01:46
<Lachy>
I'm trying to look up and book a train ticket on BritRail, for when I'll be in London over New Years. Does anyone here have any idea what the difference is between "Ticket Only", "Ticket & Reservation", and "Reservation Only"?
01:48
<gavin>
what are the differences in price between them?
01:49
gavin
predicts Ticket Only < Reservation Only < Ticket and Reservation
01:55
<Lachy>
it seems the reservation options let me choose which date and time I want to travel. But then I don't understand what it would mean to buy a ticket without a reservation
01:57
<Lachy>
they don't appear to have the time I was looking for though. My sister said she got a train on 2nd Jan deptarting 09:00/arriving 11:16, but there isn't one listed at that time.
01:58
<svl>
reservation is probably for a specific seat, with not all trains allowing seat reservation. (that's a guess; no idea if the UK even has trains with seat reservation)
02:01
<svl>
http://www.seat61.com/UKtravel.htm#Buying%20UK%20train%20tickets%20if%20youre%20from%20outside%20the%20UK might help
02:04
<Lachy>
oh, that says I can most likely just buy a ticket at the station on the day. That might be a lot easier
02:07
<Lachy>
oh, I have an oyster card from when I was in London a few months ago. Does that work on long distance trains as well, or just the suburban trains?
02:10
<svl>
oyster is london only.
02:11
<Lachy>
ok
02:11
<svl>
And yeah, unless it's a special fast long distance train, buying a ticket in advance definitely won't be necessary.
02:12
<svl>
(Just might work out cheaper.)
02:13
<Lachy>
I don't think it's a special one. I'm just travelling from London to Chester
08:29
BenMillard
is confused by http://lists.w3.org/Archives/Public/public-html/2008Dec/0088.html
08:34
<Lachy>
BenMillard, yeah, I think Joshue is just confused
08:35
<BenMillard>
Lachy, the stuff about HTTP is what threw me. :)
08:36
<Lachy>
indeed
08:36
<BenMillard>
I consider the off-screen model as being part of step 5
08:37
<BenMillard>
the way ATs interact with web pages is, as Josh describes, something of a patchwork at the moment due to inadequate APIs, bugs in applications, dynamic content and so forth
08:37
<BenMillard>
but the steps from 1-4 don't seem affected by that, afaict
08:38
<BenMillard>
Lachy, does the part Josh quoted from me make sense to you?
08:39
<Lachy>
This one isn't as clear as it could be "The UA stores them in the relevant accessibility API properties."
08:40
<Lachy>
The UA stores the associations in an implementation specific way. It doesn't really matter how. The important part is that it reports those associations to the AT through the accessibility APIs
08:44
<Lachy>
BenMillard,
08:44
<Lachy>
1. The UA runs through the <table> once using the algorithm to find and store all of the associations between cells.
08:44
<Lachy>
2. A user interacts with a cell via an AT.
08:44
<Lachy>
3. The AT obtains the associations for the cell by querying the UA via the accessibility APIs
08:45
<BenMillard>
Lachy, yeah that makes sense
08:45
<Lachy>
I cut your original step 2 and 5. Step 5 wasn't actually a step in the process, it's more of a footnote about how the AT can interact with the user
08:46
<Lachy>
step 4 would then be the AT reporting the relevant information to the user
08:48
<Lachy>
BenMillard, it could also work if the AT obtained access directly to the DOM and applied the algorithm itself to work out the associations
08:49
<BenMillard>
Lachy, my hope is that UAs will implement this since there are far fewer UAs than ATs, plus UAs have the resources and involvement in web standards to get it right. :)
08:49
<BenMillard>
but yes, ATs could do it themselves in the way you describe
08:49
<Lachy>
BenMillard, yeah, the other advantage is that its cheaper for a user to upgrade their browser than it is to upgrade their AT
08:50
<BenMillard>
Lachy, good point!
08:50
<BenMillard>
please send these thoughts to the thread, if you have the time :)
08:51
<Lachy>
I generally try to avoid contributing to accessibility related threads these days, given past experiences :-)
08:51
<Lachy>
but I suppose I could
08:52
<BenMillard>
well, your clarification that UAs can vary in how they store the information so long as ATs can query it, and that ATs can vary in how they store what they get, should help
08:53
<BenMillard>
and that the hope is for UAs to implement the association algorithm rather than ATs, for various reasons
12:56
<gsnedders>
Philip`: You are meeting me tomorrow.
