Accurate weak-lensing analysis requires not only accurate measurement of galaxy shapes but also precise and unbiased measurement of galaxy redshifts. The photometric redshift technique appears as the only possibility to determine the redshift of the background galaxies used in the weak-lensing analysis. Using the photometric redshift quality, simple shape measurement requirements, and a proper sky model, we explore what could be an optimal weak-lensing dark energy mission based on FoM calculation. We found that photometric redshifts reach their best accuracy for the bulk of the faint galaxy population when filters have a resolution R~3.2. We show that an optimal mission would survey the sky through 8 filters using 2 cameras (visible and near infrared). Assuming a 5-year mission duration, a mirror size of 1.5m, a 0.5deg2 FOV with a visible pixel scale of 0.15, we found that a homogeneous survey reaching IAB=25.6 (10sigma) with a sky coverage of ~11000deg2 maximizes the Weak Lensing FoM. The effective number density of galaxies then used for WL is ~45gal/arcmin2, at least a factor of two better than ground based survey. This work demonstrates that a full account of the observational strategy is required to properly optimize the instrument parameters to maximize the FoM of the future weak-lensing space dark energy mission.