Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
sys-auth/polkit with duktape
View unanswered posts
View posts from last 24 hours

Goto page 1, 2  Next  
Reply to topic    Gentoo Forums Forum Index Gentoo Chat
View previous topic :: View next topic  
Author Message
alamahant
Guru
Guru


Joined: 23 Mar 2019
Posts: 550

PostPosted: Fri Sep 18, 2020 9:32 pm    Post subject: sys-auth/polkit with duktape Reply with quote

Can we please have a modification of polkit ebuilds to allow the use of either spidermonkey or duktape?????
If Pentoo can do it I wonder why we cant????
Please our kind maintainers take heed.........
I tried many overlays in my quest to get rid of spidermonkey and the only one that worked was overlay "sorrow".
The others failed at the applying the patches stage
:D


Last edited by alamahant on Sat Sep 19, 2020 9:00 pm; edited 2 times in total
Back to top
View user's profile Send private message
asturm
Developer
Developer


Joined: 05 Apr 2007
Posts: 7958
Location: Austria

PostPosted: Fri Sep 18, 2020 9:45 pm    Post subject: Reply with quote

Get it upstream and then Gentoo can provide the option.
_________________
backend.cpp:92:2: warning: #warning TODO - this error message is about as useful as a cooling unit in the arctic
Back to top
View user's profile Send private message
alamahant
Guru
Guru


Joined: 23 Mar 2019
Posts: 550

PostPosted: Sat Sep 19, 2020 7:36 pm    Post subject: Reply with quote

UPDATE
I found the latest spidermonkeyless polkit in overlay::miramir.
I am now officially free from spidermonkey and......python:2_7
Many thanks to "miramir"

In package.use/polkit i used
Code:

>=sys-auth/polkit-0.118-r1::miramir  duktape

and in package.mask/polkit
Code:

sys-auth/polkit::gentoo

updated and dep-cleaned and ....such a relief!
Maybe the miramir ebuild for polkit should be included in the main repo....
:D
Back to top
View user's profile Send private message
asturm
Developer
Developer


Joined: 05 Apr 2007
Posts: 7958
Location: Austria

PostPosted: Sat Sep 19, 2020 7:41 pm    Post subject: Reply with quote

No.
_________________
backend.cpp:92:2: warning: #warning TODO - this error message is about as useful as a cooling unit in the arctic
Back to top
View user's profile Send private message
ff11
Guru
Guru


Joined: 10 Mar 2014
Posts: 591

PostPosted: Sat Sep 19, 2020 8:05 pm    Post subject: Re: sys-auth/pokit with duktape Reply with quote

alamahant wrote:
...
If Pentoo can do it I wonder why we cant????
... :D

Wow! This one even I had to check. They didn't just add the duktape, they completely removed the spidermonkey.

I worry a little about so many interested people and nobody has time to solve the small issues for the patch to be merged on upstream. But I can't say much, as I'm one of them (free time is scarce this year).
_________________
| Proverbs 26:12 |
| There is more hope for a fool than for a wise man that are wise in his own eyes. |
* AlphaGo - The Movie - Full Documentary
"I want to apologize for being so powerless" - Lee
Back to top
View user's profile Send private message
Marcih
Apprentice
Apprentice


Joined: 19 Feb 2018
Posts: 207

PostPosted: Sun Sep 20, 2020 12:57 pm    Post subject: Reply with quote

astrum,
why not explain, to the posters who may be new and not know how Gentoo works, the reasoning for your answer or provide them with some useful information regarding the question? You know that Gentoo already has a bug filed for this exact thing, you've posted in it, why not provide a link? You're posting from your dev account, you're representing Gentoo, it wouldn't hurt to at least try to be friendly.

alamahant, ff11,
Gentoo follows upstream, developers won't add any patches that deviate from that. There are bugs filed for this already, both in the Gentoo bugzilla and upstream, feel free to read through those to catch up on the situation and discuss further there.
_________________
Bones McCracker wrote:
It wouldn't be so bad, if it didn't suck.

NeddySeagoon wrote:
The problem with leaving is that you can only do it once and it reduces your influence.
Back to top
View user's profile Send private message
ff11
Guru
Guru


Joined: 10 Mar 2014
Posts: 591

PostPosted: Sun Sep 20, 2020 1:12 pm    Post subject: Reply with quote

Marcih wrote:
...
alamahant, ff11,
Gentoo follows upstream, developers won't add any patches that deviate from that. There are bugs filed for this already, both in the Gentoo bugzilla and upstream, feel free to read through those to catch up on the situation and discuss further there.


