- Ukuthuthukiswa kwamapulatifomu amaningi kuhlanganisa ukwakheka kwe-cloud-native, isitoreji esingandiswa, kanye nezinqubo zesimanje ukuqinisekisa ukubambezeleka okuphansi.
- Amasu okugcina idatha alungiselelwe kahle, imigqa yemisebenzi, kanye nezizindalwazi kusekela ukusebenza njengoba uhlelo lokusebenza lufinyelela izigidi zabasebenzisi.
- Izinhlaka ezifana ne-Flutter noma i-React Native, kanye nenkumbulo enhle, i-GPU, kanye nokuphathwa kwempahla, zivumela okuhlangenwe nakho okuseduze nendawo.
- I-AI iqhuba ukwabiwa kwezinsiza ezibikezelayo, ukuphepha okuthuthukisiwe, kanye nokwenza kube ngokwakho, okuyisihluthulelo sokuncintisana ezindaweni ezidingakalayo zeselula nezewebhu.
Ukuthuthukisa ukusebenza kumapulatifomu amaningi sekungenye yezihloko ezishisa kakhulu ekuthuthukisweni kwesofthiwe yesimanje. Abasebenzisi baqhathanisa konke nokuhlangenwe nakho kwe-AI okusheshayo, balindele izikhathi zokuphendula ezingaphansi kwemizuzwana, futhi bafuna kube bushelelezi ngokulinganayo ocingweni olungabizi kakhulu, i-iPhone Pro, ithebhulethi ye-Android, noma isiphequluli sewebhu sekhompyutha yabo ephathekayo.
Kulesi simo, "ukuyenza isebenze" nje akusanele; ukusebenza kunomthelela oqondile ekugcinweni, ekuguqulweni, kanye nasedumeni lomkhiqizo . Kusukela ezinkampanini ezikhethekile njenge-Q2BSTUDIO noma i-ITERAM kuya kumapulatifomu anekhodi ephansi njenge-Adalo, kanye namaqembu asebenza ne-Flutter, i-React Native, noma izixazululo zewebhu, ukugxila kushintshe kusuka ekuthuthukisweni ngokushesha kuya ekuthuthukisweni ngokushesha nangokuhle , ngezakhiwo ezinwebekayo, ukuqapha okuqhubekayo, kanye nesendlalelo esinamandla kakhulu sobuhlakani bokwenziwa esiza kulo lonke umjikelezo.
Ukusebenza kwezingxenyekazi ezahlukene: umongo, izinselele, kanye nezilinganiso ezibalulekile
Ushintsho lokuqala lwengqondo ukuqonda ukuthi ukwenza ngcono ukusebenza kwe-cross-platform kuyindlela eqhubekayo , hhayi umsebenzi wokugcina "wokulungisa" ngaphambi kokushicilela esitolo. Uhlelo ngalunye lokusebenza, isiphequluli, kanye nohlobo lwedivayisi lubeka imithetho yalo, kodwa umsebenzisi ulindele ukuthi uhlelo lokusebenza luzizwe lufana kuzo zonke izindawo.
Ngokombono wobunjiniyela bokusebenza, izibalo ezifana ne -Time to Interactive (TTI), i-Crash-Free Sessions, kanye ne-Frame Rendering Time zibaluleke kakhulu. Umgomo onengqondo nonamandla amakhulu namuhla ukugcina i-TTI ingaphansi kwama-500 ms kumadivayisi angu-95%, kuyilapho kuhloswe ukuthi kube nama-99,99% amaseshini angenawo ama-crash, ngisho noma ingxube yehadiwe ihlanganisa izinto ezigqokwayo, amafoni asezingeni eliphansi, amathebhulethi, namadeskithophu.
Ku-iOS, izinto eziza kuqala yi- animations ye-60fps, ukuthinta okungenalo i-lag-free, kanye nokusetshenziswa kwememori okuphumelelayo . Ku-Android, ngaphezu kokusebenza okuluhlaza, kugxilwe empilweni yebhethri, ukuhlukahluka kosayizi wesikrini, kanye nokungafani kwehadiwe. Ku-inthanethi, izihluthulelo zezinhlelo zokusebenza zewebhu eziphumelelayo isivinini sokulayisha, ukuhambisana kwesiphequluli esibanzi, kanye nokusebenza okwamukelekayo ngisho nasemanethiwekhi ampofu.
Konke lokhu kuholela embuzweni ongakhululekile kodwa odingekayo: ungafinyelela kanjani ukulingana kokusebenza kuzo zonke izinkundla ngaphandle kwezindleko zokuthuthukisa ezikhuphukayo? Yilapho-ke izinhlaka ze-cross-platform, izakhiwo ze-cloud-native, ukulungiswa kwempahla, amasu okugcina isikhashana, kanye nokusetshenziswa okuhlakaniphile kwe-AI kusebenza khona.
Isu eliyinhloko: ukwakheka, isitoreji, kanye nenethiwekhi
Ukuze uhlelo lokusebenza oluhlanganisa amapulatifomu amaningi lukhule kusukela kumakhulu kuya ezigidini zabasebenzisi, ukwakheka kwedatha kanye nenethiwekhi kubaluleke njengekhodi yesixhumi esibonakalayo . Ukusebenzisa nje ifu akwanele; kufanele uklame ukuthi idatha igcinwa kuphi, ukuthi ihamba kanjani, nokuthi yini egcinwe kunqolobane kudivayisi ngayinye.
Inhlanganisela evamile kakhulu ukusebenzisa isitoreji samafu njengomgogodla , ngedathabheyisi ephethwe, futhi uthembele kwisitoreji sendawo esisebenza kahle kakhulu kudivayisi ngayinye. Lokhu kulinganisela umthamo ongenamkhawulo ngemuva kanye nokufinyelela okusheshayo kwedatha ebalulekile ohlangothini lweklayenti, okugcina uhlelo lokusebenza lusebenziseka ngisho noma uxhumano lulinganiselwe.
Njengoba uhlelo lokusebenza lukhula, ukwakheka kwezinsizakalo ezincane ezinemodeli ethi "idathabheyisi ngensizakalo" kuba yinto ebaluleke kakhulu . Lokhu kukuvumela ukuthi ukhethe ubuchwepheshe bokugcina obuhle kakhulu bemojuli ngayinye: okuhlobene nezinkokhelo, isitoreji sezinto zokuhlaziya, izitolo zenani eliyisihluthulelo sezikhathi, noma ama-cache okucushwa. Kwenza kube lula futhi ukukhulisa ngokukhetha lokho okukudingayo ngempela ngaphandle kokuphahlaza lonke uhlelo.
Ngesikhathi esifanayo, ingqalasizinda enwebekayo (amadiski nezizindalwazi ezivumela ukulungiswa okuhlukile kwe-IOPS, i-throughput, kanye nomthamo) kwenza kube nokwenzeka ukuphendula ekukhuphukeni kwethrafikhi ngaphandle kokunikezwa ngokweqile unyaka wonke. Ukuhlanganisa ukukala okuvundlile, isitoreji esinwebekayo, kanye nokusatshalaliswa kwedatha yendawo kubalulekile ekugcineni ukubambezeleka okuphansi kunoma yisiphi isifunda.
Okokugcina, ungqimba lwenethiwekhi luguqukela kumaphrothokholi esimanje: i-HTTP/3, i-QUIC, kanye ne-WebTransport zivumela ukusakazwa okuqondile, ukulethwa okungaphandle kwe-oda, kanye nokusebenza okuthuthukisiwe ezindaweni ezine-telemetry ebanzi kanye nokuvumelanisa kwesikhathi sangempela. Ukuthuthela okuphakelayo kwedatha ebalulekile noma iziteshi kulezi zobuchwepheshe ngokuvamile kuholela ekunciphiseni okukhulu kokubambezeleka okubonwayo.
Ukuthuthukiswa kwezinsizakusebenza ezihlakaniphile: i-AI, ama-binary, kanye ne-cryptography
Kumadivayisi esimanje, i-CPU ayiseyona yodwa inkinga; ibhethri, izinga lokushisa, inkumbulo, kanye nekhwalithi yenethiwekhi konke kuthinta ulwazi lomsebenzisi . Yilapho i-AI iqala khona ukudlala indima ehamba phambili, ngamamodeli alula asebenza ngqo kudivayisi.
