IRC Conference/logs/2009-10-18

From Open Siddur Project Development Wiki

Conference log for the Oct 18, 2009 chat. See the summary for discussed topics. Questions or comments should be directed to the discussion list.

(11:15:30 AM) EfraimDF: OK -- why doesn't everyone introduce themselves and give a sentence or two about interest in the project (latter, optional)
(11:16:04 AM) EfraimDF: <-- Efraim Feinstein; I'm a developer/programmer and have my tentacles in just about every aspect.
(11:16:08 AM) realazthat is now known as Azriel
(11:16:37 AM) amg_ [i=183ca275@gateway/web/freenode/x-airraysmubthrcjx] entered the room.
(11:16:55 AM) EfraimDF: (hi amg_... doing introductions now)
(11:17:00 AM) Aharon: <-- Aharon Varady; founder of the Open Siddur, collaborator with Efraim. cheerleader and spokesmodel
(11:17:06 AM) Michael___: I am a professional reference publisher with an interest in gml/sgml/xml going back into the early 1990s.   I've been very interested personally in hebrew text processing / publishing issues, but my professional experience is with automotive and jet engine technial / engineering / control unit information processing and distribution.
(11:17:19 AM) Azadi: Ok... my name is Gella Solomon, I'm a first year rabbinical student at JTS, and an alumna of the Conservative Yeshiva in Jerusalem (was there for the past two years). My interest is... hrm, I guess in Liturgy in general. I'm here because of Aharon. He's pretty much made of awesome.
(11:17:37 AM) amg_: @EfraimDF Thanks -- did I miss much? If so, can you email be backchat at [redacted]? Thanks
(11:18:13 AM) EfraimDF: @amg Didn't miss much.  The chat logs will be on the website after the chat
(11:18:41 AM) Michael___: I see my last name didn't make it through, full name is michael bentley (i think I recognize Efraim from TBS cambridge (?)
(11:19:02 AM) EfraimDF: @Michael__ yes, that's likely 
(11:19:32 AM) Aharon: i think we'll edit email in our chat log to help avoid needless spamming :)
(11:19:37 AM) amg_: Hi, all. My name is Andrew Greene; professionally I'm a programmer focusing on page layout software (spent 11 years at Bitstream, am now in my second year at Adobe.) In that capacity I was a member of the XSLT working group for a few years.
(11:20:04 AM) EfraimDF: wow, looks like we have a lot of tech talent in the room today. :-)
(11:20:12 AM) Azriel: hi, my name is Azriel. I attend NYU Poly in Brooklyn, and I am an all around kind of guy. I administer the websites ( in my spare time ) and dabble in the XML encoding. My main focus before school started was the trancription web application
(11:20:17 AM) Azadi: I feel so lonely :P
(11:20:29 AM) Azriel: nice
(11:20:34 AM) amg_: I have been working on and off for about ten years on my own siddur, which I'm about to actually publish. And I'm starting to work on a siddur project for my children's school (The Jewish Community Day School in Watertown, MA, which describes itself as following "intentional pluralism")
(11:20:35 AM) Azadi: BROOKLYN IN DA HOUSE!
(11:20:42 AM) Azadi: Wut wut
(11:21:18 AM) EfraimDF: The original agenda focused on a few things: website friendliness, dev priorities, skill sets required for development, and building your own dev environment
(11:21:21 AM) amg_: [And since I see a mention above of TBS in Cambridge, I'll try adding to my street cred by mentioning that I was a member there in the late '80's/early '90's :-) ]
(11:21:33 AM) Aharon: azadi, you have technical experience in the siddur itself. also, capability research in primary sources
(11:21:50 AM) Azadi: Yay! I'm useful!
(11:22:13 AM) EfraimDF: Is anyone not familiar with how the project is structured?
(11:22:48 AM) Azadi: I have an idea, but could use a bit more of an idea
(11:22:49 AM) Michael___: I'm a beginner, but happy to read background materials on my own time.  have been to the wiki,
(11:23:04 AM) amg_: Glancingly familiar, but not enough to explain it :-)
(11:23:05 AM) Azadi: (ditto michael)
(11:23:28 AM) EfraimDF: There are effectively two streams: there's the text gathering and generating stream and the programming/admin stream
(11:24:00 AM) EfraimDF: text gathering requires finding appropriate base texts, scanning them, transcribing and proofreading them
(11:25:22 AM) EfraimDF: the programming/admin stream involves setting up a website/database (we're using eXist+XQuery/XSLT as our backend now), a client side web app, an XML spec (based on the TEI), and processing tools to transform that into printable text (XSLT transforming to XHTML or PDF)
(11:25:54 AM) EfraimDF: We also have some "side interests" in developing and improving high quality free software fonts, although it's not a priority at the moment
(11:26:41 AM) EfraimDF: The eXist db (no pun intended) exists.  Development there is going to be more or less responsive to the needs of the client
(11:27:06 AM) EfraimDF: We have something of an outline of an XML spec which needs some reading over to make sure that I'm not the only person in the universe who can understand it or implement it
(11:27:42 AM) Aharon: the JLPTEI (based on the TEI, Text Encoding Intiative)
(11:27:43 AM) EfraimDF: I've started working on the XSLT transforms, and can certainly use some help in that.  
(11:27:59 AM) EfraimDF: All our code is available on svn (Google code hosting)
(11:28:10 AM) EfraimDF: Any basic questions?
(11:28:57 AM) EfraimDF: Also, on topic, has anyone attempted to use the wiki to glean information about the project, have comments on what's easy to find, what's hard to find, what should be easier, etc?
(11:29:05 AM) amg_: Not from me at this point. (A few advanced questions, but they can wait until the first round of work is done.)
(11:29:56 AM) Aharon: We could probably speak a little about the client (ie, web interface) if anyone is interested
(11:30:35 AM) Azriel: I was the main developer of the web client
(11:30:44 AM) Azriel: the current design is using GWT
(11:30:49 AM) Azriel: Google Web Toolkit
(11:31:00 AM) Azriel: which allows the developer to program in Java
(11:31:20 AM) Azriel: and supplies "widgets" that gets translated into XHTML/JS
(11:31:38 AM) Azriel: abstracting from JS nitty gritty
(11:31:45 AM) Azriel: however
(11:32:18 AM) Azriel: we are considering moving to an entirely JS framework solution simply because our developer pool for JS is greater
(11:32:22 AM) EfraimDF: OT: Might it be helpful to split the chat into technical (prefix with @tech) and nontech (prefix with @nt ) discussion streams?
(11:33:02 AM) Azriel: @tech the main point of the application, is to allow transcription of our images into text
(11:33:33 AM) Aharon: EfraimDF: I think that is wise.
(11:33:55 AM) EfraimDF: @tech where images = scans of public domain or contributed siddurim
(11:34:30 AM) Azriel: @tech eventually, because of the infrastructure that the web application has to interface with the eXist db ( its in xquery ), it can be used for many other applications in manipulating the XML
(11:34:31 AM) EfraimDF: @tech @nt As of now, we have a fully scanned copy of Baer's Seder Avodat Yisrael, we're looking for other texts to use as "base texts" to build off
(11:35:23 AM) EfraimDF: @tech @nt I've been keeping a "wanted" list on the wiki, which needs research to add to, especially for non-Ashkenaz nusah
(11:37:00 AM) Azadi: @nt ok, how are we dealing with copyright issues? i.e. i could easily scan in such things as Sim Shalom and Artscroll, but fairly certain it would be problematic.
(11:37:26 AM) Azriel: good question
(11:37:39 AM) Aharon: @nt right, we are not scanning copyrighted works
(11:37:55 AM) EfraimDF: @tech How interested are you in becoming more involved in the project?  What do you think you would need to know?
(11:38:12 AM) Aharon: @nt nor are we transcribing them. our goals cannot rely on fair use. we need content that we are free to distribute
(11:38:20 AM) Azriel: @nt we usually take in pre-1923 scans
(11:38:37 AM) EfraimDF: @nt With the exception that we can accept copyrighted works *if* the author is willing to contribute to us under a free license.
(11:39:29 AM) Aharon: @nt for example, Reb Zalman Schachter-Shalomi contributed his weekday siddur and sabbath supplement
(11:39:55 AM) Azadi: ok, but certainly the prayer text itself can be transcribed from a copyrighted text so long as it's not original material, yes? As long as the layout isn't taken, etc.
(11:40:24 AM) EfraimDF: @tech Some of the issues we have are related to the textual content (eg, balancing between repeating content and using referential markup, conditionally-present text)
(11:41:02 AM) EfraimDF: @tech Some of them to XML limitations (the multiple overlapping hierarchy problem)
(11:41:34 AM) EfraimDF: @nt We'd prefer to be able to cite that the text is exactly as it appeared in some public domain edition
(11:41:51 AM) Aharon: @nt the problem with hebrew/aramaic "source" text in artscroll/koren and other contemporary siddurim is that they are actually aggregations of earlier historical siddurim, albeit without inculding their sources or what changes they made
(11:42:25 AM) amg_: @tech Are we planning to keep the corpus from each source distinct or merge them? If the latter, how do we plan to handle minor variants within the text?
(11:42:31 AM) EfraimDF: @tech The next major large components are the web interface and the transforms to get this stuff on a page
(11:42:33 AM) Michael___: there may be another track (@law?) this becomes a long conversation
(11:42:45 AM) Azriel: heh
(11:42:46 AM) Azadi: @nt hm. good point.
(11:42:53 AM) Azadi: lol
(11:43:01 AM) Aharon: @nt we have an opportunity with the Open Siddur to present nuchaot as they appeared in historical siddurim with attribution, indicating any variations in other editions, other historical siddurim, and other nuschaot
(11:43:04 AM) EfraimDF: @nt Fortunately, we have archives of ML discussions about this which can be referenced
(11:43:40 AM) EfraimDF: @tech The sources start distinct in transcription stage, then get merged in encoding
(11:44:01 AM) EfraimDF: @tech XML markup is used to differentiate what came from where.
(11:44:09 AM) Azadi: @nt ok, I actually have to split for now... I look forward to reading the rest of the conversation later, and to being a resource for source material (Library access! Bwahahahahahaha!)
(11:44:38 AM) Aharon: @nt thanks azadi/gella! shavuah tov!
(11:44:41 AM) EfraimDF: @nt If you haven't already, please join the mailing list.
(11:44:48 AM) amg_: @tech - ah, and that's where the overlapping hierarchies bite us
(11:44:55 AM) Azadi: Aharon knows how to contact me, and I'm on Facebook.
(11:45:03 AM) Azadi: Laterz, Chodesh Tov!
(11:45:12 AM) EfraimDF: @nt shavua tov/chodesh tov
(11:45:16 AM) Azadi left the room (quit: "Page closed").
(11:45:28 AM) EfraimDF: @tech Not just there.  Sometimes, the same text has overlapping hierarchies
(11:46:03 AM) EfraimDF: @tech The system is designed so that textual sources can be mixed in a final product (a siddur) to a maximal extent
(11:46:31 AM) EfraimDF: @tech Not necessarily to have the optimal way to recover an individual source text (although that is possible too)
(11:46:57 AM) EfraimDF: @tech and so a transcription mistake in one place can be corrected once.
(11:47:15 AM) amg_: @tech Makes sense.
(11:47:37 AM) Azriel: @tech I like to think our solution is very nice
(11:47:51 AM) EfraimDF: @tech Multiple overlapping hierarchies bites you any time you have poetry in the text, for example
(11:48:42 AM) EfraimDF: @tech Basically, we've decided that it's best to represent all possible hierarchies.  A processor could then decide to prioritize one or the other, or to try to mix them
(11:49:12 AM) Azriel: @tech to avoid duplication, we use pointers
(11:49:34 AM) Azriel: @tech and organize the pointers into hierarchies
(11:49:56 AM) EfraimDF: @tech Each file (aka database resource) contains a small part of the siddur text (maybe a paragraph or other logical unit)
(11:50:14 AM) amg_: @tech I realize this piece may be premature, but to what extent do we envision letting people add personal information to the processing stage? I'm thinking both about personal variants in the text and in control over layout.
(11:50:35 AM) amg_: Or is the answer to that "anyone who wants to do that can do an svn checkout and have fun on their own computer"?
(11:50:40 AM) EfraimDF: @tech To the maximal possible extent.
(11:50:53 AM) EfraimDF: @tech To some degree, the UI will lag behind the backend
(11:51:21 AM) EfraimDF: @tech So, the UI at first may only support simple kinds of changes.  As it evolves, it will have more advanced modes
(11:51:41 AM) EfraimDF: @tech The backend basically has to support as many layout variants as possible.
(11:51:59 AM) EfraimDF: @tech The transforms are being designed as a pipeline
(11:52:33 AM) EfraimDF: @tech They start with a "recipe" which is an XML file that defines a starting point and a set of user choices
(11:53:29 AM) EfraimDF: @tech The recipe is processed in all of the generic ways in the first stage of the pipeline.  That's where multiple hierarchies are joined, parallel texts from various sources are brought together, commentaries are put into place and so on
(11:53:30 AM) amg_: @tech OK, I'll have to take a look at the xml dtd/schema and transforms.
(11:54:04 AM) EfraimDF: @tech The schema docs are on the wiki (as I said, help needed with them).  The machine readable schema is in TEI ODD format
(11:54:25 AM) EfraimDF: @tech which processes to RelaxNG, DTD or XSD (I happen to like RelaxNG)
(11:54:56 AM) EfraimDF: @tech The transforms don't do much as of now.  There's an old proof of concept code which has more complete transforms, but does it on the wrong schema
(11:55:45 AM) EfraimDF: @tech The idea is to do the first transform in the pipeline on the archival form JLPTEI.
(11:56:10 AM) EfraimDF: @tech convert it to a limited-set XHTML-like language (which can display on a browser w/CSS)
(11:56:18 AM) EfraimDF: @tech and use that to convert to other formats
(11:59:00 AM) EfraimDF: @tech As an open source and web-based project, we basically have to keep our entire public db downloadable.
(11:59:34 AM) EfraimDF: @tech And even if the transforms are "finished" before the UI, they should be usable standalone
(11:59:39 AM) devpb [i=5541510f@gateway/web/freenode/x-aqjeeswzxpqjvcyw] entered the room.
(11:59:47 AM) EfraimDF: (hi devpb)
(12:00:07 PM) EfraimDF: @tech Is this answering your questions?
(12:00:49 PM) devpb_ [i=5541510f@gateway/web/freenode/x-zmfhrncfvijrxjtk] entered the room.
(12:01:28 PM) devpb left the room (quit: Client Quit).
(12:01:33 PM) EfraimDF: Did I put everyone to sleep? :-)
(12:02:01 PM) devpb_ left the room (quit: Client Quit).
(12:02:11 PM) devpb [i=5541510f@gateway/web/freenode/x-dkkocszockxvyigk] entered the room.
(12:02:27 PM) Aharon: hi devpb
(12:02:28 PM) Azriel: @tech if you want an overview of the flow, here is an old diagram: http://imagebin.org/68297
(12:03:31 PM) amg_: Sorry... had to deal with some kid-based interruptions :-)
(12:03:39 PM) EfraimDF: np
(12:03:48 PM) amg_: This is good background info, thanks
(12:04:14 PM) Aharon: devpb: our developers are giving something of a techinical summary of the projects innards. there is also a non-technical stream for open siddur contributors who want to contribute through research, transcription, outright text contribution, etc. do you have any questions about the project?
(12:04:14 PM) EfraimDF: Any other basic (or advanced) questions?
(12:04:34 PM) Azriel: @tech don't be confused by the arrows
(12:05:03 PM) EfraimDF: @tech because the entire pic is filled with arrows :-)
(12:05:39 PM) Aharon: so the two points on the agenda i really wanted to hit were development priorities and reachable milestones looking at the next month or so
(12:06:06 PM) EfraimDF: Should reference the http://wiki.jewishliturgy.org/Milestones list
(12:06:31 PM) EfraimDF: Does going through them by topic make sense?
(12:06:57 PM) amg_: Sure
(12:07:10 PM) Aharon: yes
(12:07:15 PM) EfraimDF: this won't make sense if you don't have the page open in another window 
(12:07:34 PM) EfraimDF: (1) XML format
(12:07:38 PM) devpb left the room (quit: "Page closed").
(12:08:02 PM) devpb [i=5541510f@gateway/web/freenode/x-dfmouynfarapqava] entered the room.
(12:08:02 PM) EfraimDF: I really need someone to look through the guidelines as they are and make sure they make sense to someone other than me
(12:08:34 PM) Aharon: the JLPTEI?
(12:08:43 PM) EfraimDF: I can't consider them completed until someone else has tried to implement them :-)
(12:08:44 PM) EfraimDF: yes
(12:09:08 PM) EfraimDF: One (small) project I was discussing w/Aharon offline was automated conversion of contributed texts
(12:09:33 PM) EfraimDF: These are pretty much one-time use throw-away scripts that take contributed text and convert them into JLPTEI
(12:09:56 PM) EfraimDF: There's an example (of XSLT abuse ;-) ) on svn for converting our Tanach texts
(12:09:56 PM) Aharon: is completion of JLPTEI necessary before that work can be done?
(12:10:12 PM) EfraimDF: That is what it's all about.
(12:10:19 PM) EfraimDF: Right now, I'd call it a working draft
(12:11:10 PM) amg_: I'd voluteer but I'm very overbooked for the next four weeks
(12:11:24 PM) EfraimDF: No problem.  It'll still be there in 5 weeks :-)
(12:11:35 PM) amg_: :-)
(12:11:46 PM) EfraimDF: (2) XML Database
(12:12:18 PM) EfraimDF: Revision control -- I set it up, we really need to start putting texts in the db to test this properly
(12:12:40 PM) EfraimDF: Security -- this is a big one.  There are some new docs on the eXist site about running a production eXist server
(12:13:44 PM) EfraimDF: Getting a user reg system up might make this better
(12:14:37 PM) EfraimDF: dev tools -- Been experimenting with using rsync and davfs.  It sorta works, but has some problems, mostly with eXist's use of MIME types and the fact that rsync doesn't set them
(12:15:13 PM) EfraimDF: Having a command line upload to server and set mime type utility would be good (and isn't really hard to do)
(12:15:40 PM) EfraimDF: second, a command line "sync this filesystem db image with the db"
(12:15:47 PM) EfraimDF: (which is a bit harder)
(12:16:50 PM) EfraimDF: (3) Toolkit API -- these are the XQuery functions that can be used from the client side to ask the server for specific processing
(12:17:11 PM) EfraimDF: I've been handling it as something completely reactive to what the client code needs
(12:18:10 PM) Azriel: @tech so far, I have created queries for retreiving information
(12:18:14 PM) amg_: That sounds mostly right, although I worry about versioning the queries
(12:18:23 PM) Azriel: @tech but updating is where we are up to
(12:18:27 PM) EfraimDF: There are some APIs for updates and validation (based on the eXist APIs with some implementation-specific details)
(12:18:36 PM) EfraimDF: Versioning, in what sense?
(12:19:52 PM) amg_: If a newer client requirement makes us want to change the behaviour of an existing query, how do we plan to handling that without breaking other downstream uses of that query?
(12:20:05 PM) amg_: (There are several possible answers to that question; I'm just asking which approach we're using)
(12:20:35 PM) EfraimDF: It depends on how "broken" the old way is
(12:21:40 PM) EfraimDF: If it's not too broken, we can leave the old query and deprecate it, make one with a different name for the new functionality, and after some set period of time (or version # upgrade), retire it
(12:21:55 PM) EfraimDF: (and return an error on use)
(12:22:37 PM) amg_: I like that answer
(12:22:43 PM) Azriel: also, keep in mind that the clients will be running off of our web server and can be changed in parallel
(12:22:56 PM) Azriel: although it might break foriegn clients
(12:22:58 PM) EfraimDF: At least as long as there's only one client.  
(12:23:11 PM) Azriel: right
(12:23:14 PM) EfraimDF: Once we start seeing foreign clients, we need to be more careful about these things.
(12:23:33 PM) amg_: Right, and I expect foreign clients to come up sooner rather than later.
(12:23:42 PM) amg_: Unless we build the UI in Flash. :-)
(12:23:45 PM) EfraimDF: I hope they do :-)
(12:23:58 PM) Aharonium [n=chatzill@ool-182e4348.dyn.optonline.net] entered the room.
(12:24:29 PM) EfraimDF: The UI is Java/JS based - avoiding Flash at all costs :-)
(12:24:38 PM) Azriel: lol
(12:24:50 PM) Azriel: EfraimDF is viciously anti flash
(12:25:13 PM) EfraimDF: Hey, if there were a good free implementation, it may not be so bad.
(12:25:39 PM) Azriel: I don't entirely disagree
(12:25:54 PM) EfraimDF: Next up, Transforms --
(12:26:04 PM) EfraimDF: This is really an area I'd like to see work in.
(12:26:36 PM) EfraimDF: multi-hierarchy -> single combined hierarchy code sort-of works.  We need a bit of spec-ing of the mini-XHTML format
(12:26:55 PM) EfraimDF: The selection mechanism will likely be similar to the one used in the proof of concept, but will require adapting code
(12:27:19 PM) EfraimDF: Getting to the mXHTML format is important.  
(12:27:35 PM) EfraimDF: Then, we can do all kinds of things in parallel.
(12:27:38 PM) amg_: Transforms is an area I'm very interested in -- again, if you can wait for me to wrap up my other projects over the next few weeks
(12:27:49 PM) EfraimDF: sure.
(12:28:09 PM) EfraimDF: (I'd suggest keeping tabs on the ML and maybe even sending a msg there when you're ready to jump in)
(12:28:42 PM) EfraimDF: Transforms is now my primary focus
(12:28:53 PM) EfraimDF: Content gathering...
(12:29:25 PM) EfraimDF: Lots of research needed about which texts to scan, particularly for non-Ashkenazic nusah
(12:29:56 PM) EfraimDF: Access to a high speed book scanner (and books) is necessary for scanning
(12:30:20 PM) EfraimDF: We're accepting original works, and have had some contributed
(12:30:44 PM) EfraimDF: Converting those to usable formats (as mentioned before) is a good way to practice w/JLPTEI
(12:30:58 PM) EfraimDF: Next -- content entry
(12:31:16 PM) EfraimDF: we don't have the UI set up yet, but we do have the wiki, and if people want to do it, we could start ASAP
(12:31:41 PM) EfraimDF: Actually, this will likely be the big bottleneck in the future
(12:31:54 PM) EfraimDF: so, it's best to get as much entered early as possible
(12:32:16 PM) EfraimDF: QC Guidelines for core texts -- we're just not there yet
(12:32:36 PM) EfraimDF: Azriel, milestones for developer front end?
(12:33:13 PM) Azriel: first I think we need to decide what platform to develop on
(12:33:40 PM) Azriel: from the perspective of GWT
(12:34:06 PM) Azriel: The next milestone is updating the eXist db with new content
(12:34:15 PM) EfraimDF: which has some prereqs too
(12:34:30 PM) Azriel: which requires a login system
(12:34:41 PM) Azriel: and registration system
(12:34:54 PM) Azriel: Illan has volunteered to work on that
(12:35:04 PM) Azriel: in a JS/Java agnostic manner
(12:35:32 PM) Azriel: the UI also needs some polishing to be workable
(12:36:12 PM) Azriel: http://wiki.jewishliturgy.org/OSNAT
(12:36:25 PM) EfraimDF: Anyone here have opinions on GWT vs JS+a toolkit, such as jQuery?
(12:36:35 PM) Azriel: this is where the status of the web app is kept
(12:37:27 PM) amg_: GWT lets us leave user management up to Google, right?
(12:37:49 PM) Azriel: not really
(12:37:59 PM) Azriel: its google agnostic
(12:38:08 PM) Azriel: OpenId does that though
(12:38:33 PM) Azriel: however, we have the problem of bridging our OpenId identification with the eXist login mechanism
(12:38:52 PM) Azriel: currently OpenId can be done in PHP
(12:39:02 PM) Azriel: eXist only knows simple user/pass
(12:39:31 PM) Azriel: we are looking for a better solution
(12:39:33 PM) Azriel: but right now
(12:39:54 PM) Azriel: it looks like we would get the user/pass from a php script that would id the user via OpenId
(12:40:17 PM) EfraimDF: The issue is that in order to use eXist's access restrictions, eXist has to know who you're authenticated as. Which means (I think), we need an eXist shadow login, even if we use OpenID to the public
(12:40:35 PM) Aharon left the room (quit: Read error: 110 (Connection timed out)).
(12:40:50 PM) Azriel: both eXist login and OpenId tokens bother me security wise
(12:40:58 PM) Azriel: unless its all over SSL
(12:41:05 PM) Azriel: for now we are ignoring that
(12:41:33 PM) EfraimDF: (eXist by default sends all data in clear text)
(12:41:53 PM) Aharonium: Will choosing GWT over JS+jQuery turn off some developers?
(12:42:13 PM) Azriel: it would turn off less professional developers
(12:42:26 PM) Aharonium is now known as Aharon
(12:42:26 PM) Azriel: and turn on some more professional
(12:42:28 PM) Azriel: it depends who we have more of
(12:42:58 PM) Azriel: but yes
(12:43:43 PM) EfraimDF: I'm not sure of that.  Most web devs work in pure-JS, independent of status.
(12:43:51 PM) EfraimDF: More app devs work in Java
(12:44:36 PM) Azriel: right
(12:44:48 PM) Azriel: this was the debate we were having with Illan
(12:45:10 PM) Azriel: I think we should use the most practical approach
(12:45:30 PM) Azriel: if there are enough developers that are willing to work with GWT
(12:45:33 PM) Azriel: thats great
(12:45:54 PM) EfraimDF: GWT also has a higher barrier of entry to hacking
(12:45:56 PM) Azriel: otherwise, simply for practicality, a JS framework is the way to go
(12:46:30 PM) EfraimDF: in JS, you can change code locally, GWT requires compile/run cycles
(12:46:43 PM) EfraimDF: *locally & see a RT change
(12:48:20 PM) EfraimDF: Back on topic -- are there any milestones we're likely to hit soon on the client side? (even if it means "refactor existing GWT code to jQuery" is one of them)
(12:48:21 PM) Azriel: I dc'd for a sec
(12:48:55 PM) EfraimDF: (I don't think you missed anything important)
(12:49:55 PM) Aharon: Content gathering
(12:50:21 PM) Azriel: should we be doing that?
(12:50:30 PM) Azriel: we need a decision process
(12:50:37 PM) Aharon: we collected some texts from the internet archive. very nice scans
(12:50:51 PM) EfraimDF: I think you are the decision process ;-)
(12:50:55 PM) Aharon: should we be doing what?
(12:51:11 PM) Azriel: Aharon: sorry, switching to JS
(12:51:50 PM) EfraimDF: I'd like an unencumbered digitization of the 1917 JPS translation, if I can find one
(12:52:03 PM) EfraimDF: Would be a good test for the translation alignment transforms
(12:52:24 PM) EfraimDF: (Charles Hare of the Internet Sacred Text Archive is producing one, but it's incomplete)
(12:52:58 PM) Aharon: oh this reminds me of something i wanted to discuss on the chat but not on the agenda.
(12:53:34 PM) Aharon: on the subject of non-free CC licenses or non-CC licenses in general used in concert witht he open siddur
(12:53:39 PM) EfraimDF: *correction: John Hare
(12:53:59 PM) EfraimDF: can we put that after we're done w/milestones?
(12:54:03 PM) Aharon: yes
(12:54:39 PM) Aharon: but it has to do with JPS translations for use-cases outside of printing new siddurim
(12:54:50 PM) EfraimDF: I was being OT
(12:55:04 PM) EfraimDF: (sort of, since it had to do w/content gathering)
(12:55:07 PM) Aharon: or any already copyrighted translations that some might wish us to use
(12:55:33 PM) Aharon: without changing their copyright to CC-BY-SA, CC-BY, or CC0
(12:55:52 PM) EfraimDF: Anything else in Content gathering? Any milestones we can hit?
(12:56:38 PM) Aharon: i deconstructed a number of the pdfs uploaded to the internet archive
(12:56:55 PM) EfraimDF: They also have already OCR-ed translations
(12:57:18 PM) EfraimDF: the OCR is pretty high quality
(12:57:51 PM) Aharon: like with Reb Zalman's siddur, are we now waiting on an JLPTEI XML encoding engine interface
(12:58:08 PM) EfraimDF: For those, I think you can encode w/o the interface
(12:58:14 PM) Aharon: i would enter the material in manually into a form
(12:58:34 PM) EfraimDF: throwaway transforms :-)
(12:59:32 PM) Aharon: how to avoid that when material comes to us in all sorts of different formats?
(01:00:08 PM) EfraimDF: depends on the size of the material
(01:00:38 PM) Aharon: this is at the top of my wishlist
(01:00:41 PM) EfraimDF: if it's large and pretty consistently formatted, a throwaway transform is *far* more efficient than entering it in via web interface
(01:01:18 PM) EfraimDF: if it's inconsistent and hard to make consistent, or small, it's more efficient to enter it in via web form
(01:02:32 PM) EfraimDF: The Internet archive scans are pretty consistent
(01:02:42 PM) EfraimDF: It will all need to go through a proofread stage anyway
(01:03:57 PM) EfraimDF: For an Ashkenaz mahzor *Hebrew* text, my preference would be to use Heidenheim's Seder Kerovot
(01:05:08 PM) EfraimDF: Anyway, I think we've killed this topic...
(01:06:45 PM) Aharon: i haven't heard from mendy or gabe (zemirotdatabase.org) lately
(01:06:59 PM) EfraimDF: contact them offline.
(01:07:16 PM) EfraimDF: (privately)
(01:07:33 PM) EfraimDF: You had another topic to talk about?
(01:07:53 PM) EfraimDF: (12:53:34 PM) Aharon: on the subject of non-free CC licenses or non-CC licenses in general used in concert witht he open siddur
(01:08:03 PM) EfraimDF: ^^ what aspect of it?
(01:10:26 PM) Aharonium [n=chatzill@ool-182e4348.dyn.optonline.net] entered the room.
(01:10:55 PM) EfraimDF: you there?
(01:10:57 PM) Aharonium: sorry about that folks... internet troubles in the BK... i'm on wireless
(01:11:10 PM) Aharonium: our licensing requirements make a lot of sense for the publishing of new siddurim
(01:12:28 PM) EfraimDF: and?
(01:14:06 PM) Aharonium_ [n=chatzill@ool-182e4348.dyn.optonline.net] entered the room.
(01:14:27 PM) Aharonium_: arghh
(01:14:44 PM) Michael___: thanks for letting me listen.
(01:14:53 PM) EfraimDF: bye
(01:14:54 PM) Michael___: best wishes for the week.
(01:14:58 PM) EfraimDF: hope to see you around
(01:15:06 PM) Aharonium_: thank you
(01:15:34 PM) Aharonium_: as i was saying...
(01:16:48 PM) EfraimDF: try to type w/o dc-ing :-)
(01:18:38 PM) Aharonium__ [n=chatzill@ool-182e4348.dyn.optonline.net] entered the room.
(01:18:52 PM) Azriel: heh
(01:18:54 PM) EfraimDF: Aharonium multiplies
(01:19:07 PM) Azriel: quintouplets ;)
(01:19:12 PM) EfraimDF: but does it grow linearly or exponentially?
(01:19:17 PM) Azriel: heh
(01:19:22 PM) Azriel: need a larger sample
(01:19:22 PM) Aharonium__: text formatted with a non-free license would be unavailable for inclusion in the printing process...
(01:19:41 PM) EfraimDF: which non free lic?
(01:20:01 PM) EfraimDF: I dont think we can address them all equally
(01:21:37 PM) Aharonium__: cc-nc-sa
(01:22:19 PM) EfraimDF: ok, so, CC-NC-SA is a bit weird.
(01:22:37 PM) EfraimDF: It's non-free, but it is redistributable.  It's also incompatible w/CC-BY-SA
(01:22:44 PM) EfraimDF: but compatible w/CC-BY
(01:23:37 PM) Aharonium__: interesting
(01:23:39 PM) EfraimDF: What purposes would you have CC-BY-NC-SA text serving?
(01:24:27 PM) EfraimDF: And would allowing it into the archive (1) compromise our ideals? (2) generate confusion among users and contributors?
(01:24:37 PM) EfraimDF: (3) frustrate users/contributors?
(01:25:10 PM) EfraimDF: (note: for (1) I'm talking about the public archive, not any private repositories people have)
(01:25:29 PM) Aharonium__: depends how we ultimately set up the application(s)
(01:25:43 PM) EfraimDF: I'm not sure about that --
(01:25:47 PM) Aharonium__: if an educators interface is designed as a separate platform
(01:25:58 PM) EfraimDF: Imagine the following use case:
(01:26:14 PM) EfraimDF: someone wants to set up a siddur for a class (requires distribution)
(01:26:32 PM) EfraimDF: They come to the site and see a few sets of translations and commentaries that they like
(01:26:44 PM) EfraimDF: They pick them out, then the transform returns back an error
(01:27:09 PM) EfraimDF: Can't use this recipe because of incompatible licensing restrictions
(01:27:15 PM) Aharonium__: they wouldn't be using the eduators interface
(01:27:26 PM) EfraimDF: How does which interface you're using matter?
(01:28:14 PM) Aharonium__: the educators interface might not need the transform
(01:28:24 PM) Aharonium__: only the access to the materials online
(01:28:43 PM) Aharonium__: think of the opensiddur but not oriented to pdf output
(01:28:56 PM) EfraimDF: it's still creating a derivative work
(01:29:00 PM) EfraimDF: once you combine it
(01:29:34 PM) EfraimDF: Which terms does the combined work have?  NC-SA requires it to have an NC term, SA requires it not to have an NC term
(01:29:49 PM) EfraimDF: You still need *a* transform (to XHTML)
(01:29:51 PM) Aharonium__: so the educators interface simply doesn't show user contributed work licened cc-by-sa
(01:30:15 PM) EfraimDF: so you encourage educators to contribute w/an NC term
(01:30:22 PM) EfraimDF: and lock their contributions away from everyone else?
(01:30:32 PM) Aharonium left the room (quit: Read error: 110 (Connection timed out)).
(01:30:45 PM) Aharonium__: not everyone, just a select application
(01:30:58 PM) EfraimDF: Then get a different archive!
(01:31:42 PM) Aharon left the room (quit: Read error: 110 (Connection timed out)).
(01:32:32 PM) EfraimDF: I don't see any advantage in encouraging creation of mutually incompatible content
(01:35:19 PM) EfraimDF: << looks like another connection hiccup>>
(01:35:22 PM) Aharonium_ left the room (quit: Read error: 110 (Connection timed out)).
(01:36:28 PM) Azriel: uh oh
(01:36:30 PM) Azriel: only one left
(01:36:35 PM) Azriel: :)
(01:37:48 PM) EfraimDF: conserve Aharonium 
(01:37:51 PM) EfraimDF: recycle?
(01:40:01 PM) Aharon [n=chatzill@ool-182e4348.dyn.optonline.net] entered the room.
(01:40:43 PM) Azriel: hey
(01:40:55 PM) EfraimDF: wb
(01:41:03 PM) EfraimDF: last to room:
(01:41:12 PM) EfraimDF: (01:30:45 PM) Aharonium__: not everyone, just a select application
(01:41:12 PM) EfraimDF: (01:30:58 PM) EfraimDF: Then get a different archive!
(01:41:12 PM) EfraimDF: (01:31:42 PM) Aharon left the room (quit: Read error: 110 (Connection timed out)).
(01:41:12 PM) EfraimDF: (01:32:32 PM) EfraimDF: I don't see any advantage in encouraging creation of mutually incompatible content
(01:41:46 PM) Aharon: it would be a separate archive then
(01:41:58 PM) EfraimDF: ok, now go back to the premise -- why do it?
(01:44:49 PM) Aharon: i think a more general scenario we'll experience are folks interested in an open siddur application but not comfortable with ceding as many rights as we're asking with out licenses. we can either say to them, "No" or we can have a more nuanced response
(01:45:20 PM) EfraimDF: Is this just about NC or anything else?
(01:45:52 PM) Aharon: such as, our application is open source LGPL and you can create your own archive for this material, or we can help you make an archive (for a fee obviously since our mission is different)
(01:46:01 PM) Aharon: NC was an example
(01:46:15 PM) Aharon: for an application for educators
(01:46:27 PM) EfraimDF: Is that educating educators badly?
(01:46:48 PM) EfraimDF: We may be their first encounter with free culture.
(01:48:11 PM) EfraimDF: I don't really see why educators should have less freedom than anyone else.
(01:49:05 PM) Aharonium__ left the room (quit: Read error: 110 (Connection timed out)).
(01:49:20 PM) Aharon: for sure
(01:49:59 PM) EfraimDF: I don't see a problem w/consulting on creation of proprietary archives.  
(01:50:33 PM) EfraimDF: I also don't see this as a priority (obviously the support for CC-RDF will be in the apps, so it should work for the use case w/o too many mods)
(01:50:57 PM) Aharon: i realize this is looking ahead quite a bit
(01:51:18 PM) EfraimDF: I see the primary mission of the project as a free culture project
(01:51:27 PM) Aharon: i agree
(01:51:40 PM) EfraimDF: and the support for private *undistributed* data as a compromise w/individual's privacy
(01:52:12 PM) Aharon: did we make that clear in the mission statement?
(01:52:12 PM) EfraimDF: Once you get into things like distributing non-free data through an OSP-official branded archive, it sets off alarm bells in my head
(01:52:22 PM) EfraimDF: I think so.
(01:53:06 PM) Aharon: mine too. i think when this cam up with Dan and Eileen the last time, we said something similar. i think we should write out our position
(01:53:24 PM) EfraimDF: It isn't obvious?
(01:53:57 PM) Aharon: not to folks who work for institutions and orgs unfamiliar with free culture
(01:54:06 PM) EfraimDF: Think of it as Wikipedia vs. Mediawiki
(01:54:16 PM) EfraimDF: Wikipedia is an archive.  Mediawiki is a platform.
(01:54:24 PM) Aharon: right
(01:54:40 PM) EfraimDF: The OSP-branded archive is free.  The platform software is free.  What you do with it is up to you.
(01:56:45 PM) Aharon: while we are in the position of educators for free culture, we need to provide some guidance for other orgs that are only familiar with licensing content from others
(01:56:47 PM) EfraimDF: As for whether one *should* produce an archive "for educators" that is non-free, my suggestion would be "don't do it"
(01:57:19 PM) Aharon: i disagree
(01:57:35 PM) EfraimDF: Would you think there should be a separate an incompatible Wikipedia for educators?
(01:57:51 PM) Aharon: but not specifically for educators
(01:58:51 PM) Aharon: i can understand why some content creators may wish there work to be non-commercial
(01:59:16 PM) EfraimDF: the price of the NC term is huge
(02:00:44 PM) EfraimDF: it means creating a shadow NC-universe that can't benefit from 1/2 of the free universe.
(02:01:13 PM) EfraimDF: and necessarily splitting the effort of all content creators between the NC-uni and the free-uni
(02:01:23 PM) Aharon: it promotes an academic divide
(02:02:13 PM) EfraimDF: explain
(02:03:50 PM) Aharon: the distribution of content licensed NC will follow certain streams, and remain obscure (although not inaccessible). strange how this material ends up in copyrighted journals anyways
(02:05:47 PM) EfraimDF: but there's also the NC-trap.  People who think that NC is actually a benefit
(02:08:43 PM) Aharon: the fellow who sent us his siddur for hidden Jews and portuguese/spaniards of Jewish descent
(02:09:12 PM) EfraimDF: what about it?
(02:09:30 PM) Aharon: he had CC'd his siddur with a non-free license. we should make a strong argument to him to revise his license. i was told he has an open mind about this
(02:10:57 PM) Aharon: i replied to his emailed contribution on the 16th... haven't heard from him. time for a follow up
(02:12:39 PM) EfraimDF: the argument is basically -- consider compatibility with everything else and consider the most widespread dissemination of the work.
(02:13:45 PM) Aharon: ie, your goals are furthered more by this choice of license rather than that one
(02:14:03 PM) EfraimDF: http://freedomdefined.org/Licenses/NC#The_Case_for_Free_Use:_Reasons_Not_to_Use_a_Creative_Commons_-NC_License
(02:14:36 PM) Aharon: Thanks!! ii need to get back to work emailing the recent OpSid survey responders
(02:14:47 PM) Aharon: today was a long chat
(02:15:08 PM) Aharon: sorry to everyone for my internet crapping out as it was
(02:16:20 PM) EfraimDF: ok.  I've got the chat log.  I'll send it to the list
(02:16:42 PM) EfraimDF: Remember to mark which ones you've contacted on the spreadsheet
(02:17:56 PM) Aharon: remember to edit Andrew's email address
(02:18:22 PM) EfraimDF: Btw on the free culture essay -- I'd also add -- what if a shul wants to do a fundraiser by selling printed siddurim?
(02:25:05 PM) EfraimDF: Bye everyone.  Thanks for chatting