[05:03] *** elvum_ has joined #kamaelia

[10:39] *** Lawouach has joined #kamaelia

[11:15] *** MS- has joined #kamaelia

[11:15] < MS-> greetings

[11:16] *** SamHasler has joined #kamaelia

[11:16] < Lawouach> heya MS-

[11:17] < Lawouach> how are you?

[11:29] < Lawouach> eating a clementine, feels like Xmas :)

[11:46] < MS-> heh cool

[11:46] < MS-> I'm rather tired with a rather high lag today it seems

[11:46] < MS-> I thought it was about an hour earlier than it is

[11:47] < MS-> Had some good news yesterday, and I think my tiredness is relief

[11:47] < MS-> which is all good of course in that case

[11:47] < MS-> and a small child just walked in wearing a dalek costume, so that's cool :)

[11:49] < Lawouach> :)

[11:49] < MS-> Also, I figured out a way to phone a number and get the message emailed to me. As a result I can now appear to live anywhere in the country, which is pretty neat :-D

[11:49] < Lawouach> Well sounds like it's more pressure fading away

[11:49] < MS-> yep

[11:51] < MS-> My task for today will be to create an SMS bot, and to find up a better means of recording audio. (the internal soundcard on my laptop is great for playback, but the audio picks up all sorts of electronic noise from the rest of the system)

[11:52] < Lawouach> nothing beats an external dedicated soundcard IMO

[11:52] < Lawouach> I have a M-Audio firewire solo which is quite good. The only problem is firewire and Linux can be distant friends

[11:53] *** MS- nods

[11:53] < Lawouach> For the best sound hardware deals I'd advise thomann in germany

[11:53] < Lawouach> I bought plenty of stuff from them

[11:53] < MS-> Well, for external soundcard, I'll be playing with a handful of USB devices.

[11:54] < MS-> which isn't quite the same thing, but is a slightly cheaper way of testing :-) (I only need slightly better :) )

[11:54] < MS-> It's that or sit down and figure out asterisk.

[11:54] < MS-> Or play with voip clients

[11:55] < Lawouach> I see

[11:55] < MS-> You see, one of the things work related at the moment is playing with contribution methods, and most of us have a very good way of digitising audio in our pockets - a mobile phone

[11:56] < Lawouach> true

[11:56] < MS-> Which works really quite nicely. But the thing I need to sort out is the outgoing message, and I can't see an easy way of taking the audio which I can capture that way and use it as the outgoing message

[11:56] < MS-> Which has all sorts of levels of irony :)

[11:57] < Lawouach> too many :)

[11:57] < Lawouach> anyway, FYI http://www.thomann.de/gb/line6_toneport_gx.htm

[11:57] < MS-> That's pretty dinky

[11:58] < Lawouach> yeah

[11:58] < Lawouach> cheap and straight looking

[11:58] < MS-> Nice design - audio in one side, audio out the other

[11:59] < Lawouach> I bought my Fender amp over there. They delivered it in the UK via UPS for 20 quid or something.

[11:59] < Lawouach> anyway, just fyi really :)

[12:00] < MS-> :) I have a couple of USB devices kicking about, but I can't remember how well they sound at the moment :)

[12:00] < Lawouach> http://www.thomann.de/fr/fender_72_telecaster_deluxewa.htm < -- The next guitar I want

[12:00] < Lawouach> Though guitars are the only thing I'm not buying online

[12:00] < MS-> That does look nice, but I can also understand that sentiment

[12:00] *** Davbo has joined #kamaelia

[12:01] < MS-> If you know how to play and how it should sound, then buying online is a little silly

[12:01] < Lawouach> well electric guitars are safer to buy because they usually sound the same within a given product line.

[12:01] < Lawouach> Acoustic ones are too sensitive to their environment

[12:02] < Davbo> Guitars?

