My current work environment is Linux and is on an isolated network.
You appear to be using Qt, so I would think a Linux version would not be that big of an effort. I thought I’d throw this out and see if others would like it too.
1 Like
It’s not trivial to port tens of thousands of lines of C++/Qt code to Linux. But it is doable. The main questions are:
- What Linux distribution(s) to support and how? Do we use something like https://flatpak.org/?
- Will it make us any enough extra sales to justify the effort to port the code, test it, write an installer, update the documentation and support a Linux version? In our experience Linux users mostly don’t expect to pay for software.
My one project experience was with a moderately sized app and it was surprisingly easy to port the software but I won’t second guess your situation. You, of course, know your own code and how difficult it would be for you.
Speaking for myself as a contractor, I move between jobs every few years and find that different work environments will use different operating systems. Maybe a significant number of your existing customers are contractors and would sooner or later experience a need for their favorite transforms to work in a Linux environment.
The one-time cost to port the code isn’t too big a deal. But we then have to:
-write an installer
-update the website
-update the documentation
-test the Linux version (every release)
-support the Linux version
Many Linux users are quite adept at using virtual machines for using applications that exist only for other operating systems.
Perhaps making sure that EDT runs under in a clearly described virtual environment in Linux would be a good way out of the dilemma you describe?
Running Easy Data Transform in a VM is one way. Be we haven’t worked with Linux for some years and currently don’t even have a Linux machine set up. So we aren’t very well placed to advise.
Good point, Andreas. I did find VirtualBox on my Linux machine. And I imagine others who would like EDT on Linux might have similar environments.
The only other thing to find out is how locked down the machine is and whether I could actually upload EDT to the VM.
1 Like
Perhaps a VM solution is the sweet spot between developer effort and resources, EDT availability for Linux users, and the unlocking of an additional revenue stream 
1 Like
i’d like to chime in here.
I see a potential by using batchprocessing in a linux environment, to do datawrangling in a linux server environment. (My usecase).
It will be easier to run a tomcat and a SQL database on linux than on Windows Server. So linux version would appeal to me without having to use a virtual machine like virtualbox etc.
A company i know of is currently moving every server they run to linux, and also going the way from MSSQL to Postgres, so i would assume, their clients would make good use of EDT as well.
I know that Linuxusers are going to “save” on license cost, but it would make life easier. Given, that you already provide an MacOS version, which also has a linux / unix base, i thought, it would be not that big deal.
On the other hand, if you provide a linux version for RedHat and Debian packaging, the next one claims: Hey, what about my SUSE or other distribution, nagging and sulking because the weird distribution he uses is not being supported.
They are very demanding, and some of them don’t get the idea, that a company needs to generate revenue to exist.
Of course, Mozilla does not charge for Firefox, but they are being heavily sponsored otherwise they would need to make money too.
Bottom line: You might open Pandora’s box, but i also think, that there is potential within reason.
I was searching the web literally for weeks for some solution, which does exactly what EDT does, and was glad, running over it.
I also recommend it anytime, when i run into a customer, who might need it and i usually get praise back for EDT ‘lifesaving’, ‘if i only knew about it’ and the like.
The latter one is the main problem, i think. 
We might be able to port the code across to Linux is a day or two. The documentation and website might take another day or two. That isn’t too much of a big deal. It is all the ongoing costs of testing and support that are the big deal.
All the different Linux distributions are an additional headache.
We might port to Linux at some point. I’m not ruling it out. But right now we don’t even have a Linux computer to develop on.
it would be cool, if you would provide a cli version of easydatatransform for linux, perhaps as a docker container, which could do the trick?
Then i could run my transforms scriptbased under linux.
@xeokydo Where would you expect to run this, on Windows, Mac or Linux?
In my case Debian / Ubuntu
If it is a Linux version, why would you need it in a Docker container to run it on Linux?
So edt can run under every linux flavour, if it is dockerized with alpine linux, having also a small footprint, i thought. But i might be mistaken.
How would you currently run it under debian, so i could transform files? have some really big ones and to only select the data i need before transmitting, i could possibly reduce the load by 90 % and not only save time on transfer but also being able to offload the selection process too.
Have you tried running the Windows version of Easy Data Transform on Linux using something like WINE?
Well, bringing this one up again
I talked to some guys from IT in the meanwhile. They said, that it would take quite some time to setup everything to work with WINE but they won’t invest in the effort, just because there is one softwaresolution they might need to battle with wine.
I also see a large move from ms-sql and oracle to postgres; outnumbering mysql installations in the meanwhile, running under linux. Of course, you can run ms-sql and oracle under linux as well, but given various restrictions in performance and the like, i see software developers move away.
Also, on the other hand, linux server has a smaller OS footprint in terms of ressources and on the same environment is more powerful in terms of performance, which would also something EDT would benefit from if you have large datasets to deal with. And in the end, automating stuff would also work better native than doing it over WINE.
They would like to have the dockerized version, this would make quite a lot of sense and a lot less hassle for your to provide EDT for different linux distributions.
So if you are still looking into it…
I am pretty sure, to know of a couple of companies being in need of EDT under Linux…
Best
While considerations for a production environment will be greater than the ease with which WINE is normally installed and used, I wonder whether a Linux version is the instant alternative. If EDT’s use under WINE is demonstrated by one customer, then spread to more, then a business case can be built to go native Linux.
I am not talking about just one customer. In some cases the customer doesn’t want to utiliize WINE on a Linux server, but since a lot of the servers i am working on are linux based (debian / ubuntu), i’d like a more native approach, and doing that one with docker would make it way easier, than settting WINE up on each and every server. Most of the servers i work on under linux are already dockerized, so another docker image wouldn’t create much work / hassle. We are also talking about replication of the setups across different cloud providers and hosters; which, in some cases, would rule out WINE anyway. It’s ok to use that one locally, if you are going that way, but i don’t see it practical for my specific need. Yes, you can do a lot with WINE, no doubt, but my approach is, to keep it as simple as possible without too much dependencies i have to look into and to maintain that ones as well.
This wouldn’'t be a problem with a docker image, i assume.
And it is very hard to convince IT administrators to install WINE on their precious servers, just to run EDT.
I tried…
Best
I understand WINE isn’t suitable in all cases. A Linux version remains a possibility. But it isn’t clear it would bring in many extra sales, so it isn’t a priority at present.