April 13, 2019 at 12:00 am
Comments posted to this topic are about the item The End of XP
April 13, 2019 at 4:39 pm
In our home office, the leftover of a small business venture, we're still running XP on a digital music server, Win7 on a digital picture server, and Win 10 on two desktops that we use for the more critical records and software like my SQL Server . My thought has always been to stay up to date - SORT OF.
There's a delicate balance to what is critical, both content and potential downtime, and what is not. Coincidentally, while I was writing that last sentence, the XP music server went bonkers, just hung with a seriously bad noise. So far, it is rebooting. This machine has also lost its wireless connection and is waiting for me to locate a network board in my parts storage.
So to me it boils down to how critical the machine and OS are, versus the cost and risk of upgrades just to stay current with support. And of course, how exposed the system is to security threats.
Rick
Disaster Recovery = Backup ( Backup ( Your Backup ) )
April 13, 2019 at 8:31 pm
SQL Server 2008 was version 10, not 9.
XP end of life was in 2014- the extended lifespan for XP was only available through a hack that makes your system look like Windows Embedded Industry, which happens to have an XP kernel. Any IT professional that performed a business implementation of that loophole to "support" XP should be ashamed.
April 15, 2019 at 6:58 am
I'd absolutely love to upgrade to SQL Server 2016 or higher, but MS decided to screw us over by making Reporting Services be a single instance install per server, WTF?!
We currently run an SSRS instance per customer on the same box, which in SQL 2016+ is impossible. Not to mention the fact that in SQL 2016 they change the way SSRS authentication works for custom auth - which you absolutely have to do if you are selling a web product that's not used on a domain - and thus again loads of effort.
I really wish we'd never gone near SSRS, anyone got a time machine?
April 15, 2019 at 7:29 am
I know of a system that is still active that allows a £2million revenue stream that runs on an app written in Clipper. The fundamentals of the business process haven't changed in decades, the app just works, it costs next to nothing, it isn't connected to the web, strictly speaking it is an MSDOS application, it will run without Windows. It also compiled to <128Kb so running it on today's infrastructure is like trying to light a pilot light with a 50megaton nuclear bomb. IT.....JUST.....WORKS.
I know all the arguments for upgrading, rewriting etc but IT.....JUST....WORKS. It was written before I left college, its looking like it will be running after I retire. Much like a huge volume of COBOL code that is still out there. If it ain't broke don't fix it, I'm sure there is stuff that genuinely does need fixing rather than disrupt something that chugs away generating £2million/year.
In the web/mobile world everyone is chasing the new shiny-shiny and decrying anything older than 6 months old as "legacy". Granted an OS is a complex beast and I wouldn't want to be doing new develop on top of obsolete OS' but what business progress do we actually empower when we upgrade. Are we simply plugging security holes and putting a fresh coat of paint on something that is largely unchanged?
April 15, 2019 at 4:35 pm
Granted an OS is a complex beast and I wouldn't want to be doing new develop on top of obsolete OS' but what business progress do we actually empower when we upgrade. Are we simply plugging security holes and putting a fresh coat of paint on something that is largely unchanged?
For the OS, not sure. Certainly security changes occur and there are lots of communication, storage, etc. protocols and improvements that take place.
For the database? I think lots of what most of us need are in SQL Server 2000. Prior versions had some scale issues, but from 2000 on, I think most systems could run with the functionality here. Some nice language changes in 2005 and 2012, and certainly platform changes in 2017, but most of the things I need a database to do work in 2000. We only moved forward for support reasons, and even then, very begrudingly.
I tend to think works is the most important feature. Second to that would be other stuff, but if it's still working now, why change?
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply