Citrix are pretty ubiquitous in the corporate space. Not as big a deal as they were ten or 15 years ago, but still a huge company that powers a lot of businesses.
Back in the corporate Windows desktop space of the early 2000s, the Citrix product was almost like magic. The problem it aimed to solve was reduction in cost and effort to deliver applications to users. Instead of managing a fleet of 'expensive' desktop computers, you install the desktop applications on a Citrix server and 'publish' them over a remote desktop protocol. The potential was huge! It was a real challenge to do this at the time. Citrix made it relatively easy to turn any old desktop PC or cheap terminal into a powerful workstation with the resources of a large server. The effect was extremely exciting at the time. A user was assigned a bunch of applications, so when they lpgged into the Citrix client they'd be able to run those applications no matter what device they were on. They could disconnect from running applications and reconnect from another device. There was the added bonus that it worked pretty well over slow network connections (some screen lag, but all other operations at the backend were lightning fast since the application ran in the date center on a beefy server). However, it was expensive. Either you had to have very deep pockets, or you need to dip into you desktop PC budget, fulfilling new hardware requests with existing refurbished PCs or new thin clients (at a lower cost). I never met an IT manager that was willing to do that. If it didn't work out, they'd be stuck with useless thin clients and the prospect of an expensive desktop refresh along with a backlog of applications to repackage and deploy via other means (Windows SMS Server, Group Policy etc). Also, very few people were doing remote work back then.
Citrix tried to pivot into other spaces like virtualization (Xen) but that market was flooded with alternatives, and most companies who used Xen were happy with the free version.
That's something else that had a tendency to kill a Citrix deal. But implementation details are important. In my experience, managers who worked from home loved it because they could get the complete office desktop experience at home. There'd be some visual lag for sure, but visual latency aside, file and print operations were snappy and reliable. Trying to open an Excel spreadsheet over a dial-up network drive took ages. Print jobs were worse. Most installations I saw were POCs that got used as a remote work solution for managers. Any time they decided to deploy an app to all office staff the thing would predictably fall over, because it wasn't sized for that.
My anecdote is that it works much better than any X Windows setup that I have used, and since around 2011 it is either Citrix or RDP on most of our projects.
If we get projects with customers that will rather hand us low quality Dells with 4 / 8 GB with HDD, to access their infrastructure, I will always advocate for a Citrix or RDP based setup instead, due to how bad those laptops tend to be.
Can such a setup have a lossless presentation of the GUI? Personally I really dislike using a modern computer via a lossy video signal, just as a matter of principle. Same reason I'm not interested in cloud gaming even if the latency is great.
Citrix originally licensed the technology for the Remote Desktop Protocol (RDP) to Microsoft.
RDP was a pared back version of Citrix, and the license terms left Citrix free to sell a with a more fully-featured virtualised desktop product running on top of Windows Server.
Wikipedia says "Remote Desktop Protocol (RDP) is a proprietary protocol developed by Microsoft" and that "[the] Terminal Services Edition of NT 4.0 relied on Citrix's MultiWin technology, previously provided as a part of Citrix WinFrame atop Windows NT 3.51, in order to support multiple users and login sessions simultaneously". Based on this it sounds to me like the protocol itself was the work of Microsoft and that Citrix built an extension to the server that multiplexed it. It also later remarks that "[the] T.128 application sharing technology was acquired by Microsoft from UK software developer Data Connection Limited", so it would seem that the protocol itself is a series of extensions developed on top of an open standard that they got via acquisition.
Years ago I worked in defense contracting. The US Department of Defense uses Citrix (which is easily Google-able; don't come at me about "OpSec") and my experience with it was always abysmal compared to RDP. I have a funny feeling that even with Microsoft in the mix it'd be cheaper and better for everybody involved to just pay M$ for the RDP CALs and ditch the dinosaur that is Citrix altogether.
My understanding is that back in the early 2000s the Citrix ICA and Microsoft RDP protocols were nearly identical. Common code and all that.
One of the reasons companies licensed Citrix is because Citrix kept investing in ICA and it was about 2-3 times as bandwidth efficient. In the era of 64 Kbps ISDN WAN links this was critical.
Microsoft did something around 2008 and now RDP seems better overall. It can handle 4K at 30 fps without difficulty.
Meanwhile Citrix did weird random things to ICA that made it markedly worse.
It says a lot that all of the Citrix engineers I’ve worked with (including myself) prefer to RDP into a Citrix server instead of using its native ICA protocol!
Other deliberate breakage was Citrix deprecating SSL support directly on the session hosts. They did this to force customers to buy their overpriced Netscaler / ADC appliances. These slow down connections and can’t handle many 4K streams.
Generally they seem to have become an acquisition-driven company instead of an engineering-led one.
Don't forget that Citrix had a Terminal Server offering before Microsoft incorporated it into Terminal Server. That was the big offer in my mind, not the RDP client, but the ability to connect multiple end users to one Windows NT server back in NT 3.51/4's early days.
I worked at Citrix from 1996 thru 2000. In the early days, Citrix was licensing Windows NT (source code) from Microsoft and selling a modified version of it such that (a) one could remotely control/display Windows apps, and (b) multiple users could do this simultaneously. At a certain point (‘98?) Microsoft announced they would soon offer the same functionality as Citrix, and the stock got crushed. In the end, though, it seemed Microsoft decided the practical thing to do would be to roll their own version of “(a)” (dubbed “RDP”) and license “(b)” (back, as it were) from Citrix. At least I think it was something like that.
I have managed large Citrix farms and if you have the bandwith RDP blows it away. Lot less options and control but with decent bandwith and high settings RDP is pretty nice.
That being said I hear VMware Horizon blows away Citrix too
It took an act of God to get the Horizon client working last I used it (weirdness with client SSL libraries... on Windows no less) but it worked pretty well once I got it set up.
Citrix is typically used over WAN or internet and RDP over fatter pipes. Citrix gives a more complete desktop experience using less bandwidth. Not typical concerns where RDP is dominant.
Oh wow I had no idea about this. Always thought that RDP was something MS developed and Citrix was a more of an RDP alternative with better centralised management features, OEM integration, etc.
Spot on with Citrix not being as big a deal as they once were. 10 years ago any reasonable sized desktop app deployment went through Citrix, now it's migrating to web apps. I struggle to see the long term value proposition in Citrix but my experience is limited to finance and pharma.
There's very good money to be made in consulting if you "know" Lotus Notes. Even a lot more if you know how to migrate the thing (presumably to MS Exchange).
Citrix tried to pivot into other spaces like virtualization (Xen) but that market was flooded with alternatives, and most companies who used Xen were happy with the free version.