Pith. sign in

REVIEW 3 major objections 6 minor 18 references

Quality assurance and reporting for FLASH clinical trials:the experience of the FEATHER trial

T0 review · 3 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The FEATHER trial shows that existing proton-therapy infrastructure, extended with 1 kHz time-resolved log files and a patient-specific phantom, can ensure safe delivery and support retrospective dose and dose-rate reconstruction for…

desk verdict A genuinely useful, field-tested QA and reporting template for early FLASH trials, with the log-file reconstruction being the most novel piece; the paper's main weakness is that the reconstruction is validated on one field at one point, without uncertainty bounds. read the letter →

arxiv 2502.02677 v1 pith:MHI33PYW submitted 2025-02-04 physics.med-ph

Isabella Colizzi , Robert Schaefer , Jonas Brueckner , Gaia Dellepiane , Martin Grossmann , Maximilan Koerner , Antony John Lomax , David Meer
show 4 more authors
Benno Rohrer Carla Rohrer Bley Michele Togno Serena Psoroulas
This is my paper · ORCID
classification physics.med-ph
keywords FLASHprotontherapyultra-highdoseratequalityassuranceclinicaltriallogfilereconstructionPBS-averageFEATHER
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The FEATHER trial, a randomized phase II/III study treating cats with oral squamous cell carcinoma by proton therapy, needed a quality-assurance and reporting framework for FLASH delivery before consensus guidelines existed. The authors show that an existing clinical proton beamline can be extended with a 1 kHz time-resolved monitor readout, daily dose-monitor calibration, and a patient-specific phantom so that both conventional and ultra-high-dose-rate treatments are delivered safely and reproducibly. They further show that the resulting log files, in which each spot is modeled as a two-dimensional Gaussian, allow reconstruction of the delivered dose map and the PBS-average dose-rate map; on the demonstrated field the reconstructed dose agrees with measurement within 2% and the dose rate within 10% at the reference point. If correct, this provides a concrete, reproducible template for QA and reporting in early FLASH trials, making cross-trial comparison of delivery parameters possible.

What carries the argument

The carrying mechanism is the time-resolved delivery log file combined with a spot-wise Gaussian reconstruction model. Each irradiation spot is represented as a 2D Gaussian: mean position from Hall-sensor-derived scanning magnet currents, width from commissioning data, and weight from the logged dose-monitor counts; timestamps mark the end of each spot. Summing the Gaussians yields a 2D dose map, and ordering them in time allows the Folkerts PBS-average dose rate to be computed per voxel at 5%-95% dose thresholds. This turns the treatment log into a retrospective measurement of both dose and time structure, which is what makes the reporting protocol useful after unblinding and under any future dose-rate definition.

What would settle it

Deliver the same FEATHER test field at a gantry angle far from 90° (where the beam transport was optimized) and compare the log-file reconstructed dose and dose-rate maps against the CCD-measured maps. If the reference-point disagreement exceeds 2% in dose or 10% in dose rate, or the edge discrepancies grow beyond what the threshold placement can explain, the claim that the log files enable reliable retrospective reconstruction fails to generalize.

Watch

Extended reading notes

Core claim

The central claim is that a QA and reporting strategy based on existing clinical infrastructure can guarantee safe, accurate, reproducible delivery of transmission proton beams in both conventional and ultra-high-dose-rate modes, and that time-resolved delivery log files provide everything needed for retrospective dose and dose-rate reconstruction. The strategy extends the Gantry 1 monitoring system to sample at 1 kHz, records two independent dose monitors, Hall-sensor currents from the scanning magnets, and 100 µs timestamps, and reconstructs each spot as a 2D Gaussian whose position comes from the Hall sensors, width from commissioning data, and integral dose from logged monitor counts. Dose-rate maps are computed with the Folkerts PBS-average metric. The paper validates the chain on simulated fields: patient-specific and daily QA passed the stated tolerances (spot position 2 mm, spot size ±10%, central dose within 5%, gamma 3 mm/3% for over 90% of voxels), and log-file reconstruction matched measured dose within 2% and measured dose rate within 10% at the reference point. On this basis the authors present the first QA and reporting protocol designed for and applied in a FLASH clinical trial.

