Should Manager Desktop run in browser instead of standalone application

1

I have always been a huge fan of self hosted server or local install programs. The thing that I dislike most about the cloud is your reliance on your Internet, the program’s host server, host internet as well as whether an update on the cloud has borked up the program!

Having said that, I am idly wondering whether it makes sense for Manager to support Cloud, Server, Windows Desktop, Mac Desktop and Linux Desktop. There are clearly problems using the program with Webview 2 and certain things no longer work without Webview 2 anymore.

In addition, the crashing problems on Macs - which I am aware of, but not actively following - leads me to wonder if perhaps the developer needs to review the viability of supporting so many different operating systems. I fully agree with his initial viewpoint that Manager needs to stop supporting older operating systems like the Mac version that doesn’t seem to work properly anymore and is crashing Manager all the time. It is simply not cost effective to continue supporting old operating systems for years on end.

I am ending support for Windows 7, Server 2008, Server 2012 this year on the basis that that it is no longer viable to provide IT Support for them. I think it is perfectly for programs like Manager to stop supporting older operating systems.

However Manager needs to inform users a year in advance before retiring support for an Operating system. You can’t just cut users off.

I personally think that the time to retire the Desktop Version has come. With the problems of Webview on Windows, crashing issues on older Macs and the increasing need for businesses to be able to access data from any location on any device, a standalone solution even for a small company like myself is increasingly unviable.

Yes I am sure that many users are happy with the desktop version - I was very happy using the desktop for several years before moving to my self hosted server version. But I feel that too much effort is being expended on supporting too many different platforms for the desktop version. A cloud/Server solution would make far more sense. You don’t need webview for the server version even though the Server is on a Windows Server machine.

Anyway, this is just food for thought for the developer. I was very happy with the desktop version, am now very happy with the Server version, but even I (as an anti cloud person) can see that the desktop version is actually becoming more trouble than it’s worth.

3 Likes
2

No plans to phase out desktop edition. I don’t think it’s more trouble than it’s worth. For a lot of people, desktop edition works just fine. You just don’t get them posting topics about it.

WebView2 will be compulsory - that will resolve issues on Windows. And Mac issues should be resolved within a few days. The third-party developer has fixed the issue as far as I can see so I’m just waiting for the release.

6 Likes
3

To be clear, when I say more trouble than it’s worth, I am referring to the cost of development time to support three different operating systems and the need to support end user’s mono/.net/webview2 etc and all the related end user computer problems. You might just be spreading yourself too thin. Cloud on the other hand is trouble free. All the end user needs is a working internet and browser.

Yes I agree webview needs to be made compulsory as there a lot of issues with that. Good to hear that the Mac issues will be sorted soon. However, I would still consider ending support for older operating systems.

4

It is a bit more complex but not something I would be uncomfortable with. Also, Mono/.NET dependency has been removed last year. As a result, there are no issues due to installation problems with Mono or .NET anymore.

You will see once WebView2 is made compulsory and Mac issue is fixed, there will be again error-free experience for the vast majority of desktop users.

1 Like
5

Just bear in minds problems like this where the user could not install Webview. He ended up uninstalling webview because he couldn’t get Manager to run.

6

The force of Manager is that it runs on different OS and in different versions.

1 Like
7

If Manager was only available in a web version, I would have never considered using it.

7 Likes
8

That’s true. Me too, I have tested it out for the better part of three months on Desktop versions, however, they require a lot of work just to maintain, let alone introduce any improvements.

It’s a tricky situation when you consider you have to have a desktop version for people to try out as well as to keep the downgrade-anytime and the use-offline offerings which are very essential to manager’s success.

Idk why manager has to have it’s own “app” for desktop versions since you already can open a new instance in the browser and view the full page source. Naturally you’d assume that there are all the security measures in place.

Maybe if those desktop “apps” are reduced to applets that provide source protection for the engine and help users configure and locate their data files, while all the remaining features run on instances of a web browser, then that should help.

Also, drop any utilities that relies on servers in case of cloud and server edition, because this would save development time for desktop OS, such utilities maybe (and I am just guessing here):

  • PDF button is gone, user can always print.
  • email button code can be included with the source.
  • email attachment is a printed temporary file in a temp directory.
  • translations, gone
  • …

You get the idea, just keep the OS development to the minimum required for maintaining full compatibility with cloud and server version, everything else: either it runs in the browser or it’s out the window.

9

Manager is principally coded as a web server which provides the accounting functionality. That includes printing, email, PDF generation, language translation.

