PIT - version_tracker.txt ---------------------------- Version numbers are stored in: (1) .iss file (influences setup display, Control Panel) (2) .rc file (About box) (3) g_strVersion (4) web site whatsnew.shtml (5) help (.hm3) Version history: 1.0 ... 1.1 - Majority of development done under this number! * The "versionable schema" defined in IMPLEMENT_SERIAL screws up. Regardless of the setting, the code tries to load (a) other programs' configuration files; (b) the 'wrong' versions of its own. * Still having problems with random crashes transferring Date fields via CRecordSet to ODBC, but only in Debug mode. * Remember, if a fault is *really* random (e.g. bit flip), build clean. * If you mistype RFX_Text(pFX, _T("[m_StimulusType]"), m_StimulusType); instead of RFX_Text(pFX, _T("[StimulusType]"), m_StimulusType); the ODBC systems reports the error "Too few parameters. Expected 1." * It was the case that... when a new component starts, the locomotor beams aren't checked to see if they're on. So the locomotor beambreak time may be a bit misleading. Think about what's desired. Probably fix it. Yes. Fixed. So now the number of breaks may be 1 too high in these rare situations, but the total time is correct. * Revision 22 August 2000 After much mucking around, am implementing these changes: - option to have houselight on/off - light stimulus as any combination of {houselight, stimulus light over lever, centre stimulus light, stimulus light over the lever not being used}. Pat wants this configurable per rat, so don't use the box definition file, except to define HOUSELIGHT, LEFTLIGHT, CENTRELIGHT, RIGHTLIGHT. - noise stimulus to be either TONE or CLICKER. Have added database fields HouselightOn (yes/no), LightStimulusType (text), NoiseStimulusType (text) as free text is easy for future expansion. Adds these: BOOL m_bHouselight; BOOL m_bLightCentrelight; BOOL m_bLightHouselight; BOOL m_bLightLeverlight; BOOL m_bLightOppositelight; BOOL m_bNoiseClicker; BOOL m_bNoiseTone; * 2 Dec 2002 (RNC). 2.0 - From point of introducing "default ODBC DSN" system, 15 Aug 2002. - 22-Sep-2002 Integrated HTML Help 2.1 - missing UpdateData(TRUE) bugfix 2.2 - RNC 10-Dec-2002 Option to use stimulus as SD in "Pavlovian" phase: SD indicates that nosepokes will give you reinforcement (m_bPavlovianPhaseIsSD). Option to have houselight off during CS (if it's not part of the CS) (m_bHouselightOnDuringCS). (Database - recordset - config - dialogue - implementation - help... done.) Conversion to XML parameter file. 2.3 (22 Nov 2003) - writes version number and compilation date to summary file (About box and .iss file already had version 2.3, but latest distributable was 2.2) 2.4 (16-21 June 2006) - For Joff: I've a minor request for some programming help regarding the existing PIT v2.2 software. As you may know, this was written for Pat during her attempts to demonstrate PIT with drug-associated stimuli. I am planning to do some sucrose PIT in the near future, along the lines of Jeremy's original studies, and have noticed that there are a couple of minor problems in this respect with the current programme. DONE 1. wrt the pavlovian training, I hope to use continuous tones and 10 Hz clickers as the stimuli (at the moment it is set up as lights vs. pulsed tones/clicker, and I cannot choose two auditory stimuli) DONE 2. the option to give the neutral stimuli is currently set to be given on the 2nd of every 4 cycles, but I'd like to be able to give it on the 5th and 7th of 8 cycles (i.e. before the penultimate and ultimate CS+ presentations) DONE 3. For the transfer test, the CS+ and neutral CS are currently presented in a pseudorandom order, but sequentially (i.e. ISI, CS, CS), whereas the original study used a presentation sequence of 4 x (ISI, clicker, ISI, tone), with the clicker and tone obviously being counterbalanced as CS+ and neutral CS. DONE 4. Finally, and least importantly (as I can simply use the simple schedules software instead), the instrumental training is currently set to an RI schedule, and it would be useful to have an option for FI, or for it to be switched to FI). 2.5 (22 June 2006) - as 2.4, with minor internal alteration (boolean+radio -> just radio for neutral stimulus in Pavlovian section) i.e. removed m_bPavUseNeutral 2.6 (27 June 2006) - option: m_bHouselightOnWhenTaskEnds 2.7 (8 March 2007) - easier compilation for users 3.0 (12 Jan 2009) - Server default changed from "loopback" to "localhost" (Windows Vista compatibility and more general standardization). IN PROGRESS =========== THINGS TO DO ============ - THINGS TO CONSIDER: Static linking of MFC for the release build (Mike points out that I was doing this dodgily by including only necessary DLLs). Is this correct? This takes executable from 212k to ... ? other things to do? - version where stimulus serves as SD in "Pavlovian" phase, rather than true CS e.g. so that nosepoking during the stimulus causes drug delivery. Reason: it seems very difficult to get autoshaping to stimulus predictive of noncontingent drug. Pat wants an approach model, so therefore would be happy with a non-autoshaping version (i.e. an SD). Note also that food autoshaping is partially SD - that is, the CS predicts that if you approach the food hopper, you get food. (With IV drugs, that contingency is missing.) Note also that explicit SDs are capable of inducing PIT, bidirectionally (Overmier et al. 1971 Learning and Motivation 2: 103). - option to turn the houselight off when turning the CS on. Reason: requested by Pat, for comparison to SecondOrder.