Load-bearing premise

The whole chain assumes the simplified spot model – fixed spot sizes from commissioning, positions from magnet-current readings, and dose from logged monitor counts, with constant spot-changing times and beam intensity for the dose-rate calculation – stays accurate across all gantry angles and target geometries, although the paper demonstrates agreement on only one example field.

Editorial extensions

If this is right

  • Existing clinical proton QA can be adapted to FLASH with only modest hardware and software changes, notably a 1 kHz time-resolved readout and daily dose-monitor calibration.
  • Every treated field leaves behind a machine-readable time trace from which dose and dose-rate maps can be recalculated long after delivery, even if the dose-rate definition changes.
  • The three-part blinded reporting scheme keeps dosimetric information available to clinicians while withholding dose-rate and beam-current data until the study is unblinded.
  • Per-field verification of local dose rate above 40 Gy/s is achievable with a micro diamond detector, supporting the UHDR arm's intended dose-rate level.
  • If other centers adopt the same protocol, FLASH delivery parameters from different trials can be compared directly, which is currently impossible due to heterogeneous reporting.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper, the 2%/10% agreement should be treated as a proof of method rather than a blanket accuracy bound, because only one example field is demonstrated and the beam model is simplified; a natural extension is to validate the same reconstruction across the range of gantry angles and target sizes enrolled in the trial.
  • Beyond the paper, the sharp dose-rate variations at field edges likely come from the Folkerts threshold jumping between beam-on and beam-off phases; a continuous-time reconstruction or a threshold defined per voxel on stable plateau regions would probably smooth these maps and is testable with the same log files.
  • Beyond the paper, the assumption of constant spot-changing times (11 ms in T, 4 ms in U) and constant beam intensity is an internal source of error; since the log files record actual timestamps, fitting real per-spot delivery times instead of assumed constants should reduce the discrepancy between reconstructed and planned dose rate.
  • Beyond the paper, the FEATHER design randomizes only beam current, with identical dose plans in both arms, so the archived logs could later be used to test which PBS-average dose-rate threshold actually predicts toxicity in the animals.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper reports the development and in-practice testing of a quality assurance (QA) and reporting protocol for the FEATHER trial, a veterinary phase II/III FLASH proton therapy trial at PSI Gantry 1. The authors describe UHDR-specific and patient-specific QA procedures, a time-resolved monitor readout that generates delivery log files, and a reporting template for dosimetric, machine, and dose-rate information. They validate the approach on simulated treatments: PSQA/DQA dose comparisons pass stated tolerances, and log-file-based reconstruction of one example field agrees with the measured dose within 2% and with the measured dose rate within 10% at a single reference point. The paper concludes that the protocol ensures safe, accurate, reproducible delivery and enables retrospective dose and dose-rate reconstruction.

Significance. If the claims hold, this is a useful practical contribution: it provides one of the first implemented QA and reporting templates for a clinical FLASH trial, with concrete tolerances, a time-resolved logging concept, and a reporting structure that can inform other centers. The explicit aim to align with later consensus recommendations (refs. 2, 19) and the inclusion of a sample delivery report strengthen its value to the community. The log-file reconstruction approach, if validated across fields and gantry angles, would be valuable for retrospective analyses under evolving FLASH dose-rate definitions. The main caveat is that the validation evidence is currently limited to one example field and one reference point, so the generalizability of the central reconstruction claim is not yet established.

major comments (3)
  1. [III.B, Figs. 5 and 6] The central claim that log files permit retrospective reconstruction of delivered dose and PBS-average dose-rate maps is supported only by comparison at a single 'arbitrary position' (dose within 2%, dose rate within 10%). The reconstructed dose map visibly overestimates dose on the right side of the field, and the reconstructed dose-rate map shows sharp voxel-to-voxel variations at field edges, which the authors attribute to threshold placement and spot-position adjustment. No gamma pass rate, spatial map of disagreement, uncertainty bounds, or results for the other gantry angles/fields of the trial are provided. Because the FEATHER trial uses three different gantry angles (fields F0-F2 in the Appendix) and varied target geometries, this evidence is insufficient to support the general retrospective-analysis claim. Please quantify the edge deviations, report the fraction of voxels affected in dose-rate maps, and provide multi-field validation across the trial's gantry angles.
  2. [II.B.3] The reconstruction model depends on assumptions whose validity is not demonstrated: each spot is modeled as a 2D Gaussian with width from commissioning data, position from Hall-sensor-derived currents, and integral dose from logged monitor counts; the TPS comparison assumes constant spot-changing times (11 ms in T, 4 ms in U), constant beam intensity, and a rectangular spot grid. The authors acknowledge that UHDR spots are switched on before the planned position is reached and that beam size and integral vary across the spot map. These assumptions could systematically bias reconstructed dose-rate maps, especially for voxels near the 40 Gy/s FLASH threshold, and the paper does not show that they hold across the spot map and gantry angles. Please provide a sensitivity analysis of the reconstruction to spot-position offsets, spot-size uncertainties, and spot-timing assumptions, or direct validation of these quantities against independent measurements.
  3. [III.A and II.B.2] The QA program validation is presented for a single example field and a single treatment day. The paper claims that the QA procedures 'effectively ensure the correct and safe delivery' and that the protocol was 'successfully used during the cat irradiations,' but no statistics over multiple PSQA/DQA sessions, no interlock occurrence data, and no end-to-end tests at different gantry angles are reported. Without such data, the reproducibility and safety claims cannot be quantitatively assessed, and the paper's stated generalizability to the trial's three-field, three-fraction schedule remains unsupported.
minor comments (6)
  1. [II.A] The text says the UHDR arm dose rate is estimated to be 'above 50 Gy/s' while the PSQA acceptance criterion is 'above 40 Gy/s'; please clarify the relationship between these two numbers and how the 40 Gy/s threshold is applied per field.
  2. [II.B.3] The description of the Hall-sensor-based spot position reconstruction would benefit from a specification of how the Hall sensor readings were calibrated to magnet current and then to position, including the uncertainty of that calibration.
  3. [III.B, Fig. 5] The figure caption states the µD dose is 10.3 Gy, but the text says the reconstructed and delivered dose agree to within 2% without specifying which measured value (µD, CCD, or TPS) is the reference; please state this explicitly.
  4. [Appendix, Fig. 8] The figure caption contains grammatical errors ('the dose rate calculation two neighboring points'); please rephrase to 'at two neighboring points' and clarify the unit and definition of the plotted quantity.
  5. [References] References 14 and 15 are both listed as 'IEC 60601-2-64:2014'; please consolidate the duplicate entry.
  6. [General] The paper refers to 'additional materials' containing a full delivery report of a test patient, but only a preview is shown in Figure 7; since the reporting template is a key deliverable, consider making the full template available as supplementary material.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the log-file dose/dose-rate reconstruction is benchmarked against independent measurements, and self-citations are empirical commissioning references rather than load-bearing inputs.

full rationale

The paper's central claims are that a QA protocol ensures safe/accurate UHDR and conventional delivery and that time-resolved log files enable retrospective dose and dose-rate reconstruction. Neither claim reduces to its inputs by construction. The reconstructed dose maps are built from Hall-sensor-derived spot positions, logged monitor counts, and measured time stamps, with beam widths taken from commissioning data; these are independent inputs, not outputs of the reconstruction. The reconstruction is then compared against a CCD camera, a microDiamond detector cross-calibrated to a primary-standard-traceable reference chamber, and the TPS. The TPS comparison is explicitly acknowledged to rely on idealized assumptions (constant spot-changing times, rectangular grid, constant intensity) and is not treated as ground truth; the measured microDiamond values provide the independent validation. The self-citations (refs. 12 and 17) are empirical machine commissioning and detector characterization results from the same group, used to justify operational parameters such as daily calibration and dose-rate independence; they are not unverified uniqueness theorems or ansatz adopted by citation, and they do not by themselves force the paper's conclusions. The QA protocol's corrective actions (position offsets, dose scaling, boosting factors) are standard clinical QA practice and constitute quality control, not the renaming of a fitted parameter as a prediction. The acknowledged limitations—dose overestimation on one side of the example field and sharp dose-rate variations at field edges due to threshold placement in the Folkerts metric—are correctness risks for generalization, but they are explicitly disclosed and do not indicate circular construction. Overall, no load-bearing step reduces to the claim being derived.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

No new physical entities are introduced. The central claims rest on institutional calibration data, a simplified Gaussian spot model, and the assumption that beam behavior is stable within a treatment day. The free parameters are hand-chosen constants used in the TPS dose rate model; they are not fitted to trial endpoints, but they affect the reported dose rate maps. The axioms are mostly domain assumptions from prior PSI commissioning work, plus two ad hoc modeling choices acknowledged by the authors.

free parameters (4)
  • Spot-switching time in T direction = 11 ms
    Assumed constant in the TPS dose rate calculation (Section III.B); actual switching time varies across the spot map, contributing to dose rate map deviations.
  • Spot-switching time in U direction = 4 ms
    Assumed constant for the horizontal direction in the same TPS dose rate model; used in the PBS-average dose rate estimate.
  • Beam intensity in the TPS dose rate model = About 200 MU per 1 Gy at field center
    The TPS dose rate estimate assumes constant MU/s and roughly 200 MU for 1 Gy at the field center; the paper notes actual beam intensity varies.
  • Averaged spot size and integral per gantry angle = TPS beam model averages
    The TPS assumes constant spot size and Gy/MU across the spot map within each gantry angle; the paper identifies this as a simplification producing edge deviations (Section III.B).
assumptions (5)
  • domain assumption Beam parameters such as charge density, transmission, spot position, and spot size are stable within a treatment day.
    Invoked in Sections II.B.1 and II.B.2 to justify daily QA only before treatment; based on prior commissioning experience (ref 12) rather than new trial data.
  • domain assumption The Gantry 1 interlock chain reacts fast enough that the most probable interlock failures are type-B hazards or lower under AAPM TG-35, so the interlock chain was not altered.
    Section II.B.1; supports the safety claim and relies on TG-35 classification and more than 20 years of clinical operation.
  • domain assumption The CCD camera and microDiamond detector are dose-rate independent in the UHDR range and are traceably cross-calibrated to the institute reference chamber.
    Sections II.B.2 and ref 17; required for PSQA dose and dose rate measurements to be valid under ultra-high dose rate conditions.
  • ad hoc to paper Each irradiation spot can be modeled as a 2D Gaussian with mean position from Hall sensor readings, width from commissioning data, and integral dose from logged monitor counts.
    Section II.B.3; the core log-file reconstruction model; the paper acknowledges imperfect accuracy at field edges.
  • ad hoc to paper The TPS dose rate estimate may use constant spot-changing times, a rectangular spot grid, and a step-like cumulative dose function.
    Section III.B; these simplifications explain differences between TPS and log-file reconstructed dose rate maps.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quality assurance and reporting for FLASH clinical trials:the experience of the FEATHER trial." pith.science (2026). https://pith.science/paper/MHI33PYW

@misc{pith2026250202677,
  author       = {Pith},
  title        = {Pith review of: Quality assurance and reporting for FLASH clinical trials:the experience of the FEATHER trial},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MHI33PYW}},
  note         = {Machine review of arXiv:2502.02677}
}
read the original abstract

Research on ultra-high dose rate (UHDR) radiation therapy has indicated its potential to spare normal tissue while maintaining equivalent tumor control compared to conventional treatments. First clinical trials are underway. The randomized phase II/III FEATHER clinical trial at the Paul Scherrer Institute in collaboration with the University of Zurich Animal Hospital is one of the first curative domestic animal trials to be attempted, and it is designed to provide a good example for human trials. However, the lack of standardized quality assurance (QA) guidelines for FLASH clinical trials presents a significant challenge in trial design. This work aims to demonstrate the development and testing of QA and reporting procedures implemented in the FEATHER clinical trial. We have expanded the clinical QA program to include UHDR-specific QA and additional patient-specific QA. Furthermore, we have modified the monitor readout to enable time-resolved measurements, allowing delivery log files to be used for dose and dose rate recalculations. Finally, we developed a reporting strategy encompassing relevant parameters for retrospective studies. We evaluated our QA and reporting procedures with simulated treatments. This testing confirmed that our QA procedures effectively ensure the correct and safe delivery of the planned dose. Additionally, we demonstrated that we could reconstruct the delivered dose and dose rate using the delivery log files. We developed and used in practice a comprehensive QA and reporting protocol for a FLASH clinical trial at the Paul Scherrer Institute. This work aims to establish guidelines and standardize reporting practices for future advancements in the FLASH-RT field.

