IRC log for #brlcad on 20160904

00:54.31 *** join/#brlcad teepee_ (~teepee@unaffiliated/teepee)
01:32.02 *** join/#brlcad KimK (~Kim__@2600:8803:7a85:6d00:31c9:b100:c4bc:40cd)
02:08.51 brlcad starseeker: the current attribute system is defined as a string=>string map, so introspection into the value is only arguably defined for the standard attributes (for which this tolerance setting isn't necessary, doesn't make any sense)
02:11.50 brlcad I would think it perfectly safe, perhaps even desirable, to not have -t have any bearing on attributes, to not introspect attributes -- just compare and report when they differ
02:13.45 brlcad it's not in mm because of asc -- it's because that is (intentionally) defined as the base unit for all storage and in-memory processing, that requested units merely affects presentation
02:15.19 brlcad this is done for a variety of reasons, one example being that you wouldn't want to re-encode all values just by running setting different working units. that would be a source of error accumulation and be nearly impossible to manage code-wise
02:15.24 brlcad there are other reasons.
02:16.43 brlcad so for clarity, I think your last point is spot on -- that we shouldn't do numerical difference filtering on attributes (at least until a schema defines the value as a floating point number, e.g., binary attributes)
02:30.01 starseeker brlcad: OK, I'll have to check what we do there - maybe need to do some work besides updating docs, not sure
02:35.03 starseeker yeah, structurally everything boils down to avs diffing, and I don't distinguish at the lower level between params and attrs
02:35.49 starseeker even for primitives it's a problem, if some of them have numerical parameters that aren't dimensions
02:36.47 starseeker almost begs for a per-primitive functab function that can intelligently diff two objects of the same type
02:38.04 starseeker urk
02:38.44 starseeker doing this right will take more time than I have available, certainly before the next point release or two - should we just disable the tolerance feature until we can do it right?
02:39.19 starseeker winces a little at yanking it immediately after announcing it, but it might be worse to leave it in a confusing state...
02:43.07 Notify 03BRL-CAD:starseeker * 68812 brlcad/trunk/TODO: Have man page for brlman now
02:45.17 starseeker yeah, I'm remembering now - doing "smart" primitive diffing was more than I had time to do last time as well, so I added the numerical filter as a way to at least be able to cull really small differences out of the results, however crudely
02:45.41 starseeker the unintended consequences of that might not be worth it though
03:12.06 starseeker blinks - I'll be darned, an MIT licensed cross platform text editor with both graphical and terminal modes that's written in C and lua: http://foicica.com/textadept/index.html
04:17.50 brlcad starseeker: okay, that all makes sense
04:21.14 brlcad I think the only issue is one of how to explain it in the manual page, and what that means for attrs if anything at all
04:22.52 brlcad I think saying it's unitless will be unexpected/confusing, inconsistent, and potentially conflicting down the road ; saying it's "mm" is fully inadequate and wrong for some fields
04:25.18 brlcad the simple fix might be to just state what the tolerance is intended to describe (for comparing floating point geometry parameters) and caveating that it currently also applies to all other numeric fields as well (which should be fine because integral change is usually far greater than tol)
