-
Notifications
You must be signed in to change notification settings - Fork 22
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Enhance documentation for analysing bugs #100
Comments
Well, https://forum.sailfishos.org/t/bugs-in-patches-for-patchmanager/8553/14 and https://forum.sailfishos.org/t/bugs-in-patches-for-patchmanager/8553/15 ff. are additional pieces for this, but in line with issue #85 I think we shall not add even more locations, where such documentation is spread. Furthermore, we shall clearly differentiate between "Bugs in Patchmanager" and "Issues with Patches for Patchmanager", plus also separate interfacing noobs (at FSO) and Linux / tech savvy users (here at Github). (IMO, we already reached consensus on that.) Thinking about this for a while now, I believe the Wiki here is most suitable place to gather such technical information / documentation in a structured and maintainable manner. |
Just checked all comments at Openrepos for PM3.0, these are the only three relevant places with information there:
|
It's the nature of forums and forum threads that they become unwieldy over time - I don't think anything can be done about that apart from not having them at all. Still, it's good to be visible that way on FSO. I agree that re-usable information should be in a more stable place - a Wiki is sounds like a good idea, or maybe a collection of That being said, I object to separating users into "noobs" and "savvy" people. As long as feedback isn't overwhelming, and it is not at the current pace, I don't think bug/report/feedback triaging is that much of an issue. About my posts: I agree but at the time it seemed the best place, and I like to engage complaints quickly. If an answer is useful it can quell similar reports if people read (which they don't) or search (which they also don't). So yeah, all for a wiki-type thing, and collecting information from those sources there. It will be a convenient thing to link to in whatever communication about issues happens. |
That is why I try / think we should try to avoid using a forum (-thread) for receiving bug reports.
IMO, this is too strictly black and white thinking!
I would agree, if that sentence was without "that way", because I am not sure what that is meant to imply.
A wiki at Github is exactly that, but in a separate repository, see https://github.com/sailfishos-patches/patchmanager/wiki
The Github webfrontend also does not support branches for that wiki repo, hence we should first backup what is there and transfer it into a new top-level directory (e.g.,
Why?
My concerns are not about "triaging", they are about manageability, i.e. not creating such a forum thread monster as the PM-thread@TMO. Plus avoiding the typical approach, "if that happens, just discard it and start anew", which simply recreates the same structural flaws and leads to the same result. Said in a different way: We should learn from the obvious result of Andrey stating "please file all PM-related questions there" (and how wrong a FSO "feels" for reporting and discussing bugs, if one is used to a proper issue tracker).
I fully understand that and that could have been me.
Bingo! 😞
👍
That is exactly my goal: To be able to drop a link at FSO to some wiki content here, instead of writing this content there. |
@nephros, I would appreciate very much, if you perform:
After that is done, I can wipe the wiki clean, then start building a new wiki structure and fill it by copying and adapting content from the aforementioned sources: PM3.0@Openrepos, PM-thread@TMO, README@here, your contributions at FSO etc. |
OK, done I think. The old files are now in a branch called "historic" - but that branch is only visible in checked-out clones of the wiki repo, not on Github Web. Anyway, I pushed a new, empty branch "master" which can now be used to make a new wiki structure. |
Just as a note: https://github.com/ichthyosaurus/sailfish-patch is excellent for people to start developing patches, and should be integrated at some point in the wiki. |
O.K., thanks. Good to know that the old content is saved and safe, and I can start now to create a new wiki structure and content. |
@nephros, I wanted to determine, why https://github.com/sailfishos-patches/patchmanager/wiki/Home/_history only shows your commits and cloned the wiki repo locally. Please carefully look in your local repo, if you do really have »The old files are now in a branch called "historic"«! Either I fail to see it (hope so), or it does not exist at GH. P.S.: The state at the oldest commit which still exists in the history of master does contain the old top level page, but the crucial "Patches list" sub-page (the real content, which ought to be saved) is missing, see https://github.com/sailfishos-patches/patchmanager/wiki/Home/e0c759729d831bae8dd3d311a1044d929a8d8ceb |
Well that is consistent with what you quoted above. The GH web UI only shows the master branch content.
It is certainly there, both in my local copy and on GH. Try But I seem to have misunderstood your request. I thought the goal was a clean slate without the commit history, so that is what is there now. If the goal is just to move the old pages somewhere and add other content, keeping all the history, I think we can do that simply on the original master branch.
Right, I can create a
In conclusion, I think we should undo the whole thing (especially as I kind of fumbled the "clean history" part of the new master branch, it has a lot of unnecessary commits right from the start) and simply do all changes on the old master branch, moving old stuff into |
Yes, I do see it now. I wonder why
Yes. I wonder which of my words were so easy to fundamentally misunderstand:
That is O.K., if though I did not mention anything like this.
Yes that was my original goal, plus making the old stuff visible somewhere (i.e., keeping it visible): That is why I suggested to move the old content with its history "into a new top-level directory (e.g., .wiki-backup) in the legacy branch of the code repository."
Yes, that is unnecessary, now that I know I already have it.
Well that would also work. Although I do like it less than moving the old stuff somewhere else, because I reviewed it and know that we will never need anything of it: It is simply old cruft from the "legacy" timeframe (i.e., for Patchmanager 1 / early SailfishOS 1.1.x releases). |
Right, so in conclusion:
That sound like a plan? |
Yes, absolutely the right plan from my POV. |
Both, the current wiki history and the archive of the old content in Now it is my turn to gather all the information and to structure and consolidate it in the new, empty wiki. |
I was too slow in collecting this, Andrey disabled the comments, so variants 2 & 3 are gone. |
> Just checked _all_ comments at Openrepos for PM3.0, these are the only three relevant places with information there:
>
> * WRT translations, PM's environment variable and logging:
> Three different sections of the description at https://openrepos.net/content/coderus/patchmanager-30
> * Logging variant 2: https://openrepos.net/comment/25667#comment-25667
> * Logging variant 3: https://openrepos.net/comment/25666#comment-25666
I was too slow in collecting this, @CODeRUS disabled the comments, so variants 2 & 3 are gone.
The Wayback Machine to the rescue!
https://web.archive.org/web/20200922041227/https://openrepos.net/comment/25667
|
Copy&Paste'd all extant fragments I could find into the corresponding Wiki pages at https://github.com/sailfishos-patches/patchmanager/wiki Finished pagesUnfinished pages |
Enhance documentation for analysing bugs.
There were a couple of mini-guides by Andrey, especially how to properly filter the systemd journal. But IIRC a few variants exists, and they have to be re-searched at the PM-thread at TMO, the comments at Openrepos, at TJC etc.
Then these variants should be condensed into a single, optimal one, if possible.
The text was updated successfully, but these errors were encountered: