Suggestion: using more modern build toolchain

Discussions related to using the OSE version of VirtualBox.
Post Reply
Yurii Cherkasov
Posts: 10
Joined: 9. Mar 2018, 08:14

Suggestion: using more modern build toolchain

Post by Yurii Cherkasov »

Is there any plan to migrate build toolchain from kBuild to more modern, like CMake? If so, our team is willing to participate, but we definitely need more participants, as well as developers with deep VirtualBox expertise
socratis
Site Moderator
Posts: 27330
Joined: 22. Oct 2010, 11:03
Primary OS: Mac OS X other
VBox Version: PUEL
Guest OSses: Win(*>98), Linux*, OSX>10.5
Location: Greece

Re: Suggestion: using more modern build toolchain

Post by socratis »

What would the advantages of such a migration be?
Any estimate of what it would take (roughly)?
What problems would it solve?
Why? Just because?
Are you familiar with the principle "If it ain't broken, don't fix it!"?

Moving to the "OSE" from the "Suggestions" section.
Do NOT send me Personal Messages (PMs) for troubleshooting, they are simply deleted.
Do NOT reply with the "QUOTE" button, please use the "POST REPLY", at the bottom of the form.
If you obfuscate any information requested, I will obfuscate my response. These are virtual UUIDs, not real ones.
michaln
Oracle Corporation
Posts: 2973
Joined: 19. Dec 2007, 15:45
Primary OS: MS Windows 7
VBox Version: PUEL
Guest OSses: Any and all
Contact:

Re: Suggestion: using more modern build toolchain

Post by michaln »

There is no such plan. If you want to build VirtualBox using something else, you're more than welcome to do so.
michaln
Oracle Corporation
Posts: 2973
Joined: 19. Dec 2007, 15:45
Primary OS: MS Windows 7
VBox Version: PUEL
Guest OSses: Any and all
Contact:

Re: Suggestion: using more modern build toolchain

Post by michaln »

And just to be clear, "modern" is an utterly worthless descriptor in this context. The only thing that matters is "does it work"?
Yurii Cherkasov
Posts: 10
Joined: 9. Mar 2018, 08:14

Re: Suggestion: using more modern build toolchain

Post by Yurii Cherkasov »

socratis wrote:What would the advantages of such a migration be?
Any estimate of what it would take (roughly)?
What problems would it solve?
Why? Just because?
Are you familiar with the principle "If it ain't broken, don't fix it!"?

Moving to the "OSE" from the "Suggestions" section.
Not "just because", I have enough experience to weight carefully all pro and contra of any migration. Sometimes migration is necessary to keep project alive, especially in Open Source community. I know a very good OpenSource developer, who was really interested to participate, but changed his decision with the only reason of out-of-date toolchain.
Yeah, it works right now, but some day you may wake up and realize that last developer who had expertise with your build toolchain is retired. Even your documentation states that unfortunately there are "one and a half" developer in the world who knows kmk internals with the expert level.

Choosing tools, including tools for migration, I take in account two factors - minor factor, is the technology "cool", and major - how many developers in the world familiar with that..
I do not insist on my position as the only genuine, but notice, CMake is de-facto the most popular build toolchain in the world. You declare goals to attract more Open Source developers, I'm just giving an idea how to make your project even more attractive.
Yurii Cherkasov
Posts: 10
Joined: 9. Mar 2018, 08:14

Re: Suggestion: using more modern build toolchain

Post by Yurii Cherkasov »

michaln wrote:And just to be clear, "modern" is an utterly worthless descriptor in this context. The only thing that matters is "does it work"?
It can work being written in COBOL. Do you think you can attract. many developers?
Yeah, I will try at least. If we'd be lucky enough to receive some significant result, I'll be happy to present it to the community
Whiskeyjack
Oracle Corporation
Posts: 122
Joined: 28. Aug 2017, 13:14

Re: Suggestion: using more modern build toolchain

Post by Whiskeyjack »

Personally I will be more than happy to see VirtualBox being built by CMake. That said (as fas as I know) the development team has no plans switching the built system which they have been using and improving for more than a decade. There are more pressing and critical tasks waiting attention. Also I am sure when you guys have specific questions during CMake (or whatever) migration the community will be very happy to help. But please don't ask vague questions which means at the end "I dont want do it myself, please do it for me".
socratis
Site Moderator
Posts: 27330
Joined: 22. Oct 2010, 11:03
Primary OS: Mac OS X other
VBox Version: PUEL
Guest OSses: Win(*>98), Linux*, OSX>10.5
Location: Greece

Re: Suggestion: using more modern build toolchain

Post by socratis »

I'm not a heavy duty developer, but I don't think I would care if it was built manually by a series of drone bees that were writing assembly with their dance. I couldn't care less. That's not where the development of VirtualBox is happening! You're not going to fix a HostOnly problem, or a snapshot issue in Cmake or kmk.

I believe that as a developer that you are, you're focusing on the wrong problem to solve...
Do NOT send me Personal Messages (PMs) for troubleshooting, they are simply deleted.
Do NOT reply with the "QUOTE" button, please use the "POST REPLY", at the bottom of the form.
If you obfuscate any information requested, I will obfuscate my response. These are virtual UUIDs, not real ones.
mpack
Site Moderator
Posts: 39156
Joined: 4. Sep 2008, 17:09
Primary OS: MS Windows 10
VBox Version: PUEL
Guest OSses: Mostly XP

Re: Suggestion: using more modern build toolchain

Post by mpack »

Yurii Cherkasov wrote:It can work being written in COBOL. Do you think you can attract. many developers?
I think you are operating under a mistaken assumption. I don't believe Oracle is trying to attract developers. It's an open source project yes, but not a community development project. The developers are all employees of Oracle, hired for the tool experience they need (or ability to be trained in those tools). So they have procedures which already work, you'd have to explain why it would benefit them commercially to make a big change.
michaln
Oracle Corporation
Posts: 2973
Joined: 19. Dec 2007, 15:45
Primary OS: MS Windows 7
VBox Version: PUEL
Guest OSses: Any and all
Contact:

Re: Suggestion: using more modern build toolchain

Post by michaln »

The problem is that this is completely backwards. "You" (not talking about anyone in particular) want someone else (Oracle) to do something that benefits you, something that Oracle would have to invest significant time and effort (e.g. money) into. At the end of the exercise, Oracle might end up with something better, or something worse, than what Oracle already has. What are the chances of that happening? Anyone should be able to answer that.
Post Reply