[12:02] < Lawouach> The Taylor I bought two years ago probably wouldn't sound exactly the same as exactly the same Taylor from a different shop :)

[12:02] < Lawouach> Davbo: yeap

[12:02] < Lawouach> I was showing my next purchase to Michael

[12:02] < Lawouach> http://www.thomann.de/fr/fender_72_telecaster_deluxewa.htm

[12:02] < Lawouach> well... when I'll have €800 to spare :)

[12:02] < MS-> Lawouach: /me nods

[12:03] < Lawouach> probably like ... never :)

[12:03] < MS-> heh

[12:03] < Lawouach> http://www.thomann.de/fr/taylor_314ce_western.htm < -- my acoustic Taylor

[12:03] < Lawouach> it's a beauty and its sounds live up to the look :)

[12:04] < Davbo> oooo, always liked the telecaster shape

[12:04] < MS-> Wowsers... I must admit I'd hope its sounds live up to the look (!) :-)

[12:04] < Lawouach> it does

[12:05] < Lawouach> took me 7 months to find the guitar that would sound as I wanted

[12:05] < Lawouach> finally found that one

[12:05] < Lawouach> bought it with the CherryPy's book advance :)

[12:05] < Lawouach> Thought I should get something special out of the book

[12:06] < Lawouach> http://www.flickr.com/photos/lawouach/1981457185/in/set-72157603323426589/ < -- and my current electric guitar

[12:07] < Lawouach> Epiphone Supernova (Noel Gallagher's signature)

[12:07] < Lawouach> yeah you read it right

[12:07] < MS-> Ahh, i see

[12:07] < Lawouach> I have Gallagher's signatuer on it

[12:07] < Davbo> Hah

[12:07] < Lawouach> how cooler can one get?

[12:07] < MS-> Oh that's nice :)

[12:07] < MS-> very cool

[12:07] < Lawouach> anyway... I wish I had ten more guitars :)

[12:07] *** Lawouach should stop eating and save for more guitars!

[12:08] *** Davbo has an Epiphone SG

[12:08] < Lawouach> sweet

[12:08] < Lawouach> the SG is a true rock guitar :)

[12:08] < MS-> I have a 50 quid guitar from ebuyer that I can't play, because I've never learnt. But it does look nice, and one day I'll learn to play and how bad it sounds :)

[12:08] < Davbo> hehe

[12:09] < Lawouach> :)

[12:09] < MS-> I have no illusions about my musical ability :)

[12:09] < Lawouach> cheap guitars can sound great but to be fair... you have to be good :)

[12:09] *** MS- nods

[12:09] < Davbo> it's not hard to play guitar, just hard to be good

[12:09] < Lawouach> I want to learn either the drum or the piano

[12:09] < Lawouach> just can't be bothered to actually do it

[12:09] < MS-> It's one of these, but red http://www.ebuyer.com/product/127845# :)

[12:10] < MS-> I don't think they sell it now, or could sell it :)

[12:10] < Lawouach> £27!

[12:10] < Lawouach> damn... how can it be that cheap? :D

[12:11] < MS-> Mine was about 50, but it included a small amp as well

[12:11] < Lawouach> well

[12:12] < Lawouach> ya know, if you learn to play the guitar go with Nirvana and break the instrument at the end of each gig. You can afford it at that price :)

[12:12] < MS-> heh

[12:12] < MS-> one of these days I'll learn

[12:12] < Lawouach> :)

[12:12] < Lawouach> If I could give you two advice:

[12:13] < Lawouach> 1. Make sure you find a teacher that is all about having fun with your guitar as soon as you can. Don't go with the musical theory first. It'll put you off.

[12:13] < Lawouach> 2. Don't discourage.

[12:14] < Lawouach> I have started salsa dance classes this year. The first three weeks I thought I was learning fast. Now I feel like I'm crap. But I do remember that I felt the same in the first few months I learnt the guitar.