Indlela ethuthukisiwe ukwabiwa kwezinsiza ezibikezelayo kusetshenziswa amamodeli okufunda komshini okudivayisi . Lawa mamodeli athola amasignali afana nohlobo lwedivayisi, inguqulo yesistimu yokusebenza, i-RAM yamahhala, izinga lokushisa lebhethri, iphethini yokusetshenziswa, kanye nekhwalithi yokuxhumeka, futhi alungisa ngokuguquguqukayo amapharamitha ohlelo lokusebenza njengokucindezelwa kwevidiyo, ubunzima bokunikeza, usayizi we-cache ye-UI, kanye nemvamisa yokuvumelanisa.
Ngaphezu kwalokho, inkinga yakudala "yokuqunjelwa" ezinhlakeni ze-cross-platform idinga ukuxazululwa. Ukuhlanganiswa kokuthuthumela kwesihlahla okunamandla kanye nokulayisha imodyuli enamandla kuvumela ama-binary amancane kanye nokuqala okubandayo okuthuthukisiwe, kulayishwa kuphela ukusebenza okuyinhloko ekuqaleni futhi kulethe izici ezisindayo uma kudingeka, lapho umsebenzisi ezidinga ngempela.
Enye inselele ebalulekile yi -cryptographic overhead ezweni le-post-quantum . Ama-algorithms afana ne-CRYSTALS-Kyber andisa usayizi wokhiye kanye nezindleko zokuxhawulana, okubonakalayo kumadivayisi aphakathi nendawo. Amasu ahlanganisiwe ahlanganisa i-cryptography yakudala yesivinini kanye ne-PQC ukuvikela izinto ezibalulekile, kanye nokukhipha intambo ku-key exchange, kusiza ukugcina ukuphepha ngaphandle kokuphula i-TTI (Isikhathi-so-Impact Factor).
Konke lokhu kudinga umbono obanzi wokusebenza: akukhona nje ukushefa ama-milliseconds ngesikhathi sakho, kodwa mayelana nokuklama izinhlelo eziguquguqukayo ezisabela ngaphambi kokuba umsebenzisi aqaphele nokuwohloka.
Ukugcina idatha, imigqa, kanye nezizindalwazi: ukusheshisa ukugeleza kwedatha
Uma isisekelo sesitoreji sesicatshangelwe kahle, isinyathelo esilandelayo senziwa amasu okugcina idatha kanye nezinhlelo zokufaka umugqa , okuyizinto ezisusa ingcindezi kuma-database futhi zigcine ingxenye engaphambili isebenza kahle.
Ukugcinwa kwesikhashana kwememori (isibonelo, ngeRedis) kunikeza izikhathi zokuphendula ezingaphansi kwamasekhondi ayi-millisecond futhi kungaphatha amakhulu ezinkulungwane noma izigidi zezicelo ngomzuzwana. Ngokwendlela engokoqobo, lokhu kusho ukuthi idatha efinyelelwa njalo (amaphrofayili, ukucushwa, uhlu oludumile, njll.) inikezwa ngaphandle kokufinyelela idiski, okunciphisa izindleko nezikhathi zokuphendula.
Kuzinhlelo zokusebenza zeselula nama-PWA, inhlanganisela ye -caching ekude kanye ne-caching yendawo ethile yedivayisi yenza umehluko omkhulu. Izisebenzi Zesevisi kanye ne-Cache API zikuvumela ukuthi ulondoloze i-HTML, i-CSS, kanye ne-JavaScript ukuze ulayishe ngokushesha kanye nemodi eyisisekelo yokungaxhunyiwe ku-inthanethi; kudathabheyisi yeselula, yendawo efana ne-SQLite noma i-Realm isebenza njengesendlalelo sesimo esinama-TTL anolaka aguqula ukuqala okufudumele okuhamba kancane kube ukuqala kabusha ngokushesha.
Izinhlelo zokuhlela (i-RabbitMQ, i-SQS, izixazululo eziphethwe ngamafu, njll.) zisingatha konke okungabalulekile ekusebenzisaneni okusheshayo : ukukhiqiza imibiko emikhulu, izibalo eziyinkimbinkimbi, ukuhlanganiswa komuntu wesithathu, noma ukucubungula amafayela amakhulu. Iphethini evamile ukugcina amafayela kwisitoreji se-blob, ukubuyisela i-URL esayiniwe ngaphambilini ngokuqondisa kabusha, bese uvumela ukulanda kuphathwe ngaphandle kwesendlalelo sohlelo lokusebenza oluyinhloko.
