Talk:AUR helpers

From ArchWiki
Revision as of 12:06, 1 October 2018 by Alad (talk | contribs) (Show which unsafe flags are used by default: rm closed)
Jump to: navigation, search

"Reference" implementation

This is an alternative to #Reliable Updater. Instead of an arbitrary set of test packages, we could write up a "specification" on what a reliable AUR helper should do. This should also be more helpful for potential AUR helper writers who otherwise have to wade through complex, fully-featured AUR helpers.

I propose a minimal reference implementation with the following points:

  • No client-side workarounds for upstream limitations. In particular, a reference implementation does not need to score full points on split packages, as makepkg --pkg was removed with pacman 5.
  • Minimal language constructs in e.g. a scripting language like dash.
  • Prefer simplicity of implementation over being fully featured. In particular, an implementation may only support git clone and not git diff.

My initial plan was to keep such an implementation in a man page aurhelper(7) (hosted as part of aurutils), but we can consider including on a sub-page of this article. It could be then linked from the comparison table. Thoughts? -- Alad (talk) 13:28, 8 March 2018 (UTC)

Generally agree with the idea, but I don't think there is a way around a set of PKGBUILDs that could be used to test helpers in a local AUR instance. F.e., I wouldn't define a "reliable" helper that doesn't handle split packages well. Since helpers are tolerated rather than supported, upstream limitations of the AUR might be temporary or permanent, meaning the limitation would actually be in the helper itself (f.e. like regex support). Also, I'd use pseudo code for such a reference as the actual implementation itself doesn't matter, unless you'd like to write a new minimalist helper. Spyhawk (talk) 15:26, 8 March 2018 (UTC)
Apart from FS#56602, I can't think of a case where upstream opposed removing limitations, even if helpers directly benefited. cf. the regex support discussed in [1] or the exit codes finally introduced in makepkg 5.1 which made automatic building significantly easier imo. To me it seems that the main reason we have these AUR limations is due to the minimal interest of helper writers in contributing upstream, and upstream itself having different priorities. Not sure why former is the case, the PHP codebase may play part in it - at least it does for me.
You can keep dash close enough to pseudo-code, I guess less so if you want a complete example rather than exemplary code blocks. For the PKGBUILD set, I use this: [2] -- Alad (talk) 18:34, 8 March 2018 (UTC)
My understanding is that changes that aren't invasive will be accepted upstream, but otherwise might be rejected (see [3]). One prominent example that comes to mind is FS#48796. It's not really relevant anymore since x86 has been officially dropped, but the solution would involve duplicating DB tables on the server, which isn't trivial to implement/migrate. Many of the feature requests involve non-trivial code change, which is the main reason nobody pushed patches; I dislike PHP but the language itself isn't too hard either. For regex, see the bottom of [4], which is the follow-up of your link above.
Your testsuite seems interesting (thanks for the link), but one advantage of having a fixed set of packages is that these packages might be updated and change, making these edge cases difficult to test. This happened quite a few times with my own list of test packages in the past and this was rather annoying. Spyhawk (talk) 20:20, 8 March 2018 (UTC)

Expand Secure criteria to include other (non-PKGBUILD) bundled files

[5], in particular [6]

The new criteria would be as follows:

  • PKGBUILD, no other files -> Partial
  • Other subset of files that includes the PKGBUILD -> Partial
  • No PKGBUILD -> No
  • All files in the git repo or tar archive -> Yes

Similar to the Diff view column. -- Alad (talk) 16:32, 4 July 2018 (UTC)

good idea, you also mentioned this for aurman a few months ago, see: really a good idea to implement it in a way, so that changes of all known files are being shown Polygamma (talk) 17:07, 4 July 2018 (UTC)
"All files in the git repo or tar archive -> Yes" What exactly do you mean by all files? Build files often contain non text files such as images. Git diff is smart enough to hide these but then you could consider that partial because not all files are covered.
In my opinion all a helper has to do to be secure it pause and allow the user to read the build files. The helper does not even need to offer to open them for you that's the user's responsibility. Anything more than that is nice to have but not strictly needed. Morganamilo (talk) 20:25, 4 July 2018 (UTC)
If this qualifies as "nice to have", there has to be an explicit warning that a green entry in the "Secure" column does not cover other files, files which may cause more harm than the PKGBUILD itself (such as .install files or exectuables called from the PKGBUILD). In either case it's misleading, since you either give the impression that viewing PKGBUILDs alone is sufficient (with the current criteria), or include a warning that diminguishes the value of the criteria in the first place.
Latter is similar to "Native pacman", in that you have a warning at the article top warning against any sort of pacman wrapping, and criteria in the table that ignore this warning, or even reward behavior which goes against it. -- Alad (talk) 17:07, 8 July 2018 (UTC)
That's a fair point, what about changing the name to "show files before sourcing" or something? Seems more accurate. Then it would make sense that not showing .install files to be partial. The only problem I see that it's not as hard hitting as "secure". Morganamilo (talk) 20:11, 8 July 2018 (UTC)
It cuts both ways: it's an effective deterrent against broken helpers, but it also gives the impression that using a "Secure" helper makes usage of the AUR safe, which it definitely doesn't. I'm not sure on what different name to use, though. -- Alad (talk) 17:25, 14 July 2018 (UTC)
I guess "File view" could work. -- Alad (talk) 17:44, 14 July 2018 (UTC)
The column name was updated to "File review". Are there remaining helpers that only display the PKGBUILD? (trizenAUR springs to mind) -- Alad (talk) 15:30, 23 August 2018 (UTC)

proper batch interaction 2 and 3

Is there even a way to implement this feature correctly?

I do not know much so correct me if I'm wrong but the proper way to do 2 and 3 for something like wrapper repopkg aurpkg seemingly is:

  1. copy sync db and perform sync on copied db, avoiding possible partial upgrades in the system
  2. find repopkg and aurpkg using copied sync db and AUR backend respectively
  3. [user] package confirmation
  4. get all AUR package files of aurpkg and its AUR deps
  5. [user] batch inspection
  6. loop: if new AUR deps introduced get package files for those too and do another batch inspection
  7. calculate solutions using up-to-date copy of sync db
  8. [user] batch queue updates and resolutions
  9. sync sync db with its copy and perform system upgrade
    1. calculate solutions again with sync db (if sync db is different from its copy)
    2. [user] batch queue updates and resolutions again (if solutions are different)
  10. perform other package operations in resolved order
  11. build and install AUR packages in resolved order

-- 13:50, 26 August 2018 (UTC), last edited -- Svito (talk) 16:43, 26 August 2018 (UTC)

For 2/ the biggest pitfall to avoid is to calculate the solution for AUR deps on the basis of completely updated repo packages, not on the basis of the currently installed packages (f.e., the way it is done in yaourt can lead to some issues. From what I've seen, all of the other helpers (aurman, pikaur, yay) have some issues in their latest iterations with 2/, though aurman does better than the rest). -- Spyhawk (talk) 14:27, 26 August 2018 (UTC)
I agree. Maybe it is not clear but I assumed points 1-8 have to be done with up-to-date copy of sync db, not sync db and not local db (both of which remain unchanged up to point 9). -- Svito (talk) 16:08, 26 August 2018 (UTC)
Good then. This said, your solution above seems a bit over complicated to me (f.e. points 6 that should be done in 5, or 7 that should be done before 4), but I might miss something since I have never implemented batch interaction 2/ myself - I rejected the idea because of its complexity. -- Spyhawk (talk) 09:26, 27 August 2018 (UTC)
Is there anything left to discuss that affects the article, or can this be closed? -- Alad (talk) 10:56, 18 September 2018 (UTC)
I think this discussion is certainly is not "solved" but it is not really about article itself but rather something to better discussed with helper developers as part of some issue or pull request. With that in mind, I'm closing this discussion. -- Svito (talk) 18:54, 18 September 2018 (UTC)

Pacman wrappers Specificity

I wanted to introduce the manual AUR package selection specificity for pacman wrappers. For example yay and pakku only allow -Syu where they update system packages and AUR packages at the same time. pikaur when doing -Syu also updates system packages and AUR packages at the same time but prompt user for manual selection for AUR packages, because for example I want to update to the new version of X program but not the Y lib (AUR) that is used only by Z program (AUR) so until Z is not updated I don't want to update Y. And yaourt when doing -Syu updates only system packages, -Syu --aur is needed for updating AUR packages as well, and a manual selection for AUR packages is available. That's why I wanted to introduce the manual AUR package selection specificity. -- Noraj (talk) 11:16, 9 September 2018 (UTC)

Why just not specify --ignore ylib? It is problem with zprogram if it does not use versioned deps and allows library update that breaks it. Helper should just fail to satisfy AUR deps in correct case, bail out and user should specify --ignore manually explicitly IMO.
Additionally yay asks you what AUR packages to not upgrade that also provides solution to this XY problem.
All-in-all if you can no longer build some package from AUR correctly because of dependency update that package is broken and should be fixed or be removed instead. And helpers should ideally not provide workarounds for problems that broken package introduce.
-- Svito (talk) 13:06, 9 September 2018 (UTC)
It's not really about broken programs. It could just as easily apply to some package that you're sure will compile fine, but you also remember takes 45 minutes to build, which you don't want to spend right now.
Also I really don't consider using versioned deps, to be the correct case. -- Eschwartz (talk) 14:45, 9 September 2018 (UTC)
That's exactly my use case. Whether to allow it as a specificity or not? I'd probably say sure. But I also think there should be a cap on how many specificities an entry may have. I'd say max 4 or 5, otherwise the column may as well be called feature list. Morganamilo (talk) 14:57, 9 September 2018 (UTC)

Baseline AUR helper features

Some proposals about section names:

  • Search-only -> Fetch only
  • Build and search -> Fetch and build
  • change introduction sentence
    • from: Most helpers automate the process of retrieving a package's PKGBUILD from the AUR and building the package.
    • to: AUR helpers provide functionality of searching for packages in AUR and fetching their PKGBUILDs. Some helpers additionally automate build and install process.

Also about package-queryAUR. I can't check now because github is down but if it does not retrieve PKGBUILDs (judging from N/A under Git clone column) shouldn't it be under Libraries? Why does it qualify as AUR helper?

-- Svito (talk) 16:04, 11 September 2018 (UTC)

The recent move to Pacman/Tips and tricks is absurd. When AUR automation is not considered, as it is everywhere in the wiki but AUR helpers, package-query offers no benefit over existing tools like expac. This is in to contrast repoctl which contains non-trivial local repository functionality, and has some AUR support as a secondary feature. -- Alad (talk) 14:55, 13 September 2018 (UTC)
I reverted my last mess-ups. I don't understand what I'm doing lately so I'm taking a break. -- Svito (talk) 23:01, 13 September 2018 (UTC)
Renames I like, although package-queryAUR is truly search only. The points in the intro are definitely better, the grammar is a little messed up though.
package-queryAUR Works as a command line tool. Libraries would be #included or the language's equivalent.
Morganamilo (talk) 21:15, 11 September 2018 (UTC)
I suppose one could call it a forking bash library if one really wanted. Like expac, it's really only particularly useful as a printf-style formatting parser for the package database, except that package-query merges this with the AUR's json RPC interface. It may not be a library in terms of C programming usage semantics, but terms of the intent behind its usage it very much is a library tool.
I guess the same could be said of coreutils tools like cut, join, expr, tsort, etc.
Ha, someday when I Make Yaourt Great Again™ I should also make package-query a dual shared library/PIE and a bash loadable builtin. Gotta be more libraryish in order to avoid those expensive forks, yo. -- Eschwartz (talk) 03:48, 12 September 2018 (UTC)
Please review Special:Diff/540990, Special:Diff/540997. Additionally section changes:
  • Search-only -> Search and fetch
  • Build and search -> Fetch and build
-- Svito (talk) 19:11, 18 September 2018 (UTC)
By that definition ("helpers that fetch"), package-query could be included in the official repositories, which I'd say is a highly unlikely event. I'm mostly neutral on the rest, though grammar should be checked and "fetch" is perhaps not the most formal term. -- Alad (talk) 11:46, 20 September 2018 (UTC)
Done without "helpers that fetch" part and with "download"/"retrieve" instead of "fetch". Closing. -- Svito (talk) 00:40, 28 September 2018 (UTC)

Use definition list for legend

They are the correct semantic HTML element to use and the bold styling of the definition terms makes the list more readable. Demo:

