IRC Conference/logs/2009-11-22

Conference log for the Nov 22, 2009 chat. See the summary for discussed topics. Questions or comments should be directed to the discussion list.

 [Open Siddur Live Chat -- http://wiki.jewishliturgy.org] [11:02] == Aharon [i=4c74f8f2@gateway/web/freenode/x-kkdzgtocstfzoctl] has joined #jewishliturgy [11:02]  hey [11:02]  boker tov, y'all [11:03]  (back) [11:03]  boker or! [11:03]  I hope Azriel will make it... he was committing code about once an hour until 7AM [11:06] == Jonah [i=18b8f05e@gateway/web/freenode/x-doyvlczydutbotkt] has joined #jewishliturgy [11:06]  he's on his way [11:07]  in a minute or so, we'll get started [11:07]  Hi all; sorry I'm late. [11:07]  ah, didn't start yet? [11:07]  you're on time in JST :-) [11:07]  haha good [11:08]  Hi Refoyl. Are you in Kentucky? [11:08]  yes. [11:09]  i'm from Cincinnati, originally [11:09]  I'm from Chicago. [11:09] <Aharon> the new editor of Zeek is also from Lexington [11:09] == AYFNB [n=AYFNB@pool-151-204-135-181.ny325.east.verizon.net] has joined #jewishliturgy [11:09] <EfraimDF> meanwhile, why don't we start going around and everyone can tell us who they are and what their interests are? [11:10] <AYFNB> hi all [11:10] <Refoyl> I had never heard of Zeek. But I just looked it up.  Yes, I know JoEllen Kaiser! [11:10] <EfraimDF> I'm Efraim Feinstein, lead developer of the Open Siddur Project.  I'm involved in all aspects (and would like to hand over some of them to other people... hint, hint ;-) ) [11:10] == realazthat [n=realazth@pool-151-204-135-181.ny325.east.verizon.net] has joined #jewishliturgy [11:11] <Aharon> my interests are in creating useful resources for engaging Jewish spiritual practices [11:11] <Refoyl> I'm Raphael Finkel. I'm a professor of Computer Science at the University of Kentucky. I have built an OCR program that works ok for Yiddish and can be expanded to deal with Hebrew, with nekudes. I teach Yiddish on the side and have long been active in the international Yiddish community. [11:11] <Jonah> Jonah here: I'm from central-ish Long Island, I'm a senior in List College studying Music at Columbia and Jewish Music at the Jewish Theological Seminary, and I'm currently applying for rabbinicl schools. I enjoy siddurim a lot apparently and like grammar and textual variants and Jewish history. [11:12] == mode/#jewishliturgy [+o realazthat] by ChanServ [11:12] == mode/#jewishliturgy [+o realazthat] by ChanServ [11:12] <Jonah> sorry for the bad typing ast my end... im quite tired... [11:12] == mode/#jewishliturgy [+o EfraimDF] by ChanServ [11:13] <Aharon> @Jonah have we met at Yeshivat Hadar? [11:13] <Jonah> hmm.... I know some YH people but I've actually never been to YH. [11:13] <Jonah> Do you know Andres? [11:13] == EfraimDF [n=efraimdf@c-98-216-111-79.hsd1.ma.comcast.net] has quit ["Leaving."] [11:13] == EfraimDF1 [n=efraimdf@c-98-216-111-79.hsd1.ma.comcast.net] has joined #jewishliturgy [11:13] == EfraimDF1 has changed nick to EfraimDF [11:13] <EfraimDF> (that was an inopportune time for my Internet to blink) [11:14] == mode/#jewishliturgy [+o EfraimDF] by ChanServ [11:14] <Aharon> @Jonah yes i do. stop by when you have a spare moment [11:14] <Jonah> i do intend to; YH does great things [11:14] == EfraimDF changed the topic of #jewishliturgy to: Open Siddur Live Chat -- http://wiki.jewishliturgy.org [11:14] <AYFNB> I'm Azriel Fasten, I currently help a little bit in many aspects of the project [11:14] <Jonah> I have so little tfree time unfortunately [11:14] == AYFNB has changed nick to Azriel [11:15] <Azriel> Jonah: same here ;) [11:15] <@EfraimDF> (where we define "a little bit" as "writes lots of code") [11:15] <Jonah> im actually multitasking as i type now [11:15] <Jonah> Hi Azriel, did u ntroduce ursef? [11:16] <Aharon> I sent Refoyl two scans to try out his experimental OCR, one with hebrew/yiddish from the FOLSHTENDIGER LINYENSIDER LE-VATE-SEFER VELA-AM and the other from Baer's AVODAT ISRAEL [11:17] <Azriel> Jonah: I apologize, it was a bit short [11:17] <Refoyl> I looked at one of them. I will have some problems with it, but they aren't insurmountable.  