Ngasohlangothini lwesizindalwazi, ukulungiswa okuvamile kusalokhu kubaluleke kakhulu: ukukhomba kahle kanye nokuhlanganisa uxhumano kunganciphisa izikhathi zombuzo ngaphezu kuka-70% futhi kuthuthukise ukubambezeleka kokuthengiselana ngamaphesenti afanayo. Ezinhlelweni ezifana ne-Firebase Realtime Database, izakhiwo zedatha ezisicaba, izilaleli ezibekwe kahle, kanye nemibuzo esekelwe kukhiye esikhundleni semibuzo yensimu yengane kukhulisa kakhulu ukusebenza.
I-Cross-platform front-end: izinhlaka, i-UI kanye nezithombe ezinyakazayo
Umsebenzisi ubona kuphela ungqimba lwesixhumi esibonakalayo, ngakho-ke ukwakheka kwe-front-end okucatshangelwe kahle kubaluleke njenge-back-end enamandla . Yilapho i-React Native, i-Flutter, i-.NET MAUI, i-WebAssembly, kanye ne-Progressive Web Apps, phakathi kwezinye, ngisho nezinhlaka ezifana ne-Lazarus , zisebenza khona.
I-React Native ingafinyelela ukusebenza okuseduze nendawo ngokuphatha kahle ibhuloho le-JavaScript-native , ukunciphisa ukusebenza okungadingekile, nokunikeza ukuphakamisa okunzima kumamojula endabuko. Ukulungiswa kwezingxenye, ukusetshenziswa kwezinhlu ezi-virtualized, kanye nokuphathwa kwesimo okusebenzayo kwenza umehluko omkhulu phakathi kohlelo lokusebenza oluvilaphayo noluguquguqukayo.
I-Flutter ikwiligi ehlukile maqondana nokusebenza kwe-interface ngoba ihlanganisa i-Dart ibe yikhodi yemvelo bese ipenda ngqo ku-canvasI-Dart 3 iletha ukuthuthuka ekuthayipheni, ukungasebenzisi amandla, kanye nokuhlanganiswa kwe-AOT okunciphisa izindleko zokusebenzisa imali eningi futhi kuthuthukise izikhathi zokuqalisa, ngenkathi kusetshenziswa kakhulu i- const kanye nezinto ezingaguquki kunciphisa ukwakhiwa kabusha kwamawijethi okungadingekile.
I-Impeller, injini entsha yokuhumusha ye-Flutter, yenzelwe ukuqinisekisa ukuthi amapayipi okuhumusha aqondakala kalula, ama-stutter ambalwa, kanye nokusetshenziswa okungcono kwe-GPU , ikakhulukazi kumadivayisi anezinsiza zehluzo ezilinganiselwe. Lokhu kuphumela ekugqwayizeni okubushelelezi kanye nokusetshenziswa kwezinsiza okuzinzile.
Ngaphandle kohlaka, kunezimiso ezibanzi: gwema ukudonsela ngokweqile, hlukanisa izinto ezimile zibe yizingqimba, ubeke phambili ukusebenza okubonwayo, futhi uqinisekise ukufinyeleleka . Ama-UI e-Skeleton, ukulayishwa kombhalo okuqhubekayo, ukusebenzisana okuncane nama-animation, kanye nempendulo ye-haptic efihla izikhathi zokulinda ezimfushane kuthuthukisa umuzwa wesivinini ophelele kakhulu kunokunciphisa ukubambezeleka kwekholi yenethiwekhi ngama-ms angu-20.
Imemori, i-GPU, kanye nokuphathwa kwamadivayisi angafani
Kuhlelo lwe-ecosystem lapho ama-iPhone ane-RAM engu-16GB ehambisana khona nama-Android athuthukiswe kakhulu ane-3GB, ukuphathwa kwemithombo yememori kanye nehluzo kuyitshe elibalulekile lokusebenza . Ukungabalwa kahle lapha kuyindlela yokuphahlazeka, ukuminza kwe-GPU, kanye nokuphelelwa yibhethri.
