Must-Match Definitions
Sleeping Beauty · must-match requirements · from 90,646 competitor reviews · 9 September 2026

What “must match” actually means, in users’ words

The review synthesis listed six things Sleeping Beauty must match before it can compete on anything else. This page says what each one actually means, in the words of the people who use the competitor apps, and turns it into behaviour we can test.

6 requirements23 behaviours86 verbatim quotes, each verified against the archive27 acceptance criteria

Overview

The six requirements are the rows of the “must match” column in the review synthesis. For each one this page gives a definition written from the reviews, the three or four behaviours it decomposes into, the failure mode users report when a competitor gets it wrong, verbatim quotes with links, and the acceptance criteria that would let us say Sleeping Beauty matches it. A short note says where the build stands today.

Reading order matters less than the pattern across all six: users reward apps that do a small number of things with total reliability and punish the same apps within days when reliability slips. The requirements are ranked by how much damage a failure does, not by how hard they are to build.

  1. The alarm rings, reschedules and cancels correctly. Every scheduled ring happens, no unscheduled ring ever happens, and stopping it always works.
  2. The physical mission works under real morning conditions. Recognises the real thing on the first try in a real bathroom, rejects shortcuts, and has a deliberate way out when the thing is not there.
  3. Blocks start and end automatically, and essential apps stay available. The block is on when it should be, off when it should be, holds when pushed, and never takes the apps people need.
  4. Users choose useful friction and sound behaviour. Loud enough for the sleeper, quiet enough for the household, and a dial for how hard the morning is.
  5. Clear commercial terms and uninterrupted dismissal. Nothing stands between the ringing alarm and the mission, the price is said before the effort, and billing behaves.
  6. The schedule fits the user’s week. Different days, different times; skip one morning without disarming; real work hours win over the app’s ideal.
1

The alarm rings, reschedules and cancels correctly

Every scheduled ring happens, no unscheduled ring ever happens, and stopping it always works.

Users describe this as one job. An alarm that rings 29 mornings out of 30 has failed, because the 30th was the exam. The same people are equally angry when the app rings after they turned it off, or keeps re-ringing after a completed mission, because that means the app, not the user, is in control. The third face of the same requirement is that when something is genuinely broken, a camera, a sensor, the user’s own memory, the alarm still stops in a bounded time.

Share of substantive reviews touching thisAlarmyWaykSuperAlarmOpalBrainrotPillowRISE
keyword match, ≥60 chars10% iOS6% Android3% iOS2% Android12% iOS2% Android2% iOS2% Android0% iOS1% Android1% iOS1% Android2% iOS1% Android

1.1It fires every time, at the set time, without the app open or the phone unlocked.

Failure mode. A silent morning after months of reliability is the single most damaging failure in the corpus; it converts long-term fans into one-star reviewers.

“recently stopped going off completely in the mornings, causing me to miss university exams”

“an alarm app's main purpose is reliability. If the alarm doesn't ring, users can miss important tasks, classes, meetings, or other responsibilities”

“the alarm won't go off if you're not using the app or if your phone is off”

“App does not ring. Alarm doesn’t go off and I purchased a year subscription”

1.2Off means off. A cancelled or disabled alarm never rings; a dismissed one never rings again.

Failure mode. Re-ringing after a disable or a completed mission is read as the app being untrustworthy, and the only fix users find is deleting it.

“I turned my alarm off this morning before it went off. My lock screen still had the “alarm soon” pop up but it was for the wrong time. I turned it off anyway. 2 minutes later the alarm started blaring”

“Deleting the alarm doesn’t stop the alarm from ringing again”

“I disabled the alarm the night before but it still rang the next day. But when I tried to complete the mission, it would time out”

“if I dismiss my alarm but leave the app open in the background, then go back to it, the alarm goes off again. very annoying when I've already left the house”

1.3It survives the app being killed, the phone restarting, silent mode, Focus and battery saver, and it says what it needs.

Failure mode. “I configured every setting it asked for and it still didn’t ring” is the pattern; blaming the OS does not restore trust.

“Every time I wake up and check my phone I get the "alarm didn't go off?" notice even though I've configured all the settings it asks”

“the app specifically crashes during the night while running in the background”

“I consistently use the do not disturb sleep function on my watch and pillow then just works in the background. I don’t need to worry about anything”

1.4It stops when it should: ringing is bounded, and there is a working stop when the device or the situation is broken.

