2008-04-24.log

--- Day changed Thu Apr 24 2008
kanzureso that agx-get specifies the exact gui/cli library, but provides functions that the scripts call00:00
kanzureto do #2 I want to do a review of common gui/cli libraries and the input parameters that they need00:00
kanzurethis way I don't screw us over00:00
kanzureso once I do that, I'll write up a python interface layer for that purpose and publish the specs etc.00:09
kanzurethen, uh, we'll start to write some interpreters for parts of the (skdb package file)'s metadata file. 00:09
kanzurewe'll also release the first agx-get code 00:09
kanzurei think.00:09
kanzureanother idea I had was to use yaml to serialize io objects .. but then you start getting into weird territory that might be fun to try out with, I dunno, GNU Hurd or something about the time Duke Nukem Forever comes out ... 00:10
fennyou mean use yaml as a communication protocol?00:11
kanzurenot so much00:11
kanzureuh, it's a stupid idea, I'm kinda embarrased to explain00:11
kanzurebut00:11
kanzurewhat if we could assume a java vm or something00:11
kanzureand then we could have yaml serialized objects that encode the interfaces themselves for user input and so on00:12
kanzureit's not too implementable.00:12
kanzureI'm going with the other method.00:12
kanzuresome good places to review: the HTML form specs - what are all of the types of form elements? You have text, big fields, select, radios, buttons, character input, line input, file input (stdin), ...00:13
kanzureheck, scripted input00:13
fennwow there's a lot of traffic on hplusroadmap00:13
kanzureare you being sarcastic?00:13
fennno00:13
kanzurebecause it's mostly me :)00:13
fennright00:14
kanzureother gui libraries worth reviewing?00:15
kanzurencurses, I suppose00:15
kanzuregtk, maybe00:15
fennw3c is the place to read about html form specs00:15
fenndunno why you want that though00:15
kanzurehttp://www.gtk.org/documentation.html00:16
kanzurehttp://www.cs.utk.edu/~vose/c-stuff/ncurses.html00:16
fenni like gtk00:16
kanzuredoesn't matter if it's HTML or GTK or Java Swing really00:16
kanzurethe point is that there is a set of parameters that they all use00:16
kanzureso I want a broad list of them00:16
kanzurehttp://www.w3.org/TR/html401/interact/forms.html00:16
kanzurehttp://www.jformdesigner.com/ heh00:16
kanzurehttp://www.jgoodies.com/ aha00:16
fennyou want a ui abstraction layer00:17
kanzureyes00:17
fenni dont think it's going to happen any time soon00:18
kanzureno reason to demand that all yaml-interpreters use just GTK or some bullshit like that00:18
kanzure(I happen to like gtk too)00:18
kanzurebecause remember00:18
kanzureI want cli too00:18
fennwhat you can do instead is provide an api, and link various ui's to that00:18
kanzureI think there's a gtk-cli version00:18
kanzurewait, what?00:18
kanzurehow is that different00:18
kanzurewe'd have to write a binding for GTK to the API, right?00:19
kanzureand then the scripts would use api calls of course00:19
fenna ui abstraction layer would be to design the ui and then export it to various formats, like desktop application (gtk or wxwindows) and html00:19
fenni guess00:19
fenni might be making this all up00:19
kanzureI see00:19
fennanyway, publishing an API is the standard way to do it00:19
kanzureyep00:19
kanzureso that's what I want to do00:19
kanzureso that we don't have to commit to one gui/cli lib00:19
kanzurethis way that's left up to the end user00:20
kanzurebut at the same time, it's stupid to restrict it 00:20
fennto restrict what?00:20
kanzureso we can at least look at the current standards across the gui/cli libraries00:20
kanzurerestrict parameters that the api accepts00:20
kanzurefor example, you don't just say00:20
kanzure'ask for user input'00:20
kanzureyou might say 'getchar'00:20
kanzureor 'get a line' 00:20
fennof course00:20
kanzureand if you are in a GUI mode, it's even more complicated00:20
fennthat doesnt have anything to do with the API though00:20
kanzuresure it does00:20
kanzuregetaline() versus getaline(max length = 5)00:21
fennAPI is like, a list of functions and the type of variables they accept00:21
kanzureso what if you don't have max length as a pararameter00:21
kanzureso the idea is to make sure we accept common parameters00:21
kanzurewhere 'common' is defined by the other specs I've been linking to 00:21
