I have some comments, mostly regarding language. General note: kilohertz is kHz, not KHz page1: sentence: ...detector aims to sample PbPb collisions up to a 50 kHz interaction rate. An upgraded TOF should, then, be able to handle such rates both at the detector level and at the data readout level. 2nd paragraph: 1st sentence ...luminosity, a 7Hz... 3rd sentence: well affordable-->certainly achievable last sentence: MRPCs have been demonstrated to tolerate fluxes up to 1kHz/cm2, so there are no concerns regarding the counting rates of the TOF detectors. 2 The TOF Readout If this document is meant to be for a broader audience than just the TOF experts, you should define SM, DRM, TRM, HPTDC...(probably VME and DAQ are OK ;-) ) last paragraph: Measurements done in the experimental site...do you mean Measurements performed with the existing TOF in situ? 2.1 1st sentence: ...HPTDC, when... 2nd sentence: As shown in [3], the HPTDC can handle a maximum input rate of 500 kHz. (combine the first and second paragraphs) 2nd paragraph, 2nd sentence: ...implemented, the time... 4th paragraph: ...average number of hits per TRM... “it corresponds to”-->corresponding to 2.2 2nd paragraph, 2nd sentence: So far, the trigger...ALICE has always been... 3rd sentence: ...throughput, but we need... 4th sentence: ...in pp, the multiplicities will be lower, but an... first bullet: ...non--replaceable slave cards. (also, performances should be performance in both cases) second bullet: 3rd sentence, change to: We might also consider packaging the... last sentence: ...as indicated by the dotted line... 2.3 2nd paragraph: Presently,... As shown in Fig 3, with only 1...which results... last sentence: remove “actually” 3rd paragraph, 1st sentence: This issue, which seems related to the small data size of the TOF event fragment shipped from the DRM to the LDC, will now be further investigated by the TOF and DAQ groups. 4th paragraph: Using just one DDL, we also tested the... 10000 hit->10000 hits ...Fig 4, demonstrating the remarkable performance of the TOF readout...almost empty events, a ... 2.4 ..., we foresee replacing the existing DRM cards, installing up to two DDL links per DRM card, and... We anticipate reaching a trigger output rate of 200-400 kHz. As explained above, this target is achievable through the combination of new DDL specifications, more performant PCs as LDCs and, on the firmware side, improved.... 2nd para, last sentence: As is the case for other detectors, the ... 3rd para: In PbPb collisions, the... In pp, ...kHz are required. last sentence: An initial approach will be to downscale...possibility of more efficiently selecting events... 3. Currently, the ... 2nd para: In the new trigger scheme, the possibility of shipping the entire macropad’s information to the FTP, which will then be able to combine these data with data coming from other triggering detectors (typically the ITS), will be explored. ...confirm an increase in the efficiency of selecting... From the hardware point of view, two... ...in the laboratory and at the experimental site. penultimate sentence: As anticipate in [3], with ...the expected bandwidth to the DAQ...