Well friends, as I said last night I am to report as follows:
1) The semi auto plugin didn't work as usual even after N+ was upgraded the day before yesterday. I called N+ today and they spent some time with TeamViewer and suggested to call up gdfl which I did
2) gdfl accessed my system remotely and ran their sample strategy successfully-the orders went to the API for me to PLACE. So that leaves my afl as the culprit for not being able to deliver orders to the API.
3) I am thinking of writing my own strategy, through some coder, in afl which would filter out most of the bad trades in the gdfl strategy. Now, is that a good or feasible idea? What do the experts think?
Glad to learn that you have a grip on the problem.
If I may say so, the folks that support AFL related requests here in TJ are some of the most selfless, active and finest coders you will find anywhere. If you can use their services, and offer more than electronic thanks, it is certainly a great idea. They may, or may not, have integrated AFLs on the GDFL offering earlier, but trust me, a little elbow space and they will be on top.
The only point of worry, and this applies to everybody, is that there is no simulation server; you cannot do dry runs. [Idea :!]
You have to test on a live market; and it is like repairing an car engine while it is running or doing a heart surgery. At the risk of digressing, Omnesys does offer a staging server but I do not know details.