100% correct capitalisation is likely to be unattainable, since the OS data stores all the street names in upper case (*), so the version of the name you get to see has to be case-corrected by my script, which means it needs to rely on some simple rules which are always applied.

For this example, is it always true that -y- is followed by a capitalised name fragment, or at least true often enough such that adding this in as a blanket rule would then mean having to decapitalise fewer names than currently need capitalising?


* annoyingly so, since all the other text data is stored in mixed case...

At long last, an update... 1.3 has been a while in the coming, not because there's anything earth shatteringly amazing contained in the update, but simply because I was waiting for my hosting company to finalise the setup of my secure server... Now this has been done, I'm happy to announce that 1.3 now grabs the OS data from a https: link rather than the http: one used previously, which means Chrome should no longer gripe about having to load insecure content into a supposedly secure page.

Also for ditchi and anyone else editing in Wales, I've tweaked the case-correction function so that it (hopefully) knows what to do when encountering a -y- or -yr-...

http://userscripts.org/scripts/show/137249

One of those absurdly simple "why didn't I think of that before" ideas now makes v1.4 more likely to return the correct name at the top of the search results list...

http://userscripts.org/scripts/show/137249

As with Timbones' highlighting script, the advanced mode sometimes doesn't get enabled if WME times out during its page load sequence, requiring you to reload the page manually. I'll see if I can come up with a workaround for this, because given how wonderfully stable WME tends to be, it's a problem that isn't going to go away any time soon...

dknight212 wrote:Thanks, my productivity has just increased three fold!

Back to the top of the weekly scoreboard for you then

Timbones wrote: There's an event in loginManager you can register an event with. See my latest version for how to do it.

via mobile

Ooh cool, I'll have a look at that tonight, thanks!

Oh great, they've gone and moved everything around so it now looks not all that different to the original UI - some of us have probably been thinking it would have been a good idea if they'd laid out the new UI this way round all along instead of flipping it arse about face, so I guess in the grand scheme of things this is a good move on the part of the developers. Shame they had to tease us for so long with the mirror-image layout before having this moment of clarity

Now fully operational again in WME beta - I've moved the external site links into a section just below the OS results box, which now opens on clicking the [open] link rather than on mouseover. Also tweaked the advanced mode initialisation, so (fingers crossed) there shouldn't be any need to reload WME. Finally, now shows a "Loading new OS data" message whilst waiting for any uncached data to be pulled down from the server.

http://userscripts.org/scripts/show/137249

Just uploaded 1.5.1 which should fix this issue.


I disabled the results mouseover because, with the repositioned external site links in the beta, it was too easy to accidentally swipe across the results box as you moved the pointer over to the links. To be honest, even prior to this change, I was considering disabling it, because there would be times in the public and older beta layouts when the minimised results box could end up being pushed under the mouse pointer due to WME shuffling its own interface contents around during a segment edit, causing it to re-open unexpectedly.

As much as I like the idea of minimising the number of mouse clicks required to name segments, I equally dislike having to cancel unwanted/unintended UI state changes, and going back to the [show] link seems to push the balance of clicks back towards "like".

Gnaaahhhhh, they've only just released the latest beta version and it's gone public already??? I know the temptation to release the latest shiny shiny is quite strong, but you'd think they would at least give the beta test team a chance to break it before deciding it's good enough for public consumption

Oh well, another night of script debugging beckons, it seems...