The compound Ch3-CO-Ch2-CO-Ch3 gives a positive iodoform test where the carbon involved in formation of iodoform comes from the active methylene group. This is well explained and justified supported by a mechanism.
What intrigues me though is that the compound Ch3-CO-Ch2-CO-OEt does not give a positive iodoform test. The reason being that ( according to the book), OEt is better leaving group than -CI3 and thus instead of CI3 anion being relased the ethoxide ion is removed. But then, shouldn't the ethoxide ion participate in further haloform reaction. Ethanol can after all give iodoform test and once we have ethoxide ion in the solution, why doesn't it leads to the fomation of CHI3?

I have pictures of mechanisms but I seem to unable to upload them. ( Error 413)be

Your explanation of the mechanism seems a little garbled. All that is happening is the OH^- is hydrolysing the ester in preference to deprotonating at the alpha position. The next part I think is because the carboxylic acid you get as a result contains an acidic hydrogen and the ethoxide will react with that instead of the alpha hydrogens required to give a positive iodoform test.