Mastodon Feed
Boosted by fromjason ("fromjason.xyz ❤️ 💻"):
dazo@infosec.exchange ("🔗 David Sommerseth") wrote:
I remember the early 2000, with so-called "thin clients" which where essentially graphic terminals which rendered images from a centralised server.
This was claimed to be the future for office workers. And was actually just a reinvention of the X11 capable graphic terminals from the earlier 90s and the text terminals from the 80s.
When thin-clients didn't get traction, software based solution surfaced (RDP, Citrix, VNC, SPICE, etc).
All have the same goals ... Less system administration work, higher security, unified desktop experience for all users, better utilisation of computing resources - since all can be managed centrally by the sysadmins. And lower costs and higher ROI.
The common denominator through all these periods - as long as the acquiring cost of individual computers are at an acceptable cost level, the centralised terminal approach will not be prioritised.
Of the stronger arguments for individual systems is that replacing lots of individual systems over time is an easier cost to plan for than upgrading high performance servers at regular intervals, and the setup time and preparation of a new server can take more time. It also gives greater flexibility in computing power, power users gets more capable hardware while others get cheaper hardware. In addition, a centralised approach is more vulnerable for downtime if the server or networking fails, all users are affected at once. A situation individual computers are far less prone to experience.
The centralised computing approach will surface again and again, every 10-15 years, with similar arguments. And I still believe it will still "fail" for most use cases, as long as the cost of independent computers are available at a reasonable price.