13:57
<Philip`>
gsnedders: Oh, okay
13:57
<Philip`>
gsnedders: At what time?
13:59
<gsnedders>
Philip`: T19:00:00Z/PT5m
14:00
<gsnedders>
Philip`: If technology permits, I will remind you tomorrow too.
14:19
<Philip`>
gsnedders: Okay, I think I should be able to remember that now :-)
15:06
<gsnedders>
hmm…
15:06
<gsnedders>
Do I need to take my really thick warm anorak with me?
15:17
<Philip`>
It's not overwhelmingly warm here
15:17
<Philip`>
but I imagine you're used to the cold :-p
15:34
<neatnik>
gsnedders: did you ever work things out with that girl?
15:40
<gsnedders>
neatnik: The one I was tweeting about?
15:40
<neatnik>
gsnedders: not sure, the one we chatted about, oh, a year or so ago
15:40
neatnik
= ajnewbold
15:40
<gsnedders>
ah, that makes more sense
15:40
<gsnedders>
neatnik: pm?
15:40
<neatnik>
sure
15:41
Philip`
wonders if there are good places to get access to a Windows machine with decent RAM/CPU/disk, for compiling some software, for maybe an (unpredictably scheduled) half hour a day, for as little money as possible
15:41
<gsnedders>
I don't really want to get into discussions about this sort of stuff here (or in any other logged channel :P)
15:41
<gsnedders>
Philip`: Scrounge off friends?
15:41
<gsnedders>
Philip`: That normally works for me
15:41
<Philip`>
(The only thing I've heard of is Amazon EC2, which may or may not be suitable, but there ought to be better options)
15:41
<neatnik>
gsnedders: ah, gotcha; I never even think about logs
15:41
<neatnik>
surely that'll come back to haunt me someday
15:42
<gsnedders>
:)
15:42
<Philip`>
gsnedders: All my friends run Linux ;-)
15:42
<gsnedders>
Philip`: I'm not your friend!? :'(
15:43
<Philip`>
gsnedders: and I need something fairly reliable and long-term, which probably wouldn't be possible if I was borrowing spare resources off people
15:43
<gsnedders>
Philip`: ah. I don't run Windows anyway.
15:43
<Philip`>
gsnedders: s/run Linux/don't run Windows/, then :-p
15:44
<gsnedders>
Philip`: :)
15:44
<Philip`>
I used to just run a machine in my parent's house, but that's no so good in terms of electricity cost or noise or bandwidth
15:44
<Philip`>
s/no/not/
19:58
<dave_levin__>
Does the html5 spec imply that on right click a mouseup event should be sent before the contexmenu event (http://www.whatwg.org/specs/web-apps/current-work/#context-menus)?
20:34
<yecril71>
The cost of traversing a tree is proportional to the number of nodes.
20:36
<yecril71>
Even if we strictly follow the prescription of ID maintenance, the cost of inserting an identifier is proportional to the depth of the inserted node.
20:36
<yecril71>
So it is usually much cheaper.
20:36
<yecril71>
The structure of the DOM tree is prescribed by the document,
20:37
<yecril71>
it cannot be balanced or unbalanced at the implementor�s will.
20:38
<yecril71>
There are use cases for efficient scoped identifiers;
20:38
<yecril71>
since the implementors must deal with them anyway, why not make the interface available for scripting?
20:50
<yecril71>
Inserting a node into an empty element is not so special.
20:50
<yecril71>
It is as bad as inserting an unsupported child node, which can happen to other elements as well.
20:51
<yecril71>
Why should an element with no text child nodes be illegal?
20:56
<Hixie>
dave_levin: depends on the platform (will be either after mousedown or after mouseup/click, probably)
21:17
<Lachy>
I found this awesome search plugin to fix my problem with Google's annoying redirection :-) http://mycroft.mozdev.org/search-engines.html?name=google+(no+country+redirect)
21:56
<dave_levin>
hixie: So the events around the right click (which results in contextmenu) are platform dependent and basically not defined in html5.
21:59
<dave_levin>
I ask this because webkit on Windows sends mouseup and on OSX, it doesn't. I'm trying to figure out if html5 helps me to make a case for which way to go (which I would then go and fix for OSX). (It does appear that if mouseup is sent that it should be sent before the contextmenu or else some websites don't work properly.)