November 2006 - developers-list

[Monetdb-developers] *start() functions
by Niels Nes 08 Nov '06

08 Nov '06

MonetDB developers As the pfstart() is now depricated, it might be a good moment to say goodbye to all these *start functions. We could simply start mapi by default in the mapi module (local or remote connections should be controlled by an monetdb env var then). Any objects to push this cleanup through the release branch? Niels -- Niels Nes, Centre for Mathematics and Computer Science (CWI) Kruislaan 413, 1098 SJ Amsterdam, The Netherlands room C0.02, phone ++31 20 592-4098, fax ++31 20 592-4312 url: http://www.cwi.nl/~niels e-mail: Niels.Nes(a)cwi.nl

3 3

Re: [Monetdb-developers] Monetdb-developers Digest, Vol 6, Issue 12
by Peter Boncz 08 Nov '06

08 Nov '06

Hi Torsten, Because Jens pointed put possible amiguities in the lighter variant (in the general case of any destination expression), we had decided to let it as it was. It turns out that the braces actually complicate parsing in the light variant (with respect to no braces at all, which is unambiguous, for which I provided a 'tidy' patch, but to which Jens also objects to for reasons I still do not understand -- but frankly I have given up on arguing for). Personally, I find it confusing to allow leaving out braces sometimes but requiring them in other cases. So I had accepted the current situation, and (sorry to say) agree with Jennie that your change is no improvement. Notwithstanding the fact that I do appreciate your effort to help out. Peter

1 0

[Monetdb-developers] Compilation problems MonetDB <-> Pathfinder
by Jens Teubner 08 Nov '06

08 Nov '06

Hi all, after a fresh checkout and install of buildtools and MonetDB, I'm getting the following compilation error in the Pathfinder runtime module: /opt/buildtools/bin/mel -I. -I/opt/MonetDB/include/MonetDB -I/opt/MonetDB/include/MonetDB/common -I/opt/MonetDB/include/MonetDB/gdk -I/opt/MonetDB/include/MonetDB/monet -I/opt/MonetDB/include/MonetDB/plain -I/opt/MonetDB/include/MonetDB/contrib -I/opt/MonetDB/include/MonetDB/mapi -I/opt/MonetDB/include/MonetDB/C -I/usr/include/libxml2 -I../compiler/include -I./../compiler/include -proto /opt/MonetDB/include/MonetDB/plain/constant.m > /opt/MonetDB/include/MonetDB/plain/constant.proto.h /bin/sh: /opt/MonetDB/include/MonetDB/plain/constant.proto.h: Permission denied make[3]: *** [/opt/MonetDB/include/MonetDB/plain/constant.proto.h] Error 1 make[3]: Leaving directory `/home/teubner/pf/cvs-head/runtime' make[2]: *** [all] Error 2 make[2]: Leaving directory `/home/teubner/pf/cvs-head/runtime' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/teubner/pf/cvs-head' make: *** [all] Error 2 teubner@notekemper14:~/pf/cvs-head> man test It turns out that constant.m in the installation directory of MonetDB is (slightly) newer than constant.proto.h. (Touching the latter solved the problem here.) Two questions arise: (1) Why is constant.m newer than constant.proto.h anyway, given that the latter apparently depends on the former? (Note: I did *not* do any dirty tricks here, such as parallel builds or so.) (2) Why is there a dependency in the Pathfinder Makefiles that try to rebuild files outside the Pathfinder module? If I'm compiling code, I'm not expecting to pollute anything outside the respective build directories. Note that I had installed MonetDB under the superuser account and, hence, wasn't able to overwrite constant.proto.h. Otherwise, the file silently would have been overwritten. The problem certainly won't be easy to reproduce, as it depends on exact timestamps here. But there's something wrong in our build setup that allows such situations to occur at all. Jens -- Jens Teubner Technische Universitaet Muenchen, Department of Informatics D-85748 Garching, Germany Tel: +49 89 289-17259 Fax: +49 89 289-17263 Computers are like air conditioners: They stop working properly if you open windows.