04:26.36 brlcad so they'll have to be really cautious when working with tol > 1
04:27.16 brlcad I can take a stab at it later and you can let me know how it reads
04:27.34 brlcad textadept looks really cool
04:28.57 brlcad also, fyi, I'd rather see a json-style callback that would let us universally diff any two objects
04:29.20 brlcad there's lots of good examples of json diffing that could be modified for use here
04:40.56 *** join/#brlcad boj (~boj@2001:250:3c00:2074:b1e8:ee29:561b:e3e2)
06:58.49 *** join/#brlcad ickby_ (~stefan@x5d847781.dyn.telefonica.de)
07:22.30 *** join/#brlcad ickby (~stefan@x5d847781.dyn.telefonica.de)
07:31.15 *** join/#brlcad ickby (~stefan@x5d847781.dyn.telefonica.de)
08:05.27 *** join/#brlcad Caterpillar (~caterpill@unaffiliated/caterpillar)
09:53.07 *** join/#brlcad ickby_ (~stefan@x5d847781.dyn.telefonica.de)
09:53.53 *** join/#brlcad ries (~ries@D979C7EF.cm-3-2d.dynamic.ziggo.nl)
09:56.32 *** join/#brlcad ries (~ries@D979C7EF.cm-3-2d.dynamic.ziggo.nl)
10:55.07 *** join/#brlcad amarjeet (~Amarjeet@101.211.217.123)
11:10.44 *** join/#brlcad ries (~ries@D979C7EF.cm-3-2d.dynamic.ziggo.nl)
11:13.05 *** join/#brlcad ries (~ries@D979C7EF.cm-3-2d.dynamic.ziggo.nl)
11:17.47 *** join/#brlcad amarjeet (~Amarjeet@101.211.217.123)
11:31.05 *** join/#brlcad amarjeet (~Amarjeet@101.211.217.123)
11:48.36 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
11:55.24 *** join/#brlcad amarjeet (~Amarjeet@101.211.217.123)
12:40.26 *** join/#brlcad teepee_ (~teepee@unaffiliated/teepee)
12:46.46 *** join/#brlcad amarjeet (~Amarjeet@101.211.244.182)
13:17.03 *** join/#brlcad ChanServ (ChanServ@services.)
13:17.03 *** mode/#brlcad [+o ChanServ] by tepper.freenode.net
13:27.06 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
13:41.19 *** join/#brlcad amarjeet (~Amarjeet@101.211.244.182)
13:57.38 *** join/#brlcad amarjeet (~Amarjeet@101.211.244.182)
14:19.04 *** join/#brlcad ickby_ (~stefan@x5d847442.dyn.telefonica.de)
14:36.52 *** join/#brlcad amarjeet (~Amarjeet@101.211.216.193)
14:42.59 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
14:47.02 *** join/#brlcad merzo (~merzo@195.128.182.31)
14:52.33 *** join/#brlcad merzo (~merzo@195.128.182.31)
15:03.35 *** join/#brlcad amarjeet (~Amarjeet@101.211.243.86)
15:41.16 *** join/#brlcad ickby_ (~stefan@x5d847442.dyn.telefonica.de)
15:44.10 *** join/#brlcad amarjeet (~Amarjeet@101.211.243.86)
16:02.24 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
16:20.53 *** join/#brlcad merzo (~merzo@195.128.182.31)
17:02.57 *** join/#brlcad m5w (~matthew@108-203-5-240.lightspeed.rlghnc.sbcglobal.net)
17:05.24 *** join/#brlcad amarjeet (~Amarjeet@101.211.243.86)
17:10.32 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
17:27.55 *** join/#brlcad m5w (~matthew@108-203-5-240.lightspeed.rlghnc.sbcglobal.net)
17:52.44 *** join/#brlcad amarjeet (~Amarjeet@101.211.243.86)
18:28.45 *** join/#brlcad m5w (~matthew@108-203-5-240.lightspeed.rlghnc.sbcglobal.net)
18:37.13 *** join/#brlcad amarjeet (~Amarjeet@101.211.230.106)
19:00.38 *** join/#brlcad amarjeet (~Amarjeet@101.211.230.106)
19:21.06 *** join/#brlcad m5w (~matthew@108-203-5-240.lightspeed.rlghnc.sbcglobal.net)
19:44.15 *** join/#brlcad shubham (6719e766@gateway/web/freenode/ip.103.25.231.102)
20:10.43 *** join/#brlcad KimK (~Kim__@2600:8803:7a85:6d00:31c9:b100:c4bc:40cd)
23:19.46 *** join/#brlcad teepee_ (~teepee@unaffiliated/teepee)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.