| 00:00 | <dglazkov> | Hixie: queue a task, but what's the task source |
| 00:01 | <dglazkov> | dom manipulation seems close |
| 00:06 | <Hixie> | dglazkov: the task source is a different issue than deciding when it fires |
| 00:06 | <Hixie> | dglazkov: can you tell me more about these callbacks? |
| 00:09 | <dglazkov> | they are creation callbacks. When HTML parser runs and finds a custom element, it creates the instance, but the element's initialization code is deferred until the parser finishes -- with the use of a callback. These callbacks are part of the element definition. Once parser finishes, we invoke all the callbacks. |
| 00:11 | <dglazkov> | typical uses for these callbacks is wiring up events in the shadow tree, starting xhr requests, etc. |
| 00:11 | <Hixie> | dglazkov: you probably want an algorithm that asynchronously waits for the document to be "Ready for post-load tasks" and then for each callback "queue a task" to invoke the callback. The DOM manipulation task source is probably the right one for hte tasks, yeah. |
| 00:12 | <Hixie> | dglazkov: remember though, some documents never finish loading, so waiting for load is not something i'd always recommend |
| 00:12 | <dglazkov> | yeah. also, won't this fail for foo.innerHTML = 'markup goes here';? |
| 00:12 | <annevk> | Hixie: cross-origin image with CORS because readable right? |
| 00:15 | <Hixie> | annevk: probably, just the spec, the ImageBitmap stuff is live |
| 00:15 | <Hixie> | s/just/check/ |
| 00:15 | <Hixie> | dglazkov: well if the document is already "ready for post-load tasks" you'd just queue a task, not wait first, is that what you mean? |
| 00:17 | <dglazkov> | oh i see. I think I've got it. let me read up on "ready for post-load tasks" and see how it fits. |
| 00:18 | <Hixie> | k |
| 05:12 | <GPHemsley> | Wow. admin@wiki is already getting spam. |
| 06:40 | <ruby_on_tails> | hello |
| 06:41 | <Hixie> | hi |
| 06:43 | <ruby_on_tails> | anyone alive enough to want to try out a new app i am making ? |
| 06:51 | <cgcardona> | ruby_on_tails: whacha got? |
| 06:51 | <ruby_on_tails> | cgcardona: http://thecodeplayer.com/play/add |
| 06:53 | <cgcardona> | so i type code and then it plays it back character by character ? |
| 06:53 | <cgcardona> | ruby_on_tails: ^ |
| 06:53 | <ruby_on_tails> | cgcardona: yes |
| 06:53 | <cgcardona> | wow that's really cool |
| 06:53 | <ruby_on_tails> | along with output, you can try some walkthroughs already made on the home page |
| 06:54 | <cgcardona> | this is really cool. can I spread this url? |
| 06:55 | <ruby_on_tails> | you can spread the home page, i might close down the editor after weekend for fixing things if issues come up in testing |
| 06:56 | <cgcardona> | sure. this really is great. nice work |
| 06:56 | <cgcardona> | i'm watching http://thecodeplayer.com/walkthrough/make-a-particle-system-in-html5-canvas |
| 06:56 | <ruby_on_tails> | cool |
| 07:58 | <SimonSapin> | TabAtkins: ping |
| 08:45 | <matjas> | wooooo http://trac.webkit.org/changeset/133966 |
| 08:58 | <annevk> | that code player thing is nify |
| 08:58 | <annevk> | nifty |
| 08:59 | <annevk> | matjas: aaahh "requited" :) |
| 09:00 | <annevk> | matjas: have you discussed that spec on es-discuss btw? I wonder what TC39 thinks |
| 09:03 | <annevk> | matjas: also if you want and it's unlikely this spec will be made redundant somehow by TC39, you can have whatwg/javascript on GitHub and javascript.spec.whatwg.org |
| 09:17 | <ruby_on_tails> | annevk: yea it is :D |
| 09:48 | <annevk> | SVG path syntax is simpler than I thought |
| 10:01 | <annevk> | and now Fullscreen API has a better logo http://fullscreen.spec.whatwg.org/ |
| 10:09 | <SimonSapin> | annevk: is there a unicode combining character for this? ;) |
| 10:09 | <annevk> | heh, I doubt it |
| 10:26 | <annevk> | http://url.spec.whatwg.org/ has one too now |
| 10:27 | <SimonSapin> | nice |
| 10:27 | <annevk> | if anyone has ideas for the others, https://www.w3.org/Bugs/Public/show_bug.cgi?id=19795 |
| 10:28 | <annevk> | dom / xhr / notifications / fetch (CORS) are still lacking |
| 10:29 | <SimonSapin> | annevk: how about ↓ for fetch? |
| 10:29 | <annevk> | sweet :) |
| 10:29 | <annevk> | now how to create a good path for that one... |
| 10:49 | <annevk> | SimonSapin: http://fetch.spec.whatwg.org/ |
| 10:52 | <annevk> | oh yes |
| 10:52 | <annevk> | Notifications API can be a "!" |
| 11:00 | <annevk> | http://notifications.spec.whatwg.org/ |
| 11:45 | <karlcow> | sniffing hmm could be 👃 or 💩 if emoji were more common |
| 11:46 | <Velmont> | I only see the encoding standard icon for both of those, karlcow :P |
| 11:47 | <karlcow> | yes it's why I said if emojis were more common. |
| 11:47 | <Velmont> | But most of the icons are pictures anyways(?), so it can work I guess. |
| 11:50 | <karlcow> | (⸪) |
| 11:52 | <karlcow> | ⌘ ← place of interest sign (found often in Norway) |
| 11:52 | <karlcow> | which could be a bit like parsing. |
| 12:27 | <annevk> | SVG path syntax is interesting, but it's kinda annoying there's no processing model |
| 12:28 | <annevk> | E.g. you can omit the final "z" in implementations, but whether that's conforming does not seem to be defined in http://www.w3.org/TR/SVG/paths.html |
| 12:59 | <annevk> | Hixie: in "If image is an ImageData object" you don't have the explicit "return" step, although I'm not sure that step is at all necessary in all cases |
| 13:00 | <annevk> | Hixie: also again, I really think we should have a better solution than putting this on the global object |
| 13:01 | <annevk> | Hixie: I thought we decided static methods was the way forward when we added createObjectURL; seems kinda weird for you to then introduce new factory methods |
| 15:20 | <GPHemsley> | maybe a question mark in a file-like square? |
| 15:22 | <annevk> | for MIME Sniffing that might make sense |
| 15:23 | <annevk> | maybe for MIME in general too if we eventually tackle that whole subject |
| 15:23 | <annevk> | it's about resource labels after all |
| 15:24 | <annevk> | IDNA notes: http://wiki.whatwg.org/wiki/URL#IDNA |
| 15:26 | <GPHemsley> | annevk: BTW, why did you make me read that whole post about colons? |
| 15:26 | <annevk> | for the memes |
| 15:26 | <annevk> | how do I use the obsolete template on the wiki? |
| 15:27 | <GPHemsley> | {{obsolete|spec=<link to spec>|see=<link to related document>}} |
| 15:27 | <GPHemsley> | all parameters optional |
| 15:27 | <GPHemsley> | Perhaps I should write that down on the wiki |
| 15:29 | <annevk> | nah it's k |
| 15:29 | <GPHemsley> | annevk: Do you have a list of things that would go into "MIME in general"? |
| 15:30 | <annevk> | I guess mainly parsing a MIME type |
| 15:30 | <annevk> | not sure what else is there really, would have to take a look at those unmaintained RFCs |
| 15:34 | <annevk> | oh also, not sure everyone agrees with me that's a good plan |
| 15:36 | <GPHemsley> | annevk: BTW, if you find that the spec and see parameters aren't enough, let me know. |
| 15:36 | <GPHemsley> | annevk: Given how vilified MIME types seem to be, I don't know if it's necessary. |
| 15:36 | <annevk> | vilified? |
| 15:37 | <GPHemsley> | IDK, people seem to not like them. |
| 15:37 | <GPHemsley> | Or maybe that's just you. |
| 15:37 | <GPHemsley> | ;) |
| 15:39 | <annevk> | well, they're there, and are used in a number of our formats, and we should define how those work |
| 15:39 | <annevk> | e.g. <Script type="TEXT/javascript;bogus=tralal;™=test"> |
| 15:39 | <annevk> | what does that mean? |
| 15:40 | <annevk> | the wiki should have a category for pages such as http://wiki.whatwg.org/wiki/SVG_and_canvas |
| 15:40 | <gsnedders> | annevk: That's surely a quite simple case? |
| 15:41 | <annevk> | gsnedders: browsers were not interoperable on stuff like that, might still not be |
| 15:41 | <gsnedders> | I think the varience was in how unknown parameters were handled, not the parsing of them |
| 15:48 | <annevk> | gsnedders: well e.g. ™ does not match the production in the MIME RFCs, should it not apply? |
| 15:48 | <annevk> | I think they're very much related |
| 15:51 | <GPHemsley> | it could arguably nullify the entire value, then, no? |
| 15:51 | <GPHemsley> | bad parameter = no valid MIME type |
| 15:52 | <annevk> | sure, but I doubt that's what happens in practice |
| 15:52 | <annevk> | there's similar things |
| 15:53 | <annevk> | e.g. "Content-Type: text/html;" is invalid |
| 15:53 | <annevk> | but works perfectly fine |
| 15:53 | <annevk> | MIME RFCs don't tell you that |
| 15:53 | <GPHemsley> | ah |
| 15:53 | <annevk> | it's been a while since I looked into this, but there's a lot of little gotchas like that |
| 15:54 | <GPHemsley> | annevk: BTW, if something is "in the spec now", it'd be useful to link to what spec and where :) |
| 15:54 | <annevk> | maybe next time :) |
| 15:54 | <annevk> | it's all HTML I think |
| 15:58 | <GPHemsley> | also, for the future (not sure if people still do this): only use level-2 headers and below; level-1 headers are reserved for the page title |
| 15:58 | <GPHemsley> | (it seems to be common on some of the old pages) |
| 16:04 | <hsivonen> | researched the history of polyglot and was reminded that Sam gaa been pushing this stuff since 2006 |
| 16:37 | <MikeSmith> | hsivonen: so it's been 6 years and you're still just not getting it |
| 16:37 | <MikeSmith> | obstinacy in the face of plain reason |
| 16:56 | <annevk> | http://whitepeoplemourningromney.tumblr.com/ :) |
| 17:11 | <matjas> | annevk: well I’ve been filing most of the new stuff in the JS spec as ECMAScript bugs, and the editor seems to listen to my arguments |
| 17:12 | <matjas> | annevk: the latest ES6 draft already include some of the changes, which is kinda cool |
| 17:13 | <matjas> | annevk: but i haven’t really asked on es-discuss |
| 17:13 | matjas | adds to TODO list |
| 17:13 | <matjas> | annevk: also, javascript.spec.whatwg.org + github repository sounds cool. where do i sign up? |
| 17:15 | <WeirdAl> | Hi, folks - are there any semi-standard API's on how to format whitespace inside a DOMDocument as we're serializing, when whitespace is insignificant? I'm thinking I'd like to serialize the document, but preserve the general formatting so svn blame and hg blame show only the relevant changes. |
| 17:15 | <WeirdAl> | s/blame/diff/ |
| 17:26 | <WeirdAl> | (I have to leave for a bit, but I'll be back in probably about an hour - I'll check IRC logs to see if there's an answer. Worst case, I try to write my own.) |
| 18:04 | <annevk> | WeirdAl: http://domparsing.spec.whatwg.org/ ? |
| 18:06 | <annevk> | matjas: I added you to whatwg on GitHub, feel free to create a repo |
| 18:06 | <annevk> | matjas: we'll have to wait for Hixie to set up a domain then I can set up syncing between the repo and the domain |
| 18:07 | <annevk> | gotta go |
| 18:21 | <AryehGregor> | paul_irish, Selection.toString() is a complete mess. There's like zero interop on any nontrivial case. WebKit has some algorithm that depends on the internal representation of text it uses for layout after CSS has been applied, as I understand it. I think Gecko uses a basically DOM-based view of things with hacks to account for display: none. Other browsers are in between. I tried writing a spec once, but it was far too much of a headach |
| 18:21 | <AryehGregor> | e and it didn't look like browsers wanted to converge. |
| 18:22 | <AryehGregor> | The basic problem is: a purely DOM-based view is thoroughly inadequate because it doesn't match user expectations at all. So you have to figure in CSS somehow if you want to be at all sane. But then there are massive numbers of details you could choose to account for, or not. |
| 18:22 | <paul_irish> | :( |
| 18:22 | <AryehGregor> | CSS is 2D, and you have to serialize it. |
| 18:22 | <paul_irish> | Yeah. |
| 18:22 | <AryehGregor> | You can't do anything sensible with, e.g., non-static positioning. |
| 18:22 | <AryehGregor> | Or flats. |
| 18:22 | <AryehGregor> | floats. |
| 18:23 | <AryehGregor> | I believe Gecko doesn't look at CSS at all except for display: none -- e.g., <p style="display:inline">foo</p>bar will still serialize as "foo\nbar" or such, not "foobar" as in some other browsers. |
| 18:23 | <AryehGregor> | This is the abortive spec I wrote: http://aryeh.name/spec/innertext/innertext.html |
| 18:23 | <AryehGregor> | (in the context of innerText, but it's the same idea) |
| 18:24 | <AryehGregor> | That was one of the first specs I wrote. |
| 18:24 | <AryehGregor> | (note date) |
| 18:28 | <cgcardona> | "introduced by Microsoft in Internet Explorer sometime in the mists of history" |
| 18:28 | <cgcardona> | nice |
| 18:38 | <Hixie> | what domain am i setting up? |
| 18:43 | <Ms2ger> | Hixie, polyglot.s.w.o |
| 18:43 | Ms2ger | runs |
| 18:43 | <Hixie> | o_O |
| 18:44 | <Ms2ger> | Hixie, javascript., actually |
| 18:44 | <Hixie> | matjas: we're taking over from ECMA now? :-P |
| 18:47 | <MikeSmith> | it's an "extension spec" |
| 18:47 | <Hixie> | what does it extend? |
| 18:47 | <MikeSmith> | the EcmaScript spec |
| 18:47 | <Hixie> | does it exist yet? i'm very confused |
| 18:47 | <MikeSmith> | it exists |
| 18:47 | <Hixie> | url? |
| 18:47 | <MikeSmith> | http://mathias.html5.org/specs/javascript/ |
| 18:48 | <Ms2ger> | http://mathias.html5.org/specs/javascript/ |
| 18:48 | <Ms2ger> | Damn you, MikeSmith! |
| 18:48 | <MikeSmith> | heh |
| 18:48 | <Ms2ger> | Shouldn't the sake slow you down? :( |
| 18:48 | <MikeSmith> | I'm drinking pink lemonade tonight |
| 18:48 | Ms2ger | raises an eyebrow |
| 18:49 | <Hixie> | dudes that spec is the awesome |
| 18:49 | <Hixie> | matjas++ |
| 18:49 | <MikeSmith> | さすがmatjas |
| 18:50 | <Hixie> | ok created subdomain |
| 18:50 | <SimonSapin> | About http://mathias.html5.org/specs/javascript/ , are greater and smaller implicitly "strict" in english? (ie: not greater or equal) |
| 18:51 | <Hixie> | the captcha i got was "selffula surrounds" which sounds like one of those phrases evil forces make their minions repeat to each other to spread their mythos |
| 18:51 | <Ms2ger> | Alright |
| 18:51 | <Ms2ger> | selffula surrounds |
| 18:51 | <Hixie> | selffula is coming! |
| 18:51 | <Hixie> | selffula surrounds! |
| 18:52 | Hixie | prostrates himself |
| 18:52 | <Ms2ger> | Get back to work, Great Master |
| 18:52 | <Hixie> | selffula is the great master! selffula surrounds! |
| 18:55 | <cgcardona> | interesting |
| 18:55 | <AryehGregor> | SimonSapin, that's the accepted convention outside of advanced math, in my experience. |
| 18:56 | <AryehGregor> | If you're a math major or grad student, at least in pure math, people often are sloppy about the difference between strict and loose inequality because it's often not important and it's extra info to track when doing a proof. |
| 18:57 | <AryehGregor> | (in particular because certain operations switch between strict and loose inequality and it can be annoying to keep track of them, and it makes no difference if your result doesn't care about sets of zero measure or whatever) |
| 18:57 | <SimonSapin> | AryehGregor: it’s ambiguous in French if you don’t specify otherwise |
| 18:57 | <AryehGregor> | Interesting. |
| 18:57 | <Hixie> | vraiment? |
| 18:57 | <AryehGregor> | I think in the context of the spec, it's unambiguous. |
| 18:57 | <SimonSapin> | At least that’s what I was taught in school. Many people don’t really care |
| 18:58 | <WeirdAl> | annevk: that was the first place I looked. |
| 18:58 | <Hixie> | what's the ambiguous phraseology? |
| 18:58 | <SimonSapin> | English might be different from French of course, it is for "positive". |
| 18:58 | <SimonSapin> | Hixie: greater than, smaller than |
| 18:58 | <Hixie> | i meanin french |
| 18:58 | <SimonSapin> | supérieur à, inférieur à |
| 18:59 | <SimonSapin> | C’est plus précis de dire : strictement supérieur à, supérieur ou égal à |
| 18:59 | <Hixie> | i wouldn't say that's ambiguous |
| 18:59 | WeirdAl | is thankful for the French 1A class he took this past summer :) |
| 18:59 | <SimonSapin> | people use it to mean either |
| 19:00 | <Ms2ger> | Hixie, yeah, but you speak French like a 10 years old boy :) |
| 19:00 | <Hixie> | in particular, neither http://fr.wiktionary.org/wiki/sup%C3%A9rieur nor http://en.wiktionary.org/wiki/sup%C3%A9rieur suggest it's ambiguous |
| 19:00 | <SimonSapin> | … but of course it doesn’t matter to define min and max |
| 19:01 | <Hixie> | "greater than" seems equally unambiguous to me. 3 is obviously not greater than 3 |
| 19:02 | <Hixie> | http://fr.wikipedia.org/wiki/Table_des_symboles_math%C3%A9matiques seems to consider the term unambiguous too, though they do use the more verbose "strictement" in some cases |
| 19:04 | <Hixie> | http://fr.wikipedia.org/wiki/Nombre_positif does agree though |
| 19:04 | <Hixie> | interesting |
| 19:04 | <Hixie> | weird |
| 19:05 | <SimonSapin> | My guess is that it *should* unambiguously mean "strictement supérieur", but sometimes people get sloppy and use the short form when they mean "supérieur ou égal". So in practice in can be ambiguous, so you sometimes use "strictement" to disambiguate. |
| 19:06 | <Hixie> | yeah, i guess so |
| 19:06 | <Hixie> | weird |
| 19:06 | <Hixie> | none of the dictionaries i found say it's ambiguous |
| 19:07 | <Hixie> | the only example of it being ambiguous i found was that "positive numbers" page and it disambiguated in parentheses which makes me skeptical that the usage is really correct in the firstplace |
| 19:08 | <SimonSapin> | In French schools we’re often advised to use "strictement …" or "… ou égal", especially in math class |
| 19:09 | <SimonSapin> | anyway, the good news is that both possibility gives the same definition of max(a, b) and min(a, b) |
| 19:09 | <Hixie> | assuming that if values a and b are equal they're also identical |
| 19:09 | <Hixie> | which isn't a given depending on what the values are |
| 19:10 | <SimonSapin> | http://mathias.html5.org/specs/javascript/ defines min() and max() for use in the document but I don’t see it used at all |
| 19:12 | <SimonSapin> | same for ToInteger() |
| 19:13 | <SimonSapin> | where should I send feedback? |
| 19:13 | <Hixie> | MikeSmith: is there a bug component for this spec? |
| 19:14 | <MikeSmith> | no |
| 19:14 | <MikeSmith> | I'll make one now |
| 19:14 | <Hixie> | assuming matjas wants one |
| 19:19 | <GPHemsley> | I just discovered there's a file format where multiple sources claim its mimetype as either chemical/cif or chemical/x-cif |
| 19:19 | <GPHemsley> | (as if it's the subtype that's the problem) |
| 19:21 | <SimonSapin> | twitter shall be the bug tracker until then |
| 19:24 | <WeirdAl> | uh, that JS spec sounds like a Bad Idea - not something that is really a spec, but simply documentation of historical practices of browsers |
| 19:25 | <WeirdAl> | the defineGetter function and its friends have been obsoleted by ES5's Object.defineProperty |
| 19:25 | <WeirdAl> | and Object.getOwnProperty |
| 19:26 | <Ms2ger> | SimonSapin, those functions are defined in ES proper |
| 19:27 | <WeirdAl> | Ms2ger: are you talking to me instead of SimonSapin? :) |
| 19:27 | <Ms2ger> | WeirdAl, you mean "a spec for features a browser is required to implement to be able to browse the web as it actually exists"? |
| 19:27 | <SimonSapin> | Ms2ger: these are not JS functions but functions for the algorithms in the document: http://mathias.html5.org/specs/javascript/#abstract-operations |
| 19:27 | <Ms2ger> | WeirdAl, no, he said something about ToInteger() not being defined |
| 19:27 | <Ms2ger> | SimonSapin, exactly |
| 19:27 | <SimonSapin> | defined but not used |
| 19:28 | <Ms2ger> | SimonSapin, they are used in and defined by the ES spec |
| 19:28 | <WeirdAl> | Ms2ger: some of these problems, I'm saying, the new ES specs already handle |
| 19:28 | <SimonSapin> | Ms2ger: is there a need to repeat them in this spec? |
| 19:28 | <Ms2ger> | WeirdAl, I don't think you get what I'm saying |
| 19:29 | <gsnedders> | WeirdAl: Yes, there are ways to achieve that in ES5. But any browser that wishes to be compatible with the web must support the older, non-standard means. |
| 19:29 | <Ms2ger> | WeirdAl, the existence of Object.defineProperty doesn't magically make pages that use __defineGetter__ disappear |
| 19:30 | <WeirdAl> | must we be compatible with <blink> and <marquee>? :) |
| 19:30 | <WeirdAl> | (I know, I'm trolling with a line like that) |
| 19:30 | <Ms2ger> | Yeah, we must |
| 19:31 | <Ms2ger> | marquee in particular is required to render a lot of East-Asian sites correctly |
| 19:31 | gsnedders | wonders what effect IE9 has had on __define{G,S}etter__ |
| 19:31 | <WeirdAl> | hm, I should pick my fights more carefully, based on what I actually care about :p |
| 19:33 | <Hixie> | WeirdAl: the html spec defines <blink> (that's trivial) and <marquee> (in great depth, there's a whole section devoted to it) |
| 19:33 | <Hixie> | WeirdAl: so... yes |
| 19:33 | <Hixie> | (blink is just one line of CSS) |
| 19:34 | <Hixie> | WeirdAl: of course, i'm pretty sure we'd all be happier if ES did just spec all this stuff for us :-) |
| 19:34 | <SimonSapin> | Hixie: CSS makes conforming to just ignore 'text-decoration: blink' |
| 19:34 | <Hixie> | SimonSapin: indeed |
| 19:34 | <Hixie> | SimonSapin: that's why <blink> doesn't blink in all browsers |
| 19:35 | <Hixie> | well, at least, that's the formal reason why it's ok that it doesn't blink in all browsers |
| 19:35 | <SimonSapin> | I guess that’s fine |
| 19:35 | <SimonSapin> | but I can imagine how not supporting <marquee> would break stuff |
| 19:39 | WeirdAl | wonders idly where the whatwgmemes website is |
| 19:39 | <Hixie> | you mean separate from http://w3cmemes.tumblr.com/ ? |
| 19:39 | <Hixie> | cos that's what we use :-) |
| 19:39 | <WeirdAl> | yeah :p |
| 19:40 | <WeirdAl> | I need moar funny |
| 19:41 | <Ms2ger> | Sorry, we don't do crazy in here ;) |
| 19:41 | <Hixie> | o_O |
| 19:42 | <WeirdAl> | gotta go - my flight's boarding |
| 19:42 | <Ms2ger> | ttyl |
| 19:47 | gsnedders | would still like Web ECMAScript to define most of it in ES code. |
| 19:47 | <gsnedders> | Obviously parts can't be, but all the librar yparts can. |
| 19:48 | Ms2ger | wonders what Hixie is working on at the moment |
| 19:51 | <Hixie> | canvas in workers |
| 19:52 | <Ms2ger> | Oh, right |
| 19:59 | <Hixie> | do i still need to have spec prose to handle the opera-2d context? |
| 19:59 | <Hixie> | or can i change the spec so that it assumes that each canvas can only ever have one context at a time? |
| 20:01 | <Hixie> | i'm thinking i'm going to throw opera-2d under the bus |
| 20:06 | <Ms2ger> | sgtm |
| 23:31 | <Hixie> | heycam|away: any way i can have an enum but instead of throwing if the value is not a valid enum value, the method just returns null? |