[12:14] < Lawouach> Now 15 years later I'm just glad I sticked with it :)

[12:14] *** MS- nods

[12:14] < Davbo> I learnt with a teacher when I was about 9, now i just pick up bits when i feel like it

[12:15] < MS-> I think that's a common thing with learning anything to be honest

[12:15] < Lawouach> Plus, MS-, playing the guitar will make you look younger and cooler.

[12:15] < Lawouach> :D

[12:15] < MS-> a) have fun, focus on practical stuff

[12:15] < MS-> b) remember you will fall on your face

[12:15] < MS-> There is that

[12:15] < Lawouach> Yeah. Didn't want to sound condescending :)

[12:16] < Lawouach> It's just I've seen many people who wnated, started and quickly lost there motivation.

[12:17] < MS-> Nah, I never (or try to never) take advice as condescending these days. (Unless I _know_ the person is deliberately being so :), which I find is rare )

[12:17] < Lawouach> One note nonetheless. Starting with a cheap instrument is good to check that you really want to continue but you'll quickly find the guitar is the one letting you down.

[12:18] *** MS- nods

[12:18] < MS-> I kinda guessed that

[12:18] < Lawouach> I would advise putting a bit of money into a better guitar whenever you could

[12:18] < MS-> But also figured that a cheap guitar is like prototyping

[12:18] < MS-> "buy one to throw away" :)

[12:18] < Lawouach> exactly

[12:18] < Lawouach> epiphone's are great for that

[12:18] < Lawouach> They build quality guitars at a fair price

[12:18] < MS-> cool

[12:23] < Lawouach> well anyway it'll depend on the style you feel you like playing

[12:23] < Lawouach> I personally can't stand Stratocaster

[12:23] < Lawouach> The shape your guitar has :)

[12:23] < Davbo> I've been thinking of getting a Classical guitar for ages now

[12:24] < Lawouach> it can be sweet to play

[12:24] < Davbo> Yeah, I feel a bit bored of the rock sound

[12:25] < Davbo> went to see Nitin Sawhney the other week too, totally amazing guitarist

[12:25] < Lawouach> A friend of mine has started 3 years ago gipsy jazz guitars and he wouldn't come back to anything else

[12:25] < Lawouach> I cannot play his guitars. It's a guitar so different in sound I feel like a newbie again

[12:25] < Davbo> Yeah

[12:26] < Lawouach> I've always a great admirer on how Mark Knopfler would gently use various kind of guitars into his own style

[12:27] < Davbo> Ah yes, i know who you mean now. Always forget that guys name :-)

[12:28] < Davbo> He did a few songs with Dylan iirc

[12:30] < Lawouach> he goes that way back probably

[12:30] < Lawouach> Eraly Dire Straits songs are just amazing in subtetly

[12:33] < Davbo> I've only got Dire Straits' self titled album i think

[12:33] < Davbo> although i believe that was an early one

[12:33] < Lawouach> an oldy :)

[12:33] < Lawouach> their first

[12:33] < Davbo> It has Sultans of Swing on it which is an awesome song ;-)

[12:34] < Lawouach> it is :)

[12:34] < Davbo> you should check out Nitin Sawhney if you've not heard any of his stuff

[12:34] < Lawouach> I will

[12:34] < Davbo> Right, i'm off, back later chaps

[12:59] *** Uraeus has joined #kamaelia

[13:36] *** SamHasler has joined #kamaelia

[14:05] *** vmlemon_ has joined #kamaelia

[14:07] < vmlemon_> Hi

[14:19] *** Davbo has joined #kamaelia

[14:54] *** MS- has joined #kamaelia

[18:17] *** Davbo has joined #kamaelia

[18:21] < Davbo> MS-: you learn FORTH with this book? http://globalnerdy.com/wordpress/wp-content/uploads/2007/09/forth_on_the_atari.jpg

[18:42] *** vmlemon_ has joined #kamaelia

[19:37] < MS-> OK, that's part one - I can connect an arbitrary unix programme to a particular AIM conversation. Now I need to do something more like what happens with ServerCore - wait for a message, demux and wire in.

[19:37] < MS-> break for now

[20:06] *** vmlemon_ has joined #kamaelia

[21:12] < Davbo> Anyone used Zotero before? it's awesome..who would have thought Firefox could be used like this

[21:13] < Davbo> for those that haven't, Zotero is a tool for managing references and citations etc. but actually makes it easy given resources like Google scholar

[21:14] < MS-> never heard of it

[21:15] < Davbo> open source too

[21:20] < MS-> It fails to explain why someone would want to use it, because it fails to explain the problem it solves

[21:22] < vmlemon_> Sounds mighty useful ;)

[21:22] < Davbo> Don't think so MS-, it manages your citations that's pretty much it

[21:23] < MS-> It doesn't explain why this is a problem

[21:24] < MS-> It assumes you have a problem already and will view this as the solution

[21:25] < elvum_> perhaps it assumes that people with the problem it solves will recognise it for what it is

[21:25] *** elvum_ is now known as elvum

[21:25] < Davbo> Yeah

[21:25] < Davbo> I don't see your point MS-

[21:26] < elvum> so if you don't have the problem, you won't understand :-)

[21:26] < MS-> Davbo: you asked if anyone had used it. I looked at it and thought , OK that looks pointless

[21:26] < MS-> elvum: I've heard that argument about software before and I believe it to be total BS

[21:26] < MS-> :)

[21:27] *** elvum is unsurprised ;-)

[21:27] < MS-> There was a piece of software on teh Amiga that was always described that way "If you don't know what ARexx is, you don't need to know what it is"

[21:27] < MS-> which was completely wrong, and the same applies here

[21:27] *** elvum isn't defending

[21:27] *** MS- nods

[21:27] *** MS- just types relatively quickly

[21:28] < Davbo> Well, I don't really care about their marketting decisions to be honest

[21:28] < MS-> Davbo: My point is I can't see what problem it solves. I can see it's a database, and I just think "whoop-de-do".

[21:29] < Davbo> well it has some nice features, grabbing all the data and presenting it in whatever format you want. Academics are very irritating about citation formats so it's nice to not have to even think about it

[21:30] < MS-> I'm sorry, still doesn't appear to have any value

[21:30] *** vmlemon_ has joined #kamaelia

[21:30] < elvum> heh

[21:31] < MS-> Formatting text is rather trivial at the end of the day

[21:31] < Davbo> I'm not trying to sell it you, you obviously don't need it if you don't see its value

[21:32] < MS-> I ignored python for 5 years, because it sold itself badly, and I find that rather useful these days. I rarely dismiss something these days because it sells itself badly - esp if someone else says "this looks useful"

[21:32] < MS-> It is helpful if a site bothers to explain why it's useful though :)

[21:32] < elvum> I see this as the citation equivalent of an electronic address book

[21:33] < MS-> elvum: That's useful if it's shared/collaborative, but this appears to be "local scrapings"

[21:33] < MS-> Mind you I never found electronic address books exciting either :-)

[21:33] < elvum> well, citation databases (eg bibtex) have had an enthusiastic following in academia for decades

[21:33] < vmlemon_> Hmm, what advantage does it give over Google Notebook, out of interest?

[21:34] < vmlemon_> (Which isn't that great itself, but it's still somewhat handy)

[21:34] < MS-> elvum: Indeed, that was why I was wondering what this gives over that sort of thing

[21:34] < elvum> adding browser integration and automatic field-guessing (if it works well) sounds *more* exciting to me, relative to whatever level of excitement one had to start with

[21:35] < MS-> It just strikes me as unnecessary browser integration. ie the sort of thing that should be hosted off browser really

[21:36] < MS-> so that it works better structure based on numbers of users

[21:36] < Davbo> it's not really guessing elvum, look at Amazon, Google Scholar/Books it's all there, it's just a case of parsing it correctly

[21:37] < Davbo> I don't see how it's unnecessary MS-, all research takes place online these days..

[21:37] < MS-> Davbo: The thing is that app would be infinitely more sensible if it wasn't hosted on people's machines

[21:37] < vmlemon_> Who knows, it could be the citation world's answer to Twitter or Windows Vista - it has no obvious purpose, but it becomes popular anyway ;)

[21:37] < MS-> The stuff greyed out is the stuff they haven't implemented by the looks of things, and is what would actally make it useful

[21:37] < Davbo> MS-: Zotero's preview release does just that i believe

[21:38] < elvum> anything that parses html and makes assumptions about its structure is guessing, imo

[21:39] < Davbo> it depends who's html you're parsing :-)

[21:39] < MS-> elvum: That's what the bundle of neurons in our head does too of course

[21:39] < vmlemon_> I bet it would choke on the "HTML" produced by something like Adobe PageMill or FrontPage Express

[21:40] < vmlemon_> (Not so much HTML as a nasty tag soup)

[21:40] < MS-> Davbo: looking at http://www.zotero.org/support/sync_preview - it doesn't appear to do what I think would be useful

[21:40] < Davbo> MS-: Maybe you should be looking at this then: http://www.zotero.org/support/dev/start :-)

[21:41] < MS-> Um no

[21:41] *** Davbo was joking.

[21:41] < MS-> k

[21:41] < MS-> :)

[21:41] < vmlemon_> Maybe MS- should show them how to market it ;)

[21:42] < Davbo> open source never knows how to market

[21:42] < MS-> vmlemon_: I'm the last person to tell anyone how to market anything

[21:42] < vmlemon_> (Hint: A cuddly mascot, hexagons and a detailed description, and you'll get it right away) ;)

[21:42] < vmlemon_> Heh

[21:42] < MS-> Davbo: technical people rarely know how to market anything

[21:42] < vmlemon_> "You know you need this, but we won't tell you what it is"

[21:44] < elvum> MS-: the bundle of neurons in my head tends to do a better job of parsing text and ending up with a result I like :-)

[21:44] < MS-> elvum: indeed :)

[21:45] < MS-> vmlemon_: I'm still learning how to relate that sort of thing with kamaelia really too tbh

[21:45] < MS-> You get too caught up in solving stuff without talking about what problem is being solved, hence making it difficult for people to see what the problem is

[21:46] < MS-> at the moment though enough people are worried about concurrency, so that's an easy hook

[21:46] < Davbo> Anyway, I've been using Zotero to abstract citations from me and just generating them and it's proved really useful. YMMV :-)

[21:46] *** vmlemon_ has joined #kamaelia

[21:46] < MS-> Davbo: That's more information on their website

[21:47] < MS-> :)

[21:47] < Davbo> MS-: I had a lecture on what was basically STM the other day but she decided to call it "timestamping" and make it stupid.

[21:47] < MS-> timestamping gives you the same thing - increasing numbers

[21:48] < MS-> and timestamping is arguably a good idea too

[21:48] *** Davbo agree's

[21:48] < MS-> but it's less obvious a metaphor IMO than "version control for variables"

[21:48] < MS-> especially now that I've tested that metaphor on a roomful of people :)

[21:48] < vmlemon_> Hah

[21:49] < MS-> and seen the body language change from "wtf are you talking about" to "oh, is *that* what it is"

[21:49] *** vmlemon_ was just thinking about the bicycle or car naming sketch for some reason...

[21:49] < Davbo> Anyway, it was rather amusing, separate timestamp for when you read a value and when you modify it...

[21:50] < Davbo> I didn't really see the value

[21:50] < Davbo> It was something like that anyway, the explanation of locks was god awful though (it was a lecture on concurrency with respect to databases)

[21:51] < vmlemon_> Let me guess, you've had lectures on semaphores and mutexes, too? ;)

[21:51] < MS-> "There are only two hard things in Computer Science: cache invalidation and naming things."

[21:51] < MS-> — Phil Karlton

[21:51] < vmlemon_> Hah

[21:52] < MS-> Davbo: that's akin to the three times you have in the unix file system

[21:52] < MS-> 8 atime last access time in seconds since the epoch

[21:52] < MS-> 9 mtime last modify time in seconds since the epoch

[21:52] < MS-> 10 ctime inode change time in seconds since the epoch (*)

[21:52] < Davbo> didn't really think about that

[21:54] < Davbo> locks are pretty easy to explain really, so long as you don't bombard people with all the different types

[21:54] < Davbo> My class got bombarded, now they don't even get the basics

[21:55] < MS-> Davbo: Which means none of then will be able to write concurrent properly.

[21:55] < Davbo> Heh, they'll be in the majority there

[21:55] < MS-> (if they use locks this is)

[21:56] *** vmlemon_ wonders who decided upon the specific date and time used in the UNIX epoch (00:00:00 UTC on January 1, 1970) and why...

[21:56] < vmlemon_> (Sounds like a dumb question, but after reading the Wikipedia article, I'm now curious)

[21:56] < MS-> vmlemon_: Interesting q, someone *must've* decided it, even if they were being arbitrary

[21:57] < vmlemon_> I find it interesting that the classic Mac OS epoch is different again

[21:57] < vmlemon_> (Not sure about the Windows/DOS one though)

[21:58] < elvum> "The Unix epoch is the time 00:00:00 UTC on January 1, 1970. There is a problem with this definition, in that UTC didn't exist in its current form until 1972; this issue is discussed below."

[22:00] < Davbo> MS-: FWIW this module is really easy, proper Concurrency gets its own module next year. I'll let you know how that goes when I get to it :)

[22:01] < MS-> Davbo: seems rather ironic given the level of concurrency in your GSOC project...

[22:01] < vmlemon_> Damn, Google is full of references to time on Mac OS X, but not to the previous versions based on a proprietary kernel

[22:02] < Davbo> She asked what SQL stood for, to which I replied "Scary Quirky Language" as Jacob Kaplan Moss called it in his talk at Pycon :P

[22:03] < MS-> what did she think it stood for?

[22:03] < Davbo> she also denies it stands for Structured Query Langauge, claiming "thats what people call it now, but nobody really knows.."

[22:03] < MS-> there's two possible versions

[22:03] < Davbo> Ah

[22:03] < Davbo> I guessed as much

[22:03] < vmlemon_> January 1, 1904 turns out to be Classic Mac OS's epoch

[22:03] < elvum> Semi-standardised Query Language? :-)

[22:04] < MS-> Structured Query Langauge is the common version, but it was originally Structured English Query Language -

[22:04] < vmlemon_> January 1, 1980 for MS-DOS

[22:04] < MS-> hence why it's officially pronounced "sequel"

[22:04] < vmlemon_> January 1, 1601 for NTFS for some reason

[22:04] < Davbo> Ah, cool

[22:04] < Davbo> not that it really matters

[22:05] < MS-> (unless you're MySQL - which is officially OK as "micycle")

[22:05] < vmlemon_> Haha

[22:05] < MS-> But I believe it got summarised as sql due to 8.3 naming conventions

[22:05] < MS-> since it was born in the '70

[22:05] < MS-> s

[22:05] *** vmlemon_ still pronounces it as Ess Queue Ell (is that particularly incorrect?)...

[22:06] < MS-> If she "denies it stands for Structured Query Langauge" what did she say it really stood for?

[22:06] < Davbo> these database lectures have taught me one thing though: database people are weird.

[22:06] < Davbo> "claiming "thats what people call it now, but nobody really knows..""