Thanks friend. I'm already referenced in the Gentoo bug in question (inside bug #698910). And you can see me in the comments of upstream too (the "DanielDevBR @ferion11").

I'm kind of out of time for this, but sometimes taking a look at the code. And it doesn't seem that hard to make the modifications required for the patch to be upstream accepted. As most of it, is only about refactoring the code (add some return value, separate files in .c /.h ...), and the "runaway-killer" is already supported by duktape (using DUK_USE_INTERRUPT_COUNTER and DUK_USE_EXEC_TIMEOUT_CHECK pointing to the function that will be be the callback).
_________________
| Proverbs 26:12 |
| There is more hope for a fool than for a wise man that are wise in his own eyes. |
* AlphaGo - The Movie - Full Documentary
"I want to apologize for being so powerless" - Lee
Back to top
View user's profile Send private message
Marcih
Apprentice
Apprentice


Joined: 19 Feb 2018
Posts: 207

PostPosted: Sun Sep 20, 2020 2:07 pm    Post subject: Reply with quote

ff11 wrote:
And it doesn't seem that hard to make the modifications required for the patch to be upstream accepted. As most of it, is only about refactoring the code (add some return value, separate files in .c /.h ...), and the "runaway-killer" is already supported by duktape (using DUK_USE_INTERRUPT_COUNTER and DUK_USE_EXEC_TIMEOUT_CHECK pointing to the function that will be be the callback).

I haven't tried to look into that so I'm glad to hear it should be straight-forward. Looks like what's holding back the MR is the radio silence from the requestor, which is a shame, I'm looking forward to this getting merged.
_________________
Bones McCracker wrote:
It wouldn't be so bad, if it didn't suck.

NeddySeagoon wrote:
The problem with leaving is that you can only do it once and it reduces your influence.
Back to top
View user's profile Send private message
axl
Veteran
Veteran


Joined: 11 Oct 2002
Posts: 1072
Location: Romania

PostPosted: Sun Sep 20, 2020 2:27 pm    Post subject: Reply with quote

asturm wrote:
No.


I get it that you are fed up with users, but this attitude is not gonna help anyone.

Let me ask another question. Would you guys be willing to consider making a spidermonkey-bin package?
Back to top
View user's profile Send private message
Ionen
Veteran
Veteran


Joined: 06 Dec 2018
Posts: 1247

PostPosted: Sun Sep 20, 2020 2:40 pm    Post subject: Reply with quote

axl wrote:
Let me ask another question. Would you guys be willing to consider making a spidermonkey-bin package?
Does mozilla even provide binary packages for those? Bare a few exceptions the majority of *-bin packages weren't compiled by gentoo maintainers but are just upstream binaries.
Back to top
View user's profile Send private message
axl
Veteran
Veteran


Joined: 11 Oct 2002
Posts: 1072
Location: Romania

PostPosted: Sun Sep 20, 2020 2:45 pm    Post subject: Reply with quote

That's more great news. Great!!! :(

I can't even get that new 78.2.0 compiled. On one machine it complains about undefined reference to "RustMozCrash", on another: "note: distcc[416729] (main) CRITICAL! distcc seems to have invoked itself recursively!". I never ever seen any of these errors before. Well happy digging axl. Happy digging to me.
Back to top
View user's profile Send private message
Ionen
Veteran
Veteran


Joined: 06 Dec 2018
Posts: 1247

PostPosted: Sun Sep 20, 2020 2:58 pm    Post subject: Reply with quote

78.2.0 changed a lot of things around and ebuild is fairly recent, so likely a few things to sort out still (report issues as need be if anything, likely will stay in ~testing for a while). On the good side it does get rid of the py2 requirement but trades it for rust.