Ukuthuthukisa amapulatifomu ahlukahlukene kuhilela ukuqonda ukuthi uhlaka ngalunye kanye nepulatifomu kuphatha kanjani ukwabiwa, ukuqoqwa kukadoti, kanye nemijikelezo yokuphila . Amasu anjengokuprofayili yenqwaba ngezikhathi ezithile, ukulandelela okujulile kwezinto ezigcinwe ngemva kokugeleza okuthile, kanye nokuphoqelela imikhawulo yesikrini kusiza ukuthola ukuvuza kwememori okungabonakali ngokushesha kodwa okungabonakala njengokumiswa okungahleliwe ngemva kwamasonto ambalwa.
I-GPU ingenye insiza ebalulekile: ukuhambisa izibalo ezithile (ukucubungula izithombe, izihlungi, ukuguqulwa okunzima) ukuze kubalwe ama-shaders kusetshenziswa i-Metal Compute ku-iOS noma i-Vulkan ku-Android kuvumela ukuthi intambo eyinhloko inikezelwe ku-UI, igcine ama-fps angu-60 ngisho nangesikhathi sokusebenza okunamandla.
Ngesikhathi esifanayo, ukwenza ngcono izimpahla ezibonakalayo kunciphisa ukusetshenziswa kwememori futhi kuthuthukisa izikhathi zokulayisha. Ukusebenzisa amafomethi esimanje (i-WebP, i-AVIF), ukukhiqizwa okuzenzakalelayo kobuningi obuningi be-iOS (@1x, @2x, @3x) kanye ne-Android (mdpi, hdpi, xhdpi, njll.), kanye nokusebenzisa ihluzo zevektha lapho kufaneleka khona kusiza ukugcina izinhlelo zokusebenza zilula futhi zibukhali kunoma yisiphi isikrini.
Okokugcina, ukuhlukahluka okukhulu kwehadiwe kumele kucatshangelwe. Amasu okuzivumelanisa nezimo alungisa ikhwalithi yokuthungwa, ukulungiswa kwevidiyo, noma ubunzima bokugqwayiza ngokwekhono ledivayisi avumela okuhlangenwe nakho "okuphezulu" kumadivayisi aphezulu ngaphandle kokulahla abanye abasebenzisi kuhlelo lokusebenza olungabekezeleleki.
Ukusabalala kwangempela: izimo zokusetshenziswa kanye namapulatifomu anekhodi ephansi
Uma kukhulunywa ngokukhuliswa kwezigidi zabasebenzisi, kuyasiza ukubheka labo asebevele belapho. Isibonelo, uSlack, wasuka ku-MySQL ehlukanisiwe waya ku-Vitess , ekwazile ukusingatha izigidi zemibuzo ngomzuzwana ngokubambezeleka kwama-milliseconds kuphela futhi exazulula izinkinga ze-hot shard. Lolu hlobo lokwakha, oluphindaphindwayo ezifundeni eziningi, lubonisa ukuthi i-backend ingakhula ngaphandle kokunqotshwa.
Ngasohlangothini olungaphambili, uSlack usebenze ekugcinweni kwesikhashana kanye nokwenza ngcono ukuqalisa, ekwazi ukunciphisa izikhathi zokuqalisa ngesigamu futhi ethuthukisa kakhulu ukuqalisa okushisayo . Futhi, umyalezo ucacile: akuyona indaba yoshintsho olulodwa oluyisimangaliso, kodwa yezinqumo eziningi ezincane, ezigxile kahle.
I-Airbnb, yona, ibhekane nenselele yokuvumelanisa isimo kumapulatifomu amaningi ngesakhiwo esihlukaniswe ngokwezigaba, lapho iklayenti ngalinye liphatha khona isimo salo sendawo futhi lixhumanisa ngesendlalelo esiphakathi esiqinile. Ukusetshenziswa kwama-ledger ahlelekile kwenza kube lula ukuguquka kwesistimu ngaphandle kokuphazamisa abathengi ngabanye.
Ngakolunye uhlangothi, sinamapulatifomu afana ne-Adalo, avumela amaqembu angenalo ulwazi oluningi lwengqalasizinda ukwakha izinhlelo zokusebenza zewebhu, i-iOS, ne-Android kusuka ku-codebase eyodwa , esebenzisa i-modular backend ekwazi ukucubungula amashumi ezigidi zezicelo zansuku zonke ngokutholakala okuphezulu. Indlela yabo yamathuluzi entengo abikezelwayo kanye ne-AI yokukhiqiza izikrini, ukugeleza, kanye nezakhiwo zedatha isheshisa kakhulu isikhathi sokumaketha.
Into ebalulekile lapha ukuqonda ukuthi, kungakhathaliseki ukuthi uyithuthukisa ngokwezifiso nge-Q2BSTUDIO noma i-ITERAM noma usebenzisa ikhodi ephansi, imikhuba emihle yokusebenza, ukuphepha, kanye nokukhula iyafana : ama-cache, imigqa, izizindalwazi ezibhalwe kahle, ukuqapha okuqhubekayo, kanye nokwakhiwa okuvumela ukukhula ngaphandle kokwenza konke kabusha.
I-AI, ukuphepha kanye nekusasa lokuthuthukiswa kwamapulatifomu ahlukahlukene
Uma sibheka phambili, umugqa phakathi kobunjiniyela bokusebenza kanye nobuhlakani bokwenziwa uya ngokuya ufiphala. I-AI ayikhiqizi nje ikhodi; inquma ukuthi izosetshenziswa kanjani futhi nini izinsiza , ibona amaphethini angajwayelekile, ibikezela ukuphakama komthwalo, futhi iphakamise izinguquko zokucushwa ngesikhathi sangempela.
Amathuluzi okwenza amakhodi afana ne-GitHub Copilot noma abasizi bengxoxo asheshisa intuthuko, kodwa iphuzu elithakazelisayo kakhulu likumamodeli asebenza ngaphakathi kohlelo lokusebenza ngokwalo noma emaphethelweni : abela izinsiza ngokubikezela, enze ngezifiso ulwazi ngokuya ngokuziphatha komsebenzisi, futhi alungise ikhwalithi yokuqukethwe ngokuguquguqukayo ngokuya ngenethiwekhi noma idivayisi.
Ekuphepheni, inhlanganisela ye- Zero Trust, ukuqinisekiswa okungenaphasiwedi (ama-Passkey, ama-biometric), kanye nezinhlelo zokuthola ama-anomaly ezisebenzisa i-AI kuqinisa izinhlelo zokusebenza ezibalulekile njengebhange, ukunakekelwa kwempilo, kanye nokukhiqiza. Isihluthulelo ukuhlanganisa ukuphepha kuyo yonke impilo kusetshenziswa imikhuba ye-DevSecOps, ukwenza ngokuzenzakalelayo ukuhlolwa, nokuqinisekisa ukuthi intuthuko esheshayo ayibi yindawo yokuzalela ubuthakathaka.
Ngakolunye uhlangothi, ukuxhumana ne -IoT kanye ne-edge computing kuphoqa ukuthi kwenziwe ngcono kakhulu: ukucubungula idatha eduze komthombo, ukuhlunga idatha engabalulekile ngaphambi kokuthumela noma yini efwini, nokugcina ulwazi olungenamthungo ngisho noma kunenethiwekhi engalingani. Imikhakha efana ne-logistics, i-telemedicine, kanye ne-home automation kakade incike kulezi zinhlobo zezakhiwo.
Indima yonjiniyela uqobo lwayo iyashintsha futhi: amaphrofayili onjiniyela wezinhlelo eziningi, ochwepheshe bokusebenza kwe-IT, ochwepheshe bokuhlola ukuzenzekela, kanye nabakhi bezakhiwo abasebenzisa amafu bayadingeka kakhulu. Ukuqeqeshwa kwamanje kuhlanganisa ukuthuthukiswa kwemvelo kanye ne-hybrid, ukuphathwa kwedatha, ukuphepha, ukuzenzekela, kanye, nokwenza ngcono ukusebenza kusukela ekuqaleni kuze kube sekupheleni.
Lonke leli phazili—ukwakheka kwe-cloud-native, ukuphathwa kwedatha, i-UI yesikhulumi esibanzi, i-AI, ukuphepha, kanye nokuhlola—kuhlangana emgomweni owodwa: ukwakha izinhlelo zokusebenza ezizizwa zishesha, zithembekile, futhi zihambisana kunoma iyiphi idivayisi , ngenkathi kugcinwa izindleko zilawulwa futhi kuvumela ukuphindaphinda okusheshayo. Ukuklama kusukela phansi ngokusebenza, ukukhulisa, kanye nolwazi lomsebenzisi engqondweni yikho okuhlukanisa izinhlelo zokusebenza "ezikhona" nje emikhiqizweni eba yizilinganiso emakethe yazo.