I'll describe in full when you want. [11:17] <@EfraimDF> I'd be happy to hear about them [11:17] <Aharon> Azriel I was instrumental in brining Efraim's Jewish Liturgy project and my dormant Open Siddur project together. [11:18] <Azriel> I live in Brooklyn, NY, I attend NYU Poly. A year back or so I  was thinking of starting a similar project as an open source developer [11:18] <Azriel> and found Efraim's project [11:18] <Jonah> ah very cool [11:18] <@EfraimDF> since this is our apparently first attempt at OCR that is even close to working (?) [11:18] <Azriel> and Aharon's designs [11:18] <Azriel> and merged them [11:18] <Jonah> I have often thought of, but lacked all computer skills for creating, a very Customizable Siddur from an online library of Siddur Variants. [11:18] <Jonah> So I was very excited to hear of this project. [11:19] <Azriel> Refoyl: I'm very impressed with the possiblity of an OCR [11:19] <Refoyl> Hey, it must have been one of you who helped with the birkas hamazon text for my daughter Penina's bas mitsve in 1999; we wanted the sephardic "u-ma sheokhalnu" at the end. [11:19] <Jonah> And I'm happy to contribute whatever little grammar or textual knowledge I can offer. [11:19] <Jonah> in whatever little time i can offer. [11:19] <@EfraimDF> wasn't me -- I started on this in 2006 [11:20] <Aharon> since then, we've made some progress in publicizing our efforts and vision for a collaborative publishing platform for printing personalized siddurim (or for teaching jewish liturgy, sharing prayers, or collaborating on translations, etc.) [11:20] <Refoyl> I also dabble in computational linguistics: morphology, and have a complete description of biblical Hebrew verbs, including all the weak cases. [11:20] <@EfraimDF> *that* may be very useful [11:20] <@EfraimDF> @Refoyl -- have you been following the discussion list? [11:20] <Jonah> Refoyl, I am a mere hobbyist. You sound like you will be far better for certain tasks that Efraim. [11:20] <Jonah> I mean [11:20] <Jonah> than me*. [11:21] <Refoyl> No, I just joined today. [11:21] <Jonah> Wow [11:21] <@EfraimDF> take a look at the older threads about the sheva na and qamats qatan [11:21] <Refoyl> My morphology generator gets both of those right. [11:22] <@EfraimDF> I'd love to know how you did it! Are you willing to open source it? [11:22] <Refoyl> It's GPL. http://www.cs.uky.edu/~raphael/linguistics/KATR.html [11:22] <@EfraimDF> cool [11:23] <@EfraimDF> i'll take a look at it and we should have some discussions about integrating it [11:23] <Jonah> Refoyl, you're a life-saver! [11:23] == david__ [i=18d1e042@gateway/web/freenode/x-iqckpswkutmgvfds] has joined #jewishliturgy [11:23] Hi, just testing [11:23] <@EfraimDF> as that can bring our tanach text up to standards [11:23] <@EfraimDF> and potentially help with all incoming texts [11:24] <@EfraimDF> (If you're just joining in, please introduce yourself and your interest in the project) [11:24] <@EfraimDF> I'm salivating over this code :-) [11:24] <Refoyl> When I checked my program against Genesis, there were differences; the tanakh doesn't always follow consistent forms, and there are some very unusual exceptions. [11:25] <@EfraimDF> can the program detect the exceptions and mark them? [11:25] <Aharon> fascinating! [11:25] <@EfraimDF> turning it from a multi-thousand word problem to a few hundred words would definitely make our work easier. [11:26] <@EfraimDF> (missed that test -- yuvan, I see you) [11:27] ok thanks, but not sure how to jump in here [11:27] <Refoyl> A complete description would take a long time. Here's a brief summary.  The program is written in KATR and generates forms.  It does not parse forms.  To test against Genesis I used a marked-up text (Westminster Hebrew Bible) and only looked at verbs in the perfect or imperfect.  My program generates vov-consecutives, but not always right. [11:28] took me some time just to figure out how to type in [11:28] <@EfraimDF> OK -- so, 2 things we didn't know before -- [11:28] <@EfraimDF> scratch that... [11:29] <@EfraimDF> how much prior information does it need? [11:29] <Refoyl> It needs the shoresh and the form you want, such as imperfect masculine 3 sg piel. [11:30] <@EfraimDF> is there any way we can get it to work backwards from a word? [11:30] <Refoyl> Not easily. That's part of the long-term plan, but we (Greg Stump, a linguist here, and I) have moved to other things. [11:31] <@EfraimDF> that's what we really need [11:31] <@EfraimDF> although something that could help a transcriptionist determine the correct form would certainly help [11:31] <Refoyl> I can imagine using it in a brute-force way, having it guess the shoresh and trying all forms until it finds one that works. [11:32] <@EfraimDF> all of this will only happen 1x/text, so I don't think processing time is really much of an issue [11:32] <Refoyl> Are you imagining using this for spell checking primarily? [11:32] <@EfraimDF> no [11:32] <@EfraimDF> Our problem is this: [11:33] <@EfraimDF> We have texts coming in that are transcribed [11:33] <Aharon> Good morning David___. Thanks for joining us [11:33] <@EfraimDF> Most of our source texts don't differentiate between sheva/qamats forms, but modern prayer books are expected to [11:33] <@EfraimDF> (a separate problem is that Unicode doesn't have a sheva na symbol, but that's another issue) [11:34] <@EfraimDF> we would like our texts to differentiate in the archival form [11:34] <@EfraimDF> (by the way, other groups can have other topic chats) [11:34] <@EfraimDF> (not everyone has to sit by discussing hebrew grammar processing) [11:34] <Jonah> (I enjoy listening) [11:35] <Aharon> @david__ feel free to introduce yourself [11:35] <@EfraimDF> For now, the largest repository of texts we have is the Westminster Tanach [11:36] <@EfraimDF> which has extra grammatical clues (trope) [11:36] <Refoyl> I see the problem. If you know the grammatical construction, you can get the right answer. The surface form is not always enough, but it is enough except for rule gimel. [11:36] <@EfraimDF> yup [11:36] <Azriel> Jonah: are you familiar with CSS? [11:36] <Azriel> (looking at the mailing list) [11:37] <Azriel> or anyone else here? [11:37] <Jonah> Hey Azriel, sorry I don't know CSS. [11:37] <@EfraimDF> I'd actually be happy to get it right most of the time [11:37] <@EfraimDF> especially if the program *knows* when there's an ambiguity [11:37] <Refoyl> The trop helps when it shows you the accent, needed for rule gimel. [11:37] <@EfraimDF> we can then send ambiguous cases to humans with knowledge [11:38] <@EfraimDF> most of our texts will not have trope [11:38] <@EfraimDF> but if we have them, they could be used to disambiguate when possible [11:38] <Refoyl> that works. It shouldn't be too hard to build something that works in all cases except when it needs to ask an expert. Then it can store the exception so it never needs to ask again. I don't imagine the same surface form will have two renderings. [11:39] <Azriel> Jonah: we made some progress on the xhtml output [11:39] <Azriel> Jonah: I was wondering if someone was interested in styling it [11:39] <Aharon> i think you mean, when a user wants trope for tanakh source texts the option is available for them [11:39] <Jonah> Hmm... Azriel, what do you mean by styling? [11:39] <Jonah> This does sound intriguing. [11:39] <Aharon> @efraimdf correct? [11:40] <Jonah> And Efraim, I'd be happy to help with being a human who can look at particular ambiguous cases. [11:40] <@EfraimDF> @Aharon -- it doesn't matter -- grammatical processing takes place in the archive, not in the display stage [11:40] <Aharon> @Jonah CSS (cascading style sheets) [11:40] <Jonah> (Time-permitting as always) [11:41] <@EfraimDF> @Aharon, Refoyl -- the expert thing will require yet another web interface [11:41] <@EfraimDF> @Aharon, Refoyl but if we can get to a point where we need it, we'll have made huge progress [11:41] <Refoyl> However, reasonable grammarians have disagreed over whether a particular shva is na or nakh. There isn't a standard answer in all cases. [11:41] <Aharon> we are looking for more web interface hackers [11:41] <Jonah> @Aharon I'mnot sure if I understand what "cascading style sheets" means exactly. [11:41] <@EfraimDF> In that case, we have to encode the ambiguity [11:42] <@EfraimDF> So, say there are 2 rule sets (I know I'm simplifying here) [11:42] <Azriel> Jonah: xhtml gets produced, but it sort of looks like lots of text dumped out with no structure and style [11:42] <Azriel> Jonah: if you don't know CSS its fine [11:42] <Azriel> Jonah: I was just wondering [11:42] <Jonah> Ah, got it. [11:42] <Refoyl> @Jonah: it's a way of getting XHTML to look good in a browser by setting up rules for how to display various parts. You can change the entire appearance by modifying the rules. [11:43] <@EfraimDF> while processing, say -- according to rule set 1, this word has na, according to rule set 2, this word has nach [11:43] <Aharon> @Jonah CSS is a standard by which the design of a web page (or a whole web site) is defined by a stylesheet, ie. a list of rules for the layout of the information and its appearance [11:43] <@EfraimDF> @Refoyl there's an JLPTEI (our XML format) mechanism for doing there. [11:43] <@EfraimDF> *doing that [11:44] <@EfraimDF> @Refoyl The key is to know which form goes with which rule set and to write it consistently [11:44] <Jonah> Gotcha [11:44] distinguishing betw 2 rules? [11:45] <Aharon> @Jonah for example, this is the css for opensiddur.net (our developer blog) http://opensiddur.net/wp-content/themes/plaintxtblog/style.css [11:45] <@EfraimDF> can you give a quick example of a word with 2 forms like that? [11:45] <Refoyl> OK. Where do you stand right now on software that takes surface forms and generates Unicode with the correct qomets, correct shva, a meteg for non-final accent? [11:45] <@EfraimDF> right now, we're nowhere -- the most I have is a transliterator which attempts to figure it out and is right most of the time when it knows the qamats type [11:46] <@EfraimDF> but it's only for display, not for manipulating the archive [11:46] <Refoyl> Here's an example: בְּעָלְמַָא [11:47] <@EfraimDF> ok -- [11:47] <@EfraimDF> (doesn't the language of the word disambiguate that?) [11:47] <@EfraimDF> (anyway, let's assume it doesn't) [11:47] <Aharon> @Jonah even if you're not familiar with web design or coding, we still need help in 1) transcribing texts, 2) research on siddur texts and translations, and 3) locating texts that are in the public domain for scanning [11:48] <@EfraimDF> in JLPTEI, we can do something like this: [11:48] <Refoyl> If we assume that Aramaic follows the same rules as Hebrew (it might not), the ayin might have a short qomets, and then the lamed is nakh.  Or the other way. [11:48] <Jonah> @Aharon, I'd be happy with (1), (2) and (3). [11:48] <Jonah> Also as I'm at JTS currently, I have access to a number of texts that we may need. [11:48] <Jonah> in case we're looking for anything unusual. [11:48] <Aharon> it would be nice, for example, to integrate Rav Kook's commentary on the siddur into the Open Siddur [11:50] <@EfraimDF> <tei:choice xmlns:tei="http://www.tei-c.org/ns/1.0" xmlns:j="http://jewishliturgy.org/ns/jlptei/1.0"><j:option xml:id="option1"><tei:w></tei:w></j:option><j:option xml:id="option2"><tei:w></tei:w></j:option></tei:choice> [11:50] <@EfraimDF> (Unicode BIDI is messing up the actual texts of form 1 and 2 and making it hard to type :-) ) [11:51] <Jonah> Aharon, I might be able to help with writing Rav Kook's commentary up; hopefully I'll soon be done with devoting all of my time to the Rabbinical Assembly's new Mahzor. [11:51] <Jonah> Once that happens, I'll be able to coordinate my time to volunteer some free time here and there for the Open Siddur project. [11:52] <@EfraimDF> for representing sheva na, we can use a tei:g element (until Unicode gets their act together) [11:52] <Aharon> @Jonah cool beans! we can coordinate offline, or via facebook [11:53] <@EfraimDF> @Aharon, @Jonah -- might be helpful to have some texts before commentary? [11:53] <Refoyl> For our project, we used a different Unicode altogether for the shva na. I think we settled on rofe, but that's a bad choice in the long run. [11:53] <@EfraimDF> I'd rather keep the display option up to the user [11:53] <@EfraimDF> I don't want to store sheva na as sheva-rafe in the archive [11:54] == Refoyl [n=raphael@gloucester.cs.engr.uky.edu] has left #jewishliturgy [] [11:54] <Aharon> i'd define "helpful" as what anyone is passonate about contributing [11:54] == Refoyl [n=raphael@gloucester.cs.engr.uky.edu] has joined #jewishliturgy [11:55] <Jonah> Efraim: I agree with you. [11:55] <@EfraimDF> agreed there ... but in terms of what will get the project farther fastest (and get the work out faster), probably texts will do better *in the short term* [11:55] <Jonah> I'd be happy to coordinate offline or via Facebook: best way to reach me is E-mail. [11:55] <Refoyl> Sorry, a computer glitch. I can most likely write a Perl script that takes Hebrew Unicode and usually notices when a qomets needs to be qaton and when a shva needs to be nakh. [11:56] <Aharon> my email is aharon@opensiddur.net [11:56] <@EfraimDF> of course, if someone were to say "I want to transcribe *this text and nothing else*" - certainly, I'd say, go ahead and do it) [11:56] <@EfraimDF> @Refoyl -- my last lines to you were: (11:54:14 AM) EfraimDF: I'd rather keep the display option up to the user [11:56] <@EfraimDF> (11:54:27 AM) EfraimDF: I don't want to store sheva na as sheva-rafe in the archive [11:57] <@EfraimDF> That would be a great script to have [11:57] <@EfraimDF> If it could also mark when it doesn't know the answer, even better [11:58] <Refoyl> Proper transcription also requires dagesh/mapiq, of course. I agree that storing shva na with a rafe is a bad idea; we can certainly find better representations.  I'll look into building such a script.  If I have one, it will appear on the KATR page I pointed to earlier as another language tool for Hebrew. [11:58] <Jonah> So, whoever is in charge of assigning particular texts to be written up, I would be happy to write up those texts when given a due date and particular text to write. [11:58] <@EfraimDF> Can you keep us updated on jewishliturgy-discuss? [11:59] <@EfraimDF> @Jonah nobody's in charge of assigning texts [12:00] <Refoyl> Being new to this group and to IRCs in general, I don't know how best to contact folks. How about I email to Aharon? [12:00] <Jonah> Ah got it [12:00] <@EfraimDF> @Jonah for now, you sign yourself up on the Transcription portal on the wiki [12:00] <Jonah> Ah okay. [12:00] <@EfraimDF> jewishliturgy-discuss is an email list [12:00] <@EfraimDF> http://groups.google.com/group/jewishliturgy-discuss/?pli=1 [12:01] <Aharon> @refoyl definitely [12:01] <Refoyl> OK, I see it. [12:02] <@EfraimDF> it will get to all developers and a lot of other interested or potentially interested people [12:04] <@EfraimDF> @Jonah if you have any questions about transcription, ask (the list is the best place, b/c chances are someone else who tries it will have the same question) [12:04] <Aharon> it being 12 noon i just want to thank everyone for joining us this morning, and a gut vokh and shavuah tov [12:04] <Jonah> @Efraim Will do [12:05] <Azriel> we have a bit of a surprise [12:05] <Azriel> first [12:05] <@EfraimDF> @refoyl -- basically, a good transcription will include all vowels [12:05] <Refoyl> I'll try to write up my OCR stuff and put it on the email list. The short answer is that it's great for Yiddish, but the large number of Hebrew combinations (letter + dagesh + vowel + meteg) makes it not very good for prayerbook Hebrew. [12:05] <Azriel> there is a demonstration of Ruth in XHTML, somewhat styled [12:06] <Azriel> you'll need fonts installed [12:06] <Azriel> http://jewishliturgy.org/dump/jlpdemo/Ruth.htm [12:07] <@EfraimDF> (if you want to see what info we have in a text, it's a great place to look) [12:08] <Refoyl> Looks good! Better than http://www.mechon-mamre.org/c/ct/c0.htm. [12:08] <@EfraimDF> and biblical hebrew fonts can be downloaded from http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&id=EzraSIL_Home [12:10] <@EfraimDF> In case anyone's been wondering what Azriel has been up to lately... [12:10] <Azriel> theres more [12:10] <Azriel> it'll be ready in a few [12:11] <Azriel> so hang in [12:11] <Azriel> *there [12:11] <@EfraimDF> ok, hanging [12:11] <@EfraimDF> so, I'll go into what I've been up to lately [12:12] <Refoyl> Bye, everyone; zayt gezunt; shalom; shavue tov! [12:12] == EfraimDF1 [n=efraimdf@c-98-216-111-79.hsd1.ma.comcast.net] has joined #jewishliturgy [12:12] == EfraimDF1 [n=efraimdf@c-98-216-111-79.hsd1.ma.comcast.net] has quit [Read error: 104 (Connection reset by peer)] [12:13] == EfraimDF1 [n=efraimdf@c-98-216-111-79.hsd1.ma.comcast.net] has joined #jewishliturgy [12:13] == EfraimDF [n=efraimdf@c-98-216-111-79.hsd1.ma.comcast.net] has quit [Read error: 104 (Connection reset by peer)] [12:13] <EfraimDF1> my Internet just went out again... [12:13] == EfraimDF1 has changed nick to EfraimDF [12:13] <EfraimDF> so, anyway... our archive is stored as JLPTEI [12:14] <Aharon> hi yuvan [12:14] <EfraimDF> it's a format that's designed for archiving texts with large numbers of variants