make navigation in the conversation itself, for me, easier, though I've probably not spent enough time using the conversation threads in pliney to work that out in my head yet

The new version of Pliny (Pliny the Younger), which is being tested at the moment, has the "go to the last unread post" feature.

But it is not ready to be released yet.

2Legs, here is an example of an entry with the second Pliny Display Bug:

Entry A189407

It displays OK in Ripley, but gives an error when you try to go to the page in Pliny. Is it the same for you?

This error is caused by an opening P tag which is uppercase where the matching /p tag is lowercase. I'll fix it once you've looked at it.

I can load the page up alright in Ripley, and get the contents of the entry, etc, in pliny I think (if I typed the address in correctly), I get the 'oops' page, and the 'calming image'
(the links I used, for ripley, then for pliney, are

A189407

http://www.h2g2.com/approved_entry/A189407

> He says writing tags in upper case gives the site an old fashioned look and will lower its search rating on search engines because of this.

Old fashioned, yes, if anyone cares to look. But would it really affect search engine ranking? I wouldn't have thought so. Certainly it's not breaking any official standards. HTML has always been case insensitive, and this remains true in HTML5. Uppercase is unusual these days, but perfectly valid.

> the HTML tags are passed on to the world

They could easily (?) be lowercased at this point, if necessary.

TRiG.

I've scanned 1400 entries for the two display bugs. My auto scanner was really very unreliable so I'm using a spreadsheet with clickable links to the entries in Pliny. I've only found about 15 errors so far.

I've scanned 2,000 entries - that's 20% of the edited guide. I've found and fixed 14 entries with invalid GuideML. I suspect the frequency will drop off as I get to the later ones.

I've now scanned 37% of the Edited Guide. The errors are getting less frequent. Since it is not possible to create an entry with invalid GuideMl, every one I fix is one more permanently removed from the Guide.

I hope you really meant 'every one I fix is one more {ERROR} permanently removed from the Guide'

The way I first read your post I sounded like the Entries were being removed



F S

You're right: that was badly phrased.

Roughly 50% of the Edited Guide now checked.