2 1

Re: [Monetdb-developers] [Monetdb-pf-checkins] pathfinder/runtime serialize_dflt.mx, 1.19, 1.20 serialize_sax.mx, 1.8, 1.9
by Stefan Manegold 08 Nov '06

08 Nov '06

Jan, this as well as you previous checkin went to the development trunk, hence they will NOT be in the release! Stefan On Wed, Nov 08, 2006 at 09:48:26AM +0000, Jan Rittinger wrote: > Update of /cvsroot/monetdb/pathfinder/runtime > In directory sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv20415/runtime > > Modified Files: > serialize_dflt.mx serialize_sax.mx > Log Message: > here comes the cleaning woman :) > > > Index: serialize_sax.mx > =================================================================== > RCS file: /cvsroot/monetdb/pathfinder/runtime/serialize_sax.mx,v > retrieving revision 1.8 > retrieving revision 1.9 > diff -u -d -r1.8 -r1.9 > --- serialize_sax.mx 7 Nov 2006 17:02:35 -0000 1.8 > +++ serialize_sax.mx 8 Nov 2006 09:48:23 -0000 1.9 > @@ -166,7 +166,6 @@ > static bool > sax_init (XqueryCtx* ctx, char* args, char *restag) > { > - (void) ctx; > (void) args; > ctx->driverWs = (void *) restag; /* store result tag */ > sax_init_isspecial(); > @@ -180,8 +179,6 @@ > static bool > sax_handle_startDocument (XqueryCtx* ctx) > { > - (void)ctx; > - > if ( SAXOPHONE ) > return emit_SAX_header(ctx); > else > @@ -197,7 +194,6 @@ > static bool > sax_handle_endDocument (XqueryCtx* ctx) > { > - > if ( SAXOPHONE ) > return emit_SAX_footer(); > else > @@ -239,7 +235,6 @@ > static bool > sax_handle_endElement (XqueryCtx* ctx, str prefix, str name) > { > - > if ( SAXOPHONE ) { > if ( !emit_SAX_event(ctx, "endElement",NULL,0) ) > return PROBLEM; > > Index: serialize_dflt.mx > =================================================================== > RCS file: /cvsroot/monetdb/pathfinder/runtime/serialize_dflt.mx,v > retrieving revision 1.19 > retrieving revision 1.20 > diff -u -d -r1.19 -r1.20 > --- serialize_dflt.mx 7 Nov 2006 17:02:35 -0000 1.19 > +++ serialize_dflt.mx 8 Nov 2006 09:48:23 -0000 1.20 > @@ -683,8 +683,6 @@ > static bool > xml_seqItemSeparator (XqueryCtx* ctx, int kind_before, int kind_after) > { > - (void)ctx; > - > if ((XTRACT_KIND(kind_before) < ATOMIC) && > (XTRACT_KIND(kind_after) < ATOMIC)) > stream_write (ctx->out, &s_, 1, 1); /* ' ' */ > @@ -747,7 +745,6 @@ > typed_xml_seqItemStart (XqueryCtx* ctx, int kind) > { > char *item_start = 0; > - (void)ctx; > > switch (XTRACT_KIND(kind)) { > case BOOL: > @@ -813,7 +810,6 @@ > typed_xml_seqItemEnd (XqueryCtx* ctx, int kind) > { > char *item_end = 0; > - (void)ctx; > > switch (XTRACT_KIND(kind)) { > case BOOL: > @@ -908,7 +904,6 @@ > soap_xml_seqItemEnd (XqueryCtx* ctx, int kind) > { > char *item_end = 0; > - (void)ctx; > > switch (XTRACT_KIND(kind)) { > case BOOL: > @@ -977,7 +972,6 @@ > static bool > seq_seqItemStart (XqueryCtx* ctx, int kind) > { > - (void)ctx; > (void)kind; > > /* SQL-ish mode: write start of tuple and open string */ > > > ------------------------------------------------------------------------- > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Monetdb-pf-checkins mailing list > Monetdb-pf-checkins(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins > -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

2 1

Re: [Monetdb-developers] [Monetdb-pf-checkins] pathfinder/runtime pathfinder.mx, 1.270, 1.271
by Stefan Manegold 08 Nov '06

08 Nov '06

FYI: Here's the status of `Mtest_pathfinder -r runtime/ testing/` of the CVS head at three timestamps: Tuesday, Nov 07 2006, 17:59 CET (tag information-functions_and_checkpointing-thread_before in pathfinder) 90 out of 665 tests produced SIGNIFICANTLY different output Tuesday, Nov 07 2006 23:59 CET (tag XQuery_0-14_root in pathfinder; identical with initial XQuery_0-14 release branch and tag information-functions_and_checkpointing-thread_after) 570 out of 665 tests produced SIGNIFICANTLY different output Wednesday, Nov 08 2006 9:00 CET (CVS head still/again idential with release branch) 417 out of 665 tests produced SIGNIFICANTLY different output Stefan On Wed, Nov 08, 2006 at 01:29:34AM +0000, Peter Boncz wrote: > Update of /cvsroot/monetdb/pathfinder/runtime > In directory sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv5740 > > Modified Files: > pathfinder.mx > Log Message: > - fn:doc works sgain and that should clear up quite some tests in the testweb > > > > Index: pathfinder.mx > =================================================================== > RCS file: /cvsroot/monetdb/pathfinder/runtime/pathfinder.mx,v > retrieving revision 1.270 > retrieving revision 1.271 > diff -u -d -r1.270 -r1.271 > --- pathfinder.mx 8 Nov 2006 01:00:29 -0000 1.270 > +++ pathfinder.mx 8 Nov 2006 01:29:32 -0000 1.271 > @@ -63,7 +63,6 @@ > .END pathfinder; > @mil > if (clientid() != 0) { > - drop("_pathfinder"); > ERROR("pathfinder module must be loaded in Mserver console first."); > } > module(mapi); # remote client access > @@ -464,8 +463,30 @@ > .append("_qn_prefix_uri_loc") > .append("_qn_histogram"); > > +PROC pf_checkpoint(BAT[void,str] commitBAT) : bit > +{ > + if (count(commitBAT) = 0) return false; > + commitBAT := commitBAT.access(BAT_WRITE); > + commitBAT.append("xquery_catalog"); > + commitBAT.append("xquery_seqs"); > + commitBAT.append("xquery_snapshots"); > + commitBAT.append("collection_name"); > + commitBAT.append("collection_size"); > + commitBAT.append("doc_name"); > + commitBAT.append("doc_location"); > + commitBAT.append("doc_collection"); > + commitBAT.append("doc_timestamp"); > + commitBAT.append("uri_lifetime"); > + var ok := false; > + lock_set(pf_wal); > + CATCH(ok := subcommit(commitBAT)); > + lock_unset(pf_wal); > + if (not(ok)) ERROR("XQDY0062: checkpoint failed, query aborted.\n"); > + return true; > +} > + > # the logger background thread > -PROC xquery_logmanager() : void { > +PROC pf_logmanager() : void { > var cnt := 0; > while(true) { > if ((logger_changes(pf_logger) > 100000) and ((cnt :+= 1) <= 10)) { > @@ -478,7 +499,7 @@ > sleep(90); > } > } > -fork(xquery_logmanager()); > +fork(pf_logmanager()); > > # the set of document collections (may only be accessed holding pf_short) > var collection_name; # bat[oid,str] collection name > @@ -513,9 +534,8 @@ > doc_location := new(oid,str).persists(true).rename("doc_location"); > doc_timestamp := new(oid,timestamp).persists(true).rename("doc_timestamp"); > uri_lifetime := new(str,lng).persists(true).rename("uri_lifetime"); > -} > -if (not(doc_collection.exist(TEMP_DOC))) { > doc_collection.insert(TEMP_DOC, DOCID_MIN); # hack: store next oid as in invalid TEMP_DOC tuple > + pf_checkpoint(bat(void,str).append("uri_lifetime")); > } > > # runtime collection acccess control (TRANSIENT) > @@ -1819,39 +1839,6 @@ > } > > > -@- commit stuff > - > -Update queries use the WAL for committing their changes rapidly in a single bulk I/O. > -They leave the modified master BATs dirty. > - > -A checkpointing mechanism flushes the BATs to disk periodically (or forced when under > -memory pressure -- but performance-wise it is better to do it pro-actively). > - > -The shredding of new documents uses checkpointing (i.e. direct flush and no WAL) > -directly. This is a reason why update-queries may not contain fn:add_doc() or fn:del_doc() > -statements; such queries would need two different kinds of commits, making them non-atomic. > -@mil > -PROC pf_checkpoint(BAT[void,str] commitBAT) : bit > -{ > - if (count(commitBAT) = 0) return false; > - commitBAT := commitBAT.access(BAT_WRITE); > - commitBAT.append("xquery_catalog"); > - commitBAT.append("xquery_seqs"); > - commitBAT.append("xquery_snapshots"); > - commitBAT.append("collection_name"); > - commitBAT.append("collection_size"); > - commitBAT.append("doc_name"); > - commitBAT.append("doc_location"); > - commitBAT.append("doc_collection"); > - commitBAT.append("doc_timestamp"); > - commitBAT.append("uri_lifetime"); > - var ok := false; > - lock_set(pf_wal); > - CATCH(ok := subcommit(commitBAT)); > - lock_unset(pf_wal); > - if (not(ok)) ERROR("XQDY0062: checkpoint failed, query aborted.\n"); > - return true; > -} > > > @- adding documents > @@ -2147,9 +2134,9 @@ > oid coll_oid, > BAT[lock,bat] runtime, > oid docid_base, > - bit doCommit, > lng pageFree, > timestamp ts, > + bit doCommit, > Stream s, > lng wsid) : oid > { > @@ -2266,7 +2253,7 @@ > shred_into_docBAT(docBAT, > idx.leftfetchjoin(idx_locations), > idx.leftfetchjoin(idx_names), > - colname, coll_oid, runtime, oid_nil, true, percentage, timestamp_nil, s, wsid); > + colname, coll_oid, runtime, oid_nil, percentage, timestamp_nil, true, s, wsid); > > if (isnil(coll_oid)) { > # register new bats with the logger > > > ------------------------------------------------------------------------- > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Monetdb-pf-checkins mailing list > Monetdb-pf-checkins(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins > > -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

1 0

Re: [Monetdb-developers] [Monetdb-pf-checkins] pathfinder EXTENSIONS, 1.5, 1.6
by Ying Zhang 08 Nov '06

08 Nov '06

Hello Teggy, I'd like to ask the reason why you are adding this "lighter" variant. Did I miss part of the discussion on this issue? First of all, I need to mention that, yesterday, Peter and I had (again:)) some discussion about the XRPC syntax with Jens. According to Jens, he has tried to remove the braces arround the 'Expr', since that was the decision we made during the conference call last week. But removing those braces can result in some ambiguity. So, the decision of yesterday's discussion is to just keep those braces. This should be the final decision, I think (I hope:)). I didn't have time to mail everyone about this, so I do it now. Thus your lighter variant should be in the form: execute at {URILiteral} {FunctionCall} Further, I'm wondering how useful it is to add this lighter variant explicitly. Isn't 'URILiteral' just an 'Expr'? Do you want to add this variant because the time needed for compilation can then be significantly reduced if the URI is a simple string, instead of a complex Expr? Regards, Jennie On Tue, Nov 07, 2006 at 10:10:30PM +0000, Torsten Grust wrote: > Update of /cvsroot/monetdb/pathfinder > In directory sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv15478 > > Modified Files: > EXTENSIONS > Log Message: > Added support for the syntactically ''lighter'' variant of XRPC > invocation. May only used if a literal URI is given: > > execute at URILiteral { FunctionCall } > > > > Index: EXTENSIONS > =================================================================== > RCS file: /cvsroot/monetdb/pathfinder/EXTENSIONS,v > retrieving revision 1.5 > retrieving revision 1.6 > diff -u -d -r1.5 -r1.6 > --- EXTENSIONS 31 Oct 2006 08:47:13 -0000 1.5 > +++ EXTENSIONS 7 Nov 2006 22:10:28 -0000 1.6 > @@ -120,7 +120,10 @@ > will execute the function call given in the second pair of curly > braces at the URI specified in the first pair of braces. The URI > expression may be any expression that can sensibly be interpreted > -as a string. > +as a string. A syntactically lighter variant using a literal URI > +is acceptable as well: > + > + execute at "<<URI>>" { <<FunctionCall>> } > > XQuery RPC functionality is currently only available in Pathfinder's > "milprint_summer" back-end. > > > ------------------------------------------------------------------------- > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Monetdb-pf-checkins mailing list > Monetdb-pf-checkins(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins

2 1

Re: [Monetdb-developers] [Monetdb-pf-checkins] pathfinder/compiler/parser scanner.l, 1.40, 1.41 parser.y, 1.49, 1.50
by Stefan Manegold 08 Nov '06

08 Nov '06

Teggy, I wasn't done with the branches, yet, hence your checkins are somewhere in "no man's land" and in particular mess-up my tagging schemes ... |-( As said, I'll let you know once I'm done with the branches --- and whether I manages to fix my tagging schemes which are (a.o.) crucial for allowing convenient propagation of bug fixes from the release branches to the development trunks... Stefan On Tue, Nov 07, 2006 at 10:10:30PM +0000, Torsten Grust wrote: > Update of /cvsroot/monetdb/pathfinder/compiler/parser > In directory sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv15478/compiler/parser > > Modified Files: > scanner.l parser.y > Log Message: > Added support for the syntactically ''lighter'' variant of XRPC > invocation. May only used if a literal URI is given: > > execute at URILiteral { FunctionCall } > > > > Index: parser.y > =================================================================== > RCS file: /cvsroot/monetdb/pathfinder/compiler/parser/parser.y,v > retrieving revision 1.49 > retrieving revision 1.50 > diff -u -d -r1.49 -r1.50 > --- parser.y 7 Nov 2006 07:53:59 -0000 1.49 > +++ parser.y 7 Nov 2006 22:10:28 -0000 1.50 > @@ -357,7 +357,7 @@ > %token every_dollar "every $" > %token except "except" > %token excl_equals "!=" > -%token execute_at_lbrace "execute at {" > +%token execute_at "execute at" > %token external_ "external" > %token following_colon_colon "following::" > %token following_sibling_colon_colon "following-sibling::" > @@ -2779,8 +2779,16 @@ > /* end of the Pathfinder recursion extension */ > > /* Pathfinder extension: XRPC */ > -XRPCCall : "execute at {" Expr "}" "{" FunctionCall "}" > - { $$ = wire2 (p_xrpc, @$, $2, $5); } > +XRPCCall : "execute at" "{" ExprSingle "}" "{" FunctionCall "}" > + { $$ = wire2 (p_xrpc, @$, $3, $6); } > + | "execute at" URILiteral "{" FunctionCall "}" > + { $$ = wire2 (p_xrpc, > + @$, > + (c = leaf (p_lit_str, @2), > + c->sem.str = $2, > + c), > + $4); > + } > ; > /* End Pathfinder extension */ > > > Index: scanner.l > =================================================================== > RCS file: /cvsroot/monetdb/pathfinder/compiler/parser/scanner.l,v > retrieving revision 1.40 > retrieving revision 1.41 > diff -u -d -r1.40 -r1.41 > --- scanner.l 24 Oct 2006 15:20:02 -0000 1.40 > +++ scanner.l 7 Nov 2006 22:10:28 -0000 1.41 > @@ -604,8 +604,8 @@ > yield (document_lbrace); } > "element"{_}?"{" { pushState (OPERATOR); pushState (DEFAULT); > yield (element_lbrace); } > -"execute"{_}"at"{_}?"{" { pushState (OPERATOR); pushState (DEFAULT); > - yield (execute_at_lbrace); } > +"execute"{_}"at" { pushState (OPERATOR); pushState (DEFAULT); > + yield (execute_at); } > "attribute"{_}?"{" { pushState (OPERATOR); pushState (DEFAULT); > yield (attribute_lbrace); } > "text"{_}?"{" { pushState (OPERATOR); > > > ------------------------------------------------------------------------- > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Monetdb-pf-checkins mailing list > Monetdb-pf-checkins(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins > > -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

2 2

[Monetdb-developers] Feature Freeze for upcoming MonetDB-4.14, MonetDB/SQL-2.14, MonetDB/XQuery-0.14
by Stefan Manegold 08 Nov '06

08 Nov '06

Dear all, the *feature freeze* for the upcoming release of MonetDB-4.14, MonetDB/SQL-2.14, MonetDB/XQuery-0.14 is scheduled for tomorrow, Tuesday Nov 07 2006, 18:00 (6:00pm) CET As usually, the feature freeze will be enforced by creating new release branches in CVS. After the above deadline (i.e., as soon as the new release branches have been created), *all* *bugfixes* (and *only these*) must be checked in to these new release branches, while *all* *new features* must be checked in to the development trunk (and *only there*). Please accept my/our apologies for the rather short term announcement, but I simply forgot to publicly announce the feature freeze on time... Kind Regards, Stefan -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

1 3

Re: [Monetdb-developers] [Monetdb-pf-checkins] pathfinder/runtime pathfinder.mx, 1.268, 1.269
by Peter Boncz 07 Nov '06

07 Nov '06

Hi Fabian, I have no idea, and your help is appreciated. thank you. Peter -----Original Message----- From: monetdb-pf-checkins-bounces(a)lists.sourceforge.net [mailto:monetdb-pf-checkins-bounces@lists.sourceforge.net]On Behalf Of Fabian Groffen Sent: 07 November 2006 21:18 To: monetdb-pf-checkins(a)lists.sourceforge.net Subject: Re: [Monetdb-pf-checkins] pathfinder/runtime pathfinder.mx, 1.268,1.269 Peter, How can a XQuery server be started which is bound to the ANY interface (0.0.0.0) instead of the safe, but rather useless localhost (127.0.0.1) interface with these changes? Second, how does this relate to the SQL module's way of doing things? If I recall correctly, Niels didn't do this "streamlining" because that would be problematic in some way. I guess, however (I like the streamlining behaviour) that for consistency sql_server (and maybe mapi/mil) should behave in the same way before the release. On 07-11-2006 18:11:30 +0000, Peter Boncz wrote: > Update of /cvsroot/monetdb/pathfinder/runtime > In directory sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv19894/runtime > > Modified Files: > pathfinder.mx > Log Message: > - streamlined startup (pfstart() is no longer needed) > NOTE: you must start pathfinder (i.e. 'module(pathfinder);') now in the Mserver always!!! > doing this in MIL programs on each MapiClient session only (and not beforehand in the server) > will produce errors on multiple/concurrent sessions > > - added a checkpointing thread ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Monetdb-pf-checkins mailing list Monetdb-pf-checkins(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins

1 0

Re: [Monetdb-developers] [Monetdb-pf-checkins] pathfinder/runtime pathfinder.mx, 1.268, 1.269
by Fabian Groffen 07 Nov '06

07 Nov '06

On 07-11-2006 21:17:51 +0100, Fabian Groffen wrote: > Peter, > > How can a XQuery server be started which is bound to the ANY interface > (0.0.0.0) instead of the safe, but rather useless localhost (127.0.0.1) > interface with these changes? I see I was to quick, my question has been answered already, it's now in the MonetDB.conf file. This feels like a much better place for it, but brings me to the question whether or not this should have been a general mapi setting, and not a setting for each "module" separately. Since the all are under the same port, I would expect that this setting also can only be controlled from "mapi" or whatever the wrapper is. > Second, how does this relate to the SQL module's way of doing things? > If I recall correctly, Niels didn't do this "streamlining" because that > would be problematic in some way. I guess, however (I like the > streamlining behaviour) that for consistency sql_server (and maybe > mapi/mil) should behave in the same way before the release. > > > On 07-11-2006 18:11:30 +0000, Peter Boncz wrote: > > Update of /cvsroot/monetdb/pathfinder/runtime > > In directory sc8-pr-cvs7.sourceforge.net:/tmp/cvs-serv19894/runtime > > > > Modified Files: > > pathfinder.mx > > Log Message: > > - streamlined startup (pfstart() is no longer needed) > > NOTE: you must start pathfinder (i.e. 'module(pathfinder);') now in the Mserver always!!! > > doing this in MIL programs on each MapiClient session only (and not beforehand in the server) > > will produce errors on multiple/concurrent sessions > > > > - added a checkpointing thread

1 0