Failure mode. An alarm that cannot be silenced for hours is described as the phone being held hostage; users will factory reset or drive to a repair shop.

“it didn't stop for 3 hours until I tried to get professional help”

“There is no emergency exit, no sensor check, and no safe fallback. The only guaranteed escape is a factory reset”

“I forgot I set an alarm and left early, had to pull over and find something to take a picture of to stop it”

Acceptance criteria for Sleeping Beauty

  1. Seven consecutive mornings on a physical iPhone with zero missed rings, including nights with the app force-quit, the phone restarted, Silent on, and a Sleep Focus active.
  2. Disabling the routine before bed produces no ring, no escalation, and no stale shield the next morning.
  3. Tapping Stop never produces a second ring unless the still-in-bed escalation is armed, and that escalation is announced on screen the moment it is armed.
  4. A time-zone change and a DST change each move the alarm correctly, verified by a schedule read-back.
  5. Every ring stops within a bounded time by design, and the fallback path is reachable without typing or a camera.

Where the build stands, 9 Sep 2026

The daily alarm and the still-in-bed escalation are both AlarmKit system alarms, so they fire with the app killed. Cancellation clears both. What we have not done is the seven-night device run, or the Focus, time-zone and restart cases.

2

The physical mission works under real morning conditions

Recognises the real thing on the first try in a real bathroom, rejects shortcuts, and has a deliberate way out when the thing is not there.

The mission is why people choose these apps over the built-in clock, and it is praised when it does one thing: it gets them out of bed and keeps them out. The praise is conditional on three things. Recognition has to work in the actual room, in the actual light, while an alarm is blaring and the person is half asleep. The mission has to be un-cheatable enough to stay meaningful. And when the object is genuinely unavailable, because they travelled, left early, or the camera broke, there has to be a way out that is hard but possible. Apps that get the first right and the third wrong are described as holding people hostage.

Share of substantive reviews touching thisAlarmyWaykSuperAlarmOpalBrainrotPillowRISE
keyword match, ≥60 chars19% iOS7% Android16% iOS9% Android14% iOS9% Android0% iOS1% Android1% iOS4% Android12% iOS4% Android5% iOS4% Android

2.1It recognises the real object, first try, in real light, while the alarm is ringing.

Failure mode. A failed scan under a blaring alarm is the most stressful moment in the product; users report panic, restarts and deletion.

“It doesn't recognize my squats. I tried different angles but it always says try again”

“I have set push-ups as a task, but it does not recognize it on the first attempt”

“The photo scanner half the time doesnt work. And the written sentence one NEVER works”

“trying to do the 20 steps and wouldnt read so my alarm was going off rampant”

2.2It rejects shortcuts, so the mission stays meaningful.

Failure mode. Once a loophole is found it is used every morning, and the app’s value collapses; users report this in the same breath as praising the concept.

“once I discovered a loophole, things changed I couldn’t use it effectively anymore. If the alarm rings and you still want to sleep, you can simply keep pressing the middle button of your phone”

“if I click the add and then exit it, it will take about 30 seconds to reload the alarm which is the exact amount of time it takes to delete the app”

“I love how you can remove the alternative typing option when the challenge is scanning”

2.3When the object is not there, there is a deliberate way out: hard, but possible without a camera or a keyboard.

Failure mode. Travel, leaving early and driving are the recurring cases. An escape that requires typing three lines perfectly is not an escape.

“there needs to be a way to override it. Some complicated way, but there needs to be a way”

“you have to type 3-4 lines perfectly without missing a single space which seems almost impossible in real emergency situations”

“I can't type while driving, so I have to sit and listen to an alarm go off for however long my drive is”

“if I’m already awake and I’m in my car driving obviously I’m not in my house and I can’t take any pictures of my refrigerator”

2.4The mission produces the outcome: out of bed, and not straight back in.

Failure mode. Users are honest that a completed scan is not the same as staying up; the ones who stay up credit a mission that moves them somewhere else.