and overlapping structural hierarchies [12:14] <Aharon> @yuvan feel free to introduce yourself [12:14] Hiaharon [12:15] <EfraimDF> with the tradeoff being that it's verbose in XML tags, but minimally verbose in actual texts [12:15] my unicoding is way behind, just listening [12:15] <EfraimDF> what that means is that it tries to collapse as many texts with commonalities between them into one place [12:16] <EfraimDF> good for fixing errors, not so good for display [12:16] <EfraimDF> the project software has to convert this format into something that can be displayed [12:17] <EfraimDF> with the ultimate goal being more than one display target [12:17] <EfraimDF> (eg, XHTML for web, PDF for print) [12:17] <Aharon> the open siddur will also need some more font faces besides Ezra SIL. i'm looking for help bringin free hebrew fonts up to speed with the latest unicode [12:17] <EfraimDF> everyone with me so far? [12:18] == Refoyl [n=raphael@gloucester.cs.engr.uky.edu] has left #jewishliturgy [] [12:18] == Jonah [i=18b8f05e@gateway/web/freenode/x-doyvlczydutbotkt] has quit ["Page closed"] [12:18] <Aharon> i'm still with you @EfraimDF [12:19] <EfraimDF> we just had a few disconnects [12:19] <EfraimDF> was wondering if I disconnected again too ... :-) [12:20] <EfraimDF> in svn, there's now some XSLT code to transform JLPTEI files (the tanach) to a form of XHTML [12:20] <EfraimDF> that's all html:div tags [12:20] <EfraimDF> html : div [12:21] <EfraimDF> the code does not yet do conditional parsing (like the original proof of concept did) [12:21] <EfraimDF> that's the next step [12:22] <EfraimDF> what it does do is precisely what the siddur needs as a first step -- [12:22] <EfraimDF> take a file with a list of references into other files and incorporate them into a larger output file [12:23] <EfraimDF> so, for example, the book of Ruth in Azriel's example is actually 5 combined JLPTEI files [12:23] <EfraimDF> which were all converted to XHTML in-order [12:24] <EfraimDF> my last 2(?) weeks were writing the XSLT to get that working [12:24] <EfraimDF> the next step is restoring the conditional processing [12:25] <EfraimDF> I actually think that with the current XSLT, you could generate a file containing most of pesukei d'zimrah or hallel [12:26] <EfraimDF> and that's the programming front from my end [12:26] <EfraimDF> On other fronts, I obtained a 1940 printing of Siddur Torah Or (Chabad Nusah Ari), which appears to be in the public domain [12:27] <EfraimDF> it is fully scanned and uploaded to the website (where the transcription app will reside), but not yet to the wiki [12:28] <EfraimDF> the web-quality JPGs are at [12:28] <EfraimDF> http://shell.jewishliturgy.org/base/sources/ShulzingerBros-Siddur_Torah_Or/ [12:29] shalom, laitratov ... my head is swarming. Need to let some of this process from a distance [12:29] <EfraimDF> ok, chat logs will be up on the wiki [12:30] will check out torah or, thanks [12:30] <EfraimDF> not much you can do w/it yet [12:30] <EfraimDF> if you're interested in transcribing, let me know and I'll get some of it on the wiki [12:31] == yuvan [i=4d823e6a@gateway/web/freenode/x-qptdnnogpknopveo] has quit ["Page closed"] [12:32] <EfraimDF> anything else we should discuss while we're here? [12:32] <EfraimDF> (and waiting for Azriel's second announcement, which will probably have to go on the ML) [12:34] <Aharon> Just received a message from an open siddur fan to check out this service http://www.dimdim.com/ for our open chats [12:35] == realazthat[2] [n=realazth@pool-151-204-135-181.ny325.east.verizon.net] has joined #jewishliturgy [12:35] <EfraimDF> what's wrong w/IRC? [12:35] <EfraimDF> it's free and open. [12:35] == realazthat [n=realazth@pool-151-204-135-181.ny325.east.verizon.net] has quit [Read error: 104 (Connection reset by peer)] [12:38] == yuvan [i=4d823e6a@gateway/web/freenode/x-ouaqfvmmypkgsypo] has joined #jewishliturgy [12:38] <EfraimDF> there are lots of chat services out there [12:39] <EfraimDF> btw Aharon, can you deal w/the chat logs b/c my connection was pretty messed up this time around (problem was on my side) [12:40] I see what you mean about torah or, thanks [12:40] == yuvan [i=4d823e6a@gateway/web/freenode/x-ouaqfvmmypkgsypo] has quit [Client Quit] [12:50] <Azriel> funnny [12:50] <Azriel> dimdim uses gwt [12:53] == realazthat [n=realazth@pool-151-204-135-181.ny325.east.verizon.net] has joined #jewishliturgy [12:53] <EfraimDF> wb [12:53] == realazthat[2] [n=realazth@pool-151-204-135-181.ny325.east.verizon.net] has quit [Read error: 104 (Connection reset by peer)] [13:20] == AYFNB [n=AYFNB@ool-182e43a5.dyn.optonline.net] has joined #jewishliturgy [13:48] == Azriel [n=AYFNB@pool-151-204-135-181.ny325.east.verizon.net] has quit [Read error: 110 (Connection timed out)] [13:48] == AYFNB has changed nick to Azriel [13:54] == Rob___ [i=ced4ec51@gateway/web/freenode/x-qkswcorwdhqzblau] has joined #jewishliturgy [13:54] <EfraimDF> hi [14:03] == rtanner [n=chatzill@206.212.236.81] has joined #jewishliturgy [14:04] connect [14:04] <Azriel> ur connected [14:04] <Azriel> heh [14:05] <EfraimDF> i see you [14:05] Just figuting out how to use the client [14:06] <EfraimDF> ok [14:06] <EfraimDF> will happily ignore anything you say that looks like it's supposed to go to a computer [14:21] I haven't actually done anything yet so I really don't have anything to contribute. My intention is just to listen in right now and see what others have to say, but it seems rather quiet. Is my IRC session not working quite right, or is this normal? [14:24] <EfraimDF> you're actually about 3 hrs late [14:25] <EfraimDF> I just have myself still logged in [14:25] <Azriel> post the logs [14:25] <EfraimDF> I got disconnected a few times -- Azriel - do you have logs? [14:26] <EfraimDF> (I may have missed a few lines of text here and there) [14:26] <Azriel> most of them [14:26] <EfraimDF> If I remember correctly, you know something about Java, so you may be able to help w/the issue we're having now? [14:27] Oh! 11am Eastern time and I'm on the Pacific side of the world. Yes, I am a java coder -- I do it for a living. [14:28] <Azriel> my logs are here: http://codepad.org/GLmpIgWK [14:28] <Azriel> EfraimDF: you can fill in any blanks [14:28] <Azriel> EfraimDF:and I came 10 minutes late [14:29] <Azriel> rtanner: any experience with applets? [14:30] A very little bit. I'm generally a JSP and servlet guy and just getting into GWT (working on my seconf GWT project) [14:30] <Azriel> thats interesting [14:31] <EfraimDF> Azriel: I'll fill in the first 10m [14:31] <EfraimDF> can you fill rtanner in on current Java stuff? [14:31] <Azriel> yes [14:31] <Azriel> are transcription application is in GWT (client side only) [14:31] <Azriel> but we are on the verge on switching to regular jquery/other frameworks [14:32] <Azriel> because there is a larger pool of available volunteers [14:32] <Azriel> that would be able to help out [14:32] <Azriel> but thats off topic for today [14:33] <Azriel> you can see more info at http://wiki.jewishliturgy.org/OSNAT [14:33] <Azriel> which pretty much outlines that project [14:33] <Azriel> right now we are working on a demonstration of the new transforms [14:33] I think I said before sometime (probably in a private email) that I really have no jQuery experience but I'm certainly willing to learn on the job, so to speak. [14:33] <Azriel> thats great [14:34] <EfraimDF> (transforms are the XSLT to convert JLPTEI XML -- our archival form of XML -- to printable formats like XHTML) [14:34] <Azriel> the transforms convert encoded information into practicle xhtml [14:35] <Azriel> I think we took a unique approach [14:35] <Azriel> basically [14:35] <EfraimDF> not that unique :-) [14:35] <Azriel> the applet loads a local webserver on the client side [14:35] <Azriel> that *is* unique [14:35] <EfraimDF> oh, you mean for the demo app ... [14:35] <Azriel> heh [14:35] <EfraimDF> that's .. special [14:36] <Azriel> why don't you try it [14:36] <Azriel> it isn't ready yet [14:36] <Azriel> and we are unsure if it works online as an applet [14:36] <Azriel> some success [14:36] <Azriel> some failure [14:36] <Azriel> http://jewishliturgy.org/dump/jlpdemo/jnlp/lib/jlpdemo.htm [14:36] <EfraimDF> it does work as a standalone application [14:36] <EfraimDF> if you're willing to pull from svn [14:41] I need to get an SVN client whicj I can do today and pull a copy of the app [14:41] <EfraimDF> do you know where the repos is? [14:41] <EfraimDF> what we're trying to do is get this working off the web [14:42] <EfraimDF> as an applet [14:43] It's probably on the wiki but you can always give me the URL here. I'd love to play with the app to see what I can do. [14:43] <EfraimDF> azriel, can you give a bit of a wider perspective as to why this app and why now? [14:44] <EfraimDF> yeah its on the wiki and at jewishliturgy.googlecode.com [14:44] <Azriel> basically, Efraim redid the entire encoding scheme to be able to handle "overlapping hierarchies" [14:45] <Azriel> which means that we can have a sentence span over a paragraph etc. [14:45] <Azriel> something which cannot normally be expressed in xml [14:46] <Azriel> we decided that the transforms should primarily target a small subset of xhtml [14:46] <Azriel> which would contain leave as configuration as possible to CSS after the transform [14:46] <Azriel> from the box model of xhtml [14:47] <Azriel> we would then transform it to other formats [14:47] <Azriel> and in the meantime gain the ability to have a nice demonstration in XHTML [14:48] <Azriel> so this application was designed to obtain the XML from our XML database [14:48] <Azriel> (an eXist database) [14:48] <EfraimDF> some of this is coming from experience with an earlier proof of concept that converted our XML->TeX->PDF [14:48] <EfraimDF> which has the nasty problem that it has a lot of huge dependencies [14:48] <Azriel> the application can just as easily get it from a filesystem ( I think ) [14:49] <EfraimDF> yes, it can. i do all my development on the filesystem first [14:49] <Azriel> and then, using saxon, performs the transforms [14:49] <Azriel> with the amazing XSL sheets that efraim designed [14:49] <EfraimDF> saxon = Java based XSLT processor [14:49] <Azriel> (also in the database) [14:50] <Azriel> and produces awesome XHTML [14:50] <Azriel> then the application serves these pages to the client via a local webserver [14:50] <Azriel> sort of like Google Desktop's interface [14:51] <Azriel> http://jewishliturgy.org/dump/jlpdemo/Ruth.htm [14:51] <Azriel> here is some sample output, saved to our webserver [14:51] <EfraimDF> Rob, you picking this up or are we just skipping over essential info? [14:53] That all sounds fairly straight forward and a useful overview. I'll want to dig around a while and then I'll probably have some questions, details that I really don't have a handle on yet. [14:54] <Azriel> sure [14:54] <EfraimDF> feel free to ask questions -- the ML will definitely get both me and azriel [14:55] <EfraimDF> i mostly work on the XSLT/XQuery, Azriel has been doing the Java side [14:56] Will do. So, I'll go ahead and download the app and start with that. Sound good to you? [14:57] <EfraimDF> the Java app(let) is stored in svn at trunk/code/JLPDemo [14:57] <EfraimDF> the build should work like a normal ant Java build [14:58] <EfraimDF> it can be started as an app either through javaws or with java command + setting classpath [14:59] <EfraimDF> as far as I can tell, it's working as a standalone application (it's supposed to serve up books and chapters of tanach that are on-the-fly converted from the archive on the database) [14:59] <EfraimDF> we're having some search path issue when loading as an applet [15:00] <EfraimDF> if it's fixed, you'll see an announcement on the ML [15:00] <EfraimDF> (since this is our first demo of the transforms) [15:00] Having source of working code is always a good place to start. Path issues are always a bear because normally it's very straight forward and when it isn't it ISN'T. [15:00] <EfraimDF> incidentally, the reason we're doing this as an applet and not a web application is because the transforms take a lot of processing power and i'd rather a client do it than the server [15:01] That makes complete sense. [15:03] <EfraimDF> for things like the transcription app, a web app makes more sense (I think) [15:03] Unless there's more you need to add, I'm going to go ahead and sign off now. [15:03] <EfraimDF> the current demo app probably will someday morph into the end-user/developer processor [15:03] <EfraimDF> ok, thanks for stopping by [15:04] <EfraimDF> we'll post the chat logs on the wiki [15:04] <EfraimDF> (it's mostly about hebrew grammar) [15:04] Thanks. [15:06] == rtanner [n=chatzill@206.212.236.81] has left #jewishliturgy [] [15:25] == david__ [i=18d1e042@gateway/web/freenode/x-iqckpswkutmgvfds] has quit ["Page closed"]