What’s New

The current release of the CASTR application: a small fix to Import .pri. Your existing projects and their results don't change. For everything it does, see Features and the User Manual.

Fixed

• Import .pri keeps a row that has a Ref # but no name. A new row not yet filled in used to be skipped, so the same file could show one more risk in the CASTR Spreadsheet than in the app. It now comes in at Include = No, so the two counts agree and no run changes.

Goes with CASTR Spreadsheet 75.7.0: both now call a risk's permanent number its Ref #, and keep it through a .pri file. Existing projects and their results did not change.

New

• Ref #. A risk's permanent number is now called Ref # everywhere: on Risk Identification (it was #), on the Risk Register (it was Reg #), and in titles, reports, exports and messages. The numbers themselves stay the same.
• A count of the risks in each table. The Standard and Triage views of Risk Identification and every view of the Risk Register show it below the table, on the left, or "Showing 5 of 37 risks" when a filter is on. A deleted risk's number is never used again, so the highest Ref # isn't the number of risks.

Improved

• A deleted risk's number stays retired through a .pri file. Export .pri writes the highest Ref # issued as the file's last line, and Import .pri numbers new risks above it.
• Minor-risk groups keep their Ref # through a .pri file. Import .pri used to give them new numbers. It now keeps the ones in the file, and Export .pri writes them back.
• A notice when no earlier projects are found. If a first start of CASTR finds no earlier projects, the project list says where CASTR 2.17 and earlier kept them and how to copy them in. New users can dismiss it.

Fixed

• A correlation on a minor-risk group is no longer lost. Import .pri left out a correlation pair on a group's Register row, even one CASTR's own Export .pri had written, which could change results. It now stays on its group.
• A project exported to a .pri and imported again now runs the same for the same seed. Renumbering its minor-risk groups could change which random draws each risk took.

Phase-segment charts choose what they show, a project with one segment no longer keeps a phase Project Wide, and imports warn of text where a number belongs.

Before you install (users of 2.18)

• A project with one segment no longer has a phase Project Wide. With one segment, "CN / Project Wide" is the same as "CN / Seg 1". The first start of 2.19.0 folds any you have into the segment: the phase's Project Wide row adds its cost estimate and contingency to the phase's own row, and its risks move to that phase-segment. Their delays then move that phase-segment's dates, so such a project's results can change.

New

• Distributions: Current Year, Year of Expenditure or Completion Date for each phase-segment. Each phase-segment chart has a dropdown to choose what it shows, and Print / Export follows each chart's choice.
• Imports warn of text where a number belongs. Import .pri names each such cell first on its card, and the Excel risk import warns on the row, instead of quietly reading 0.

Improved

• "(assigned risks)": the report-only card and chart for a phase's Project Wide, for example "CN / Project Wide (assigned risks)", now say that they hold only the risks assigned to it, none of the phase's base cost.
• One segment: Import .pri and a backup's import fold a phase's Project Wide into the segment, and deleting a segment down to one is refused while a phase's Project Wide still has a row or a risk.

Released together with CASTR Spreadsheet 75.6.0: both follow the same delay, escalation and convergence rules.

Before you install (users of 2.17 and earlier)

• Your projects move to your Windows user folder (%LOCALAPPDATA%\CASTR). The first start of 2.18.0 copies your old database there once. Extract the new zip over your old CASTR folder, or copy _internal\castr.db from the old folder by hand.
• Reg # can change once on upgrade, for projects with excluded risks. From then on a risk's # and Reg # never change and are never reused.
• The "Use Base Variability" switch is gone: base variability is always on.
• A risk needs a Phase-Segment Impacted to sync. A Risk Register risk with none, or on the Project Wide of a phase with no Base Variability row, is left out of runs and named on the Simulation page before them; it used to run as a whole-project risk.
• A Minor risk on Project Wide (the whole project) now runs as Yes once you agree (Risk Identification and Sync ask); Import .pri changes it with a note.
• Results can move slightly. Dates keep the fraction of a day (a 2-month lag is 60.83 days), so existing projects' dates can move by up to about a day per linked phase-segment, and some escalation months with them. Projects with project-wide risks follow the new delay and escalation rules below.

New

• Import and Export .pri. Open a CASTR Spreadsheet input file as a new project, and save any project as one, for Spreadsheet 75.5.0 and later (Project Wide included) or in the older layout for 75.4 and earlier.
• Tail Driver Analysis, rebuilt on the CASTR Spreadsheet's method. Every project and phase-segment output is analysed, cost outputs list the delay risks behind each tail, and an Explain panel reads the tables and tests the whole distribution for two humps.
• Joint Confidence tab: cost and completion date together, as the spreadsheet's sheet.
• What-If Sensitivity recomputes the run's own iterations with the risks you switch off, for every output, and lists the dependent risks that follow a switched-off parent.
• Inflation Point per phase (Segments & Phases): where in each phase-segment its costs are escalated to.
• Indirect Delay Cost per phase-segment (Review Staging), link lag (Predecessors) and the Fixed value risk type.
• Minor risks: Include = Yes / No / Minor on Risk Identification, and a computed Minor Risks page.
• Market Conditions: Impact BtP / WtP columns, rows built from the impacts and Base Cost Variability, and Min / ML / Max.
• Two bundled projects (the CASTR Example Project Demo and the Rural Highway Widening demo), and the trial holds five projects, the two bundled projects included.

Improved

• Project-wide delays raise each phase-segment's escalation by when they fall in their span, a project-wide risk's own cost escalates over its delayed span, and a phase's Project Wide row no longer sets the completion date. Projects without project-wide risks give the same results as before.
• Convergence: Cost CY, Cost YOE and Duration are checked with 95% intervals for the mean and the budget percentile, and the run stops at the first check within the tolerance from 1,000 iterations on. A run that doesn't converge says how far it got and about how many iterations it needs. The default is 20,000 iterations.
• Total Project Duration runs from the earliest risked start of any row.
• The same engine as the spreadsheet: dependent risks occur per iteration, the planned schedule follows the links, and the Tornado ranks each risk by its realized impact.
• Results: the project-wide risks' output is All Project-Wide, and each phase with a risk on its Project Wide gets a report-only card.
• Risk Register: unchecked risks stay out of the run unless locked, and the Contingency Envelope counts locked rows.
• Desktop launcher: a second copy started while CASTR is running opens the running one, and the window stays open on a start-up error.
• License Agreement 2.1 (free license, trial and refund terms), and the support email and phone in the app.

Fixed

• A simulation no longer fails with "database is locked", and a failed run keeps the engine's own error.
• Project ids never repeat.
• What-If Sensitivity with nothing switched off shows the run's own budget percentile exactly.

A patch release with one fix.

Fixed

• The Renew links in the coverage banners open the current renewal page.