“I literally have to make my bed then take a photo of my bed made to turn the alarm off. Now I’m forced to stay out of bed”

“I have to go scan my toothbrush before it gets extremely loud again lol I’m up in record time”

“For my earliest mornings, I have it set where I need to scan a product in my bathroom”

“even after scanning the item I sometimes again go to sleep but that's on me”

Acceptance criteria for Sleeping Beauty

  1. Setup rehearses the user’s actual toothbrush in their actual bathroom and will not enable the routine until a scan has passed there.
  2. Recognition passes within ten seconds in a dim bathroom with the light off and a phone torch on, across at least three toothbrush types, and rejects a hand, a pen and a photo of a toothbrush.
  3. The gate cannot be completed by leaving the app, locking the phone, or killing the app; the escalation alarm covers the killed-app case.
  4. A fallback exists that needs no camera and no typing, takes deliberate effort, is logged as “fallback” rather than “scan passed”, and is visible in the morning summary.
  5. The still-in-bed escalation is the answer to “back to bed after the scan”: it is armed on Stop and stood down only by steps or the scan.

Where the build stands, 9 Sep 2026

The toothbrush detector requires three consecutive matches above a confidence threshold, and the still-in-bed escalation now covers the return-to-bed case. The alarm gate can still be closed without a scan through the exit button, which the review report flagged; that path needs to become a deliberate, labelled fallback rather than a silent completion. No bathroom-condition testing has been done yet.

3

Blocks start and end automatically, and essential apps stay available

The block is on when it should be, off when it should be, holds when pushed, and never takes the apps people need.

For blocker users the schedule is the product. A block that only activates when they open the app, or stops at midnight, or defaults to AM, is the same as no block. Once it is on, it has to hold: a second tap should not open the app, and an overlay is not a block. At the same time the block must never swallow the apps someone needs for work, navigation or music, and an exception they set must stay set. The last thing users ask for is a middle ground between a trivial bypass and a day-long lockout: a break that costs something.

Share of substantive reviews touching thisAlarmyWaykSuperAlarmOpalBrainrotPillowRISE
keyword match, ≥60 chars6% iOS3% Android6% iOS2% Android3% iOS0% Android44% iOS39% Android23% iOS3% Android2% iOS3% Android4% iOS3% Android

3.1It starts and ends on schedule without opening the app, across midnight and AM/PM.

Failure mode. “Only works when the app is open” and “stops at midnight” are the two most common blocker complaints in the corpus.

“app blocking doesn’t always work. Sometimes, it seems to overlook the usage until i am already over my daily threshold. Once I open the app, only then it recognizes that the app should have been blocked”

“the bed time timer stops at midnight so can't handle across the calendar date”

“there's no option for AM or PM, so it always defaults to AM”

“The apps that should be blocked at the same times everyday are only sometimes blocking”

3.2It holds the line. A block that opens on the second tap is not a block.

Failure mode. Users who need a blocker say, in their own words, that they cannot be trusted with a soft one; the praise goes to the app that “holds the line”.

“It actually holds the line when you try to get around it, which is what made everything else I tried fail”

“It stops me from opening an app once and if I click it a second time it just opens”

“please fix this by adding a feature to kick out of the app when app is blocked and not just a screen block on top”

“Unless you have a lot of self control (which you probably don't if you're getting this app to help), it's way too easy to bypass the settings that block the apps”

3.3Essential apps stay reachable, and an exception stays set once it is set.

Failure mode. Losing an “always allow” entry, or having to re-arm the block every night, is reported as the app deciding for the user.

“there are a couple of apps that I need for work, and despite repeatedly adding them to the "always allow" category, Opal would still decide that they were distracting me”

“whenever I used chatgpt and forest for studies it would count 1 hr in them making me not use youtube”

“I didn’t realize I had to reset it every night. I wish that feature saved for every night, and it was a little difficult to figure out what meant blocked vs allowed”

“I couldn't access any if my apps. This made me realise how dangerous it is to give an app all access to my phone”

3.4There is a middle ground: a break that costs something, not an ad, and not a day-long pause.

Failure mode. Both extremes get one-star reviews: five-second waits are too weak, and a whole day as the minimum break is too blunt.

“I would strictly prefer a timer that increase from time to time for taking a break”

“I just need to open something for like five minutes, why would you make an entire day the minimum break time?”