kanzuredoes it make sense00:23
kanzureif not, shoot me down now, I have to log off (but not before saving those other specs to the hdd for in school viewing and maybe finally getting some work done)00:23
fennit just sounds like you're fuzzy on what an API is00:23
kanzurewe'll provide a set of functions00:24
kanzuregetchar, getline, gettext, getfile, stdin, whatever00:24
fennthose aren't the sort of functions though00:24
kanzurebut the method of implementation (gui, cli) will be up to agx-get00:24
kanzurehm00:24
kanzurewhat functions are you considering?00:24
fennits more like, updatedb(); check-keys(); parse-file()00:24
kanzurethat's agx-get00:25
kanzuresigh00:25
kanzureokay, so it tries to parse the file00:25
kanzurecatches the exception00:25
kanzuregoes looking for the plugin to fix it, which would provide the class (the yaml is the serialized object that fits that class)00:25
fennwait, what is "it"?00:25
kanzureit = agx-get00:25
kanzureyou're describing agx-get00:25
fenn~[6~[6~[6~[6~[6~[6~[6~[6~[6~what's the gui for?00:25
kanzurenot for me, 00:26
kanzureI like cli00:26
kanzurebut that's my preference00:26
fenni mean, its an interface to what00:26
kanzureto agx-get00:26
kanzureagx-get will read skdb metadata files00:26
kanzureright?00:26
fennyeah, and agx-get presents the basic usage functions as its API00:26
kanzuremaybe - let me type out what I was thinking first, then I'll consider that00:27
kanzureonce agx-get installs a plugin to read the yaml file, the plugin gains control and gets to ask the user questions or whatever (this can be bypassed by parameters of course)00:27
fenngetline is part of the client code, which then gets parsed and turned into function calls to the API00:27
kanzureso it would have a way of saying "well, there's this list here, which one(s) would you like? or all of them?"00:27
kanzureso the plugin would call getChar() or whatever to get the information that the user types in00:28
kanzurehm00:28
kanzurenow, in your switch-aroo,00:28
fennew00:28
kanzurewhy would getline "then get[s] parsed and turned into function calls to the API"00:29
kanzurewhere getline is a part of the client code00:29
fennthat's not very flexible, what if you switch to a radio button instead? then you have to go and change getChar() in every plugin00:29
kanzureand the API is supposed to be the ui stuff00:29
kanzurehuh?00:29
kanzureI don't understand that00:29
kanzurethe plugins are individually responsible for their interface00:29
fennno, API is not the UI it's the interface between UI and library code00:29
kanzurehow is it that you think that00:30
kanzureerm00:30
kanzureswitching to a radio button?00:30
kanzurewhy would you do that00:30
kanzureohh00:30
kanzureyou mean00:30
fennbecause you're writing a GUI00:30
kanzureswitching a plugin's representation00:30
kanzurewould mean it changes for all of them that rely on that plugin00:30
kanzurefor example00:31
kanzurepackage chair has (conveniently?) a dependency on a plugin to parse chair-type metadata files00:31
kanzureand another package entirely, ed302941 also has that !! yaml statement for the chair-type00:31
fenni never really expected a plugin to directly ask the user for input (because dpkg always pissed me off when it got stuck halfway through a big install waiting for me to press enter)00:31
kanzurethen, changing the plugin's code to ask the user for stuff and so on - well - then it would change it for anybody using either chair *or* ed30294100:32
kanzurefenn: that's fixed with passing parameters to dpkg ...00:32
fennnot if you dont know you have to pass any parameters00:32
kanzureunless it's a genuine 'uh oh and halt' dpkg error00:32
fennits stuff like 'this version of libpam does blahdyblah and is a possible security risk if you are running foobar v123' and i generally have no idea what they're talking about00:32
kanzureI still think you are wrong --> "that's not very flexible, what if you switch to a radio button instead? then you have to go and change getChar() in every plugin"00:33
kanzuregetChar() in every plugin? 00:33
fennthe point is getChar shouldnt be in the plugin at all00:33
kanzurethat's not true at all00:33
kanzureright00:33
kanzureso00:33
kanzurelet's say it's gtkGetChar()00:33
kanzuresome object for a gui or something00:33
kanzurethat's not a plugin00:33
fennneither should that00:33
kanzurethat's in the interface to the API layer or whatever00:34
kanzureand then the plugins call to the APIs00:34
fennin the gtk GUI?00:34
kanzuregtkGetChar() would call an object in the GTK gui00:34
kanzureor whatever00:34
kanzureI don't know how that works00:34
kanzuregtk bindings for the ui-abstraction-layer00:34
kanzureso ui.getChar() would then be mapped to gtkGetChar()00:34
kanzurethen the python plugins that would do user io or just output or whatever would call ui.someFunction()00:35
kanzurewhere someFunction is of course dependent on what the user's setup is00:35
kanzureso gtk, or cli, or whatever00:35
fenni want to believe that your ui.someFunction is what i'm calling an API00:35
fennsee, getChar isn't appropriate to a radiobutton, but it performs the same function on a ncurses interface00:36
kanzurewait, what?00:36
fennsay you're at an interactive session00:37
kanzuregetChar would do get-character in ncurses just the same as get-character from a prompt, but the trick is that ncurses gets to choose a weird way to implement it00:37
kanzureokay00:37
fennwould you like: a, b, or c?00:37
kanzureand this could be either graphical or just text-based00:37
kanzureright?00:37
kanzurethe script is fundamentally asking me for those options, but my API knows that I want in gui form for example00:37
kanzurebecause that's what I've configured/told it to do00:37
fennin the text-based version you're type 'a' but in the gui version you'd click the box next to a00:37
kanzure(where I am the user)00:37
kanzurewell00:37
kanzureI'm not so sure about that, that's more of a question for configuration I think, right?00:38
kanzureI mean, that might be a popular configuration for end-users00:38
kanzurebut personally, intuitively, I'd think that even in cli you should still be able to do a key press00:38
kanzurebut I dunno.00:38
kanzureoops00:38
kanzure*even in gui00:38
kanzureI think that is something reasonably configurable, right?00:38
fennyeah, sure, most gui options have key shortcuts00:38
kanzurebut I mean whether or not to make it a radio button00:39
kanzurewhether or not to make it a radio button should be up to *something* on the end-user end00:39
fenninstead of what, a box you type in? gimme a break00:39
kanzurenot the python script itself00:39
kanzuredunno00:39
kanzurewhatever00:39
kanzureI don't care about it, that's up to the end-user00:40
kanzureor the gui/cli library00:40
kanzureokay, so I think that part sounds solid00:40
fennhrmm00:40
fennwhat do you think would be the API for agx-get?00:40
kanzureI mentioned that I'd like to do a review of html form specs, gtk, and ncurses to see common parameters so that I don't screw us over with a bad API-layer/thingy first off.00:41
kanzureit would be in python obviously00:41
-!- Aulere [n=dragon_d@131.229.176.252] has quit [Read error: 110 (Connection timed out)]00:41
kanzureand I'm hoping there's python-gtk bindings or python-ncurses bindings out there to make things easier to test out and debug or show mockups etc00:41
fennsure00:42
kanzureI'll dump a few notes on to the wiki re: this stuff, then download some docs/specs, and get to work tomorrow00:43
fenni will sit on my bum and feel sorry for myself00:43
kanzurehttp://library.gnome.org/devel/references which reference manual do I want?00:44
kanzurewhy's that?00:44
fennno idea00:44
kanzurehttp://library.gnome.org/devel/gtk/2.6/ ?00:44
fennuse pygtk.org00:44
kanzureconvenient00:45
kanzurethe python guys always seem to keep simple websites00:45
kanzureI suspect this is just due to the python syntax style00:45
fennthe python code is much easier to read than c with funky type declarations everywhere00:45
kanzuremy typed text kinda have those funky type declarations everywhere like in c :(00:46
kanzure*has those funky00:46
kanzurehm, glade00:46
kanzureI remember trying to use glade once or twice00:46
fenni have mixed feelings about glade00:46
fenni think it just makes things more complicated in the end00:47
fennauto-generated code is a big nono in my book00:47
kanzureI remember being it barely usable 00:47
kanzurewait, what about gcc and its automated code generators?00:48
kanzureand compiler compilers in general?00:48
fennbut even with libglade, then you have to sort through a big mess of buttons and windows to figure out where to put your code00:48
fennkanzure: as long as i dont have to see it00:48
kanzuremore than fair :)00:48
fennas long as someone cant possibly mistake it for source code00:48
fennthe core communication API of emc was written with a code generator.. been suffering ever since00:49
fennyou cant just rip it out00:49
fennoh well00:49
fennmodularity, abstraction layers.. the more the merrier00:50
kanzurehm, a code generator to communicate with what?00:52
kanzurewith the machines?00:52
kanzureoh, right00:52
kanzureg-code is like a language00:52
kanzureso I maybe see what they were thinking?00:53
fennno, nothing to do with g-code00:53
kanzurewtf?00:53
fennstuff like, turnSpindleOn(), popupStupidErrorMessage()00:54
fennthousands and thousands of these, all with their own c++ types00:54
kanzureso was a code generator a lazy man's answer to a problem ?00:54
kanzureeek00:54
fennit was because they couldnt figure out a way to send messages between machines with different architectures (besides c++ types)00:55
fennsurely there's a better way (and it wouldnt even have been an issue with an interpreted language)00:55
fennplain RPC is what I would have done00:55
fennoh well00:55
fennsee, back in the day, computers werent fast enough to handle controlling the table and the i/o at the same time (?) so they split up the different parts of the system between computers00:56
fennbut it was a govt project so of course they had to support a bunch of incompatible computer types00:57
fennnow, it's been scaled back to just x86 machines running rtai-linux, but the inter-process messaging layer is the same00:57
kanzurehey, uh, how hard would it be to convert between git, svn, cvs, monotone, etc.?00:58
fennnot too hard, but then you have to interconvert all the time00:59
fenn(i've never done it)00:59
fennyou dont want to use cvs00:59
kanzurewell01:00
kanzuremy main woe at this time is that01:00
kanzureapt-get is supposedly our cover-story, I mean, what we use as an example01:00
kanzure"See how well it works?" or at least the concept01:00
fennheh01:00
kanzureand if they can't apt-get install git-core01:00
kanzurethen WTF are they going to be thinking?01:00
kanzurebut otherwise git looks like an awesome choice01:00
fennum.. by the time we have anything useful released, nobody will be using debian-etch01:00
kanzureI guess we can say "go get unstable"01:00
kanzureeh01:00
fennsorry01:01
kanzureit took me four years to get around to upgrading my last knoppix box01:01
kanzureit was this terribly transformed knoppix->debian->ubuntu?->debian box from random sources.list additions/deletions01:01
kanzure(newbish ;))01:01
fenni think git-core is supposed to work on etch, but somehow you borked it by installing GNU git01:02
fennso, apt-get is not perfect, but you have to admit it works pretty well most of the time01:04
fenngit can be a deb-dependency of our package if we use it, but i cant realistically see most end-users learning how to use a command-line version control system01:06
fennpackages can be downloaded a zillion other ways01:06
kanzuresure01:06
kanzureagx-get would probably automatically clone some git repos ;)01:06
kanzurethat's what I'm thinking. 01:06
kanzurebut on the other hand, I'm not so sure01:07
kanzurethat'd be more like a parameter feature or something01:07
kanzuresince the user doesn't necessarily want to download all 10 MB associated with a project, just certain data files or just the human readable stuff at the moment01:07
kanzurewhatever.01:07
fenni may be confusing this project with my theoretical cad software which uses a git repo to save files01:08
fennbased loosely on this idea: http://www.bruno.postle.net/neatstuff/draft/01:08
fenni like files :)01:09
kanzurefenn: I don't see how the two are different. CAD software is the same thing methinks. But ultimately you still need to have that 'social knowledge' somehow01:10
kanzureor something01:10
kanzuretoo tired01:10
kanzuredon't make much fo that01:10
kanzure*of that01:11
fenngit-rebase in a parametric cad .. mmmmm01:13
kanzureguys in #python suggesting a "cli markup language that can then be converted to glade"01:16
kanzureeek, glade references.01:16
kanzure(2008-04-24 00:20:23) xif: you just need something that says <this is a form that lets the user input a string>, and then translate to the same thing in Gnome.01:16
fennno, workflow is different in a command line program01:17
kanzurehttp://ubuntuforums.org/showthread.php?t=630693&page=8 cli markup language01:18
kanzureoh01:18
kanzurehttp://linux.pte.hu/~pipas/CUI/index.html01:18
kanzurecomposite ui01:19
fennoh wow, its like a terminal except it's not!01:20
kanzurehrm01:20
kanzureterrible idea01:20
fenni think switching from cli to gui means a total rewrite and re-think is in order01:21
kanzure(2008-04-24 00:24:51) xif: kanzure: what I'd probably do, is define a small set of widgets that I actually need, then write an adapter that reliably represents each widget in a CLI markup and glade.01:21
kanzurehm01:21
kanzurebut then how does debian do it01:21
kanzurethey have graphical frontends01:21
kanzureooh01:21
fennthe dialog thingy? blue background and red buttons01:21
kanzurethey put them in "shut the fuck up" mode01:22
kanzurenope01:22
kanzurethat's not it01:22
kanzurethey have gtk versions01:22
kanzureof aptitude and so on01:22
kanzurebut I think they are in STFU mode01:22
kanzuremaybe.01:22
fennsynaptic?01:22
kanzureyes01:22
fennits a frontend to apt, but it uses apt's API01:22
fennthe stuff defined in libapt or whatever its called01:23
kanzureso01:23
kanzurein some cases the configuration scripts require user input01:23
kanzuredoes this translate up to gtk window popups ?01:23
kanzureand if so, how? are they doing the api layer thing that I have been suggesting?01:23
fenni dont know01:24
fenni cant think of any cases where we would need user input halfway through an operation01:24
fennideally they'd just press 'go' and it would spit out the result01:25
fenndem compyootah's better not gimme any back-talk01:26
kanzurewhat if they didn't know about the parameters for that project01:27
kanzurethey need to be told or something01:27
kanzurelike "Hello, this is your first time using this data type, so here's a quick walkthrough" or something? (yuck?)01:28
fennhmm that ubuntu thread. "training wheels for the CLI" sounds like a good idea01:28
kanzureyeah, hold on01:28
kanzurescrew gui01:28
kanzurewhy do I care this much01:28
kanzurehttp://surfraw.alioth.debian.org/  Surfraw - Shell Users' Revolutionary Front Rage Against the Web01:29
kanzureSurfraw abstracts the browser away from input. Doing so lets it get on with what it's good at. Browsing. Interpretation of linguistic forms is handed back to the shell, which is what it, and human beings are good at. Combined with netscape-remote or incremental text browsers, such as lynx, links or w3m, along with screen a Surfraw liberateur is capable of navigating speeds that leave GUI tainted idolaters agape with fear and wonde01:30
fennyeah its a great vision but the software is useless'01:30
fennat least google provides a decent set of "command-line" style search queries (instead of making you click boxes in some advanced search page)01:31
kanzurehurray for box clicking01:32
kanzure:(01:32
kanzurewasted hours01:32
kanzureprobably ten million man hours by now01:32
fennbut 99% of the website designers out there could care less01:32
kanzureit's not that they don't care, I think they just don't understand01:32
fennah, no, its that they dont care, and it wasnt part of the assignment01:32
fennthe person giving the assignment doesnt understand01:33
-!- Aulere [n=dragon_d@131.229.176.252] has joined #hplusroadmap01:33
fennif you're writing dynamic web code you are probably using a command line at multiple points in the process01:33
kanzureback when I was doing 1k~ of dynamic web code a night, the only command line I knew about was MSDOS01:34
kanzureand it was *useless*01:34
kanzurealthough I guess I could have used EDIT and then FTP01:34
kanzurebut really.01:34
fennhrm. i suppose it could be possible that one could write code without using a command line... (are my GNU horns showing?)01:35
kanzurenot sure01:36
kanzureback in my day, we programmed by smoke signals01:37
fennfwiw i learned to program on macintosh quickbasic and then ms-somethingorother C++ IDE01:37
kanzurenone of these damned 'electrons' and 'fields force'. 01:37
kanzuremy intro to programming was, sadly, html + js01:37
kanzurebut I was young enough for it to not scar me01:37
fennwell, javascript is sorta useful01:38
kanzuresure01:38
kanzurebut not when you think anybody who "knows" HTML to be a demigod01:38
fennin 1996 it was pretty much a requirement for participating on the (http) web01:38
kanzure'night01:39
-!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit ["Leaving."]01:39
-!- AulereV [n=dragon_d@131.229.176.252] has joined #hplusroadmap03:46
-!- Aulere [n=dragon_d@131.229.176.252] has quit [Read error: 110 (Connection timed out)]03:55
-!- marainein [n=marainei@220-253-110-236.VIC.netspace.net.au] has joined #hplusroadmap03:57
-!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap07:24
-!- marainein [n=marainei@220-253-110-236.VIC.netspace.net.au] has quit [Remote closed the connection]07:27
-!- marainein [n=marainei@220-253-110-236.VIC.netspace.net.au] has joined #hplusroadmap07:32
-!- AulereV [n=dragon_d@131.229.176.252] has quit []07:40
-!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit ["Leaving."]08:05
-!- marainein [n=marainei@220-253-110-236.VIC.netspace.net.au] has quit ["Ex-Chat"]08:41
-!- marainein [n=marainei@220-253-110-236.VIC.netspace.net.au] has joined #hplusroadmap10:31
-!- marainein [n=marainei@220-253-110-236.VIC.netspace.net.au] has quit [Read error: 110 (Connection timed out)]12:20
-!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap17:48
-!- marainein [n=marainei@220-253-113-232.VIC.netspace.net.au] has joined #hplusroadmap17:49
kanzureconclusion:17:49
kanzurefenn: no reason for these scripts to need input17:49
kanzureI was wrong.17:49
kanzureSo I was talking with Greg Egan this morning.17:49
-!- Splicer [n=p@h35n1c1o261.bredband.skanova.com] has joined #hplusroadmap18:56
kanzureHey Splicer.18:58
Splicerhello19:02
SplicerI was listening to a berkley genetics podcast earlier today, about the Recombinant DNA Debate...19:11
kanzureDebate??19:12
Splicerit started among scientists, some of which were very notable.. spread to the public..19:12
kanzurewhat debate.19:12
kanzure" Ready are you? What know you of ready? For eight hundred years have I trained Jedi. My own counsel will I keep on who is to be trained. A Jedi must have the deepest commitment, the most serious mind. 19:12
kanzureThis one a long time have I watched. All his life has he looked away... to the future, to the horizon. Never his mind on where he was. Hmm? What he was doing. Hmph. Adventure. Heh. Excitement. Heh. A Jedi craves not these things. You are reckless. "19:12
Splicerheh19:13
SplicerThe recombinant DNA debate, there is a thread about it on DIYbio.. sonething about berkley laws, the debate started there.19:14
Splicernot berkley... the other place19:14
Splicer'Cambridge DNA laws' is the name of the thread19:15
kanzureThere were only three or four emails on the diybio mailing list abotu that.19:15
Spliceryes19:15
kanzureSo how is it a debate. 19:15
kanzureThere's nothing *to* debate. We *know* how biology does it and we know it occurs, randomly, billions of times each day, perhaps billions of times each moment.19:16
Splicermmm... the berkley professor describes it better than I can:19:16
kanzureif you can link me to the transcripts to the genetics podcast, that'd be great19:16
Splicerdon´t have transcrips, sorry: http://webcast.berkeley.edu/media/f2007/mcb140/mcb140_2007-09-24.mp319:17
kanzurewtf19:17
Splicer?19:17
Splicerhehe.. no trascript sorry19:18
SplicerIt starts 33.10 in19:18
SplicerIt ends with the mayor of Cambridge banning the research and the mayor of berkley banning it for 2 years and setting up an advisory board consisting of firemen and nurses.19:20
kanzureHm.19:23
kanzureListening.19:23
kanzureErwin Chargaff19:24
Spliceryup.. he got carried away.. the good stuff is to come19:24
kanzure"... whether we have the right to put a fearful load on the future people. You can recall a bomb, you can recall a war; you can't recall a new organism." 19:24
kanzureAh, so it's just the grey goo argument.19:24
Splicerand he wasn´t just anyone19:25
maraineinbah...just try recalling a war19:25
kanzuremarainein: Your men.19:26
maraineinI meant in practical terms19:27
kanzure"And so they banned recombinant DNA research." Hahah. Try banning natural biology. Yes, that's a great idea. Sure. Just try.19:27
kanzure"We will create the perfect race, citing Adolf Hitler" - yawn, eugenitical approach to transhumanism19:27