Figures

Figures reproduced from arXiv: 2502.02677 by the authors.

Figure 1
Figure 1. In-house QA phantom designed for the QA procedure: CCD mounted to a rotation [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. The five-spot pattern is delivered to evaluate spot positioning and beam size (left). [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. 2D dose distribution of the reference and test fields, including the dose difference [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: (left) µD detector current readout during a UHDR beam delivery. (right) The cumulative sum of the dose delivered. III.B. Dose and dose rate reconstruction from log files In [PITH_FULL_IMAGE:figures/full_fig_p013_4.png]
Figure 5
Figure 5. Figure 5: Dose map recalculated from log files (left), measured with the CCD camera (cen [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: Dose Rate map recalculated from log files (left), and calculated by our TPS (right). [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Screenshot of the delivery report for an example patient. [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: Difference between the dose rate calculation two neighboring points. Even if the [PITH_FULL_IMAGE:figures/full_fig_p020_8.png]
Figure 9
Figure 9. Figure 9: Comparison between log file reconstructed dose rate using a continuous function [PITH_FULL_IMAGE:figures/full_fig_p020_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

18 extracted references · 16 canonical work pages

  1. [1]

    Favaudon, L

    V. Favaudon, L. Caplier, V. Monceau, F. Pouzoulet, M. Sayarath, C. Fouillade, M.-F. Poupon, I. Brito, P. Hupé, J. Bourhis, J. Hall, J.-J. Fontaine, and M.-C. Vozenin, Ultrahigh dose-rate FLASH irradiation increases the differential response between normal and tumor tissue in mice, Science Translational Medicine 6 (2014)

    work page 2014

  2. [2]

    Medical Physics Dataset Article: a database of FLASH Murine In-Vivo Studies

    M. Toschini, I. Colizzi, A. J. Lomax, and S. Psoroulas, Medical Physics Dataset Article : a database of FLASH Murine In - Vivo Studies , 2025, arXiv:2501.16906 [physics]

    work page Pith review arXiv 2025

  3. [3]

    Colizzi, M

    I. Colizzi, M. Toschini, A. J. Lomax, and S. Psoroulas, A database of FLASH Murine In - Vivo Studies , 2024

    work page 2024

  4. [4]

    E. S. Diffenderfer, B. S. Sørensen, A. Mazal, and D. J. Carlson, The current status of preclinical proton FLASH radiation and future directions, Medical Physics 49 , 2039--2054 (2022)

    work page 2022

  5. [5]

    P. A. Taylor, J. M. Moran, D. A. Jaffray, and J. C. Buchsbaum, A roadmap to clinical trials for FLASH , Medical Physics 49 , 4099--4108 (2022)

    work page 2022

  6. [6]

    W. Zou et al., Framework for Quality Assurance of Ultrahigh Dose Rate Clinical Trials Investigating FLASH Effects and Current Technology Gaps , International Journal of Radiation Oncology, Biology, Physics 116 , 1202--1217 (2023)

    work page 2023

  7. [7]

    Spruijt, S

    K. Spruijt, S. Mossahebi, H. Lin, E. Lee, J. Kraus, A. Dhabaan, P. Poulsen, M. Lowe, A. Ayan, S. Spiessens, J. Godart, and M. Hoogeman, Multi-institutional consensus on machine QA for isochronous cyclotron-based systems delivering ultra-high dose rate ( FLASH ) pencil beam scanning proton therapy in transmission mode, Medical Physics 51 , 786--798 (2024)

    work page 2024

  8. [8]

    Garibaldi, S

    C. Garibaldi, S. Beddar, N. Bizzocchi, T. T. Böhlen, C. Iliaskou, R. Moeckli, S. Psoroulas, A. Subiel, P. A. Taylor, F. V. d. Heuvel, V. Vanreusel, and D. Verellen, Minimum and optimal requirements for a safe clinical implementation of ultra-high dose rate radiotherapy: A focus on patient’s safety and radiation protection, Radiotherapy and Oncology 196 (2...

    work page 2024

Show all 18 references
  1. [9]

    M. M. Folkerts, E. Abel, S. Busold, J. R. Perez, V. Krishnamurthi, and C. C. Ling, A framework for defining FLASH dose rate for pencil beam scanning, Medical Physics 47 , 6396--6404 (2020)

    2020

  2. [10]

    B. S. Sørensen, E. Kanouta, C. Ankjærgaard, L. Kristensen, J. G. Johansen, M. K. Sitarz, C. E. Andersen, C. Grau, and P. Poulsen, Proton FLASH : Impact of Dose Rate and Split Dose on Acute Skin Toxicity in a Murine Model , International Journal of Radiation Oncology*Biology*Ph...

    2024

  3. [11]

    S. Lin, T. Boehringer, A. Coray, M. Grossmann, and E. Pedroni, More than 10 years experience of beam monitoring with the Gantry 1 spot scanning proton therapy facility at PSI , Medical Physics 36 , 5331--5340 (2009)

    2009

  4. [12]

    K. P. Nesteruk, M. Togno, M. Grossmann, A. J. Lomax, D. C. Weber, J. M. Schippers, S. Safai, D. Meer, and S. Psoroulas, Commissioning of a clinical pencil beam scanning proton therapy unit for ultra-high dose rates ( FLASH ), Medical Physics 48 , 4017--4026 (2021)

    2021

  5. [13]

    J. A. Purdy, P. J. Biggs, C. Bowers, E. Dally, W. Downs, B. A. Fraass, C. J. Karzmark, F. Khan, P. Morgan, and R. Morton, Medical accelerator safety considerations: report of AAPM Radiation Therapy Committee Task Group No . 35, Medical Physics 20 , 1261--1275 (1993)

    1993

  6. [14]

    Togno, K

    M. Togno, K. P. Nesteruk, R. Schäfer, S. Psoroulas, D. Meer, M. Grossmann, J. B. Christensen, E. G. Yukihara, A. J. Lomax, D. C. Weber, and S. Safai, Ultra-high dose rate dosimetry for pre-clinical experiments with mm-small proton fields, Physica Medica: European Journal of Me...

    2022

  7. [15]

    van de Water, S

    S. van de Water, S. Safai, J. M. Schippers, D. C. Weber, and A. J. Lomax, Towards FLASH proton therapy: the impact of treatment planning and machine characteristics on achievable dose rates, Acta Oncologica (Stockholm, Sweden) 58 , 1463--1469 (2019)

    2019

  8. [16]

    Tobias Böhlen, S

    T. Tobias Böhlen, S. Psoroulas, J. D. Aylward, S. Beddar, A. Douralis, G. Delpon, C. Garibaldi, A. Gasparini, E. Schüler, F. Stephan, R. Moeckli, and A. Subiel, Recording and reporting of ultra-high dose rate " FLASH " delivery for preclinical and clinical settings, Radiothera...

    2024

  9. [17]

    , " * write output.state after.block =

    ") INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.all := #1 'mid.sentence := #2 'after.sentence := #3 'after.block := STRINGS s t FUNCTION output.nonnull 's := output.state mid.sentence = ", " * write output.state...

  10. [18]

    write newline

    " write newline "" before.all 'output.state := FUNCTION fin.entry add.period write newline FUNCTION new.block output.state before.all = 'skip after.block 'output.state := if FUNCTION new.sentence skip FUNCTION not #0 #1 if FUNCTION and 'skip pop #0 if FUNCTION or pop #1 'skip ...

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.