“Allows you to take break and has option to make it even harder for the next one, so you just don’t get addicted to breaks”

“I like that I can strict block apps where I can't access them at all, can't even opt for a little break, until the end of the strict block period”

“I can’t figure out how to pause a rule for things like vacation or holidays”

Acceptance criteria for Sleeping Beauty

  1. With the app force-quit at 9 p.m., the shield is up at the scheduled start and down at the scheduled end, on a schedule that crosses midnight, for seven consecutive nights.
  2. A blocked app stays blocked on the second, third and tenth open, and after a phone restart.
  3. The allowed list survives app updates, reinstalls and routine off/on cycles, and the dashboard shows tonight’s start, end, allowed apps and armed state before bed.
  4. There is exactly one documented break rule, it is shown on the shield, and it never involves an ad or a full-day pause.
  5. Turning the routine off leaves no shield behind, verified by opening a blocked app immediately afterwards.

Where the build stands, 9 Sep 2026

The bedtime shield is a DeviceActivity schedule from seven hours before the alarm until the alarm, with an allowed-apps list, plus the optional post-wake block and the custom shield with a Dismiss button. The schedule is system-owned, so it should hold with the app killed, but the cross-midnight and restart cases are unverified on device, and there is no break rule at all yet: the shield is absolute until the scheduled end.

4

Users choose useful friction and sound behaviour

Loud enough for the sleeper, quiet enough for the household, and a dial for how hard the morning is.

Two people in the same bed want opposite things from the same alarm, and reviewers say so. Heavy sleepers praise volume and vibration and ask for a follow-up awake check; light sleepers and partners ask for merciful sounds, silence during the mission, and a ramp that really starts from silent. What both groups agree on is that these are choices the user makes while awake, that the choice must stay made, and that the dismiss control must not be something a sleeping hand can hit.

Share of substantive reviews touching thisAlarmyWaykSuperAlarmOpalBrainrotPillowRISE
keyword match, ≥60 chars33% iOS11% Android16% iOS2% Android17% iOS11% Android8% iOS3% Android0% iOS2% Android9% iOS2% Android5% iOS2% Android

4.1Loud enough for a heavy sleeper, and the volume the user set is the volume that plays.

Failure mode. The single most upvoted review in the whole corpus is about the alarm becoming quiet after an update.

“after the recent update the alarm volume is extremely low, even when my phone’s alarm/media volume is set to maximum. This defeats the purpose of an alarm app”

“with a combination of vibrations and noise, I'm up like a bat, and the math forces me to open my eyes and use my brain”

“sometimes the volume doesn't even match the volume I set it at”

4.2Not more than the room can bear: merciful sounds, and silence during the mission, and the setting stays set.

Failure mode. A mission-silence toggle that resets itself overnight wakes the rest of the house every morning; a paying customer threatened to leave over exactly this.

“Every night before I go to bed, I switch off the button in settings, “keep alarm during mission.” Yet every morning it’s switched back on”

“too loud, scary and just bad sounds with no easy accesses to stop the app, i either oversleep (i became even a heavier sleeper...) or i cant just peacefully fall asleep any longer... Can you please add more merciful options”

“if i ever wake my ehole house again cause i cant turn off the alarm someone gonna find out the hard way”

4.3Friction is a dial, not a switch: snooze limits, difficulty, a wake check, and a dismiss that cannot be hit by accident.

Failure mode. Users who want it strict ask for no-snooze and a follow-up check; the same users say an easy-to-tap dismiss defeats everything.

“Options to check if you're awake later, add a task you have to complete to clear alarms, customize snooze, etc.”

“I recommend turning on no snooze so it just keeps going on until you decide to complete your mission”

“Dismiss is way too easy to accidentally hit. It should be swipe or press and hold”

“Each time I set up a new alarm I change the alarm tone, disable "gradually increase volume" feature, set up the same missions, and turn on Wake Up Check feature”

4.4Sound choice, including their own music, random tones so the brain does not habituate, and a ramp that starts from silent.

Failure mode. “Gentle wake-up” that starts loud is called out as not gentle; custom sounds are the most common feature request for Wayk.

“Please add a feature to make custom alarm sounds”

“the 'gentle wake-up' setting needs changing as it should start from silent and then gradually increase to full volume”

“the random noise alarm is great so you don’t get used to the ringtone”

Acceptance criteria for Sleeping Beauty

  1. At least three bundled sounds plus the user’s own file, a ramp that starts from silent, and a preview button that plays the real sound at the real volume before bed.
  2. A “quiet during the scan” option that persists across nights and app updates, verified by reading it back the next morning.
  3. The still-in-bed escalation delay (Off, 3, 5, 10 minutes) is presented as the friction dial, alongside a snooze policy that defaults to none.
  4. The Stop control on the gate requires a deliberate gesture, and the escalation is armed the moment it is used.

Where the build stands, 9 Sep 2026

One bundled sound, Morning Joy, for both the alarm and the escalation; AlarmKit controls volume; the gate loops the tone until the scan passes. The escalation delay is the only friction setting. There is no sound choice, no ramp, no quiet-during-scan option and no snooze policy.

5

Clear commercial terms and uninterrupted dismissal

Nothing stands between the ringing alarm and the mission, the price is said before the effort, and billing behaves.

This is the category’s most frequent complaint by volume, and it has two halves. The first is the morning: an ad, a rating prompt, a subscription screen or a motivational dashboard between the ring and the mission is experienced as an attack on a half-asleep person, and it is also the loophole cheaters use. The second is trust around money: users are not against paying, and say so, but they want the price before the questionnaire, a trial that does not charge on day one, a cancel button that works, and increasingly a one-time price for a clock. The apps that are praised for their free tier are praised as much for honesty as for generosity.

Share of substantive reviews touching thisAlarmyWaykSuperAlarmOpalBrainrotPillowRISE
keyword match, ≥60 chars18% iOS13% Android35% iOS38% Android29% iOS17% Android11% iOS16% Android55% iOS39% Android32% iOS39% Android52% iOS39% Android

5.1Nothing between the ring and the mission, and nothing after it: no ad, no upsell, no rating prompt, no dashboard.

Failure mode. Ads in the dismissal path are both hated and exploited; they let a sleeping user out of the mission.

“It trying to show me ads when I'm barely awake, before I can turn my alarm off, is infuriating”

“Ads are awful, they pop up when you press the screen trying to dismiss alarm and banners too- I've opened multiple while half asleep”

“clicking the ad during alarm effectively gets you out of the alarm screen, bypassing all countermeasure”

“I just want to close the Alarm and continue my work. i don't want the screen which comes after i dismiss an alarm. i don't want any weather information or motivation quotes”

5.2Say the price before the setup effort, and let people try the real thing before they pay.

Failure mode. A paywall after the questionnaire, or after granting permissions, is called deceptive even by people willing to pay.

“as I was selecting times, ring tones and such I was then hit with a pay wall/subscription, before I could try the actual app, it just left a sour taste in my mouth”

“Only tells you that you need to pay after you give them access to all your data. If I need to pay to use the app at least tell me right away”

“I could’ve reduced my screen time if you told me it was paid only from the beginning”

“you expect me to download this app and immediately rate it before I’ve even used it???”

5.3Trials and billing behave: no charge after cancelling, and cancel is findable.

Failure mode. Being charged after cancelling a trial produces the word “scam” more than any other event in the corpus; a fast refund converts it back to five stars.

“I was told my card would not be charged until the free trial was up and despite this, my card was charged instantly”

“charged me after i cancelled free trial”

“My biggest gripe is the cancellation process. They make it SO difficult”

“I was still charged despite cancelling, but I reached out to support and they handled it immediately”

5.4People pay for a product that delivers, and they keep asking for a one-time price for a clock.

Failure mode. A subscription for an alarm is resented on principle; ads shown to someone who already paid is resented more.

“You have to pay a fee, but they make it very affordable at only $30/year”

“You're also no longer able to pay a one time payment to remove them, you need to hop on a subscription service. I'm not paying a monthly subscription for my alarm clock”

“please give me a one time payment option”

“I'm getting ads for some features, which is unacceptable for a product I already bought”

Acceptance criteria for Sleeping Beauty

  1. A written rule in the codebase: no modal, sheet, prompt or navigation can appear between the alarm firing and the scan passing, and the gate view has no entry point for any of them.
  2. After the scan, the app returns to a calm summary with no quote, weather, streak or upsell; the user can leave within one tap.
  3. If a price exists, it is shown on the dashboard before the routine can be enabled, and the trial, if any, does not require a card.
  4. Store review prompts, if ever used, fire only after a completed morning and never within the first week.

Where the build stands, 9 Sep 2026

There is no monetisation, so the morning path is clean by default. The requirement is to keep it that way when a price is added, and to decide the pricing shape with the one-time-purchase demand in mind.

6

The schedule fits the user’s week

Different days, different times; skip one morning without disarming; real work hours win over the app’s ideal.

One daily time is not how anyone’s week works. Reviewers ask for weekday and weekend times, alternate weeks, specific dates, and a way to skip tomorrow without turning the alarm off and forgetting to turn it back on. The sleep apps get the harshest version of this: a nurse on three twelve-hour shifts, a parent of a toddler, a service worker, and someone who must be out of the door at the exact time the app calls ideal all say the same thing, that the app’s schedule has to bend to their obligations, not the other way round.

Share of substantive reviews touching thisAlarmyWaykSuperAlarmOpalBrainrotPillowRISE
keyword match, ≥60 chars4% iOS2% Android2% iOS1% Android2% iOS1% Android3% iOS1% Android1% iOS8% Android4% iOS8% Android6% iOS8% Android

6.1Different times on different days, with weekends off, not one daily time.

Failure mode. Long-term users leave over a missing bi-weekly or per-day option even when everything else works.

“the only reason I'm abandoning it is it still can't do a bi-weekly alarm, in other words you can have it ring at a different time every other Thursday”

“I also don't like that it couples your weekday awake time and weekend awake time”

“I use it to get up for work, and to not sleep the day away on the weekends!”

“there needs to be a way to set an alarm for a specific future date. Right now you can only set an alarm to go off within the next 7 days”

6.2Skip one morning without disarming the routine, and pause for a holiday with an obvious way back.

Failure mode. Turning the alarm off for a day off and forgetting to turn it back on is the failure users are trying to avoid.

“I would like to be able to skip the alarm for one day. Instead of turning it off completely. So when I have a day off I can't forget to turn it back on”

“I accidentally forgot to turn it off in a day where I dont have school I end up crashing out and crying while solving math problems in pure rage”

“I can’t figure out how to pause a rule for things like vacation or holidays”

6.3Real obligations win: fixed work times, shifts, parents and time zones are inputs, not exceptions.

Failure mode. An app that tells a night-shift nurse to sleep during her shift, or cannot enter a daytime sleep window at all, is judged useless regardless of its science.

“it says my ideal wake time is 7am however, I have to be out the door for work at 7am and there's no way to adjust it”

“It'll work for 9-5ers weekends off but I'm a nightshift nurse 3 12hour shifts a week some weekends off”

“I work night shift and sleep during the day. I cannot enter my sleep schedule into the app because the time isn't even available”

“It also doesn't know my timezone”

Acceptance criteria for Sleeping Beauty

  1. Weekday and weekend wake times as a minimum, with per-day times as the target, and the bedtime block computed from each day’s own alarm.
  2. A “skip tomorrow” action on the dashboard that disarms one night only and shows the next armed night.
  3. The block lead time is a visible setting labelled as app blocking, not a sleep prescription, and can be shortened for late work nights.
  4. A daytime wake time (for night workers) schedules correctly, and a time-zone change is reflected in both the alarm and the shield window.

Where the build stands, 9 Sep 2026

One wake time, repeated every day of the week, with a fixed seven-hour block before it. There is no weekend setting, no skip, no lead-time control and no time-zone handling beyond what the system alarm does on its own. This is the least-covered of the six.

Method and caveats

Source: the 90,646-review archive collected on 7 September 2026 (2,115 iOS, 88,531 Android; Alarmy dominates the Android half). Two keyword scans were run over every review of at least 60 characters: a broad pass per requirement, and a targeted pass for each behaviour. The footprint table under each requirement shows the share of an app’s substantive reviews that matched the broad pass. It is a keyword match, so it counts praise and complaint alike, and it says nothing about prevalence in the user base.

Quotes were chosen from matches since September 2025, favouring reviews other users marked helpful, and balanced across apps and star levels. Each is a verbatim substring; the page builder resolves the review by ID and refuses to render a quote it cannot find in the archived text. Links open the public store feed, which can change; the archived text is in the research folder. Android reports are not evidence of iOS behaviour, and nothing here was reproduced on a device.

The acceptance criteria and the “where the build stands” notes are ours, written against the code as of 9 September 2026, after the still-in-bed escalation, the post-wake block and the custom shield were added.