kanzurehttp://heybryan.org/transhumanism_def.html19:27
maraineinneed to ban sex too...don't forget about homologous recombination during meiosis19:28
* Splicer sings "We will not be cloooooned"19:28
kanzuremarainein: A few days ago fenn was mentioning how Texas banned 'sex toys' up til, what was it, 2007?19:28
* kanzure lives in Texas and didn't know :-(19:28
kanzure"You don't need to understand things to argue about them." erm...19:29
maraineinwell, technically that's true19:29
kanzureif you want to be grounded.19:30
Splicerpublic debate follow the laws of memetics not reason19:30
kanzuredoesn't matter, they are wrong19:30
Splicerof course, and technology beats politics, but they can make things miserable for a while19:31
Spliceri like his conclusion, that noone cares about that particular debate now... every student is doing it first thing19:32
-!- marainein [n=marainei@220-253-113-232.VIC.netspace.net.au] has quit ["Ex-Chat"]19:48
Splicerhmm.. the biology channel seems quiet19:55
Splicertime to sleep19:56
Splicercu19:56
-!- Splicer [n=p@h35n1c1o261.bredband.skanova.com] has left #hplusroadmap []19:56
-!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit [Connection timed out]20:26
-!- Aulere [n=dragon_d@131.229.176.252] has joined #hplusroadmap20:48
-!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap21:00
kanzurehttp://www.davidzindell.com/id16.html21:50
Aulereis he a favorite author of yours?21:53
Aulerelooks interesting21:53
kanzureYes, I have a copy of his book on http://heybryan.org/docs/22:01
kanzureI am also wikifying a copy of the book.22:29
kanzurehttp://heybryan.org/mediawiki/index.php/Neverness/122:30
Aulerenice22:40
Aulerehe has a certain poetry to his prose22:41
kanzureSure.22:42
kanzureAlright fenn, so I realized that all of that configuration talk was just bullshit. The only thing we have to do is download the plugins that provide the classes. These classes can be configured, yes, but ultimately it's only two ways. (1) Through the parameters. (2) Default config as provided by the package.22:44
kanzureHere's why. Let's say that you type in agx-get install --recursive chair to get anything and everything related to a chair22:45
kanzureIt will pick up the yaml plugin for the chair datatype. So it might need to do some configuration, right? Not really. What's there to configure? All that you really care about is installing it, so it'll just do default stuff for you. If you don't give it anything *specific* to 'chair'-type then it will not really do much.22:46
kanzureso you'd have to wait22:46
kanzureif you're going to be waiting there, you might as well quit the program and just type in some parameters22:46
kanzurethis way, even better, you can *repeat* the configuration steps and show other people22:46
kanzurebut it's wrong to call it configuration22:46
kanzurethis is more like: agx-get install chair22:46
kanzure> Hello, you're downloading a chair. Did you know you can specify --specific <id> to get a specific model? If you don't know the ID, it's okay, just run chairFinder to help you out.22:47
kanzurethat sort of thing22:47
kanzurehttp://www.cosmographica.com/gallery/index_main.html22:58
-!- marainein [n=marainei@220-253-10-191.VIC.netspace.net.au] has joined #hplusroadmap23:01
fennmarainein: is your nick based on the iain banks language?23:49
[Users #hplusroadmap]23:49
[ Aulere] [ drazak] [ fenn] [ kanzure] [ krebs_] [ marainein] 23:49
-!- Irssi: #hplusroadmap: Total of 6 nicks [0 ops, 0 halfops, 0 voices, 6 normal]23:49
maraineinno...what makes you ask?23:50
fennbecause then i'd know you'd read some banks novels :)23:51
* marainein has read some of banks's books, but doesn't recognise the reference23:51
fennmarain is the language of Culture23:51
maraineinah...didn't know that23:51
maraineinalthough I've read some culture books23:51
fenndid anyone watch the 1968 engelbart demo?23:53
fenni've been geeking out on engelbart since i did. he's all about bootstrapping23:54
fennits really amazing all the different things they had to invent, and how spot-on they got the whole system and interface the very first time23:56
fennlike, there was no such thing as a monitor that showed words electronically23:56
kanzurenot yet 23:59
kanzureah23:59

Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!