FreeBSD based Japanese Enterprise System and Tukubai Method
This is a modal window.
Das Video konnte nicht geladen werden, da entweder ein Server- oder Netzwerkfehler auftrat oder das Format nicht unterstützt wird.
Formale Metadaten
| Titel |
| |
| Untertitel |
| |
| Serientitel | ||
| Anzahl der Teile | 26 | |
| Autor | ||
| Lizenz | CC-Namensnennung 3.0 Unported: Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen. | |
| Identifikatoren | 10.5446/19169 (DOI) | |
| Herausgeber | ||
| Erscheinungsjahr | ||
| Sprache |
Inhaltliche Metadaten
| Fachgebiet | ||
| Genre | ||
| Abstract |
|
1
2
7
8
10
11
13
14
15
18
19
20
21
00:00
FreewareEDV-BeratungUnternehmensarchitekturSystemprogrammierungWebDAVSoftwarewartungW3C-StandardDatenverwaltungServerSchreiben <Datenverarbeitung>SystementwurfVorzeichen <Mathematik>Dienst <Informatik>ProgrammverifikationSystemplattformDigitalsignalElektronischer ProgrammführerGebäude <Mathematik>ExpertensystemDreiEreignishorizontInformationSoftwareentwicklerFreewareUnternehmensarchitekturEDV-BeratungServerSystemprogrammierungBenutzerbeteiligungDatenverwaltungIn-System-ProgrammierungDateiverwaltungSupercomputerCASE <Informatik>Vollständiger VerbandWeb ServicesGebäude <Mathematik>SoftwarewartungAutomatische HandlungsplanungRelativitätstheorieVertauschungsrelationFormation <Mathematik>HypermediaElektronisches BuchKonstruktor <Informatik>Elektronische PublikationSoftwareentwicklerSystemplattformEindeutigkeitProdukt <Mathematik>Familie <Mathematik>Prozess <Informatik>App <Programm>VerkehrsinformationEntwurfsautomationNeue MedienQuellcodeExpertensystemDatenmissbrauchE-MailEntscheidungstheorieSprachsyntheseEreignishorizontBitTelekommunikationt-TestNichtlinearer OperatorStichprobenumfangWeb SiteEnergiedichteMultiplikationsoperatorAdditionGüte der AnpassungBildschirmmaskeInformationsverarbeitungKoordinatenFrequenzTermExogene VariableSoftwareentwicklungZahlenbereichRechenwerkÜberlagerung <Mathematik>Gemeinsamer SpeicherZellularer AutomatData Mining
06:34
FreewareTechnische OptikCAN-BusFreewareWeb SiteEntscheidungstheorieGüte der AnpassungGesetz <Physik>PunktWeb ServicesMomentenproblemGefrierenMultiplikationsoperatorVorzeichen <Mathematik>SystemprogrammierungProgramm/QuellcodeXMLUML
07:31
UnternehmensarchitekturDatenverwaltungSystemprogrammierungSoftwareSoftwarewartungOrakel <Informatik>Personal Area NetworkDatenstrukturSchreiben <Datenverarbeitung>Codierung <Programmierung>DruckspannungQuellcodeSoftwareentwicklungNabel <Mathematik>ProgrammMetropolitan area networkMIDI <Musikelektronik>COMSystemprogrammierungSoftwarewartungZahlenbereichMomentenproblemSoftwareDistributionenraumSoftwareentwicklungHomepageBildgebendes VerfahrenInhalt <Mathematik>StrömungsrichtungPunktAuswahlaxiomResultanteInterleavingWeb SiteSoftwareentwicklerDatenverwaltungQuellcodeQuick-SortMultiplikationsoperatorStichprobenumfangDatenstrukturMeterEnergiedichteTeilmengeUnternehmensarchitekturWort <Informatik>Inverser LimesKonditionszahlSchlussregelForcingLoginModulare ProgrammierungRelativitätstheorieVersionsverwaltungCodeProjektive EbeneEindeutigkeit
11:27
FIESTA <Programm>DreiSinusfunktionUnternehmensarchitekturSystemprogrammierungSkriptspracheNabel <Mathematik>Notepad-ComputerSoftwareentwicklerRechenwerkInformationSoftwareentwicklungEnergiedichteSelbst organisierendes SystemCoprozessorSoftwareProgrammProzess <Informatik>Schreib-Lese-KopfWiederkehrender ZustandComputerunterstützte ÜbersetzungSteuerwerkSystemplattformSpezielle unitäre GruppeQuick-SortOperations ResearchMatchingElektronische PublikationFreewareCOTSSchlüsselverwaltungCOMSpeicherabzugBefehlsprozessorSkalierbarkeitMultiplikationLesen <Datenverarbeitung>DatensatzSummierbarkeitZahlenbereichTabelleZählenCAN-BusTemplateHorizontaleOffene MengeVersionsverwaltungAlgebraisches ModellKernel <Informatik>SystemaufrufVerzeichnisdienstUnternehmensarchitekturSoftwarewartungMereologieTabelleProzess <Informatik>SystemprogrammierungNabel <Mathematik>SoftwareentwicklerOffene MengeDatenverwaltungSpeicherabzugEindeutigkeitSkriptspracheRegelkreisVersionsverwaltungCodeDatenbankKommunikationssystemCASE <Informatik>ImplementierungKartesische KoordinatenGrundraumTelekommunikationUnternehmensmodellDateiformatSystemplattformSchlüsselverwaltungElektronische PublikationEchtzeitsystemProjektive EbeneGemeinsamer SpeicherInformationsspeicherungGrenzschichtablösungSchreib-Lese-KopfArchitektur <Informatik>VerzeichnisdienstEnergiedichteKernel <Informatik>Gewicht <Ausgleichsrechnung>Wald <Graphentheorie>Quelle <Physik>Objekt <Kategorie>InstantiierungRootkitSoftwareentwicklungNotepad-ComputerFormale SemantikMultiplikationsoperatorDelisches ProblemSprachsyntheseDifferenteManufacturing Execution SystemSoundverarbeitungGesetz <Physik>GraphRechenwerkGeradeBildschirmmaskeZellularer AutomatTypentheorieRechter WinkelOffice-PaketDatenstrukturMinkowski-MetrikKugelkappePhysikalischer EffektJSON
19:45
Lipschitz-StetigkeitSchlüsselverwaltungDatenfeldComputerunterstützte ÜbersetzungE-MailBinärdatenCodeStichprobeFunktion <Mathematik>Inklusion <Mathematik>System-on-ChipQuick-SortSummierbarkeitLESWiederkehrender ZustandSimulationMAPTotal <Mathematik>Spezielle unitäre GruppeElektronischer FingerabdruckNeuronales NetzVerzweigendes ProgrammBildverarbeitungStammdatenCASE <Informatik>Ext-FunktorURNMetropolitan area networkQuantorMereologiePropagatorDifferenteSkriptspracheLesen <Datenverarbeitung>DatenbankUnternehmensarchitekturSystemprogrammierungBenutzerbeteiligungProgramm/QuellcodeXML
20:44
Selbst organisierendes SystemSoftwareStrategisches SpielGrenzschichtablösungElektronische PublikationNabel <Mathematik>SkriptspracheEin-AusgabeSoftwareentwicklerProgrammfehlerHardwareBenutzerfreundlichkeitTaskZurücksetzung <Transaktion>DistributionenraumKartesische KoordinatenAnalysisRechter WinkelDistributionenraumPhysikalische TheoriePunktRechenwerkSchlüsselverwaltungSoftwareGrenzschichtablösungBildverstehenGemeinsamer SpeicherGesetz <Physik>Elektronische PublikationSoftwareentwicklungSchreiben <Datenverarbeitung>HardwareSoftwareentwicklerEindeutigkeitFunktion <Mathematik>Bildgebendes VerfahrenJSON
23:10
MathematikFlächeninhaltTemperaturstrahlungLastProgrammStrategisches SpielZentralisatorDistributionenraumFunktion <Mathematik>SkriptspracheTouchscreenExplosion <Stochastik>Ordnung <Mathematik>DatensatzDatenflussEin-AusgabeDatenverwaltungURNHypermediaElektronische PublikationKartesische KoordinatenVarietät <Mathematik>MAPFunktion <Mathematik>TouchscreenZahlenbereichProzess <Informatik>SkriptspracheTransaktionDatenverwaltungEin-AusgabeSoftwareentwicklerEindeutigkeitComputeranimation
24:12
MAPDatenverwaltungRohdatenPunktSystemprogrammierungQuantenzustandSelbst organisierendes SystemDean-ZahlExplosion <Stochastik>SpezialrechnerLogarithmusCAN-BusCASE <Informatik>PhasenumwandlungMetropolitan area networkAdressierungWeitverkehrsnetzSpezielle unitäre GruppeFokalpunktCAMDatenflussExogene VariableSupremum <Mathematik>MehragentensystemSoftwareentwicklerProgrammVakuumAppletAnalogieschlussGrenzschichtablösungSkriptspracheNabel <Mathematik>Eigentliche AbbildungSchlussregelCOMGeradeProgrammierumgebungVariableE-MailElektronische PublikationProzess <Informatik>AggregatzustandBefehl <Informatik>MultiplikationsoperatorFunktion <Mathematik>SemaphorSicherungskopieSystemaufrufVersionsverwaltungInklusion <Mathematik>DimensionsanalyseQuellcodeGruppoidModallogikKonfigurationsraumProgrammschemaStapeldateiInformationWeb SiteInhalt <Mathematik>SoftwareentwicklungAusnahmebehandlungZeichenketteSystemplattformCodierungstheorieComputersicherheitLesen <Datenverarbeitung>CodeFehlermeldungTransaktionMAPQuellcodeDatenverwaltungNabel <Mathematik>SoftwareentwicklerGrenzschichtablösungMultiplikationsoperatorFehlermeldungKartesische KoordinatenEindeutigkeitInhalt <Mathematik>SkriptspracheMinkowski-MetrikWechselsprungWeb SiteImplementierungZeichenketteProgrammierumgebungVariableSoftwareentwicklungAbstandSystemprogrammierungSystemplattformUnternehmensmodellComputersicherheitElektronische PublikationDatenflussProdukt <Mathematik>AusnahmebehandlungCodierungGruppenoperationWärmeausdehnungCodierung <Programmierung>RuhmasseSpezielle unitäre GruppeSpannweite <Stochastik>MereologieQuantorDivisionComputerspielMultiplikationFunktionalDreiecksfreier GraphBildschirmmaskeProzess <Informatik>Projektive EbenePhasenumwandlungFreewareGrundraumQuelle <Physik>TermDichte <Stochastik>EinfügungsdämpfungProgramm/QuellcodeJSON
33:08
BinärdatenStichprobeComputerunterstützte ÜbersetzungFunktion <Mathematik>FehlermeldungTermProgrammbibliothekVariableSpieltheorieLogarithmusElektronische PublikationWärmeausdehnungZweiSpezielle unitäre GruppeGruppoidKartesische KoordinatenProdukt <Mathematik>Isolation <Informatik>MereologieDickeMAPKernel <Informatik>BimodulBaum <Mathematik>AdressraumVerzeichnisdienstBridge <Kommunikationstechnik>Schreib-Lese-KopfSchaltwerkImplementierungSystemprogrammierungSystemaufrufDefaultGenetischer AlgorithmusVerschlingungLaufzeitfehlerCASE <Informatik>Wiederkehrender ZustandKontrollstrukturRechnernetzSoftwareentwicklungPolarkoordinatenFreewareQuick-SortUnternehmensarchitekturBefehlsprozessorOktaederCAN-BusServerSummierbarkeitRootkitEmulationLie-GruppeMetropolitan area networkCodeQuellcodeIntelSpeicherabzugThreadDreiLesen <Datenverarbeitung>Nabel <Mathematik>SinusfunktionSkriptspracheSoftwareentwicklungSchlüsselverwaltungSystemprogrammierungMereologieFunktion <Mathematik>MultiplikationsoperatorZahlenbereichMathematikKrümmungsmaßSpielkonsoleRuhmasseSpannweite <Stochastik>ForcingDateiverwaltungWärmeausdehnungPhasenumwandlungSystemaufrufSchlussregelArchitektur <Informatik>DickeAuswahlaxiomElektronische PublikationGruppenoperationWhiteboardZeichenketteMinkowski-MetrikSymmetrieAggregatzustandRelativitätstheorieDifferenteTransportproblemBildschirmmaskeEuler-WinkelEinfügungsdämpfungNichtlinearer OperatorQuellcodeDatenstrukturKernel <Informatik>Überlagerung <Mathematik>ImplementierungMaßerweiterungAnwendungsspezifischer ProzessorMotion CapturingStatistische SchlussweiseMomentenproblemFernwartungSichtenkonzeptBefehlsprozessorInformationProzess <Informatik>ProgrammierumgebungBitKartesische KoordinatenGüte der AnpassungCoprozessorUnternehmensarchitekturSoftwareentwicklerCodeProgrammfehlerDatenbankAbfrageNabel <Mathematik>VerschlingungFehlermeldungSymboltabelleVariableSchreib-Lese-KopfGraphBimodulDefaultLaufzeitfehlerThreadSpeicherabzugZweiMessage-PassingServerSpider <Programm>SoftwareProgrammiergerätAttributierte GrammatikProgrammschemaPunktFormale SemantikJSON
42:04
SummierbarkeitDateiformatAdressierungMengenlehreStichprobeComputerunterstützte ÜbersetzungExt-FunktorTotal <Mathematik>E-MailBinärdatenPASS <Programm>Nichtlineares ZuordnungsproblemCAN-BusSpezielle unitäre GruppeProzess <Informatik>BenchmarkMetropolitan area networkQuick-SortFlächeninhaltSechseckRechenwerkKnoten <Statik>MehrwertnetzPrinzip der gleichmäßigen BeschränktheitURNFormation <Mathematik>PortscannerSchlussregelServerTransaktionRankingLokales MinimumMittelwertElement <Gruppentheorie>DickeGeschlecht <Mathematik>GruppenkeimProgrammfehlerMathematikKreisringWhiteboardDemoszene <Programmierung>ResultanteProzess <Informatik>Kartesische KoordinatenZweiNabel <Mathematik>SkriptspracheNichtlinearer OperatorCoprozessorDatensatzRandomisierungBenchmarkSystemprogrammierungDatenfeldPunktTrennschärfe <Statistik>TelekommunikationMaschinelles GrundbuchMathematikRelativitätstheorieDatenbankZahlenbereichFormale SemantikFlussdiagrammProgramm/QuellcodeXML
45:10
CAN-BusTransaktionPlastikkarteLastAppletLogarithmusSystem-on-ChipStapeldateiProzess <Informatik>SinusfunktionMarketinginformationssystemCADTypentheorieCOMBeobachtungsstudieCASE <Informatik>SystemprogrammierungSystemprogrammSuchmaschineMengenlehrePersonal Area NetworkMittelwertServerFlip-FlopSpezielle unitäre GruppeBefehlsprozessorWinkelZahlenbereichCodierungstheorieGeradeCodeNabel <Mathematik>SkriptspracheSupremum <Mathematik>DatenbankInformationsspeicherungPaarvergleichNichtlineares ZuordnungsproblemWiederkehrender ZustandSCI <Informatik>SummierbarkeitReelle ZahlMaß <Mathematik>Kartesische KoordinatenVerschlingungRechenwerkHill-DifferentialgleichungSpeicherabzugLokales MinimumDigitale PhotographieCAMGeräuschÄquivalenzklasseSoftwareentwicklungDedekind-SchnittURNZählenGesetz <Physik>UnternehmensarchitekturEDV-BeratungSystemplattformFreewareAuswahlaxiomChaostheorieDienst <Informatik>InformationProgrammverifikationInhalt <Mathematik>ComputersicherheitProdukt <Mathematik>Patch <Software>IntelDivergente ReiheWeb SiteOffene MengeTranslation <Mathematik>SystemprogrammierungServerUnternehmensarchitekturSkriptspracheFreewareSpeicherabzugWeb ServicesEDV-BeratungEntwurfsautomationNabel <Mathematik>AuswahlaxiomSystemplattformCASE <Informatik>InformationPlastikkarteVollständiger VerbandDigitales ZertifikatSoftwareentwicklungProgrammverifikationTranslation <Mathematik>Divergente ReiheAbfrageZentrische StreckungService providerRadikal <Mathematik>DateiverwaltungDesign by ContractVerzweigendes ProgrammBildschirmmaskeGeradeKategorie <Mathematik>DatenmissbrauchNichtlineares GleichungssystemVerzweigung <Mathematik>UngleichungGraphTouchscreenPatch <Software>Selbst organisierendes SystemBitrateElektronische PublikationGüte der AnpassungGrundsätze ordnungsmäßiger DatenverarbeitungObjekt <Kategorie>Rechter WinkelShape <Informatik>BitKurvenanpassungQuelle <Physik>ComputeranimationFlussdiagrammJSONUMLXML
51:16
Selbst organisierendes SystemFreewareUnternehmensarchitekturTranslation <Mathematik>MultiplikationsoperatorNichtlineares ZuordnungsproblemKernel <Informatik>BenutzerfreundlichkeitSystemplattformStrom <Mathematik>BildschirmmaskeProgrammverifikationDienst <Informatik>Produkt <Mathematik>ServerSuchmaschineMengenlehreAnalysisWeb logAdressraumSpezielle unitäre GruppeSondierungTreiber <Programm>SystemprogrammierungUnternehmensarchitekturInformationsspeicherungDigitales ZertifikatKernel <Informatik>MultiplikationsoperatorBitrateService providerFreewareElektronische PublikationWhiteboardSelbst organisierendes SystemTranslation <Mathematik>FehlermeldungHidden-Markov-ModellStrömungsrichtungFunktionalQuick-SortGemeinsamer SpeicherSoftwareentwicklungSchlüsselverwaltungRechter WinkelComputervirusWeb-SeiteLineare RegressionPlotterGruppenoperationDesign by ContractTaskSprachsyntheseRechenwerkSoftwareentwicklerPunktPhysikalische TheorieEreignishorizontComputerspielEinfügungsdämpfungZellularer AutomatInternetworkingWasserdampftafel
56:54
CodeGeradeCOMMehrwertnetzCAN-BusNabel <Mathematik>SkriptspracheZahlenbereichBenchmarkServerMatchingTransaktionElement <Gruppentheorie>MittelwertGraphGruppenkeimGeschlecht <Mathematik>ElementargeometrieRegulärer Ausdruck <Textverarbeitung>ProgrammfehlerGewöhnliche DifferentialgleichungRechnernetzKernel <Informatik>MathematikQuellcodeSoftwareentwicklungKontrollstrukturReverse EngineeringGruppoidSpezielle unitäre GruppeSummierbarkeitProgrammbibliothekBinärdatenVariableSpieltheorieEntscheidungsmodellWeb SiteInhalt <Mathematik>Eigentliche AbbildungProgrammierumgebungE-MailProzess <Informatik>Tablet PCBefehl <Informatik>Funktion <Mathematik>MultiplikationsoperatorGrenzschichtablösungVersionsverwaltungSystemaufrufSchlussregelElektronische PublikationSicherungskopieSemaphorMAPDatenverwaltungPunktRechenwerkTouchscreenExplosion <Stochastik>System-on-ChipURNOrdnung <Mathematik>DatensatzDatenflussEin-AusgabeHypermediaPASS <Programm>TemperaturstrahlungLastÜbersetzer <Informatik>ProgrammStrategisches SpielDistributionenraumSchlüsselverwaltungStichprobeSCI <Informatik>Disjunktion <Logik>Ext-FunktorElektronischer FingerabdruckWiederkehrender ZustandTotal <Mathematik>Computerunterstützte ÜbersetzungQuick-SortDatenfeldInklusion <Mathematik>MultiplikationsoperatorHidden-Markov-ModellCoxeter-GruppeZusammengesetzte VerteilungDiagrammFlussdiagrammProgramm/QuellcodeXML
57:48
Nabel <Mathematik>SystemprogrammierungSystemaufrufKernel <Informatik>VerzeichnisdienstOffene MengeProzess <Informatik>COTSSpeicherabzugBefehlsprozessorSchlüsselverwaltungQuick-SortVollständiger VerbandMatchingKontrollstrukturOperations ResearchDatenverwaltungElektronische PublikationFreewareSoftwareUnternehmensarchitekturSkriptspracheProgrammZählenAdditionTabelleDatensatzLesen <Datenverarbeitung>Computerunterstützte ÜbersetzungZahlenbereichTemplateSummierbarkeitNichtlineares ZuordnungsproblemE-MailBinärdatenStichprobeCodeWiederkehrender ZustandFunktion <Mathematik>Total <Mathematik>Inklusion <Mathematik>DatenfeldElektronischer FingerabdruckLastMathematikStrategisches SpielDistributionenraumHypermediaVirtuelle MaschineResultanteRechenschieberJSONProgramm/QuellcodeXMLComputeranimation
58:43
CAN-BusMAPDatenverwaltungPhasenumwandlungPortscannerRohdatenPunktSystemprogrammierungQuantenzustandSelbst organisierendes SystemDean-ZahlSpezialrechnerExplosion <Stochastik>LogarithmusMetropolitan area networkSystemaufrufVersionsverwaltungOpen SourceRankingSystemprogrammierungMereologieElektronische PublikationRechter WinkelWort <Informatik>VideokonferenzInstantiierungMultiplikationsoperatorIntegralFestplatteCoxeter-GruppeSichtenkonzeptEinfacher RingKernel <Informatik>SoftwareentwicklungVerschlingungDickeJSONFlussdiagramm
Transkript: Englisch(automatisch erzeugt)
00:03
Welcome to my session. I'm Daichi Koto. Today in my session, I'll talk about one of my privacy business in Japan. About Japanese companies, enterprise system, development, and privacy use cases.
00:21
How I offer privacy as their base platform and what I developed for them to improve their business values. Before I get started my talk, I have to have a story to have to say. My English is lesser. I am not good at English, especially for me.
00:46
It is hard to listen. To listen English is hard for me. If you have some questions, email me after the decision is better for each others. No matter what you want to talk to me on the fly, would you please slow down your speaking.
01:08
Assembly. At first, I'll introduce myself, my job, and privacy related activities. Next, I'm going to introduce USP Labs, Universal Shell Programming Laboratory.
01:24
They are a rapid growing IT system development company. The company size is small or middle size. USP Labs has very unique technique to develop enterprise system in a day and age.
01:41
Called it unique age development method. Very fuzzy method. Then, I will introduce two products what I made for them. Enterprise specialized shell, called U-Shell, and new file system, Google file system.
02:01
And I will talk about my work to construct a privacy-based high-performance computing platform called USP Boa. At last, I suggest a problem now we've faced in our business. Yes, let's get started. Introduction.
02:21
My name is Daichi Goto. Daichi is first name, Goto is last name. Goto is not programming syntax. I'm not a fourth-round family man. I'm a port committee, and a source committee. Originally union file system and Japanese port is my work.
02:44
I have two small companies. My first and even now main company is ONGS. ONGS is so-called jack-of-all-trades company. Enterprise system design, development, management and maintenance, web server design, development, maintenance and management.
03:07
I de-related news, articles, magazines and books, writing. I de-related seminar. I de-related consulting. Many, many. One year ago, I founded another company, BSC Consulting, as a co-founder.
03:28
At that time, my business partner company needed a free BSC support company. So, we founded it. Perhaps, only Japan specializes in free BSC support and consulting company.
03:50
Yeah, I was doing some free BSC related jobs. Free BSC daily topics. Daily free BSC news in Japanese.
04:01
And my partnership with Gijutsu Hyoroshi. I am writing some books, magazines and articles with Gihyo, too. And at the BSC Consulting Works, we are doing free BSC hardware verification services and consulting services.
04:20
Free BSC-based enterprise platform constructions and maintenance services, too. Yes, this is a free BSC paper book, published last year. Prolactica Free BSC Server Building App and Management Guide.
04:41
Unfortunately, the book industry in Japan is very serious situation. I guess this is the last free BSC paper book in Japan. And next. Free BSC digital book. Last year published it, too.
05:01
Free BSC Expert Digital Edition 2012. I will plan to publish four free BSC digital books per year. Writing for next published is still down. Yeah, my free BSC related activities.
05:22
Yes, I am a social media post commuter. And another of my activities is to join BSC conference. I attend to worldwide BSC conference. Yeah, BSC camp and Asia BSC conference, of course. EuroBSC camp, band summit in San Francisco.
05:42
And free BSC, let develop sound if possible. And I will report conference topics to Japanese audience. My English is lesser, yeah. And other Japanese English are terrible, terrible at English. To report some conference topics in Japanese is very valuable work.
06:07
And I have free BSC seminar per month supported by KDDI Web Communications. BSC Consulting, ONGS, and Gijutsu Hyourosha. KDDI Web Communications is one of the big ISPs using free BSC as their best platform.
06:26
KDDI Web Communications management a few thousand free BSC servers, a big ISP. Yeah, free BSC daily talks, daily free BSC news site hosting on Gijutsu Hyourosha.
06:45
This is a same free BSC seminar. The Loom supported by KDDI Web Communications, very Korean. Good conference room.
07:01
And my free BSC seminar, main teacher is Sato-san, yeah, HLS. And sometimes me and Asada-san. Asada-san is attending this BSC camp now. He is developing around beehive, beehive native bias he is developing now.
07:20
At the moment, free C committers praise law as a free C teacher. So, I will move on the main point. There are too many industries. I will talk about retail industry, retail distribution industry, and related industries.
07:44
I guess the no cares about many Japanese companies have no old IT sectors. They always outsource their system development, maintenance, and management.
08:05
They love to outsource IT vendors. As a result, big companies depend on big IT vendors. How about middle small-sized companies? They use imported and translated commercial software packages.
08:31
Yeah, so many software used in Japan as foreign software, just translated into Japanese.
08:47
Domestic IT vendors like Enoishi, Hitachi, and Fujitsu use imported and translated software from YAKOLAKO, SAP, Microsoft, Amazon. And Japanese enterprise IT solution is sub-sub-sub-sub-sub-contractor structure.
09:08
Bigger vendors praise only many manager. Sub-contractors praise as project manager. Sub-sub-contractors write specifications.
09:20
Sub-sub-sub-contractors write Excel word documents. And leaf like code. As a result, high cost and low efficiency. So many enterprise system development projects are not working well.
09:42
Too many costs, too many people, too many time, too many unnecessary documents. Especially Japanese logs Excel document. Too many time, too many unnecessary source code, and too many stressful workers. All not happy.
10:00
And so, USP Lab, Universal Share Programming Laboratory. Very unique laboratory-gluing enterprise system development company. Owns my first company established in 2002.
10:21
USP Lab established in 2005. Very young company. If you have interested in this company, a few months ago they created English version website.
10:40
So please visit their homepage and leave the English version website. So USP Lab's current number of employees is from 20 to 13. A small company.
11:00
And they are growing rapidly, very rapidly. The main enterprise system is sales accounting system, payroll accounting system, corporate system, MRC system, merchandising system, enterprise system self-manufacturing, etc.
11:28
Yes, they are not all one of their customers. Many, many companies there. George knows well, I guess, very famous companies in Japan.
11:46
What USP Lab made for their customers? For example, for Ryohinkaku, internal manufacturing mentorship, merchandising system, head store communication system for world,
12:02
equipment management system, share energy for world. Stores sales management system for Hanshoku, mission critical system for Seijoichi, sales forecast system, investment control system for Tokihans, store management system, body shop.
12:23
Yes, body shop is Canada too. For body shop, BI tool system. For trial company, mission control system. For your Seijoichi, real time sales and marketing support system. So too many, too many, too many systems they made.
12:42
It is a part of their works. From 2005 established it. It is just seven or five years. USP Lab made for many, many enterprise system.
13:03
So they have very unique tools to develop their enterprise system in our day. The Commons and ShareScript. Commons and ShareScript. They have specialized command called USB TSUKBAI. USB TSUKBAI, around 50 selected commons survived among from several thousands of commons they developed in past years of their business.
13:34
Yeah, it looks like the 40 years old Unix style system development.
13:40
Very unique nowadays. Okay, so their business. Many business folks and developers at first feel disturbed and laugh at their development style to spawn. I know it's unnatural. However, USB develops enterprise system in very quick, very, very quick.
14:06
A few days in some cases and the system works very well. Surprised. And develop cost. Cost is very reasonable, very cheap. And development, their development is very flexible because their application code is very few.
14:29
And approach is very easy. At last, customers company could do self-manufacture. First I said, many Japanese companies have no IT sector, but at last the customer have their IT sectors.
14:49
Because development is very easy. This is one of their research project.
15:03
So they have research joined with Tokyo University, Keio University, Waseda University, Nageji University. Famous universities in Japan. Unique age development method. Unique age development method is a software development method using Unix of its frequency, text file, commands, and shell script.
15:33
It includes from weakest listening and human communication to system architecture design,
15:43
platform constructing, data design, agile programming, coding manner, allocation, documentation, and maintenance. Comprehensive development method.
16:00
Yeah, the key tools are text file, commands, and pipeline. They use inexpensive PC, common PC, as their base platform. And the data is only text file, white, space-separated, plain text file, coded, feed, format text.
16:21
And they use Unix common commands. Set, oak, tear, graph, echo, cap, head, tail, etc. And customized command. USP Tsubai command. Self, join0, join1, join2, sum2, sum4, sum5, jo, keta, marme, dilsin, so on.
16:47
And it's joined by pipeline and shell script. Yeah, every day we use this command. Plus some Tsubai command.
17:05
Yeah, that's all. Their system development method, that's all. Makes a data file and a common pipeline, common pipeline, common pipeline, pipeline processing.
17:27
And this model is very familiar with much core and many core systems. Process, too many processes use so many cores at the same time.
17:44
And this is a table of Unix command. So USP Tsubai has many Japanese-named commands because their business is in Japan. And they are preparing English version USP Tsubai now.
18:05
I guess USP version become familiar with you. And I, so USP Lab opened a license-free version of Tsubai commands. Open USP Tsubai, that is written in Python.
18:23
So Komeka USP Tsubai commands a written in C for get first speed. Open USP Tsubai are written in Python. But it is enough about the speed. Open USP Tsubai aimed at expanding the ideas of unique age development method.
18:46
And some companies are doing their job using Open USP Tsubai. Open USP Tsubai implementation are enough, well, that works very well.
19:05
So why unique age development method? So first, no middleware, no Oracle, no database. And Shell use kernel feature directory system called pipe, fork, EXEC, wait, open.
19:26
So no middleware realize the first processing. And applications are very short, a couple dozen lines, very very short.
19:40
Yes, this is application. They call, this is application. Yes, first I see in the script, they called it application.
20:03
Application? It is just some pipeline. But this works very well for enterprise system. This is Aman's example. This is web application, that's all.
20:25
This script act as database update and database search. Name read join zero mojihame at Tsubai command.
20:48
Why flexible? Because it's simple and easy to run and customize. Important and main ideas of USP are unique age development methods.
21:06
It's data, data separation and data distribution. This is key point. First, it's separate. Yes, two separate is 2N star.
21:23
That is USP's idea. And the kanji wakeru means separate and understand. Akanji have two meanings, separate and understand.
21:45
So Japanese thinking to separate is to understand. Separated by business, separated by organization, separated by software and law.
22:04
Next is distribution and unique age method. Data are copied to all software and distributed to everywhere.
22:21
Any programs don't share one file. Files copied to all applications and not override. Applications just read a file and write into another new file.
22:43
This is key point two. So for distribution and no override is a robust for an expected accent. Wrong data input, software back, hardware back. Then the work can inspect easily because they are just only text file.
23:06
And this realize easy lower back the images of distribution and full sharing.
23:20
The data are the same, same data. They copy all files for all applications. And this picture shows unique age development method data flow. So first input data is from POS or all the data, master data, transaction data, something like that.
23:51
Then they refine the data to the text file. So the text file separated to five layer data management.
24:03
Five layer data and script processing and output result. That's all. Yes, they use a level one, level two, three, four, five data separation.
24:22
Level one is a low data, just copy from POS. Yeah. Yeah, but low data must have low data. Level two, level two has been formally put into the system at a particular point of time.
24:46
And the level three, currently application use level three. Level three data is, what should I say, normalized data data to treat by command.
25:15
And that is plain text file.
25:20
And level four is specialized for application, some kind of application. And level five is PDF, Excel or something like that. This is a more detailed picture to show the five layer data management.
25:40
Level one, low data from POS or low data or some other application output, sales data, production masters and transaction data. It is translated to level two data and above level three data.
26:08
And the task, five data, English, who, when, where, who, when, how, why, exactly.
26:34
Next.
26:45
How, yeah, yes. Yeah, thank you. And this picture shows a comprehensive development lifecycle.
27:03
Unique Edge development method has two develop phase, implementing phase. And first time is discussion, listen to customer request and design.
27:20
Yes, 70% important for this phase. And next is short-term development. Yes, application is very small, so development time is very small. And the third phase, they redefine the design.
27:45
And the fourth stage, they finish their implementation. Then they find the application, application is a little, just quick, just quick.
28:02
Yeah, they use shell scripts, so shell scripts are extremely flexible. So there are many restricted coding models. There are models, but models are not documented.
28:27
Usually, they use agile programming. So daily work, mentor, programmer, educate, partner, the coding models, documentation.
28:49
Unique Edge has little documentation. So because application source code very few, so they don't need many documentation.
29:09
So required documents are the application IO API and the management source code data, original data description.
29:21
And system purpose, business flow manuals. So these seven sentences are very important for them.
29:45
Unique Edge development method is based on the Unix philosophy. Do you know the Unix philosophy? There are 40 years old philosophy. Small is beautiful. One program should only do one thing.
30:03
Proximity should be first as possible. Proximity takes presence over efficiency. Date is stored as plain text, commonly used as letters. Applications are written in shell scripts or programs are designed as filters.
30:25
They're doing Unix philosophy. That's all. And they have a unique site, UEC, Unique Edge Engineering Community Site.
30:43
The site for the unique edge engineers. So it's very crazy. All contents are specialized for shell script programmings. The world's most crazy shell script site. So unfortunately, in Japanese only, I guess Google Translate helps you.
31:08
So next, I will introduce my work, U-Shell. So USB uses shell script. Shell script is their key technology.
31:29
So I developed new shell for them. I suggest them. Send OS and bash, so it's good.
31:46
Previously, a shell is better for you. And they loved FreeBSD and A-Shell, so I developed an A-Shell-based customized shell. Remove unneeded features, unnecessary features, and add new features they want.
32:07
Debugging feature, exception handling, brace expansion, and string handling. Now U-Shell is USB space platform. U-Shell, cutting robustness. No unnecessary feature to improve coding quality.
32:29
No features relating to social security vulnerabilities, especially on the environment variables. And I added some new features.
32:43
For example, this is error handling code. A-Shell has no ERR error command. I added to U-Shell error command. So error command logistic send function.
33:07
Shell command fails, jump to function, and run, and return a false error.
33:21
And this error command traverses the sub-shell and sub-shell. This fork and sub-shell and pipe make sub-shell, sub-shell, sub-shell, and a false.
33:41
But U-Shell checks the error of that. And A-Shell has a high query. U-Shell has a high query too. And if high query is set at error time, it exists automatically.
34:05
Next, what was the output for debugging? Previously, A-Shell has no number output. So developers cannot understand where, describe where, U-Shell outputs the number of code.
34:30
So developers can quick find the program point. U-Shell has no export, because environment variables lead some programs for beginner programmers.
34:56
And log command, U-Shell has log command.
35:01
It's like EXCC, but EXCC and EXCC space two, bigger than space, file name, and semantic wrong. EXCC? Why does it lie into the file?
35:23
Beginner program cannot understand. So we renamed the command. And brace expansion. USB uses brace expansion well, so I added this feature to U-Shell.
35:41
Crawling is very hard. And sub-string operation. Yes, A-Shell has sub-string operation, sharp, sharp, sharp, percent, and percent, percent. That is not easy for beginners. So I added sub-string operation like Oak.
36:01
It is good for beginner programmer. And next, bug on file system. Bug on file system create a new file, which is part of a SAM file, without the I-O, copy in I-O.
36:21
And this feature implemented as kernel module. So we're going to test the I-O, the kernel module. And look, 43 GB file, this is the master data file.
36:47
Yes, for example, primary key, data, SAM handle data, blah, blah, blah. The link. Yes, it looks like common symbolic link, but cat, just only 34 bytes.
37:10
It looks like symbolic link file, but works as part of some other file.
37:20
YabuboFS is a system called Hawk Magic implementation. YabuboFS implements all related system called to put BugoFS feature into place. YabuboFS kernel module replace previous default system called to use BugoFS system calls at runtime.
37:49
So we tried some kind of implementation. I choose to use symbolic link file as a trick of a bug file system, because it is the most of forces. I try to edit a file with attribute information, head link file, and symbolic link file.
38:10
Symbolic link are the most forces. Use case, one big master, 10 GB, edit common C for enterprise system.
38:22
Some applications want to use a part of the master file. For example, 100 applications lead from the master file and graph it and make new file very slow. If you use bug file system, you can create millions more files in seconds.
38:45
No separate to separate your lead and the light new file. Very fast. Database view? Yes, exactly.
39:01
Yes, yes. Yes, exactly, yes. Next step. They wanted more debugging feature, remote control and network programming. Bash has port, port pass.
39:21
Do you know that? Bash support pass? SuraDev, SuraPorts, Sura80s, blah blah blah. That is the Bash extension, like including netcat feature in shell.
39:40
And Gattai file system, the robust feature of Google file system. A file constructed any other files. But a file is like a bug. Next, big data treatment.
40:00
USB ball. This is previously based big data treatment appliance. I made it for USB lab. One master, 5 threads PC that connected via 10 gigabit NIC.
40:23
This is the architecture, but simpler architecture. So I made this system three times. First time. Yes, it failed.
40:41
First of all, I chose Mellanox Technologies InfraBand ConnectX2 for network. That works. Yes, great, but unstable. And InfraBand network transportation is very unstable.
41:01
So lack of time. I cannot change the current source code. So we should choose another choice. To buy another device. Second. Second, I tried to use Intel 10 gigabit NIC. That's good.
41:21
But we need more processor power. We need more impact to other customers. And third, server wall success. So we replace all massive threads CPU to
41:40
first is iCore 7 and common Xeon core. At that time, a little bit cost over. But that's good enough impact I get.
42:02
Yes, next. Show the application. Used it on the USB board. CrustaFoo. CrustaFoo. CrustaFoo. CrustaFoo is USB board specialized command.
42:23
If you did Crusta joins out, joins out process did. Disobuted to the Slimer scenes. And automatically correct the result to the master.
42:43
I'm on the benchmark result. Yes, select. Select process from 1 billion records. Using Tempara processor. 3 seconds.
43:01
Sold. 1 billion random records. 97 seconds. 3 seconds. Some key field in 1 billion random records. 35 seconds.
43:22
Must match the operation. 22 seconds. Enjoy. Sorry. Sorry. Sorry. Compiled operation 77. These results are very fast against the Hadoop.
43:44
For example, the random record sold. Using 200 Hadoop nodes to sort 1 billion random records takes 20 minutes.
44:06
But the system does it at 100 seconds. Very fast. And another benchmark.
44:20
Yeah, big point is last one. Big data search. Huh? No, no, no. Yes, yes. Yes, last. Three big data search. Especially. So search.
44:42
A person's data from 10 billion records. It is just only 1.2 seconds. All data on TemparaFS. And each node connected via 10-gigabit NIC.
45:01
We can get disk speed. No database, no middleware. Just command, operating system, shell script. And as example. Yeah, two main example. All shows new system works very well.
45:24
Yes, yes, yes, yes, yes. There are business example. Yes, get first. Next. Get first.
45:43
Get first. Get last. Yeah, they translate SQL query to the shell script. Currently, they are developing the program to automate, translate from SQL to shell script using scubai commands.
46:04
They are working very well. So, to use SQL, SQL is very hard to get the scale to the merge core to the main core. But UNIX command and piped join UNIX command shell script
46:28
very well scaled to merge core to main core. And it looks very easy. Yeah, this approach is very good, I guess.
46:40
And next, basic consulting. I'll talk basic consulting for enterprise. Just one year ago, basic consulting established it. BSD consulting hold all the sub diary of USP laboratory. Short name is BSCC.
47:02
President is Tonaka and directed me. Yet, two years ago, Tonaka have involved me as a free BSC consultant. Then, USP found that free BSC is better choice for them as base platform.
47:24
So, until that time, they used CentOS and Bash. I pushed them free BSC and Ash. And I told it, I developed this customized shell, used shell and specialized file system for their business.
47:43
And that time, great new USP's customer hesitated to take free BSC as their base platform. Hmm, that is a common case. They said, we call the lock of the company for support of free BSC.
48:01
So, they cannot choose free BSC. Hmm, exactly. We locked the free BSC company at that time. So, we established the BSC consulting for our business. Basic consulting logo. We supported the consulting of free BSC.
48:22
Providing free BSC information in Japanese. Yeah, common Japanese, terrible in English. So, providing free BSC information in Japanese is valuable. And free BSC hardware verification service and terminal services.
48:40
Hmm, some folks attended the LGR BSC conference here. And you know the terrible in English or Japanese hardware verification service. Japanese domestic server hardware vendor like NVC, Hitachi and Fujitsu lack of free BSC support
49:06
because of the lack of free BSC company. But good news, we concluded with NVC. NVC express 5800 verification service.
49:25
So, last year, they changed their base chip from Intel to Broadcom because of the cost. Intel cost many money.
49:41
And they chose a new mega writer card that does not work with MIF on free BSC. And free BSC did not work on NSC to new express. Their customer got angry.
50:02
The BSC and NSC have a contract of free BSC support on hardware verification. So, from this spring, express series will work with free BSC. And patches, documents, information will be opened on BSC site.
50:26
And last year ago, we prepared another organization, NPO for enterprise.
50:42
So, BSTC and BSR, BST Research Chairman is Satoshi. Satoshi is now established NPO for free BSC.
51:00
And BSC Research, its core business is Asia BSC Conference, BSD certification and BSC documentation, translation. BSC and BSC are. So, one year ago, we thought that we need a fair and impartial,
51:24
certificated organization to promote free BSC to enterprise company. So, we really have BSC certification, BSC certification group. So, BSC are new NPO company cooperate with BSC CG and provide BSC certification in Japanese.
51:47
Translation is ongoing. At last, I suggest a program to be solved as soon as possible. That is Infinbam driver and off and on.
52:04
HPC need Infinbam driver. 10 gigabit NIC works fine. That works very well. But Infinibam transport 3.2 times faster than 10 gigabit NIC.
52:20
Yeah, in fact, we are constructing a new USB board with Linux because of Infinibam. So, that is three months ago. We have tried to improve OPDO on free BSC. But in the end, it failed.
52:45
At last, we realized that we implemented all Linux in API in the free BSC kernel. Linux has an API. API is some kind of kernel specialized function to develop device driver.
53:03
So, it looks like a long approach. We are considering the next approach. If you have any ideas, please contact me. We need Infinibam drivers. The lack of Infinibam drivers gives Red Hat and CentOS some advantage.
53:24
Come HPC platform. So, should we contract to contact to Mellanox Technologies? Or, should we suggest to FreeBC Foundation to develop latest OPDO subsystem?
53:42
Current big concern, big business showstopper is Infinibam. I heard some big vendors using Infinibam. And they are accelerated to use Infinibam in their data center. The other one, the other company you talked to a lot was Geologic.
54:03
Intel? Intel bought Geologic before. Intel? So, there were two big vendors of Infinibam.
54:20
Last year in June, Intel purchased and acquired Geologic.
54:51
But the market share never looks like a big win to Intel.
55:08
Will Intel release a FreeBC Infinibar driver?
55:26
My talk is done. Any questions? How exactly does the system of storing files in this sort of large file without databases,
55:42
how exactly does it handle for data integrity? For example, a big convenience store that has a thousand stores in the business area.
56:07
Common master file size is 1GB. So big retail company's master data is 10GB and so-so.
56:24
Right, but I think what he's getting to is the error rates coming off the drives. The drives do not guarantee perfect data. They guarantee nearly perfect data. And your data is so big now that reading it can result in an untracked error.
56:46
Hmm. And there is no problem because all data are copied at any time.
57:08
There are any data. So, if failed, detected, to rollback, pre-stage, and calculate again. So, whatever you're pulling in data, you're pulling it from multiple places and
57:23
then comparing them to make sure both are correct before working with them? But I'm saying how would you know if the data was bad? Hmm.
57:52
USP uses the same machine always and checks the results at last.
58:02
So, if the results are mismatched, they do it again. All right? Yes, another question? Can we have a copy of your slides?
58:31
A slide. This one? Yes, your presentation.
58:42
Sorry, copies, right? Presentation. Can you send a copy? Ah, yeah, yeah, no problem. Yes, I'll do it. Okay. Question 2. Yeah? USH and Google at best, are they available or is this proprietary for a company?
59:02
Yes, currently proprietary only. We cannot write with it. It sounds straightforward because the file name, the offset, and the length are encoded in similar. Unfortunately, we are implementing it for Linux kernel now because of USP or program.
59:29
And that will be GPL, so that will perhaps open source to the world.
59:42
The symbolic link is interesting. Always the symbolic link. Always. Because he's encoding, if you will, the select or the view. We have three. He's got 128 bytes to do something with. So we have three implementations. First, I thought, how…
01:00:00
The filtering is the fastest one, but it is not, because to create a new file, it accesses to the hard disk sometimes. And to use attribute, a file system's attribute, it writes sometimes to file system.
01:00:20
The symbolic ring is just one time light, and that is the most fastest reason. Alright, any questions? Nothing? Yeah, thank you very much.