[developers] logon demo giving 502 proxy error

[Dan Flickinger]

Stephan (and Joshua) -

I can't seem to reproduce this behavior on my machine.  When I load ERG 1214 and then execute the (tsdb::reconstruct ...) expression pasted from your email message, I get the following result:

((TSDB::ROOT_STANDARD
  (1922 TSDB::SB-HD_MC_C -3.22511 0 4
   (1891 TSDB::HDN_BNP-PN_C 0.0930572 0 1
    (1890 TSDB::N_SG_ILR 0.135806 0 1 #))
   (1921 TSDB::HD-AJ_SCP_C -3.21834 1 4
    (1920 TSDB::HD-CMP_U_C -0.350811 1 3 # #)
    (1917 TSDB::VP_SBRD-PRD-PAS_C -1.81916 3 4 #))))
 0 :SPONSOR "`ROOT_STANDARD'")

I was a little suspicious that this good result might have been to do with some setting on my laptop, so I did the same thing on the `cypriot' desktop, which has a "pure" LOGON repository, and I got the same correct behavior.

So I'm not sure how to offer help with the puzzle of the web demo.  I do find it a little surprising that the file erg/www/setup.lsp says that lkb::root_informal corresponds to "$root_standard" instead of "$root_informal", but perhaps you can explain that, and conclude that this anomaly is not related to the unwanted behavior.

 Dan

________________________________________
From: developers-bounces at emmtee.net <developers-bounces at emmtee.net> on behalf of Stephan Oepen <oe at ifi.uio.no>
Sent: Monday, October 19, 2015 2:22 PM
To: Joshua Crowgey
Cc: developers at delph-in.net
Subject: Re: [developers] logon demo giving 502 proxy error

> I'm getting a 502 "Proxy Error" when I try to use the on-line demo for the ERG:

thanks for your report, joshua.

it appears that your particular input ‘kim made sandy run.’ exposed
two suprising properties in the ERG on-line demonstration: (a) PET
returns a derivation that [incr tsdb()] and the LKB cannot
reconstruct, and (b) the visualization code failed to handle this
situation robustly (now corrected).

as for (a), it may well just be a subtle disagreement between the ERG
configurations used on either side, but i fail to see it.  so, in part
for my own records, the following derivation:fails when reconstruction
of the start symbol is requested (which is actually not necessary in
this context):

0[7]: (TSDB::RECONSTRUCT
          "(root_standard (1922 sb-hd_mc_c -3.22511 0 4 (1891
hdn_bnp-pn_c 0.0930572 0 1 (1890 n_sg_ilr 0.135806 0 1 (82 kim 0 0 1
(\"kim\" 67 \"token [ +CARG #1=\\\"kim\\\" +CLASS alphabetic [ +CASE
non_capitalized+lower +INITIAL + ] +FORM #1 +FROM #2=\\\"0\\\" +ID
*diff-list* [ LAST #3=*list* LIST *cons* [ FIRST \\\"1\\\" REST #3 ] ]
+PRED predsort +TICK bool +TNT null_tnt [ +MAIN tnt_main [ +PRB
\\\"0.37500149999999999\\\" +TAG \\\"VB\\\" ] +PRBS *null* +TAGS
*null* ] +TO \\\"3\\\" +TRAIT token_trait [ +HD token_head [ +LL ctype
[ -CTYPE- string ] +TG string +TI #2 ] +IT italics +LB bracket_null
+RB bracket_null +UW - ] ]\")))) (1921 hd-aj_scp_c -3.21834 1 4 (1920
hd-cmp_u_c -0.350811 1 3 (1908 v_pst_olr -0.30915 1 2 (121 make_prd_v1
-0.184629 1 2 (\"made\" 65 \"token [ +CARG #1=\\\"made\\\" +CLASS
alphabetic [ +CASE non_capitalized+lower +INITIAL - ] +FORM #1 +FROM
#2=\\\"4\\\" +ID *diff-list* [ LAST #3=*list* LIST *cons* [ FIRST
\\\"2\\\" REST #3 ] ] +PRED predsort +TICK bool +TNT null_tnt [ +MAIN
tnt_main [ +PRB \\\"0.69666410000000001\\\" +TAG \\\"VBN\\\" ] +PRBS
*null* +TAGS *null* ] +TO \\\"8\\\" +TRAIT token_trait [ +HD
token_head [ +LL ctype [ -CTYPE- string ] +TG string +TI #2 ] +IT
italics +LB bracket_null +RB bracket_null +UW - ] ]\"))) (284 sandy_a1
-0.115294 2 3 (\"sandy\" 64 \"token [ +CARG #1=\\\"sandy\\\" +CLASS
alphabetic [ +CASE non_capitalized+lower +INITIAL - ] +FORM #1 +FROM
#2=\\\"9\\\" +ID *diff-list* [ LAST #3=*list* LIST *cons* [ FIRST
\\\"3\\\" REST #3 ] ] +PRED predsort +TICK bool +TNT null_tnt [ +MAIN
tnt_main [ +PRB \\\"1\\\" +TAG \\\"JJ\\\" ] +PRBS *null* +TAGS *null*
] +TO \\\"14\\\" +TRAIT token_trait [ +HD token_head [ +LL ctype [
-CTYPE- string ] +TG string +TI #2 ] +IT italics +LB bracket_null +RB
bracket_null +UW - ] ]\"))) (1917 vp_sbrd-prd-pas_c -1.81916 3 4 (1916
hd_optcmp_c -0.253319 3 4 (1910 w_period_plr -1.03543 3 4 (1909
v_pas_odlr -0.550822 3 4 (320 run_v2 0 3 4 (\"run.\" 68 \"token [
+CARG \\\"run\\\" +CLASS alphabetic [ +CASE non_capitalized+lower
+INITIAL - ] +FORM \\\"run.\\\" +FROM #1=\\\"15\\\" +ID *diff-list* [
LAST #2=*list* LIST *cons* [ FIRST \\\"4\\\" REST *cons* [ FIRST
\\\"5\\\" REST #2 ] ] ] +PRED predsort +TICK bool +TNT null_tnt [
+MAIN tnt_main [ +PRB \\\"0.96471700000000005\\\" +TAG \\\"NN\\\" ]
+PRBS *null* +TAGS *null* ] +TO \\\"19\\\" +TRAIT token_trait [ +HD
token_head [ +LL ctype [ -CTYPE- string ] +TG string +TI #1 ] +IT
italics +LB bracket_null +RB bracket_null +UW - ] ]\")))))))))")
 0[7]: returned NIL

dan, in case you have read this far: is there an obvious reason why
the above should fail to unify with ‘root_standard’ in 1214?

—for the on-line demonstration, i am now working around this problem
by not requesting validation of the start symbol on the [incr tsdb()]
side.

best wishes, oe