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

In light of the hsif, esuom, racket2, and llabwons, my is getting lonely. A would be most appreciated.

Thank you for your consideration!
-FTF

Much as I'm sure this would be much easier to make than some of the other suggestions, the CommunityArtists will still have to say "no" to you.

The smiley ideas list is long enough, and the CAs have decided not to take any more requests for it. Thanks for suggesting, though!

~

Hi Owl. It's not really the CAs saying no - it's the staff.

OK... that's odd. Surely the CAs should be allowed to do as much work as they want......?

However trivial they are to create, smileys still have to be uploaded to the right BBC servers/database and the code added to recognise new ones.

Although this might not be a BBC concern, the <./>Smileys</.> page is very slow to display because of the huge number of graphics. It may become necessary to divide it into sections if more are added. The digibox users apparently have some problems with it at the moment but I don't know the precise cause. Using the A-number is said to work better and clicking on a smiley in a conversation is said to be fatal. So I would guess it is some combination of memory required for the page and the redirection of links to there.

Ahm... OK.

The code I'm just getting round to writing for my new website (which is a lot like h2g2 in a sense) just detects new smileys and adds them automatically when they're put in the smileys/ directory, based on the fact that their <SMILEY NAME="blah"/> code and their are all consistent.

I guess I'd just assumed the same amount of effort from the BBC's side of things, given that they nuked the code and rewrote from scratch at one point...

I'll volunteer to do the uploading/page-splitting etc if it means more smileys! (joke)

~

The naming isn't entirely consistent because some smileys have the original text shortcuts too. For conversations that may be parsed and substituted early on but page text is supposed to be stored as is - requiring substitution each time (not applicable to GuideML of course). There is also the business of the 2 skin types. All h2g2 skins use one of 2 directories (at least they are presented as if they are directories) and other sites sometimes have their own versions.

Again, these are all shortcomings of C++, I'm afraid. A language like Perl or Python was designed with all these string-manipulation things in mind (my code is written in Python) but, of course, they come at a cost of performance - after all, h2g2 used to be written in perl, and _someone_ made the decision to change it. I found that replicating the forum text-processing engine here in python was very straightforward, thanks to regular expressions and hashtables being easily accessible from the language. Ah well, perhaps the BBC need more techies working on the site so that random bits and pieces like this can be added (hint, hint). *cough*

I've worked on big projects before, and I know _just_ how hard it can be, so I'll shut up now.

We need to update the page anyway, to add the Christmas smileys into their proper sections. I'll talk to Abi about how we can make the page a little bit more accessible.

Is there any chance of at least getting the fixed versions of smileys which had faulty transparencies uploaded? That doesn't involve any code changes - it should be just some file copying somewhere.

Yes, we'll look into it. If it's simple we'll do it at the same time.

OK. The files are already on Yahoo but they could go in the usual place on the BBC server if that turns out to be easier for transferring to the final destination.

I've got a couple of directories on a shared drive that I put them into. Then I simply send an email with the details and sit back and wait. No hard work in smileys for me, unlike blobbing graphics.

*Sigh*
Oh well, my has found a freind, .

So long and thanks for all the responses!
-FTF