PCMTec Development Blog

[Tazzi]

Always loved the look of 3D mapping. Keen to develop something like that.
Can easily see the huge amount of work they have put into designing and structuring their application.

Looks like yours is coming together Rolls!

[rolls]

Always loved the look of 3D mapping. Keen to develop something like that.
Can easily see the huge amount of work they have put into designing and structuring their application.

Looks like yours is coming together Rolls!

Yeah. Their software obviously does a lot more than just tune cars as well, they have a massive amount of data logging and ecu building software as well. I think they got their moneys worth.

Regarding the 3d mapping if I add it in (not sure if its actually useful as I've never used it) I'll use something already existing. No point reinventing the wheel with something like that.

[Tazzi]

Always loved the look of 3D mapping. Keen to develop something like that.
Can easily see the huge amount of work they have put into designing and structuring their application.

Looks like yours is coming together Rolls!

Yeah. Their software obviously does a lot more than just tune cars as well, they have a massive amount of data logging and ecu building software as well. I think they got their moneys worth.

Regarding the 3d mapping if I add it in (not sure if its actually useful as I've never used it) I'll use something already existing. No point reinventing the wheel with something like that.

I thinks its more for visually identifying spikes and smoothing graphs easier. I recall watching a tuning video which visually showed the difference between LS1 tunes for VT through to VZ commodores using the 3D mapping, and can easily identify the differences and changes in the map.

All personal preference though.

[rolls]

I thinks its more for visually identifying spikes and smoothing graphs easier. I recall watching a tuning video which visually showed the difference between LS1 tunes for VT through to VZ commodores using the 3D mapping, and can easily identify the differences and changes in the map.

All personal preference though.

I was thinking I would introduce a derivative colour mode. Where the colour depends on the sum of the derivative of the edges. So you would get a colour change on the edge opposed to the peak to help identify lack of smoothness.

3D helps as well though.

[Tazzi]

I was thinking I would introduce a derivative colour mode. Where the colour depends on the sum of the derivative of the edges. So you would get a colour change on the edge opposed to the peak to help identify lack of smoothness.

3D helps as well though.

Thats an interesting take on it! Thinking outside the box

[VX L67 Getrag]

Yeah honestly I like where you heads at, if we can start a public sourcing project to really get you going I think you're really on the right track & not just kicking tyres!!!

[DarrylC]

Regarding the 3d mapping if I add it in (not sure if its actually useful as I've never used it) I'll use something already existing. No point reinventing the wheel with something like that.

It is definitely useful. Tuners will alter the tune and keep the basic shape. Often they will just grab different points and pull them up and down keeping the shape and in particular hills and valleys. This is what a tuner in Frankston showed me when he did tuning 101 for me whilst we had time to kill waiting for a dyno.

[Dylan]

I to use the 3D map for editing

[rolls]

Ok. Something we will definitely need to implement then. Devexpress has some 3d graphs im fairly sure, will have a play.

[rolls]

Taught myself http over the weekend. I've never actually had a need to write a http service before so thought this would be a good time to learn. I previously understood http and had implemented raw TCP services (using modbus etc) but never one over http.

Wrote a small Heroku service (they host web services so you don't need to manage security on server) that will host the templates and talk to the editor program via simple JSON over HTTP GET requests. Handles the user database, encryption, licenses etc all off site so there is nothing to protect within the actual software as it is all done offsite and protected that way.

Learnt all about storing passwords using bcrypt so you store a secure password hash only (so you don't have to handle actual passwords) and use a different salt for every single password. This makes all sorts of password attack vectors like rainbow tables etc useless.

Also implemented SQL that uses parameterised queries so the server is immune to SQL injection attacks and things like that. Server uses unique session identifiers and HTTPS so theres no possibility of a replay attack.

What is even better is due to the extremely low bandwidth and storage required it fits into the Heroku free version that doesn't even cost anything.

After reading it seems a large number of people still do not know what they are doing in this regard. Eg storing actual passwords in databases, or using constant salts. Non constant time password comparison where you can use statistical analysis on how long it took for a password to be rejected to infer how many characters of the password were correct.

Stuff like writing dynamic SQL queries which are susceptible to sql injection to get a server to spit out a list of user accounts and passwords or even drop the entire database!

Seems even back in 2012 linkedin who are a massive company did dumb stuff like using no salt on their passwords!