17:35:04 <RRSAgent> RRSAgent has joined #exi
17:35:04 <RRSAgent> logging to http://www.w3.org/2016/10/25-exi-irc
17:35:06 <trackbot> RRSAgent, make logs public
17:35:06 <Zakim> Zakim has joined #exi
17:35:08 <trackbot> Zakim, this will be EXIWG
17:35:08 <Zakim> ok, trackbot
17:35:09 <trackbot> Meeting: Efficient XML Interchange Working Group Teleconference
17:35:09 <trackbot> Date: 25 October 2016
17:35:15 <taki> zakim, this is exi
17:35:15 <Zakim> got it, taki
17:41:52 <dape> dape has joined #exi
17:44:17 <taki> TOPIC: EXI Renaming
17:44:56 <taki> TK: Most people agreed to our plan.
17:45:26 <taki> DP: Some people think EXI should not mean anything.
17:45:41 <taki> DP: Following practices of AT&T, IBM, etc.
17:46:51 <taki> DP: I personally would like to give it a name.
17:48:40 <taki> TK: I agree with you. We want to give "EXI" a meaning.
17:49:02 <taki> CB: John Cowan is advocating for acronym.
17:49:17 <taki> CB: There was not a pushback for the new names.
17:49:51 <taki> CB: John Cowan is saying "Don't change the acronym." This is my understanding.
17:51:05 <taki> TK: Let's revisit this topic after two weeks to see if there are any pushback.
17:51:36 <taki> TOPIC: Canonical EXI
17:54:00 <taki> DP: I updated SOTD section.
17:54:08 <taki> DP: I also added CHANGES section.
17:55:17 <taki> TK: I drafted a publication request.
17:55:40 <taki> DP: I looked at it. It looks ok, did not find anything confising.
17:56:01 <taki> CB: It looks good to go.
18:00:59 <taki> TOPIC: EXI4CSS
18:02:28 <taki> CB: It means W3C blog.
18:02:41 <taki> DP: Who can write a blog entry?
18:02:45 <caribou> https://www.w3.org/blog/
18:02:48 <taki> CB: WG can write it.
18:03:14 <taki> CB: Some of them are written by WG chairs.
18:03:22 <taki> CB: Still open to WG people.
18:03:55 <taki> DP: What's the easiest way to publish here?
18:05:27 <taki> DP: I could write a blog entry.
18:06:32 <taki> DP: I would like to write a short blog with link to the slides.
18:07:37 <taki> DP: In relation to JavaScript, we noticed that we never sent out information about CSS.
18:07:52 <n8s> n8s has joined #exi
18:10:16 <taki> DP: I will make a draft and send the link to the group.
18:10:47 <taki> TOPIC: JavaScript support
18:11:05 <taki> https://lists.w3.org/Archives/Member/member-exi-wg/2016Oct/0042.html
18:12:02 <taki> DP: I created a github project for it where I put a schema, and created a Java implementation.
18:12:34 <taki> DP: It is not still fully conformant. There are still several missing constructs, which I will fix in coming days.
18:13:07 <taki> DP: The result is promising. Without using GZIP, the size is as small as original JavaScript.
18:13:32 <taki> DP: With minifier, it is 60 or 70%.
18:13:44 <taki> DP: I created C implementation.
18:14:05 <taki> DP: It is not yet tuned for performance, but can be used for experimentation.
18:14:32 <taki> DP: E.g. Browser can use it to see how fast JavaScript loading is gonna be.
18:17:58 <taki> DP: For example, angular2.js 1182582 bytes becomes 388994 in EXI4JS.
18:18:28 <taki> DP: It is not just about size. It is also about processing.
18:19:50 <taki> DP: We might also want to take a look at XML schema.
18:20:54 <taki> DP: I still keep "array" to make it easy to reconstruct JSON array.
18:22:16 <taki> DP: I looked at Mozilla SpiderMonkey API.
18:22:36 <taki> DP: I took this definition to create schema.
18:22:45 <dape> https://developer.mozilla.org/en-US/docs/Mozilla/Projects/SpiderMonkey/Parser_API
18:26:32 <dape> https://github.com/EXIficient/exificient-for-javascript/blob/master/src/main/java/com/siemens/ct/exi/javascript/JStoEXI.java
18:29:19 <n8s> n8s has joined #exi
18:29:27 <taki> TOPIC: Extended String
18:31:28 <taki> DP: With a little element name replacement, such as annotation to schema, and adding type element, xerces could process example 2 easily.
18:32:19 <taki> DP: annotation to schema, appinfo to type. Slightly changed snippet, EXI could process.
18:33:22 <taki> DP: Example 2 is more verbose than example 1. Example 1, you need to yourself parse strings out of it.
18:33:34 <taki> DP: I would tend to go with example 2.
18:36:46 <taki> DP: The longer the string list is, the bits it takes becomes longer, and needs more time to process.
18:37:49 <taki> DP: We also need to think about derived types.
18:40:45 <taki> DP: EXI enumerated values are not sorted.
18:41:15 <taki> DP: If sorted, binary search is useful.
18:44:22 <taki> TK: Any other topics?
18:44:36 <taki> TK: There is no EXI telecon next week.
18:45:49 <taki> rrsagent, create minutes
18:45:49 <RRSAgent> I have made the request to generate http://www.w3.org/2016/10/25-exi-minutes.html taki
18:54:23 <brutzman> brutzman has joined #exi
19:54:38 <n8s> n8s has joined #exi
19:56:17 <liam> liam has joined #exi
19:57:02 <n8s> n8s has joined #exi
20:02:20 <n8s> n8s has joined #exi
20:03:41 <Zakim> Zakim has left #exi
20:36:28 <n8s> n8s has joined #exi
21:00:10 <liam_> liam_ has joined #exi
21:54:30 <n8s> n8s has joined #exi
23:00:28 <n8s> n8s has joined #exi
23:04:51 <taki> taki has left #exi
23:17:05 <n8s> n8s has joined #exi
23:39:04 <n8s> n8s has joined #exi