12

NVIDIA screwed up and pushed some buggy drivers, affecting many users who are coming to AU with the same issue:

Strange artifacts along window borders after waking computer from sleep mode

The question has a list of affected cards at the end:

enter image description here

Someone has proposed that this question be made a community wiki, so that this section could be more easily edited (I don't agree but I've been invited to write the meta question - see my answer below).

In chat, the person who wrote the NVIDIA question agreed with me that there was no benefit to making this post community wiki. Most of those who answered it have not, as far as I know, given their views yet.

Community wiki posts:

  • may be edited by any user with over 100 reputation
  • do not result in reputation changes (up or down) for the post author (but if a post already has votes, converting to community wiki effectively freezes rep gained or lost from it - previous rep changes are not wiped)

Should this post be converted to community wiki?

(this can only be done by mods, so last call is with them)

Note: it has also been suggested that the question should be closed as a bug report. I fear my question would be too broad if I also invited discussion on that, but I will edit if people think it should be included

  • 2
    I am confused why you asked this and then answered no to it your self, did I miss some context? I do agree it does not need to be a CW though. IMO CW is for collaborative efforts (ie multiple people worked on the post or a post was edited beyond its original scope). I do think it should be closed as a bug so people go to the bug report but not deleted, so it can then be reopened when a fix is issued from Nvidia. – Mark Kirby Apr 3 '17 at 8:01
  • the person who thinks it should be CW invited me to make a meta post @MarkKirby so I've asked the question "neutrally" so that they can add their answer in favour of it – Zanna Apr 3 '17 at 8:12
  • 2
    Just commenting about the note, as I can see people are already voting to close the question. I don't think it should be closed as a bug report in this case. I think protecting the question is enough. The problem is somewhat serious, and letting people know how they can temporarily fix their problem is helpful. – Dan Apr 3 '17 at 8:54
  • 1
    I'm inclined to agree that it's better to keep it open – Zanna Apr 3 '17 at 9:04
  • I would agree that there is no reason to convert it to a CW, closing? In it's previous form it seemed clear that the root cause of the problem is bugs in the NVIDIA driver prior to the most recent edit, I believe the answers to workaround the issue have value (at least until such time as NVIDIA resolves the issue with their proprietary drivers) All that said, Closing the question isn't the same as deleting it so the answers will still exist and still be useful even if it is closed so I'm not adamant on either side of this one. – Elder Geek Apr 3 '17 at 16:43
  • 2
    I agree that there is no good reason to make this CW. But, just a remark: The premise being tossed around here seems odd. There's actually no compelling reason to keep the affected hardware list up to date (or even include it to begin with). The post wouldn't be difficult to find without it, and even if this post isn't found there's plenty of info about the issue elsewhere ... – Jason C Apr 3 '17 at 23:15
  • 1
    ... on the internet that somebody will find instead (and in the Linux community a fair number of the issues are probably just being bug reported to various mailing lists anyways), and the bug is known and a fix will be released soon, making the whole thing obsolete anyways. The presence of a perfect list of affected hardware doesn't actually change much and isn't something to be considered in the CW-ness of this. NVIDIA uses unified drivers for all their stuff anyways so the hardware isn't even that important, the driver versions are of primary interest (fsv "interest"). – Jason C Apr 3 '17 at 23:16
  • 1
    (And even driver version isn't particularly varying; it's going to be one of the Feb 14 2017 driver branches, maybe even just 375/378.) – Jason C Apr 3 '17 at 23:22
  • @JasonC I agree with you and muru has said the same from the start. – Zanna Apr 4 '17 at 4:25
12

There's no compelling reason to make this post a community wiki.

The argument that it would make editing easier has very marginal traction I think, since anyone can suggest an edit and the Suggested Edits review queue is very fast moving (moreover, the person who last edited it to add their card does not have enough rep to edit CW posts without review, so it would have made no difference to them or to reviewers if the post were CW), and lowering the bar to edits being audited is not a particularly good idea in my opinion - we check the quality of edits for very good reasons.

The post does not have many possible answers (in this blog post on community wikis, Grace Note argues that, while there is very rarely a benefit to making questions community wiki, the status could be used to suppress a post that is growing out of control) and is now protected to prevent '(me too|thanks)!' answers being posted. The proliferation of similar answers is mainly due to another question having been merged into this one, and new, bad answers are not appearing.

The question author and people who have answered have not expressed any wish to make the post a community wiki and would be, in one stroke, prevented from gaining any further reputation from it. I don't think suppressing reputation gain makes any sense as a reason to make this post community wiki, and the affected users may well be upset.

  • 3
    Do we need to maintain a list of all known bad (driver x card) combinations? The list not going to be exhaustive, so what's the benefit in doing so? – muru Apr 3 '17 at 9:05
  • @muru I would say yes, as it helps searchability and allows users to come to that question sooner. – Kaz Wolfe Apr 3 '17 at 16:33
  • 1
    And is that worth a community wiki question? – muru Apr 3 '17 at 16:35
  • 3
    I don't think adding cards to the list is a bad idea as the information is useful to those experiencing the problem, nor do I think it needs to be a CW to accomplish that. – Elder Geek Apr 3 '17 at 16:46
8

It seems to me that the only argument for making the question a CW is that it will be easier to maintain a list of card x drivers affected by this. Let's examine that.

Of the six 12 listed currently, 5 were added by a 20k user, 6 by OP and just one was added via a suggested edit. One edit out of 7 revisions. That's not enough to argue that the post should be made CW for making editing easier. Even CW posts need 100 rep to edit, at that level, the user can easily comment on the post if they don't feel like editing, and OP or any 2k+ user can edit that in.

I would rather the list, if at all we need a list1, be maintained in an answer. A single answer can be made CW quite easily, is just as included in search indexes as the question and most importantly, doesn't affect any of the other posts in the question.


1 Just considering that 6xx models are affected implies dozens of cards spanning years would be affected. Nobody, not even the users we're so worried about, is going to maintain such a list.

-7

There's no compelling reason not to make the post Community Wiki

First off, turning a post into a community wiki makes the post belong to the community, which is something that has been done with a decent number of masterposts. As this post is also a masterpost now (look at its timeline, it has had countless questions closed and merged into it, it would make sense that this question follows suit. Grace Note's post on the SO blog is not set-in-stone policy, but only a guideline that individual sites (or even moderators/users) can choose to follow/ignore as they see fit.

Secondly, it keeps trivial edits (such as adding cards or example images) out of the review queue. Low-quality edits aren't that big of an issue -- rollbacks exist, and the masterposts that are already community wiki aren't vandalized often. While, yes, the edit queue is fast, it still isn't a good idea to clog it with edits that don't need reviews. This comes even more into effect when you realize that editors are supposed to verify that information edited into posts is factually correct (although everyone seems to ignore that rule for the sake of brevity). With a CW, there's an implied trust that users know what they're talking about (which is, in fact, the concept of reputation).

Being a community wiki doesn't encourage bad edits or answers any more than a regular question. In fact, to that regard, I've already protected this question as there has been at least one case of a "me too!" post, and I can only see that number growing as time goes on.

We're already breaking plenty of rules with this post (for example, it should be closed as a bug, we're allowing edits of new cards/examples, and so on), so it's obviously a special snowflake. According to our own edit rules, further examples and cards should be blocked as an "attempt to reply," but this behavior is tolerated here precisely because this is meant to be a canonical masterpost.

Finally, there's the question of symbolism. This question has been asked by many people, each adding their own little spin on to the original question. Putting this question into the ownership and control of the community can only benefit it and make the question better and easier to find.

  • 3
    If these edits are "useless things" then why do we need to make it easier to add them? As to reviewers needing to verify the information and potentially not doing so, how does allowing people to add such information without review help avoid the possibility of false information being added? That part of the argument makes no sense at all imho – Zanna Apr 3 '17 at 17:16
  • @Zanna Edited to make that section more clear (at least to me) – Kaz Wolfe Apr 3 '17 at 22:37

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .