Currently changing a name of a page will update the links of that page, but not queries it is used in. All queries using a renamed title will break.
Would be great to update the pages name in queries as well, in addition to updating it in links.
I see this too - Iād consider it a bug.
Filed bug report: Renaming a page doesn't update queries using the page Ā· Issue #7519 Ā· logseq/logseq Ā· GitHub
(Please put a like on github for any reported issues that affect you as well - that helps the devs triage/prioritize.)
Great, thanks. I liked it to upvote it.
On my side, this behaviour still occurs, but your bug report has been closed on github.
Iāve filed a new bug report to reopen it: Renaming a page doesnāt update the page name in queries Ā· Issue #8391 Ā· logseq/logseq (github.com)
Itās a bit annoying, as something as simple as changing the name of a page has the unintended effect of breaking queries, and without you noticing it.
Hmm sometimes I want to change the page name so it wouldnāt show up in a query.
Eg in a file system, when I change the file name because I named it wrong, I donāt want a search of the old file name to still appear. Would be really confusing.
We may mean different things. I donāt mean pages not showing up in the results of queries. But about queries themselves breaking.
Example: Say I tagged book entries with $Book and have various queries that use this, e.g. to query for books from a certain author or topic. Now I want to change the page $Book to #Books for some reason. As a result, all queries that used the old tag [$Book] will break.
But just as the (bi-directional) links are updated, so should queries be updated with a name change. How else could the queries keep working ?
Apparently, the LogSeq devs closed the bug report I filed (Renaming a page doesn't update queries using the page Ā· Issue #7519 Ā· logseq/logseq Ā· GitHub) as ānot a bugā - and that it wouldnāt be handled on GitHub.
I canāt for the world understand why this wouldnāt be a bug.
And I canāt for the world understand how any dev team of a software project can think that this is a reasonable attitude to have towards users trying to help and reporting issues. Or how such an attitude towards bugs in the software could not be detrimental for the project.
Further on this by me in the issue: Renaming a page doesn't update queries using the page Ā· Issue #7519 Ā· logseq/logseq Ā· GitHub
Unfortunately, these attitudes are in line with my other experiences with LogSeq during the few years Iāve used it.
LogSeq has a problem here. Itās not about this particular case. This is a general problem.
Related: Opinion: development focus should shift towards improving LogSeq's partially unstable core
What is the exact problem? I donāt see any rude behavior from anyone in the team in that conversation (my opinion of course). Or is your gripe that someone from the team called it a feature request instead of a bug? Because it being a bug is your opinion, as you said here:
Iām eager to hear what you think would be a reasonable attitude and how the team should respond instead in your opinionāso we can improve.
Some things are difficult to implement, especially with our limited team size. One of the team members responded with an explanation:
First: no, no ārudeā behaviour. If that was not fully clear, I want to clarify that now. I have never experienced rude language here on the forum or in the Github conversations. Note also that I have never used the word ārudeā - Iām talking about āattitudesā. Visible through actions (and non-actions), not in language.
What is the exact problem?
I explained that carefully in a comment to the GitHub issue, which I linked to in my post above - but here is the link again: Renaming a page doesn't update queries using the page Ā· Issue #7519 Ā· logseq/logseq Ā· GitHub
And while youāre at it, for context, read the few lines of comments in this issue - issue identical to mine, opened as a reaction to mine being closed: [feat request] add support for local files embeds and links Ā· Issue #839 Ā· logseq/logseq Ā· GitHub
If anything is still unclear, give me hint on what, and Iāll do my best to clarify.
Or is your gripe that someone from the team called it a feature request instead of a bug? [my emphasis]
No, not in any way. What someone ācalls itā doesnāt really bother me - itās the actions that were taken that I canāt understand, especially as they are an example of a broader problem. What happened in the issue I think will be very clear if you just read it through briefly (itās a very short issue) - if after that itās still not clear what my āgripeā was, again, tell me whatās unclear and Iāll try to be more specific.
Because it being a bug is your opinion, as you said here:
Iād consider it a bug.
Thatās kind of a strange way to try to mark my words (however I assume no ill intent). I canāt see that it carries any meaning or has any relation to what weāre discussing here. But for what itās worth, youāre making an invalid inference: my wording doesnāt mean what you imply. For example Iād consider 1+1 being equal to 2. That doesnāt mean I have said that itās ājust my opinionā.
Because it being a bug is your opinion
I think (I could be wrong of course) you, once having read the issue, shares that opinion with me.
For the un-convinced, Github user @fschuman provided this motivation, in a comment posted to the issue a few hours ago:
Unless Iām missing something, this behaviour makes the combination of queries and renaming page names unreliable ALWAYS AND IN ALL CASES. EVERY single renaming of a page will break ALL queries that it is part of !!!
Itās strange if one has to decide between using functions as fundamental to Logseq as queries and renaming pages. The conclusion would be (1) either never rename pages, or (2) never use queries. (Iām assuming here that manually tracking if a page was used in a query is unfeasible).
I would consider that a major bug, given what Logseq is ??
Iām eager to hear what you think would be a reasonable attitude and how the team should respond instead in your opinionāso we can improve.
I believe that could be clear now - again, if not, tell me so.
I actually havenāt so far critizised the bug not yet being fixed, or not being prioritized - thatās obviously not what made me annoyed.
But I do take the opportunity now to say that I, e.g. from @fschumanās motivation above, canāt understand how such a bug could not be given some reasonable prio.
But this is actually not about this particular issue. Although it is a very telling example. Itās about seemingly not: seeing value in users submitting bug reports, caring about the core functionality, about bugs, about actual user experience.
Sorry for the experience. But we may not underestimated the underlying engineering challenges. I replied in Renaming a page doesn't update queries using the page Ā· Issue #7519 Ā· logseq/logseq Ā· GitHub and discussion is welcome.
Also see Opinion: development focus should shift towards improving LogSeq's partially unstable core - #5 by Junyi_Du
Feel free to share insights if you have engineering experience on the field of query DSL.
1 Like
Sorry that I had interpreted your words wrong @and_yet_it_moves, thatās my mistake.
I understand your frustration that this issue isnāt handled yet, but as @Junyi_Du explained on GitHub and here there are complexities at play that make it a difficult issue to tackle. What may seem an easy fix from the outside, may turn out to be a very complex change in the code base.
Knowing the team, I know that they definitely appreciate bug reports and users sharing their experience. While some bug reports are more useful than others (because they contain more context), all reports are appreciated. But with a team of 10 engineers who are working hard to also release new features, there is very limited capacity to tackle all bugs.
The bugs that get attention are the ones that affect lots of users or make Logseq unusable. But that also means that some quality of life bugs do not receive as much attention.
Is this an issue? Iād say so, yes. But until our team grows considerably or more community members help out with their coding knowledge, there will always be bugs that fall through the cracks or do not receive the attention that they deserve.
But once again, that is not what my frustration is about.
I once again refer to Renaming a page doesn't update queries using the page Ā· Issue #7519 Ā· logseq/logseq Ā· GitHub for that.
Iād want to add another motivation to this: Iāve started to atdd emjoiās to my project pages to have a more modern feel and easier navigation. Luckily - I realised early this means that ALL the queries that entail the changed project name would brake by doing this.
It seems a small and āmodernā change that essentially could ruin oneās entire system. Similarly by adding emojiās to tags such as ābooksā and so on.