Remote desktop vs. browser based web apps. (3 Viewers)

BlueSpruce

Active member
Local time
Today, 00:54
Joined
Jul 18, 2025
Messages
387
I have a prospective user who thinks RDP is just as slow, or slower, than browser based remote access applications, such as TeamViewer, and LogMeIn. I explained that RDP is generally faster than those web based apps, unless there's an issue with the internet connection. RDP sends much less data to render a screen than a web browser. The beauty of RDP is that if you loose the connection, the screen will be exactlty were you were left off whe you re-connect and nothing will be corrupted. I will setup an RDP dog and pony demo so the prospect can judge for himself.

The prospect also wants remote users to locally use devices to capture and save images, scan barcodes, and print labels. I explained that web based apps have a very hard time accessing hardware devices, such as cameras, scanners, and the local file system because web apps are stateless and unbound. Hardware is generally isolated in browser based applications for security reasons. A web browser, for example, cannot automatically grab files from your computer and upload them, unless a user manually selects file(s) to upload, or a user installs a dangerous addin on the browser.

The prospective user also thinks standard consumer PC's can be used as web servers. Desktop Windows editions only allows a max of 10 external connections. A server OS, such as Windows Server, and several Linux flavors don't have that limitation. Desktop PC's don't run IIS, Apache, or nginx web servers. IIS tends to be pre-installed with Windows Server. Another is support for Terminal services (the remote virtual desktop VM system). This can’t run on a regular PC. The TS feature allows 15 users on the road using the server remote and each user gets their “own” vitual desktop, so terminal services includes a complete remote solution, one that handles multiple users at the same time.

Are my statements accurate?... Can you add anything I may have left out?... Thanks!
 
Last edited:
"The prospect also wants remote users to use portable devices to capture and save images, scan barcodes, and print labels..." PowerApps does that well.

See the videos in this series on YouTube, such as this one on scanning barcodes for Northwind Developer.

I also take photos of book covers and upload them to Azure Blob Storage using a PowerApps app.

Printing labels. Not so much.
 
If TS stands for Terminal Server, why only 15 users? The number of users depen on the number of licenses you buy.
That's correct. The prospect wants to start out with 15 users. I am leaning towards proposing TS RemoteApp so the Access FE's transparently appear and behave as if they're locally installed on each user's remote workstation. That way they can also leverage devices that are connected to their workstations.
 
"The prospect also wants remote users to use portable devices to capture and save images, scan barcodes, and print labels..." PowerApps does that well.

See the videos in this series on YouTube, such as this one on scanning barcodes for Northwind Developer.

I also take photos of book covers and upload them to Azure Blob Storage using a PowerApps app.

Printing labels. Not so much.
Thanks, but I will stick with what's already proven to work and are mature products. The prospect already has a mature Access app and using RDP eliminates the need to rewrite or modify his existing Access app.
 
I don't know what's the deal with so many people here claiming web = slow, it's not 1995. No idea what sort of web app you've made that was slow, but they are not inherently slow.
 
I don't know what's the deal with so many people here claiming web = slow, it's not 1995. No idea what sort of web app you've made that was slow, but they are not inherently slow.
Microsoft's RDP (Remote Desktop Protocol) is faster than browser based remote desktop apps, like TeamViewer, RemotePC, AnyDesk, for accessing and controlling a remote computer.
 
The only time an RDP is slow is if you are putting too much through it - i.e. the exact problem that every OTHER network connection has. BUT if you have a bunch of people banging away via RDP such that 15 users are on the system at once, that system will be 15 times slower than a single-user RDP setup. Just remember that when comparing apples to apples, the NUMBER of apples counts for something.
 
The only time an RDP is slow is if you are putting too much through it - i.e. the exact problem that every OTHER network connection has. BUT if you have a bunch of people banging away via RDP such that 15 users are on the system at once, that system will be 15 times slower than a single-user RDP setup. Just remember that when comparing apples to apples, the NUMBER of apples counts for something.
True, but a decently configured server running 15 concurrent TS sessions will provide good performance.
 
True, but a decently configured server running 15 concurrent TS sessions will provide good performance.

Yes... if the DB and user directory structure on that server are ALSO well-design. Not trying to bust your chops here. Just pointing out where you might have problems. The most common failing I have seen in the past is for the server's admin to not understand that isolation of users - and particularly, user FE files - is crucial.
 
Classic Outlook only downloads messages and their attachments. Web and New Outlook will have the same number of bits to download and display on your computer. Classic Outlook might have a problem if you have an address book the size we had for the NMCI network, measured in thousands of addressees who are members of the U.S. Navy or USN Reserves, five divisions of same, and literally several hundreds or even a couple of thousand departments. The routing tables get a little touchy when they get that big.
 
Are you anticipating the 15 users logged into the Terminal Server concurrently during the work day? Or will their use be more sporadic?
What is the anticipated growth of users - over 6 months - 1 year - 2 years - 3 years?
 
Are you anticipating the 15 users logged into the Terminal Server concurrently during the work day? Or will their use be more sporadic?
What is the anticipated growth of users - over 6 months - 1 year - 2 years - 3 years?
Yes, this prospect receives large volume freight shipments that require special handling, aka "white glove", and has a fleet of 50 trucks that delivers on a daily basis. Each item received must be inspected, photographed, barcoded, warehoused for later delivery, or immediately loaded onto specific trucks, delivered, and photographed again in case customers file damage claims.
 
Yes, this prospect receives large volume freight shipments that require special handling, aka "white glove", and has a fleet of 50 trucks that delivers on a daily basis. Each item received must be inspected, photographed, barcoded, warehoused for later delivery, or immediately loaded onto specific trucks, delivered, and photographed again in case customers file damage claims.
And estimated growth in number of users using the system?
 
And estimated growth in number of users using the system?
Growth undetermined at this time. They're currently at full capacity. They cannot take on more customers or increased shipments unless they expand warehouse space, hire more workers, and lease more trucks.
 
Last edited:
Growth undetermined at this time. They're currently at full capacity. They cannot take on more customers or increased shipments unless they expand warehouse space, hire more workers, and lease more trucks.
OK - story time from an old guy...

Back in the early 1990's (when I started with Access) remote access to Access applications was not demanded by my customers (and not really viable either). If my memory serves, it was also in the early 1990's that Microsoft first developed RDP (which was initially intended for remote administration of Windows Server). As the 1990's wore on and the new century was ushered in, "always-on" internet connections opened the door to a host of new remote control technologies. I have experimented with them all and can tell you that to this day - none of them comes close to the performance of RDP. If your customer has a preconceived notion that RDP is slow it's not technologically based.

It wasn't until 2001 that I had the first situation where I had a customer who opened a branch office overseas and needed to give an initial 6 employees access to their Access application. They didn't have a firm sense of user growth because the success of the new branch office was uncertain. I simply decided to place 6 workstations in a rack in the server room and have each employee RDP into their own workstation. It worked flawlessly - no issues. As the branch office grew, we added more workstations to the rack. After a year or so, the number of workstations became 12 and at that point, I suggested to the customer that they budget for a Terminal Server. We implemented a properly spec'd Terminal Server that could support up to 24 concurrent users and selected 2 of the best end-users to migrate to it for testing. We had issues to iron out - both with the Access application running in a Terminal Server environment and some issues common to Terminal Server (remote printing was the bane of my existence back then). We ultimately migrated everyone to the Terminal Server and repurposed the old RDP workstations for use in the remote office for future remote users. We later went on to deploy a 2nd Terminal Server to host additional remote users at the branch office as it grew.

Because of my own experience above, a dozen users became sort of my "line in the sand" for deciding on moving from individual workstations to a formal TS environment. Your customer is at 15 and no estimate as to future user growth. Your customer could pick up off-lease HP EliteDesk mini's for about $200 each (so around $3K for 15 of them). I have not spec'd out a new server in a long time - but I would be willing to bet that a decently spec'd Terminal Server (for 20-25 concurrent users) would have to be $10K by the time you take into account Terminal Server licensing. In the end - the decision will be made, in part, by the guy signing the check - it always works that way.

Let me share another thought. When my customer had the dozen workstations in the rack for their remote users - when one of the workstations went down - we had a hot spare ready to go and the user was inconvenienced for an hour while we configured a new profile for them on the new workstation. When the Terminal Server went down (which it did on rare occasion) the customer had the entire remote office sitting around doing nothing while we got it back up.

... old guy story over.
 
OK - story time from an old guy...

Back in the early 1990's (when I started with Access) remote access to Access applications was not demanded by my customers (and not really viable either). If my memory serves, it was also in the early 1990's that Microsoft first developed RDP (which was initially intended for remote administration of Windows Server). As the 1990's wore on and the new century was ushered in, "always-on" internet connections opened the door to a host of new remote control technologies. I have experimented with them all and can tell you that to this day - none of them comes close to the performance of RDP. If your customer has a preconceived notion that RDP is slow it's not technologically based.

It wasn't until 2001 that I had the first situation where I had a customer who opened a branch office overseas and needed to give an initial 6 employees access to their Access application. They didn't have a firm sense of user growth because the success of the new branch office was uncertain. I simply decided to place 6 workstations in a rack in the server room and have each employee RDP into their own workstation. It worked flawlessly - no issues. As the branch office grew, we added more workstations to the rack. After a year or so, the number of workstations became 12 and at that point, I suggested to the customer that they budget for a Terminal Server. We implemented a properly spec'd Terminal Server that could support up to 24 concurrent users and selected 2 of the best end-users to migrate to it for testing. We had issues to iron out - both with the Access application running in a Terminal Server environment and some issues common to Terminal Server (remote printing was the bane of my existence back then). We ultimately migrated everyone to the Terminal Server and repurposed the old RDP workstations for use in the remote office for future remote users. We later went on to deploy a 2nd Terminal Server to host additional remote users at the branch office as it grew.

Because of my own experience above, a dozen users became sort of my "line in the sand" for deciding on moving from individual workstations to a formal TS environment. Your customer is at 15 and no estimate as to future user growth. Your customer could pick up off-lease HP EliteDesk mini's for about $200 each (so around $3K for 15 of them). I have not spec'd out a new server in a long time - but I would be willing to bet that a decently spec'd Terminal Server (for 20-25 concurrent users) would have to be $10K by the time you take into account Terminal Server licensing. In the end - the decision will be made, in part, by the guy signing the check - it always works that way.

Let me share another thought. When my customer had the dozen workstations in the rack for their remote users - when one of the workstations went down - we had a hot spare ready to go and the user was inconvenienced for an hour while we configured a new profile for them on the new workstation. When the Terminal Server went down (which it did on rare occasion) the customer had the entire remote office sitting around doing nothing while we got it back up.

... old guy story over.
RDP is basically streaming video P2P accross the web with a refresh rate of 30 frames per second. Each viewer's workstation has an RDP client installed which communicates with the RDP server on the host. Since my propsect wants to start out with 15 seats, I did the math on the TS CAL's and it's within the prospect's budget. We also put together a server config for under $5K that can scale up if needed. Windows Server 2022 and SQL Server 2022 Standard editions are also within budget. I'm even looking at Wyse 3040 Thin Clients for the workstations, which I can get for ~ $75/ea.
 
Last edited:

Users who are viewing this thread

Back
Top Bottom