The purpose of the paper is very important supporting the research activity but also the training activity for new surveyor. The DFM here descripted is a powerful mean that collects a large amount of image from numerous seismic events very that can be used in future to better understand the macroseismic effect of an earthquake.

The rules at the basis of the DFM are sharable; it is important to state how to establish the vulnerability of the structure. This is important to guide the novel researcher, especially in case of reinforced concrete structures, because they can be more vulnerable if the evaluation doesn’t take into account the rules and the legislation in force at the time they were realized.

It could be useful to add, when possible, the image of the structure before the event. In many cases, using Google Street View, should be possible obtain the situation before the structure has been subjected to the earthquake. Sometimes, especially for heavy damaged building, it is important visualize the original defects that subsequently cause the damage.

Fig.1      It is better to add in “Classification” also the term “Vulnerability”, otherwise “Typology” may appear as a synonymous of “Vulnerability” and we know that’s not true

Fig. 2     It has said that the data start from 1976 Friuli earthquake, but this event is not represented in this figure, why?

Table 1 is missing; in the text they make reference to table 1 but there isn’t at all.

LINE 48 the link is uncorrected because contain also “in Italian” that must be removed

The paper describes a FAIR-compliant dataset created by INGV to archive photographic documentation of earthquake effects in Italy since the late 1970s. Compiled mainly by the QUEST team during post‑event surveys, it preserves both digitized analogue photos and more recent digital images, each including metadata and damage classifications based on EMS‑98.  The database structure and technical features are described clearly and in enough details to allow for its use, and it is clearly  a potentially valuable resource for research in engineering and historical seismology, also since it is interoperable with the INGV seismic databases.

The paper is certainly worth for publication, but there are several comments observations that the authors should consider, that I´ll list in the following: 

- From the technical perspective the interface of of the database is well described, however the first earthquake listed in the platform is the 1980 Irpinia one, not the 1976 Friuli as indicated. Please check.

- The date and time of the captured photo are not indicated in the database interface. Is this information available? If so, could be useful to show it, in particular to better place the picture in the context of multiple events or aftershocks.

- There is no specific indication of the ID of the object the picture refer to. Is this even recorded? From the schema and figures is not clear but it would be very useful considering that multiple pictures refer to the same objects from different perspective and distances.

- In the map placing the localities with pictures, an overlay describing the estimated macroseismic intensity (e.g. from shake-map) could be provided, and the icons indicated the locality might have variable dimension depending on the number of surveyed objects.

- why is important to blurry any personal information in the pictures, it is not clear why the geolocalization of the pictures is not indicated, since it is a public information (e.g., google street view shows both buildings and their locations). Also, often the pictures refer to small details of the buildings and is missing an overview of the building itself, which could be useful to review the assessment or as base information for subsequent assessments (e.g. in case of multiple events).  

- Since some information is not published, is there a plan to provide interested researchers with a more detailed set of information, or a qualified entry point to the full set of information? Since the database is clearly science and research-oriented, a partial disclosure of the collected information would make it much less useful since there is no obvious use for e.g., communication purposes. 

- While the Mw 6.29 2009 "L'Aquila" earthquake lists 944 pictures, 56% of the entire database, in case of other recent and major earthquakes (e.g., the Mw  6.18 2016 -August- earthquake lists only 11 pictures) there is only an handful of localities with pictures of damaged buildings. Also in the case of 2023 Umbria earthquake there is only one locality with two pictures of two buildings. The 2023 Tuscany M5.1 earthquake is not present, etc. The authors acknowledged that "The population of the DFM is a recent and ongoing long-term effort, and the dataset is expected to grow substantially in the coming years [...]" but is not clear what roadmap does it have with respect to adding further data to the database and based on which prioritisation scheme.  Also, an indication of the percentage of considered objects with respect to available data would be useful to estimate coverage and completeness of the database. This is an important aspect that should be addressed more carefully, also to highlight the long-term sustainability of the database. Also, is there in place a protocol/workflow to integrate the database in all future QUEST missions, so that a minor delay can be expected before this information is made available? 

- Would be interesting for the authors to comment on the possibility to integrate rapid damage mapping approaches such as the one described in Pittore et al. (2018) to complement the database and provide a web-based interface for EMS-98 damage mapping and semi-automated evaluation of macroseismic intensity. This would also highly enrich the scientific and operational value of the database. 

- This database could be linked or integrated with the IDEA and DADO from Eurocentre. The latter for instance provide building-by-building damage data for hundreds of thousands of objects. The authors could comment on the potential for integrating these different sources.