It also received some changes a few minutes ago (removed clang dep), but probably too early to sync.
Back to top
View user's profile Send private message
axl
Veteran
Veteran


Joined: 11 Oct 2002
Posts: 1072
Location: Romania

PostPosted: Sun Sep 20, 2020 3:16 pm    Post subject: Reply with quote

There aren't many google results about that error: "undefined reference to RustMozCrash". Turns out it was LTO related. With LTO off, it finally build successfully. Just in case someone else runs into that error, try with LTO off. At least for now.
Back to top
View user's profile Send private message
asturm
Developer
Developer


Joined: 05 Apr 2007
Posts: 7958
Location: Austria

PostPosted: Sun Sep 20, 2020 3:24 pm    Post subject: Reply with quote

Marcih wrote:
why not explain

That was in my first answer.
_________________
backend.cpp:92:2: warning: #warning TODO - this error message is about as useful as a cooling unit in the arctic
Back to top
View user's profile Send private message
alamahant
Guru
Guru


Joined: 23 Mar 2019
Posts: 550

PostPosted: Sun Sep 20, 2020 3:31 pm    Post subject: Reply with quote

In my ignorance i tried to see how is Miramir ebuild different from Gentoo

Code:

diff miramir gentoo

15c15
< IUSE="duktape consolekit elogind examples gtk +introspection jit kde nls pam selinux systemd test"
---
> IUSE="consolekit elogind examples gtk +introspection jit kde nls pam selinux systemd test"
34,35c34
<    !duktape? ( dev-lang/spidermonkey:78[-debug] )
<    duktape? ( dev-lang/duktape )
---
>    dev-lang/spidermonkey:78[-debug]
60,63d58
< QA_MULTILIB_PATHS="
<    usr/lib/polkit-1/polkit-agent-helper-1
<    usr/lib/polkit-1/polkitd"
<
68a64,67
> QA_MULTILIB_PATHS="
>    usr/lib/polkit-1/polkit-agent-helper-1
>    usr/lib/polkit-1/polkitd"
>
70,74d68
<    if use duktape ; then
<       PATCHES+=(
<          "${FILESDIR}"/polkit-0.118-duktape.patch
<       )
<    fi
111,115d104
<    if use duktape ; then
<       myeconfargs+=(
<          --with-duktape
<       )
<    fi
143a133
>

The differences are so minor.
Gentoo already uses an "elogind" patch.
Why cant we also add the "ducktape" pathc and offer USE"spidermonkey duktape"?
Is the duktape patch so difucult to get or maintain???
In my simplistic mind I cant understand why make such a big fuss about such an easily applicable convenience and choise for the users...
Why we need to ask the upstream to provide this choice if we can do it ourselves?
:D
Back to top
View user's profile Send private message
asturm
Developer
Developer


Joined: 05 Apr 2007
Posts: 7958
Location: Austria

PostPosted: Sun Sep 20, 2020 3:36 pm    Post subject: Reply with quote

What "elogind" patch do you mean?

This duktape patch adds ~1400 LOC, are you prepared to become the new upstream of spidermonkey, rebase the patch on every release and deal with bugs [and have Gentoo user bugs rejected upstream as a result of it]?
_________________
backend.cpp:92:2: warning: #warning TODO - this error message is about as useful as a cooling unit in the arctic
Back to top
View user's profile Send private message
alamahant
Guru
Guru


Joined: 23 Mar 2019
Posts: 550

PostPosted: Sun Sep 20, 2020 3:40 pm    Post subject: Reply with quote

Someone must be writing these 1400 LOC.
Is there a project that maintains this patch?
Can we not just use their duktape.patch?
Again please bear with me....
Back to top
View user's profile Send private message
Ionen
Veteran
Veteran


Joined: 06 Dec 2018
Posts: 1247

PostPosted: Sun Sep 20, 2020 3:44 pm    Post subject: Reply with quote

I have to admit that fixing the upstream MR to meet standards and have working tests seem like the better idea than accepting large patches in gentoo. Maintainers aren't necessarily programmers that can ensure this keeps working (and working properly), or have the time to even.. this is going into fork territory.

Meanwhile just keep using the overlay if you prefer, at least gentoo doesn't have to feel responsible for what overlays do.
Back to top
View user's profile Send private message
fedeliallalinea
Bodhisattva
Bodhisattva


Joined: 08 Mar 2003
Posts: 24623
Location: here

PostPosted: Sun Sep 20, 2020 3:47 pm    Post subject: Reply with quote

alamahant wrote:
Can we not just use their duktape.patch?

And what happens when package change version and patch is obsolete, who port patch to new version an test that package work fine?
Create an ebuild for a package does not imply knowing how to modify all the source code.
_________________
Questions are guaranteed in life; Answers aren't.


