On 1/21/19 4:22 PM, Chase wrote: > I did see this, but doesn't the preprocessor define instances of v_io_buf as > it's ptr version? Or is that only
Only if if C_API is defined. Otherwise v_io_buf is used. <within that file? And what about the references to dtxlate stuff from the exported version of DtMmdb? I have no idea where to start there. C_API is defined during compilation (all I did was move DtMmdb to lib and disable building of mmdb. I don't know as I've never looked into the mmdb stuff. It's clear there are things going on that you do not quite understand yet -- for example why there are two mmdb libs where one symlinks to another. You should figure out why that was done first before changing anything... Maybe they are built with different defines so that the lib exports different behavior to different software components for some reason. C_API is a clue, there may be more - read the Imakefiles and look at the options passed when building them. You will have to figure out how this is all being built and used now before you can re-implement it properly... Also, I was puzzled by this comment earlier: >>> would take me a long time to write makefile.ams for due to the fact that >>> as far as I can tell, automake doesn't have an easy way to implement >>> macros, and thus I would have to write a plain Makefile direction for What do you mean that automake doesn't have an easy way to make macros? What macros? -jon > > > Thank you for your time, > -Chase > > ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ > On Monday, January 21, 2019 4:24 PM, Jon Trulson <j...@radscan.com> wrote: > >> On 1/21/19 2:07 PM, Chase via cdesktopenv-devel wrote: >> >>> Hi all, >>> The further and further I delve into the document building system in >>> CDE, the more of a trainwreck I see that it is. >> >> :) >> >>> I have been working on the autotools port, and one point of interest I >>> am running into continuously is the issue of downsizing duplicate code. >>> I have successfully done this with the replacement of nsgmls, however >>> due to the in progress replacement of the sgml docs with xml docs, this >>> patch was rejected. Matthew said he would do this, however due to real >>> life obligations and the size of the task, I don't believe that this >>> will be done soon. No matter, nsgmls was simple enough to make a >>> makefile.am for. >> >> Good - I think as you get this stuff building, it would be unwise to >> make other changes that affect behavior.... We can address that stuff later. >> >>> However, what isn't simple is the make file for dtinfo/mmdb. To make it >>> short, this entire directory is just simlinks to dtinfo/DtMmdb, this >>> would take me a long time to write makefile.ams for due to the fact that >>> as far as I can tell, automake doesn't have an easy way to implement >>> macros, and thus I would have to write a plain Makefile direction for >>> each of the symlinked files. I tried to implement a merge of dtinfo/mmdb >>> and dtinfo/DtMmdb, but I have hit these two linker errors that I am >>> stumped on (log file attached). Could anyone help me with this? >> >> So do you know why these are two separate dirs for building? I don't, >> but usually there is a reason, like some sort of translation of symbols >> and the like. >> >> At any rate, if you look at programs/dtinfo/DtMmdb/object/pstring.h you >> will see that this member variable is not available if everything is >> built with C_API defined, which C_API should always be defined. >> >> So I am guessing that when /programs/dtinfo/dtinfo/src/Marks/Mark_mmdb.C >> was compiled, C_API was not defined as it should be. This would then >> cause pstring::v_io_buf_ptr to be used instead, which you should see if >> you run 'nm' on the programs/dtinfo/DtMmdb/libDtMmdb.so.2.1 library. >> >>> TLDR; Need help with optimization patch, can provide WIP patch if needed >> >> TLDR; ensure C_API is defined when Marks/*.C is compiled...? >> >> -jon >> >>> Thank you for your time, >>> -Chase >>> >>> cdesktopenv-devel mailing list >>> cdesktopenv-devel@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/cdesktopenv-devel >> >> -- >> >> Jon Trulson >> >> "The Party told you to reject the evidence of your eyes and ears. >> It was their final, most essential command." >> >> -- 1984 >> >> cdesktopenv-devel mailing list >> cdesktopenv-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/cdesktopenv-devel > -- Jon Trulson "The Party told you to reject the evidence of your eyes and ears. It was their final, most essential command." -- 1984 _______________________________________________ cdesktopenv-devel mailing list cdesktopenv-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/cdesktopenv-devel
