More HTML-ized code?

[bekateen]

Hi Jools,
Today I was looking at an older post and I noticed that a CLOG reference, which was typed perfectly, appeared as text rather than as a link. I verified that the CLOG is typed properly because if I click the quote icon for this old post, and use the preview tool, then the same text becomes an active link to C. aeneus.

This reminded me of another current bug being checked in which someone called it HTML-ized code, so I'm wondering if it's the same problem here. Obviously, it is not a priority to go back into all the old forum threads looking for bad links, so that's not why I'm mentioning this. What I'm hoping is that there might be some singular glitch in the current code, or on a style sheet or something like that, which causes good links to be misread as text, and that a singular fix can correct the problem globally... but since things are rarely as simple as they might appear, I'm guessing this won't be so simple, will it?

God bless you, buddy, and all the coders, for putting up with this.

Cheers, Eric

[Jools]

Eric,

You should be a software tester!

This is a entirely different issue from the one reported by Lee. The forum (PHPBB) was introduced in 2003, in 2004 I introduced the CLOG tag. In about 2008 we upgraded from PHPBB2 to PHPBB3 and I think that any use of the clog tag (or any custom PHPBB forum tags) prior to that point reverted to plain text and we just "took the hit" on that. If the post is edited and re-submitted, it will create the tag again and so display a working link/hover over.

If there are any specific ones you think would be useful to do, as an admin (or mod) I can edit and re-submit to make them work.

However, this is not a bug per se, moving to resolved.

Jools

[bekateen]

Hi Jools,

Thanks for the explanation. I find it interesting that a software program can read a string of letters and interpret it as a link in one situation, and yet the same software isn't smart enough to read the exact same string properly in a different situation.

This was the only example I've seen so far and IMHO, no, it's not critical for you or anyone to go back and fix manually - not that this is a trivial post, but the problem is just a nonfunctional CLOG. Please don't spend your time going back to alter this.

Cheers, Eric

P.S., Thanks for the compliment. I do write routines in one language called LabVIEW, and as a teacher I offer an introductory (VERY introductory) course to bioengineering students where they learn to write "virtual instruments." In order to be able to grade their work, yeah, I have to do some software testing.

[Jools]

Hi Eric,

Good question.

The software interprets the clog tag at the time it is submitted to the database. This is good design as a typical post will only be written to the database once (or sometimes a few times if edited) but can be read hundreds or thousands of times especially if we include years old posts and bear search 'bots in mind. If the "interpretation" was done at "read time", it would take up much more CPU / be slower to load etc.

I looked at writing a one off routine to convert all the old ones, back in the day but it was too hard and only of limited benefit.

Jools

[bekateen]

Hi Jools,

I don't know if what I'm about to mention is the same as what we've been talking about in this thread, or if this next observation is more HTML-ized code, or something different, but for the last two days I've been browsing many of the Cat-eLog entries (trying to play the "What am I?" game), and I'm noticing that several of the Cat-eLog data sheets (e.g.,

Pterygoplichthys ambrosettii

) have HTML-style font coding revealed when it shouldn't be: For example, "<em>Pterygoplichthys</em>" in the Identification section of the data sheet.

Personally, I wouldn't worry much about weird code problems in older forum entries, but I would imagine that we want to avoid these problems in Cat-eLog entries, which (hopefully) are accessed frequently by users and guests alike.

Cheers, Eric

EDIT: A Google search of PlanetCatfish.com finds about 4,000 instances of visible <em> codes, although not all are in Cat-eLog data sheets (some are in Articles and others are in the forums).

ANOTHER EDIT: This is odd. The Google search link above yields either 3,600 or 4,100 results, depending on whether I open the link in MS Internet Explorer or Google Chrome... Those giants - they are such petty children.

[Jools]

Hi Eric,

This is indeed an issue, the

Pterygoplichthys

being a good example. @Mats might want to join in on this one, but I will look at it when I have a few hours - it's not a simple thing - and I need to fix the root cause before I track down the major occurrences of the issue and sort them.

Cheers,

Jools

[Edit: Fix misspelled genus name --Mats]

[Jools]

As per this thread, I think we've got the second issue discussed here under control.

Jools