{"title":"ACM SIGOPS European workshop 1988: position paper","authors":"M. Wilkes","doi":"10.1145/504092.504134","DOIUrl":null,"url":null,"abstract":"Position Paper When personal computers first appeared some people said that they marked the end of time-sharing. This did not happen. Personal computers were sold on a wide scale and through them new markets were opened up for the computer industry. However, serious users-those who had heavy requirements for processor cycles and the need to share programmes and data with colleagues-continued to use time-sharing systems. They found the new, low cost, departmental time-sharing systems ideal for their needs. These low cost systems owed their existence every bit as much to VLSI as personal computers did. With the advent of more powerful personal computers-which became known as work stations-users of the class I referred to above began to use them instead of time-sharing systems. However, few such users' regard their work stations as stand-alone entities. Services available via a local area network are essential to their needs. A file server was early perceived as being of first importance, but it soon became apparent that a work station user needed all the services traditionally available on a time-sharing system except processor cycles. Even that exception requires qualification, since the hunger for processor cycles is such that work station owners are often glad to avail themselves of cycles available elsewhere on the network, for example, in other users'workstations or in a time-sharing system. A user judges a computing environment by the response he gets to the commands he types at his work station. There was once a tendency to assume that the response would always be better if the work was done in the work station itself. There is now a growing appreciation that this is not necessarily the case, and that a judicious distribution of function between the work station and computers elsewhere on the local network may give improved performance. We thus have the concept of work station style computing in which it appears to the user that everything is being done in the equipment on his own desk whereas, in fact, much of it may be being done elsewhere. It is along the above lines that I approach the topics discussed in the handout for the proposed workshop. The availability, at the present time, of work stations in various price ranges and their acceptance by the user community, gives topical importance to the issues proposed for the workshop.","PeriodicalId":246681,"journal":{"name":"EW 3","volume":"59 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1988-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"EW 3","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/504092.504134","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Position Paper When personal computers first appeared some people said that they marked the end of time-sharing. This did not happen. Personal computers were sold on a wide scale and through them new markets were opened up for the computer industry. However, serious users-those who had heavy requirements for processor cycles and the need to share programmes and data with colleagues-continued to use time-sharing systems. They found the new, low cost, departmental time-sharing systems ideal for their needs. These low cost systems owed their existence every bit as much to VLSI as personal computers did. With the advent of more powerful personal computers-which became known as work stations-users of the class I referred to above began to use them instead of time-sharing systems. However, few such users' regard their work stations as stand-alone entities. Services available via a local area network are essential to their needs. A file server was early perceived as being of first importance, but it soon became apparent that a work station user needed all the services traditionally available on a time-sharing system except processor cycles. Even that exception requires qualification, since the hunger for processor cycles is such that work station owners are often glad to avail themselves of cycles available elsewhere on the network, for example, in other users'workstations or in a time-sharing system. A user judges a computing environment by the response he gets to the commands he types at his work station. There was once a tendency to assume that the response would always be better if the work was done in the work station itself. There is now a growing appreciation that this is not necessarily the case, and that a judicious distribution of function between the work station and computers elsewhere on the local network may give improved performance. We thus have the concept of work station style computing in which it appears to the user that everything is being done in the equipment on his own desk whereas, in fact, much of it may be being done elsewhere. It is along the above lines that I approach the topics discussed in the handout for the proposed workshop. The availability, at the present time, of work stations in various price ranges and their acceptance by the user community, gives topical importance to the issues proposed for the workshop.