  • The own server version of Manager makes this code available to Manager customers

  • The NG Software cloud version puts that code an a network of servers with account login, access control, payment, backup and server update tools.

  • The desktop version runs the same code in the background (which is why desktop users can open a Manager link in a browser of there preference). The desktop version adds an installation application. It also adds a wrapper so when the user starts Manager they actually start two programs; the Manager server, and a web browser to render the user interface.

As a result once the installation type specific code is stable (which probably mostly uses third party tools) then updates the the core Manager code applies to all versions, as it is actually the same code.

10

Maybe desktop users should consider and adopt the process used by those running the server version… You only update the Manager product once you are confident by reading the forum confirming that there are no pressing issues that will impact your arrangement by upgrading to the new current version. We wait 2 to 3 months to find a good entry point where things in Manager are going smoothly and offering changes that are in fact meaningful.

11

I agree with you. IMHO The issue is not desktop vs server or mac vs windows. @lubos already told us that it doesn’t take time to keep all those “lines open”.

I think that the real issue is the fact that Manager is developed in a way that is like you are always installing a nightly version. It would be so much better to have some releases classified as stable and keep the line of nightly releases like we have today.

But this would imply to have a team of testers to establish which version should be considered stable.

3 Likes
12

Updating, entering hundreds of transactions, then finding something you rely on has been ‘broken’ is the big PIA. Especially when you can’t roll back to the last version you were happy with.
A stable version and the ability to roll back without losing day’s/weeks/months of transactions would be great.
As it stands every time you update you have to go through every type of transaction, report etc. to make sure something you rely on to run your business books isn’t broken.

13

The irony of my post is that I fully agree with you. I probably would not have started with Manager if they only had a cloud version. However, the existence of the server version would probably have tempted me anyway.

14

I fully agree with the complaints about a lack of a stable/development version.

In point of fact, I created a post specifically on that!

I actually think that you have hit on the solution! The issue of cloud/server/desktop is a bit of a red herring. The real issue is the need to add extra functionality for the desktop version to get it to work. When I install the Server - all I am really doing is extracting some files and creating a service to start the service. I don’t need to install webviewer etc. Why is this necessary for the desktop version?

@lubos would you not consider a solution that works similar to the server install for desktop users. You don’t need to “install” the program, nor worry about dependencies such as .net and webview. You have all the benefits of using the browser - eg open in new tab etc! The only thing that needs to be addressed is the install of a service like the server service.

I would highly recommend that the server version be tweaked a bit to make it easy for end users to install (the service and extract the files) and web shortcut to open in a browser and then have one product which works on cloud, server and desktop computers!

So don’t abandon the desktop version - rather merge it into the server version concept.

15

The Manager desktop version add things to decrease the barrier of entry for new users.
Maybe effort well spent for NG Software. Would have to know how much harder it would be for new users compared how must maintenance effort NG Software could save to know.

16

I think that Lubos has already answered the question. I know that the last month has been a little frustrating for many bugs but we were told that there were many changes in the engine underneath Manager.

17

Well all the server version actually is the following.

  1. You extract the files to a folder.
  2. You install another 3rd party program to create the Manager service and to start it on boot.
  3. You use something like IIS to run the webserver
  4. You open a web browser and type in the address of the Manager server.

All of the above could be incorporated into the desktop version and done automatically for the end user. This is not that difficult. Essentially you would be installing the server backend on the desktop computer and access Manager via a browser. This would reduce effort worrying about webview, Linux, Mac and Windows issues as it would all be browser based.

I think it’s doable.

18

The Free Desktop Edition is what attracted many of us to join this family. Some of the cloud users of today started on the desktop edition. Others also downgrade to the desktop edition because of financial challenges.

The Desktop Edition is the reason why people are leaving Popular accounting software for Manager.

4 Likes
19

I agree with you. I do understand the appeal of the desktop version. I myself was drawn to Manager for this very reason.

I have already suggested another direction in my above posts. I believe that abandoning the desktop is not the solution, but rather merging the desktop and server installation into one, but in a very user friendly way. This eliminates 3rd party requirements like webview etc. The Server version does not need this.

I believe the real reason why people are leaving other accounting programs and coming to Manager is partially cost (of the other providers), but also the lack of support and the bloat in other programs is driving people away. Some accounting programs you need a 2 day training course just to learn how to use the program never mind use if optimally! Manager is gaining users because of it’s simplicity.

20

Few of the desktop users would have the expertise to run a server version.

It is a little bit more complex to run than what you indicated above even if it is wrapped up in a package.