Last edited by fedeliallalinea on Sun Sep 20, 2020 3:56 pm; edited 1 time in total
Back to top
View user's profile Send private message
ff11
Guru
Guru


Joined: 10 Mar 2014
Posts: 591

PostPosted: Sun Sep 20, 2020 3:54 pm    Post subject: Reply with quote

alamahant wrote:
Someone must be writing these 1400 LOC.
...

This someone in question is Wu Xiaotian (yetist), maybe you know him as one of the main developers of the MATE Desktop.
_________________
| Proverbs 26:12 |
| There is more hope for a fool than for a wise man that are wise in his own eyes. |
* AlphaGo - The Movie - Full Documentary
"I want to apologize for being so powerless" - Lee
Back to top
View user's profile Send private message
Ionen
Veteran
Veteran


Joined: 06 Dec 2018
Posts: 1247

PostPosted: Sun Sep 20, 2020 4:15 pm    Post subject: Reply with quote

Had a quick look at patch line counts in ::gentoo, and excluding tar'ed patchsets (like for gcc, glibc, kernel), this duktape patch would be the 2nd largest single .patch in all of the tree 8O Beaten only by torque-6.0.4-gcc7.patch

Last edited by Ionen on Sun Sep 20, 2020 4:17 pm; edited 1 time in total
Back to top
View user's profile Send private message
Hu
Moderator
Moderator


Joined: 06 Mar 2007
Posts: 15967

PostPosted: Sun Sep 20, 2020 4:16 pm    Post subject: Reply with quote

Marcih wrote:
I haven't tried to look into that so I'm glad to hear it should be straight-forward. Looks like what's holding back the MR is the radio silence from the requestor, which is a shame, I'm looking forward to this getting merged.
Could you take over for the requester, or find someone interested who will take over? Presumably upstream is less interested in having the original requester handle this than in having someone resolve any remaining change requests / questions. If someone from the community stepped up to do that, and did so well, upstream might move forward despite the disappearance of the original requester.
Back to top
View user's profile Send private message
ff11
Guru
Guru


Joined: 10 Mar 2014
Posts: 591

PostPosted: Sun Sep 20, 2020 5:55 pm    Post subject: Reply with quote

By the way, the changes of the patch:
* polkit-0.116-duktape.patch to polkit-0.117-duktape.patch is: change 2 lines with [mozjs-60] to [mozjs-68]
* polkit-0.117-duktape.patch to polkit-0.118-duktape.patch is: change 2 lines with [mozjs-68] to [mozjs-78]

And while it's easy to update the patch, there will probably always be some overlay with the update.
_________________
| Proverbs 26:12 |
| There is more hope for a fool than for a wise man that are wise in his own eyes. |
* AlphaGo - The Movie - Full Documentary
"I want to apologize for being so powerless" - Lee
Back to top
View user's profile Send private message
asturm
Developer
Developer


Joined: 05 Apr 2007
Posts: 7958
Location: Austria

PostPosted: Sun Sep 20, 2020 6:00 pm    Post subject: Reply with quote

It is easy until it is not; it may apply but it may regress; as long as upstream's concerns are not addressed this is impossible to enter tree.
_________________
backend.cpp:92:2: warning: #warning TODO - this error message is about as useful as a cooling unit in the arctic
Back to top
View user's profile Send private message
ff11
Guru
Guru


Joined: 10 Mar 2014
Posts: 591

PostPosted: Sun Sep 20, 2020 6:20 pm    Post subject: Reply with quote

asturm wrote:
It is easy until it is not; it may apply but it may regress; as long as upstream's concerns are not addressed this is impossible to enter tree.

Well, I believe that this applies to many projects, including Gentoo itself, that when a simple solution is no longer working, it often just drops the ebuild for lack of maintainer (solutions can be added and removed that easily).
But solutions using overlays are very good and offer reasonable support. There is no need for it to enter the Gentoo tree without one interested maintainer.
_________________
| Proverbs 26:12 |
| There is more hope for a fool than for a wise man that are wise in his own eyes. |
* AlphaGo - The Movie - Full Documentary
"I want to apologize for being so powerless" - Lee
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Gentoo Chat All times are GMT
Goto page 1, 2  Next
Page 1 of 2

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum