From all the things said in that interview only one thing interesting (to developers only) : projects. But it means absolutely nothing to an average metatrader user (apart from the fact that it is probably going to be much, much easier to post compiled code than "projects" and then be drawn into endless explanation how to compile a project)

as for the rest : Classes without inheritance? No operators overload? Classes and structures without any debugger (I am not even saying a good debugger)? Come on. It is called "building the house from the roof" and always has one ending scenario

All I wish to see on MT5 is a easy ECN integration with no bridges needed, a level2 etc. TO GOOD TO BE TRUE. This is my wish anyway

All I wish for MT5 is an easy ECN integration with no bridges needed, a level 2 stuff etc. I know this would be too good to be true but this is what I wish anyway.

my wish too...

This is MQL5 we are talking about, the programming language, not the platform client. Plus the interview makes no mention of the platform getting an upgrade - or am I wrong?

imo, Metaquotes won't fix known issues, let alone add new features, until it sees its business threatened, either by the NFA culling its main customers (the bucketshops) and/or ECN's deciding that it might make business sense to allow MT traders access to their feeds. But I'm not holding my breath on either happening anytime soon...

Actually, there is one broker doing just that and the bridge should be done in a couple of weeks.

__________________
1/2 pip spread rebates and a free forex signals blog including OzFX based swing, only at RebateFX.com

My wishes also mostly about the tester. When you have tested things using your own framework, it's a pain in the ass to use the very limited mt4 tester.

Some improvement suggestions would be possibility to test strategies on multiple currencies at the same time, possibility to have one optimization period and one verification period, no limit on parameter combinations and better optimization strategies for genetic algos like possibility to optimize for combinations of results (50% profit, 50% profit factor) and there should absolutely be a possibility to optimize for consistency of results (based on distribution of results from different orders)

edit: and my main wish would be working Linux version and to be honest that would be that they would throw away the whole mqlX and just would give a standard api for java (or c/++, lisp, erlang, ocaml, ruby or anything because there are a lot of mature languages already so there's no point in trying to create your own). Oh and add a decent scoping too, it's absolutely crazy that inner scope variables are available outside the scope too.

There was a thread started on TSD not long ago by an ECN rep. stating that they had already developed a bridge. All seemed great till they revealed that it would not be made available to individual traders, only brokers - they even had the audacity to suggest that traders should try and get their broker to implement their software - real slime! Lets hope this isn't what you are referring to...

Exactly....

I'm curious as to which broker that might be.

Trader1