I'm trying to estimate an OLS gravity equation and I have encountered a problem regarding the model. The dataset range from 1993-2017 of bilateral trade from UK to the countries in which data is available. I am trying to include a importer, exporter and time fixed effects. A lot of the variables are being omitted, so I'm unsure what I am doing wrong and I am curious whether someone could give me some input.

reg lnexport lngdp_d lngdp_o lndist contig colony comlang_off fta_wto dimp* dyear* dyear*, robust
The PPML is most likely the way to go, but I want to sort out what is wrong with my OLS model first. Please let me know if there are more information I could provide. Furthermore, I have tried to define variables as paneldata in order to use the xtreg instead, but i am most likely doing something wrong since the variables still get omitted. I define the variables as

egen id = group(Countryi CountryJ)
tsset id Year
sort Countryi CountryJ Year