[22:07] < Davbo> I think I've had the SQL 101 lecture about 3 times now, it's really annoying

[22:07] < MS-> really

[22:08] < MS-> point her at this then:

[22:08] < MS-> http://www.almaden.ibm.com/cs/people/chamberlin/sequel-1974.pdf

[22:08] < MS-> "SEQUEL: A STRUCIURED ~NGLISH QUERY LANGUAGE"

[22:08] < Davbo> I didn't pay all that much attention TBH MS- so I could be wrong

[22:08] < MS-> Donald D. Chamberlin

[22:08] < MS-> Raymond F. Boyce

[22:08] < MS-> I think those two people would know

[22:08] < Davbo> Ah those names are familiar

[22:08] < MS-> :)

[22:08] < MS-> Yes, Boyce Codd Normal form

[22:08] < vmlemon_> Interesting, "January 1, 1904, was chosen as the base for the Macintosh clock because it was the first leap year of the twentieth century."

[22:09] < Davbo> Yeah, they were certainly mentioned

[22:09] < MS-> derivable directly from an EER diagram if you know how to draw it properl

[22:09] < MS-> y

[22:09] < MS-> It's where every attribute in every table

[22:09] < Davbo> I bet she knows, I just wasn't listening (for obvious reasons....SQL 101)

[22:09] < MS-> depends on "the key, the whole key, and nothing but the key"

[22:09] *** Davbo nods yeah I remember that stuff

[22:10] < MS-> which is the way our db lecturer used to teach us, which is a nice way of teaching it

[22:10] < MS-> she was the one who mentioned that names matter as well

[22:10] < MS-> and that good metaphors matter

[22:10] < vmlemon_> Amazed that the small amount of Macintosh System 7 source code I have here doesn't reference the Epoch

[22:11] < Davbo> ohhh!!! you'll love this MS- 1 second

[22:11] < vmlemon_> or at least doesn't in an obvious way

[22:11] < Davbo> remember the discovery method guy MS-?

[22:11] < Davbo> http://en.wikipedia.org/wiki/JWalk

[22:11] < Davbo> this is also his

[22:12] < Davbo> Click "history" you wont be surprised

[22:12] < vmlemon_> Although I don't think Google Desktop Search is well suited to searching through Pascal source files with Macintosh line breaks

[22:12] < MS-> Oh the guy who was touting his own methodology as being the worlds gift to the software engineering world, but no-one has heard of it, let alone uses it?

[22:12] *** Davbo nods

[22:13] < MS-> Shrug, there's not alot wrong with adding a page to wikipedia on a project

[22:13] < Davbo> I find it amusing how he expects us to use this JWalk

[22:14] < MS-> why so?

[22:14] < Davbo> I asked if I could use JUnit instead..which is pretty much industry standard and far more powerful

[22:14] < Davbo> and useful

[22:14] < MS-> You're probably guinea pigs

[22:14] < MS-> though to be fair, if that's the case he should split the group in two :)

[22:15] < Davbo> He wont even provide the source code to the students

[22:15] < MS-> That said, the way that page is written it appears to breach the no original research rule of wikipedia

[22:16] < MS-> I don't think I care enough :)

[22:18] < Davbo> what made me laugh is how the thing he describes JWalk as doing, "Lazy systematic unit testing" is a link to another wikipedia article which he also authored

[22:18] < Davbo> it pretty much sums him up

[22:18] < vmlemon_> Hah

[22:19] < vmlemon_> Peddling snake-oil? ;)

[22:19] < MS-> vmlemon_: No, probably desperately trying to get traction for something he thinks is a good idea, and going about it badly

[22:19] < MS-> I'm sure he genuinely believes his approach is best

[22:20] < MS-> It's very easy to be critical.

[22:20] < Davbo> Oh of course

[22:20] < elvum> it sounds interesting, but not all states can be inferred from analysis of a class's code; a quick glance doesn't reveal how he deals with that

[22:21] < Davbo> his code is closed too, which is totally the wrong way to get traction for his project

[22:21] < Davbo> for this type of project anyway

[22:21] < Davbo> I would hate to use a closed source unit testing tool

[22:21] < vmlemon_> Nothing that a Java decompiler won't "fix"

[22:22] < vmlemon_> even though the result would be less than satisfactory

[22:22] < Davbo> exactly vmlemon_, what's the point in the first place

[22:22] < Davbo> true, but still.

[22:22] < vmlemon_> Hell, he could have written it against a .NET CLI language - it's still easier than native code to reverse-engineer

[22:23] < MS-> Davbo: The problem is that academics often miss out practicality. Not all of them, but enough of them. Probably because practicality is not a necessity to academic research.

[22:23] < Davbo> a unit testing framework is one of the number one things I'd want to be open source!

[22:24] < MS-> After all, the aim of academic research is primarily to increase knowledge space

[22:24] < Davbo> Sometimes I wonder what it is MS-

[22:24] < Davbo> when it seems to impede itself at every opportunity

[22:24] < vmlemon_> There's probably very little "secret sauce" in that sort of thing

[22:24] < MS-> A secondary goal of academic research is the ability to publish more papers, in order to attract funding

[22:25] < MS-> The harder to comprehend a paper is, the greater the likelihood of publishability

[22:25] < vmlemon_> (Same deal with folks who insist on keeping Ethernet card drivers closed, although it's probably a bad analogy )

[22:25] < MS-> based on what I've seen

[22:25] < MS-> ie *documenting* work, is more important than *communicating* work. The latter results in understanding, and potential use

[22:26] < vmlemon_> Hmm, I don't think that'll go down well for Sun - there was supposedly a plan to make advertising supported versions of OpenOffice

[22:27] < MS-> as a the choice of result open/closed code really depends on how much benefit it brings academically

[22:27] < Davbo> MS-: I was looking at a project computer science did a couple of years ago, it was to analyse text to pick out anomalies for the government to catch terrorists saying they were going to "deliver to the ice cream to red house" or something like that

[22:27] < Davbo> you get the idea

[22:28] < Davbo> lets be nice and say that it worked 99% of the time...that is a massive amount of error right there it would be totally unusable

[22:29] < Davbo> they just say that they can do it and get a fat grant from the MOD it's obscene

[22:30] < Davbo> Man, I really couldn't work as an academic if I was doing shit like that

[22:32] < Davbo> I guess you don't agree MS-?

[22:33] < MS-> I'm not sure I can read what you mean :-)

[22:34] < Davbo> Sorry, it really annoys me as you can tell.

[22:37] < Davbo> Basically, they wanted to analyse text to catch anomalous word useage. For obvious reasons the government had an interest in that research so partly funded it (from what i gather)

[22:38] < Davbo> I'd argue that something which did that would need such a low error rate it's totally pointless

[22:40] < Davbo> My point being, sometimes it's far worse than just - "< MS-> A secondary goal of academic research is the ability to publish more papers, in order to attract funding"

[22:41] *** Lawouach has joined #kamaelia

[22:42] < MS-> ahh, I see

[22:42] < MS-> you get good behaviours and bad

[22:43] < MS-> I doubt the people doing that work *didn't* believe it could help/work

[22:44] < Davbo> Yeah I guess, I'd hate to work on something like that though

[22:44] < Davbo> which imo is a complete waste of time

[22:44] < MS-> even if you thought you could get it low enough error rate?

[22:44] < MS-> what if you decided to have a policy of "zero false positives"

[22:45] < Davbo> that would be different

[22:45] < MS-> That would mean you almost certainly miss positives

[22:45] < Davbo> but it would also be neigh impossible

[22:45] < Davbo> well, yeah

[22:46] < Davbo> I'm thinking of popping into ARM Sheffield and seeing if they're going to have a summer placement again though

[22:46] < Davbo> they're a nice company to work for, from what I've heard

[22:48] *** Davbo goes on a tangent :-)