Port znuny from perl 2 java or other modern language?

[hkais]

Hi Znuny-developers,

I am just wondering, if you face the same issues as we. We are having troubles to find younger perl developers.
So this is a high indicator of a dying programming language. Also discussing with junior developers, they dislike such old languages.

So do you have any ideas to port Znuny to a modern language?
E.g. Java, which is also "old" but still very relevant.
Or any other modern language?

best wishes

[Johannes]

Hi,

I doubt that someone will just fork it to a new language.
The costs are so immense, that it would not make any sense to do so.
From the business side of things.

From our experience, there are still a lot of perl devs out there, also younger ones.
In fact 3 of of our devs are < ~30

regards

[hkais]

Hi Johannes,

thank you for your feedback.
Especially the feedback about your juniors.

IMHO Perl is dying
Anyway IMHO perl is on the dying path as language. The questions is how strong the degradation of human capital is going on.
So in the last 10 years the perl server side reduced from 0.8% to 0.1% and if counting relative 87,5% less servers on which to develop. In my experience it is highly related to the "sexiness" as programming language to new developers. Combined with the fact that Perl developers are rare, it is a bad situation for considering on the long run. I have had similar situation with COBOL. The experienced developers went into pension, and left a big system on a language, which no-one (youngsters) wanted to modify anymore. Now they are hiring the elder pensioner back, to support the youngsters to re-implement in more sexy languages... This may or may not be the same case with perl, but all indicators are showing the same path...

Here some facts: https://w3techs.com/technologies/histor ... guage/ms/y
If you extrapolate the degradation of perl, there is a halfening of the servers every 3 years. So only a matter of time, except perl gets sexy again...

Other developers with same thoughts?
Also about the effort, you are right that it might cost tons of of time. And I did not expect that Znuny GmbH will be happy about it.
But there are more Znuny users with development background outside.
To all developers, what do you think, would it be a good idea to discuss about porting Znuny to a new and more modern language and thus platform?

[zzz]

Hey,

It's true that Perl is not a very popular language, but if you're a programmer (especially familiar with other scripting languages), it's very easy to learn and also fun to use. It's very straightforward, and in my opinion, it's more important to find someone who is flexible with new technologies than finding a person who is good at one specific technology.

I'd agree with Johannes that the cost of a migration would be too high. It's already challenging to modify core modules without breaking anything. Porting Znuny would require extensive testing to ensure the functions work the same way. At that point, you might as well create a new app from scratch and eliminate some old structural decisions.

Besides that, I don't think Java would a good choice anyway

A native connector to enable creating new modules in other languages, like JavaScript/Node.js, could be a more realistic approach. Similar to C++ Add-ons in Node.js or what OTRS did in versions 7/8. But that's a decision for the core team to make.

Best
Emin

[hkais]

Besides that, I don't think Java would a good choice anyway

if you start from the work hypothesis, the project is funded, so time and money would not be relevant.

What would you choose as approach?
- What language?
- What would you first change on Znuny from the software architecture perspective?
- what features would you like to have which are missing from your perspective?

[Johannes]

Hi,

as far as I can tell you either underestimate the cost of running „the project“ or you just want to abandon the current work for maintenance and features. You can not just simply outsource everything to a low cost country and hope for the best.
It would cost easily three to four times more to migrate everything to $language, then it costs to just run everything. So “is funded” is one thing… and a pretty bold statement tbh.
This is also nothing for “the community”.
Or you might just want to kill the project(starting with question of the language itself), then it would be perfect.
You need a strict project management, teams for the different parts of the project, people who really know what to deliver - which also need a brief experience in the “old” code and product. A lot of design and product questions to be handled. The migration process and so on.
And last but not least, money. A lot of money. Because you need accountability for and from those teams. If a “hobby” or “professional” open source maintainer simply has no time anymore or is bored… you as project loose.

I really don’t see a point for discussion like this here. If you want you can try it on your own or just send a business proposal to the ones you see as relevant. You can move on in this thread, but I will drop out.

One last thing: “Perl is dying” is a sentence I heard the first in the early 2000. Yet the language is still there and you would be surprised where it is used in your all day usage. Just because a language is not in the top x of a list, we will not drop it for no reason. It is not even announced EOL and is still distributed with all major distros. The percentage is dropping in this field for a reason. Projects using perl simply died. And often have not been replaced.
There are people still know how to code it and mostly: the users actually do not care. As long as it is reliable, fast and does not need 100xmore resources.

regards
Johannes