| 00:00 | <Hixie> | we're up to 992 members |
| 00:13 | <Hixie> | holy batman, check out the amount of metadata on http://www.house.gov/ |
| 00:14 | Hixie | fails to find a page that anyone would ever want to do anything with that has this metadata |
| 00:14 | <Hixie> | if anyone reading these logs finds pages with this metadata where the metadata would actually be useful, please do let me know |
| 08:58 | jgraham | particularly likes how house.gov has a <meta name=" DC.relation.conformsTo" content=""> where the content attribute lists lots of specificaions including XHTML 1.1, even though it doesn't even pass the machine-checkable conformnce criteria for that specificaion |
| 09:11 | <Hixie> | jgraham: it doesn't even pass the machine-checkable conformnce criteria for the "DC.relation.conformsTo" metadata item |
| 09:12 | <Hixie> | jgraham: so technically it doesn't _actually_ claim to support those... |
| 09:12 | <Hixie> | s/support/conform to/ |
| 09:15 | <fearphage> | is this a place i can ask xhr questions or is there a better place? |
| 09:16 | <Philip`> | DC.relation.conformsTo is a refutation of the idea that there's no value in namespaces being URIs because you can look for the terms on Google instead to find the appropriate specification |
| 09:44 | <Philip`> | https://www.bethere.co.uk/ - "Please Note: Copy & paste is not permitted in this field" - why do they go to so much trouble to intentionally disable useful features? |
| 14:03 | <Philip`> | So, Opera now has Opera Turbo to optimise bandwidth usage on slow/expensive links |
| 14:03 | <Philip`> | The question is: How long will it take until the savings from Turbo exceed the bandwidth I have to use in order to download the latest version of Opera? |
| 16:40 | <gsnedders> | http://core.trac.wordpress.org/ticket/9689 — more yay |
| 19:44 | <eric_carlson> | Hixie: ping? |
| 19:58 | <eric_carlson> | Hixie: I am going to have to run so I will leave my question here for you to find later |
| 19:58 | <eric_carlson> | Hixie: the spec says that a media element that isn't in the DOM should not be collected until there is not more audio to play |
| 19:58 | <eric_carlson> | Hixie: and mentions paused(), ended(), and rate==0 as things to consider before allowing collection |
| 19:58 | <eric_carlson> | Hixie: it doesn't say what to do with a file that is slow to load |
| 19:58 | <eric_carlson> | Hixie: eg. 'new Audio("http://....")' may not be playable before the collector runs |
| 19:58 | <eric_carlson> | Hixie: so I think the spec should mention this case, but I am not sure what the guidance should be |
| 19:58 | <eric_carlson> | Hixie: maybe at least wait until readyState >= HAVE_METADATA? |
| 19:59 | <eric_carlson> | Hixie: I will look for you later to chat |
| 20:39 | <gsnedders> | What does & do before a variable name in a function definition in C++? |
| 20:40 | <Philip`> | Makes it pass-by-reference instead of pass-by-value |
| 20:41 | <gsnedders> | What's the difference between pass by reference and passing by pointer? |
| 20:41 | <Hixie> | it's an operator that returns a pointer with the address of the variable, right? |
| 20:41 | <Hixie> | the address of its operand, i should say |
| 20:41 | <gsnedders> | (Which is what * does, right?) |
| 20:41 | <Philip`> | i.e. the argument is not (as it would otherwise be) a copy of the parameter's value; it's the same variable |
| 20:42 | <Hixie> | oh in a function definition |
| 20:42 | <Hixie> | never mind |
| 20:42 | Hixie | shuts up |
| 20:42 | <Philip`> | gsnedders: Syntax |
| 20:42 | <gsnedders> | Philip`: Then why does one function definition in moz use both? :P |
| 20:42 | <Philip`> | The implementation of "void f(T& x)" and "void f(T* x)" is identical; the only difference is that in the first case you can refer to the value as "x", in the second case you have to use "*x" |
| 20:43 | <Philip`> | gsnedders: Probably a combination of conventions |
| 20:44 | <Philip`> | e.g. it's sometimes conventional to pass output arguments by pointer (so it's explicit) |
| 20:44 | <Philip`> | and it's usually conventional to pass complex data types by reference (since it's more efficient than passing by value, and you don't have to worry about all the * syntax) |
| 20:46 | <gsnedders> | Parsing of media types is crazy. |
| 20:49 | <Philip`> | I suggest building a lookup table for all media type strings |
| 20:54 | gsnedders | doesn't quite understand what Gecko does |
| 21:22 | <gsnedders> | Hmm, there seems to be nothing to stop a media type containing multiple forward chars |
| 23:58 | gsnedders | squints |
| 23:58 | <gsnedders> | A tourist page about Switzerland referencing Lac Léman, and not Lac de Genevè |
| 23:58 | <gsnedders> | * Genève |