From gordon.dyke@bluewin.ch Wed Jul 24 13:23:29 2002 Return-Path: X-Sender: gordon.dyke@bluewin.ch X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-8_0_7_4); 24 Jul 2002 20:23:29 -0000 Received: (qmail 6365 invoked from network); 24 Jul 2002 20:23:24 -0000 Received: from unknown (66.218.66.216) by m1.grp.scd.yahoo.com with QMQP; 24 Jul 2002 20:23:24 -0000 Received: from unknown (HELO mta3n.bluewin.ch) (195.186.1.212) by mta1.grp.scd.yahoo.com with SMTP; 24 Jul 2002 20:23:23 -0000 Received: from oemcomputer (213.3.164.62) by mta3n.bluewin.ch (Bluewin AG 6.5.026) id 3D3D0B8D00038842 for lojban@yahoogroups.com; Wed, 24 Jul 2002 22:23:22 +0200 Message-ID: <018e01c23350$150a6c00$73a1ca3e@oemcomputer> To: "jboste" References: <5.1.0.14.0.20020723195058.030913c0@pop.east.cox.net> <5.1.0.14.0.20020723025544.032cba90@pop.east.cox.net> <4.3.2.7.2.20010730221611.00b10c00@pop.cais.com> <5.1.0.14.0.20020723025544.032cba90@pop.east.cox.net> <20020723103956.E28971@miranda.org> <5.1.0.14.0.20020723195058.030913c0@pop.east.cox.net> <5.1.0.14.0.20020724122649.032e7ec0@pop.east.cox.net> Subject: Re: [lojban] to-do list (was Re: New Members, Board of Directors, other LogFest results) Date: Wed, 24 Jul 2002 22:23:31 +0200 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 From: "G. Dyke" X-Yahoo-Group-Post: member; u=81437350 X-Yahoo-Profile: gregvdyke cu'u la lojbab > Alice was a somewhat larger group effort, but from what I gather, most of > the work was done by one or two people who almost certainly spent more than > 40 hours on what they did over a period of time. So? It got done, didn't it? > > >Do you think that real dictionaries are done by individuals with word > >processors or text editors? I seriously doubt it. > > Actually, at least until the last few years (which I can't testify to), > they have been done on 3x5 cards, (or automated versions thereof that use, > indeed, simple text editors or even cruder data bases - the most common > computer tool for documenting a new language has been freeware called > Shoebox that was developed by people at Summer Institute of Linguistics, > and it basically is a data base that is a computerized shoebox of 3x5 > cards). Shoebox's editing capabilities, when last I looked, are below > those of most text editors. > > The premium dictionaries that expect to produce multiple editions have gone > to developing their own in-house specialized tools suited specifically for > the needs of the project. From what I've read, these tools are major > software development efforts taking man-years. The developers also get to > dictate the hardware that will be used, or to be able to rely on management > doing so, so they don't have to design for cross-platform use; we're stuck > with people wedded to Unix/Linux, to Windows, and to Macs in mixed > proportions, with people apparently unable or unwilling to consider the > needs of those using other platforms (and I don't pretend that I am any > better about this than others). Get everyone at Logfest together and show them how to use Cygwin or their OSXTerminal ;-) > > The most common tool used to develop cross-platform stuff these days seems > to be Java, which also, so I've been told, makes net-based Java applets > among the most dangerous security threats to a home computer; we have > Internet Java disabled on all but one of our computers, and I allow them > only with a security prompt. Yeah, but writing Java is hard (especially for a team effort when different people have different Ideas about how Object Oriented the code should be) Perl is also cross-platform... > (which is the way Nora has > historically done most of HER Lojban work, on the train to and from work - > we got a laptop this year with that in mind, but she hasn't even had time > to set up a user area yet). You should sue Bill Gates for that ;-) > >I'm trying to get people to accept the idea of working on the dictionary > >in a sane, community-based fashion, with special tools designed to ease > >the burden. > > That's fine. But who will write the tools AND GET THEM DONE. LLG can't > rely on promises, and we certainly can't plan to coordinate a big volunteer > effort based on software, when we don't even have specs for what they are > supposed to do. I can't write a spec for what we need, because at this > point I don't have the imagination to figure out what would work for the > activity-level and long-term-consistency of most of our volunteers. How complicated can a dictionnary spec be? {gismu : rafsi : keyword : clue : official definition : and/or more complete definition based on official definition} or {cmavo : rafsi : selma'o** : keyword : clue-gismu : official definition : and/or more complete definition based on official definition } or {lujvo : components : gismu : place structure} followed by {examples : associated words : date last modified : person by whom it was modified : comments} Perl should make a tolerable 3x5 shoebox based on this sort of spec. Incidentally, Jay, why do you have an eat gismu/eat cmavo program associated with your flash card program? then all people have to do is workout how to use CVS and we're up-up and away oh yeah, add a page to the wiki called examples. When you think of a good example for a little-used word, put it there **incidentally, how can all gismu be of selma'o BRODA when they are not cmavo? > In that sense, the pretty version you made up IS "the dictionary". ok then, let's call it "the mini-dictionnary" and call it done. > 1997 minutes specified the "next book" will be a mini-dictionary, and > indicated its intended content. It was put as 8th on my list of priorities > in 1998. I think it's moved up a couple of notches, but Nick's work moved > his books to higher priority in the publishing stream. Resuming JL/LK is > also higher priority. The full dictionary will come after the > mini-dictionary (which I'll admit you could probably produce for us faster > than I could at this point). The minuites between 1995or6 and 2000 are not available anywhere that I've looked. Re: JL/LK I find that anything I do want to publish in/about lojban, I send to the list or put up on the wiki immediately; publishing that on a regular basis is a huge task which would be likely to anihilate all others (or end up being published yearly) > > > Now that I've pried another secret out of you, I can prepare > >that much better. It would be very easy to reproduce that beautified PDF > >with whatever corrections and errata you'd like fixed, and then you could > >put your stamp of approval on it and tell people that there is an official > >Lojban dictionary, there just isn't currently funding to print it. (And > >that would likely make it significantly easier to get donations, if people > >saw what their donating was going to do.) > > It isn't that high on the list even if it gets done. Well, actually, if it > gets done, it will de facto move up the priority list. But I don't have > time to work on corrections/errata. I just realized yesterday that Nora > was supposed to do indexing for Nick's books, which she never did and I > don't think any one else did, though I haven't looked lately. My own > priority after business matters that never get done as it is, is getting > our address list up to date so that I can put out a JL/LK that will get to > the people who have paid for it; I haven't had time for that either. What is on your priority list? apart from getting the mailing list sorted? And anyway, there is no way that any single person can "ok" the whole dictionnary in a reasonable amount of time. > > Correct. > > >But when CVS seems to be beyond a number of Lojbanists, > > You've finally realized this!!! I posted on that, but I appear to have suffered from cowanitus in that Jay is as yet the sole beneficiary of my wisdom "I don't think CVS _should_ be beyond anyone, wincvs is excessively complicated looking. Using Cygwin with xemacs (or whatever is needed for editing and viewing or with the cvs directory shortcutted to the desktop) is so easy even I figured it out; all you need to know is cvs login cvs co cvs update cvs commit along with how to get out of vim" Why is CVS beyond a number of Lojbanists? Apart from the fact that the guys who wrote it made it so intuitive that they didn't realize how complicated it seems to a beginner? > And that cooperating framework will likely have to be AOL-istically > simple. Hence flat text files. No-one said anything against flat text files, but they'll have to include a few > > > But publishing the current dictionary files in print, is far more than we > > > can afford. If publishing Nick's books increases our revenue stream, that > > > could change in a year or two. > > > >Just because something official exists doesn't mean you've got to print it > >and sell it yourself! > > If it is published, then people will send orders to me. > > >All the LLG would need to do is distribute an official PDF of the dictionary, > >and say that printed copies of the unmolested PDF are also official. Then > >individuals could go to Kinko's, or use their own printer. Or maybe some > >Lojbanist would print off the PDF on request and sell it. > > They can do that now. But it isn't official. And I don't think we should > put the "official" label on things that are ad hoc, even if they are likely > to be ad hoc for a long time. We set a standard for ourselves with the > quality of CLL, and we have to live up to that standard with the other > baseline books. Otherwise, I might as well just stick with the printed > word lists - which people have NOT found satisfactory. > > > > We could talk about publishing a set of materials on CD-ROM, but my > > > understanding is that CD-ROM dictionaries are already becoming passe > > > because on-line lookup is more convenient for those who need convenience, > > > and the download time for the current file is quite short. Then we should publish a program which makes lookup of words simple, the equivalent of searching a flat text file with a text editor, but beautified > > > >And if they don't follow through, what happens? > > The mild semblance of order in this disjoint organization, that I have > maintained with difficulty, ceases. I'm sorry, but calling "not doing anything 'cos we probably won't get it done" is no more "semblance of order" than having 50 half-done projects > >Right now, you're signed up for everything on the list, and you're not > >getting any of it done. How can things get any worse? > > At least now I know what is and is not getting done. And since I'm legally > responsible, that is something I find necessary. I'd like to know whether all this legal-shmegal business is really important, this is a problem that obviously exists, but I can't get my head around the why of it. > > >You Can Not Get Everything Done Yourself. Even if you worked Lojban as > >a paid, fulltime job, _and_ your hobby, you *still* wouldn't be able to > >do everything. > > > >Delegation is the most important thing a manager can learn. And as the > >President/CEO of the LLG, you're the manager. > > I manager with no dependable resources has nothing to manage. > If you like, we'll cludge-up a program which reminds everyone that we should try to be dependable ressources once a week ;-))))) > >So Delegate. What on earth have you got to lose? > > I've tried with less important tasks, and been bitten. But ultimately it > is up to the voting membership, if they think I can delegate more than I > am. That is one reason they are there. Such as? > What do I stand to lose? 16 years of time and emotional investment in this > project if it falls apart. I think the language will survive my personal > involvement now. I'm not sure the organization is so solid. Hey, we *are* all behind you, although we could get half-through a dictionnary and all say "*** Lojban, I'm learning thinkit's binary language instead". You still would be no worse-off than you are now It seems that you intend to "personnally" review anything before it becomes official. The only problem will be that stuff can be done quicker than you can ever review it. Greg