File review
Does not source the PKGBUILD at all by default; or, alerts the user and offers the opportunity to inspect the PKGBUILD manually before it is sourced. Some helpers are known to source PKGBUILDs before the user can inspect them, allowing malicious code to be executed. Optional means that there is a command line flag or configuration option to prevent the automatic sourcing before viewing.
Diff view
Ability to view package differences on inspection. Besides the PKGBUILD, this includes changes to files such as .install or .patch files.
Git clone
Uses git-clone(1) by default to retrieve build files from the AUR.
Reliable parser
Ability to handle complex packages by using the provided metadata (RPC/.SRCINFO) instead of PKGBUILD parsing, such as aws-cli-gitAUR.
Reliable solver
Ability to correctly solve and build complex dependency chains, such as ros-lunar-desktopAUR.
Split packages
Ability to correctly build and install:
  • Multiple packages from the same package base, without rebuilding or reinstalling multiple times, such as clionAUR
  • Split packages which depend on a package from the same package base, such as libc++AUR and libc++abiAUR.
  • Split packages independently, such as python-pyalsaaudioAUR and python2-pyalsaaudioAUR.
Clean build
Does not export new variables that can prevent a successful build process.
Batch interaction
Ability to prompt before the build process, in particular:
  1. Inspection of package files or their differences;
  2. Summary of package upgrades;
  3. Resolution of package conflicts and installations.
An asterisk denotes functionality specifically enabled by the user.
Shell completion
Tab completion is available for the listed shells.

--Larivact (talk) 04:40, 14 September 2018 (UTC)

+1 -- Alad (talk) 10:51, 14 September 2018 (UTC)
Since the terms are not part of the first sentence of the definition anymore, they should all start with a capital letter. -- Lahwaacz (talk) 07:21, 15 September 2018 (UTC)
I updated my demo. --Larivact (talk) 06:07, 17 September 2018 (UTC)
Merged, thank you. -- Alad (talk) 10:54, 18 September 2018 (UTC)

Drop "Clean Build" Column

Literally every helper apart from yaourt has this as yes. Perhaps yaourt's variable exporting should be listed under specificity or the "Unsafe Flags" column could be adjusted to unsafe actions? Morganamilo (talk) 04:54, 14 September 2018 (UTC)

Disagree, since it's a reminder to anyone writing (existing and new) helpers. Example: [7] -- Alad (talk) 10:52, 14 September 2018 (UTC)

New test cases for dependency resolution

ros-foo-desktop-meta have always been difficult to build, even more so with KDE4 libs moved to AUR (see arch-dev-public). Besides the sheer number of dependencies, they otherwise have little interesting properties either.

I propose to instead use various cross-compilation packages as test cases, e.g. mingw-w64-zlibAUR and powerpc-linux-gnu-gccAUR. These appear very efficient at exposing problems with complex dependency algorithms (see for example [8] or aurman's issues with nsisAUR) and don't take 2 years to build either. -- Alad (talk) 11:34, 14 September 2018 (UTC)

We could also add some simpler cases, like fortune-mod-all-enAUR, and add details similar to the Split packages description. That way, all existing entries with "Yes" in Reliable solver would at minimum have "Partial". -- Alad (talk) 12:01, 14 September 2018 (UTC)
I believe the mingw stuff has a bunch of circular dependencies and a bootstrapping process. Do you think AUR helper's should be expected to handle this? Morganamilo (talk) 15:25, 14 September 2018 (UTC)
No, I don't think so. There will always be cases that can only be dealt reasonably manually, simply because the involved complexity in implementation isn't worth it. See also related sicussion Talk:AUR_helpers#.22Reference.22_implementation above, where a set of fixed reference packages would be better than live packages. No idea how to dealt with than without a local AUR instance though. -Spyhawk (talk) 10:42, 15 September 2018 (UTC)
The mingw packages haven't had cycles in (global) depends for a while. As such, makepkg -r works fine for these packages without manual intervention or "bootstrapping". If helpers fail, it may be because of handling split depends contrary to PKGBUILD(5), or other flaws in their dependency algorithms. -- Alad (talk) 10:05, 18 September 2018 (UTC)

Erroneous implication that AUR helpers are supported

In Special:Diff/544344 the meaning of the warning was totally stripped away. By longstanding community tradition, AUR helpers are not supported, no ifs ands or buts. It has nothing to do with the rule against AUR helpers in the official repos -- the two concepts merely coincide. If anything, I'd argue that the link in question should be removed -- it seems almost entirely offtopic, and one developer assertion of this fact does not really add more corroborating proof than simply saying so in the wiki I guess. -- Eschwartz (talk) 07:07, 28 September 2018 (UTC)

To put an end to these unannounced edits taking excessive wiki admin time, I will be moving the comparison table to github [9], and the article remainders to the (more closely followed) AUR article. -- Alad (talk) 12:05, 1 October 2018 (UTC)