#softwaregripe — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #softwaregripe, aggregated by home.social.
-
@oddhack Huh! That is definitely a lot more resultification than I got. Maybe I missspellled it?
[does search again...]
Oh. OH. OMFG with facepalm.
The helpful AI was assuming I misspelled "neurodivergent", and ONLY gave me results for THAT.
May I observe, once again: OMFG. :blobcatupsidedown:
(Dear search companies: DO NOT DO THIS. I expect it to be fixed by tomorrow morning. Thank you.)
-
New wishlist-item for #KateEditor :
- a command to "check all open files to see if they exist; close any that do not, and never mention them again ever"
(Do you think your associates can arrange that for me, Mr. Morden?)
-
How is it that the search function on Tribler has gotten enshittified??
-
CW: Signal app gripe (substantial)
@woozle @signalapp I've just started to explore Signal, and the limited functionality using it fro desktop is why I probably won't pursue it further.
Biggest for me is that it's an accessibility problem.
My disabilities make it very challenging to type messages on my phone.
-
CW: Signal app gripe (substantial)
I just updated @signalapp and read the update message (from March 4, iirc).
I liked the first part:
Signal secure backups are here! Now you can choose to enable automatic end-to-end encrypted backups so you'll never lose a message when you get a new phone or reinstall Signal.
...but then...
You can back up all of your messages and the last 45 days of media for free, or you can pay a small monthly fee to get 100 GB of storage to back up every attachment. Tap on backups in your Signal settings to get started. Thanks to everyone who helped test this feature during the beta period!
...umm, where's the option to back it up locally, guys? This feels like extortion and customer lock-in, which should be antithetical to your mission. :kestraglow:
Worse, I clicked on the "learn more" link (buried elsewhere in the app) and found:
Signal Desktop does not support transferring message history to or from any device.
WTF. Just WTF. I am not okay with this. (Desktop should be where the most features are.)
-
CW: gripe about audio-library players.
(I'm using Kubuntu, for what it's worth. [insert ominous guitar-plucks] "There's something happening here...")
They always seem to assume:
- all of my music is on my main drive, probably under "Music" even if that folder doesn't exist
- any audio files in my home folder or under it should be added to the library by default
- any images found are probably album or single covers for something, use your best random guess
- any individual tracks found that have album information (or for which album information can be found) should be presented as an album of just that one track, rather than just showing them as individual tracks
- collections of music are best shown as thumbnail images (maybe with some abbreviated text), even when there is no image-file
- it's very important for every track to have an image, and for that image to be the first thing you see and to make it larger than the text describing the track
- all of my listening-music is under the same folder, with no other audio that I don't want to include
I just want it to wait until I tell it which folders to add, and only scan when I ask it to, and show the tracks to me in a grid with name and artist columns, and album if available, and let me add semantic data as the spirit moves me which can then also become columns (optionally)".
Why is that apparently not the way they think it should be?
(P.S. why can I still not put hashtags in the CW?)
-
I'm now wishing for the ability to search things I have boosted, specifically all media I've boosted.
The "media" feed under my profile seems to be only original media (a good default), but... it's not even in chrono order, as far as I can tell. :-P
-
It's kind of annoying that you can't just edit IP blocks. You have to delete them and then recreate them -- with no easy way to keep a record of the IP address/range involved, much less the notes. #mastodon #softwareGripe
-
Technical Note: When a site moderator deletes flagged posts in a report via the admin panel, that also deletes their content from the report.
This is bad design, because it means that deleting the posts to resolve a report also removes the record of what was in them That record is often necessary in order to:
- understand (later) why they were deleted
- notice if they're part of a pattern which the reported user is continuing
- share knowledge among multiple moderators
Does anyone know if there's a trouble-ticket for this?
-
You'd think disk-usage scanners would be smart enough not to scan mounted volumes when you scan the main filesystem, or at least offer to skip them as an option. #softwareGripe
-
(Edit: it's been pointed out to me that HTML works for both of those. I probably knew that but forgot. It still doesn't do
<small>, but I guess eh whatevs.)original post
I remain annoyed that MastoGlitch Markdown apparently only supports superscripts for connected characters (single words), and also does not appear to support subscripts.
...mainly because I'd like a way to make text smaller -- so I guess I'm actually annoyed that it doesn't support that, or at least let me fake it via the back-door of superscripts.
(Steve Martin as hacker / Ant Man: "Okay, I'm in -- now, let's get small.")
-
(UPDATE: I found a way to make the search work; see end.)
Someone made a bad UI design choice...
Just now I do a search/replace in kate (across multiple files) where the replacement string happens to include
\t(it's part of a PHP namespace).Kate replaces those two characters with a TAB, in all the matching files.
Searching for
\tdoes not find the affected text.Copying the tab from the document and searching for that does not find it either.
I can't even search for [space]
tbecause it used a TAB character, not spaces, in the output -- even though I have editing configured to always use spaces.ARRGGGH. There are probably about a HUNDRED FILES affected. Maybe more.
Update: apparently I needed to paste the tab into a longer string or something, and then it will find matches.
This is still bad UX, though: the "search" text and the "replace" text shouldn't follow different rules.
-
I swear, at some point I am going to write a web server which provides a utility that will answer the question "for this URL, show me the exact sequence of events and configuration settings (and where they were defined) which led to returning an error". #softwareGripe
-
Is it part of Google's evilness that there's literally no way to delete some types of files from your account, in order to make room?
@Harena's Google has become full -- mostly with old photos, but some of them are in "Spaces" and there's no apparent way to delete them (it's not clear if deleting the Space will even do it) -- and deleting from Photos doesn't even seem to free up space, though the UI makes it hard to tell what's going on.
It used to be -- I think? -- that you could access GDrive via WebDAV... or did I imagine that? (Not that this would solve the problem unless GDrive exposes all the other storage as folders... which also isn't clear.)
(Yes, I'm looking at
rclone, and it may be able to help -- but it has to go through the API, and the API has "limitations" including what it will let you delete.) -
Some docurantation about checkbook software: https://wooz.dev/2025/02/05/personal_money_management_software
-
Trying to figure something out is extra frustrating when what the documentation describes isn't what's actually there.
(Yes, I checked the version numbers.)
Has anyone ever actually succeeded in writing {macros or BASIC code or forms} in #LibreOffice?
-
...and yet it constantly happens that while I'm typing, I accidentally brush against the touchpad and suddenly I'm typing somewhere completely different (often obliterating a whole section of text because it interpreted the brush as a select for some reason, even though I didn't actually click).
I have to ask: what does "while typing" even mean, then? Is it, like, the 50 msec during which I'm pressing down each key, and then the tenths or hundredths of a second between keystrokes are counted as "not typing"? Why isn't there a "delay" or "hysteresis" setting? :blobcatupsidedown:
-
I really, really need a file-browser that will:
- ask before moving or copying anything in response to a drag/drop
- keep a log of recent operations, so that if e.g. I accidentally move a folder into another folder, I'll know where to find it so I can move it back
-
No, seriously -- why does MATE still default to thinking that the
.pubextension means "Microsoft Publisher"? It's the extension for public keys. Microsoft needs to butt out. #softwareGripe -
If I open a text article, the page doesn't automatically start reading it for me -- scrolling it downward slowly at the rate it thinks I read. I have to actually scroll through it, once I'm done with what's visible, to see the next bit.
Videos should behave likewise.
Just sayin'.
(This is me complaining about YT's habit of automatically starting a video when I open it in another tab for later viewing, in case that's not clear. DON'T PLAY UNTIL I PRESS THE PLAY BUTTON, you blasted bucket of binary bolts and bits!!)
-
I find it ironic that when I was writing DOS text-mode apps in the 80s and 90s, I had much more control over the screen than I do when writing text-mode apps in Linux today.
(Is this a #softwareGripe?)
-
I figured out why a thing was happening, and now I have to do a smol #softwareGripe:
If the user is saving a file somewhere, and opens folders in a tree inside the file-listing window and then selects a folder, the folder into which the file should go is the bloomin' selected folder, ye half-wit excuse fer a GUI!!
Instead, it puts it at the root of the visible tree, ignoring any navigation I've done (unless I double-clicked on a folder to make it the (visible) root.)
I think this is a KDE issue, if "Portal" is KDE's file-saving dialog service. It's not 100% clear because there is no "about" feature on the dialog.
-
It's really kinda QUITE ANNOYING that...
- #MediaWiki chose to replace
<source>with<syntaxhighlight>... - ...and provided no way to invoke it via a more memorable/typeable template.
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
<syntaxhighlight lang=php inline>{{{1}}}</syntaxhighlight>, for example, invoking the template always gets you{{{1}}}on the page. ...and there's no alternative{{#syntaxhighlight:...}}syntax, as far as I can tell.)
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
Seems like there really ought to be a way to invoke any XML-style tag in a way that means "parse this after template argument replacement, please".
- #MediaWiki chose to replace
-
It's really kinda QUITE ANNOYING that...
- #MediaWiki chose to replace
<source>with<syntaxhighlight>... - ...and provided no way to invoke it via a more memorable/typeable template.
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
<syntaxhighlight lang=php inline>{{{1}}}</syntaxhighlight>, for example, invoking the template always gets you{{{1}}}on the page. ...and there's no alternative{{#syntaxhighlight:...}}syntax, as far as I can tell.)
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
Seems like there really ought to be a way to invoke any XML-style tag in a way that means "parse this after template argument replacement, please".
- #MediaWiki chose to replace
-
It's really kinda QUITE ANNOYING that...
- #MediaWiki chose to replace
<source>with<syntaxhighlight>... - ...and provided no way to invoke it via a more memorable/typeable template.
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
<syntaxhighlight lang=php inline>{{{1}}}</syntaxhighlight>, for example, invoking the template always gets you{{{1}}}on the page. ...and there's no alternative{{#syntaxhighlight:...}}syntax, as far as I can tell.)
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
Seems like there really ought to be a way to invoke any XML-style tag in a way that means "parse this after template argument replacement, please".
- #MediaWiki chose to replace
-
It's really kinda QUITE ANNOYING that...
- #MediaWiki chose to replace
<source>with<syntaxhighlight>... - ...and provided no way to invoke it via a more memorable/typeable template.
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
<syntaxhighlight lang=php inline>{{{1}}}</syntaxhighlight>, for example, invoking the template always gets you{{{1}}}on the page. ...and there's no alternative{{#syntaxhighlight:...}}syntax, as far as I can tell.)
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
Seems like there really ought to be a way to invoke any XML-style tag in a way that means "parse this after template argument replacement, please".
- #MediaWiki chose to replace
-
It's really kinda QUITE ANNOYING that...
- #MediaWiki chose to replace
<source>with<syntaxhighlight>... - ...and provided no way to invoke it via a more memorable/typeable template.
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
<syntaxhighlight lang=php inline>{{{1}}}</syntaxhighlight>, for example, invoking the template always gets you{{{1}}}on the page. ...and there's no alternative{{#syntaxhighlight:...}}syntax, as far as I can tell.)
- (...because angle-bracket tags parse the template parameter-markup before the markup is replaced with the template arguments, so if I do a template that's
Seems like there really ought to be a way to invoke any XML-style tag in a way that means "parse this after template argument replacement, please".
- #MediaWiki chose to replace
-
(update: one of the "alternatives" does the cascading menus -- but I still don't understand why anyone would want to have them condensed by default. If you don't want them condensed, then don't make submenus.)
original post
...wait... so, if you edit your app menu in #KDE to put multimedia apps in separate folders (audio, video, players) to make them easier to find instead of having them all in one large folder, KDE still shows them all as one large folder?
The heq. 💢
-
...and it looks like uploading to YT Music is kind of useless, because:
- (a) you can't add a description or anything, much less a cover image
- (b) the track isn't shareable; when added to a playlist, it disappears from the public version of that playlist.
Guess I'll have to make a still-video for it...
-
CW: arcane but critical SSH Agent detail
TIL: Even if you can
sshinto a remote server keylessly, without specifying which keyfile to use, the SSH Agent -- which is apparently what is available via API, i.e. used by other programs that are notsshitself -- still needs to have your private keys added to it explicitly before it can connect the same way.Fortunately, this is easy to do:
ssh-add ~/.ssh/{filename}I guess I'm just not understanding why
sshand SSH Agent don't use the same keys by default.(Is this a #softwareGripe?)
-
Somehow, I feel like I shouldn't have to spend hours adding debugging code in order to figure out that the reason a site isn't loading is because it's unable to create a file because the disk-quota has been exceeded.
If it had even told me that it was having trouble creating a file, that would have been helpful -- but no, just blank screen. #phpbb #softwareGripe
Between this and recent events, I've decided that disk-quotas are a daft/capitalistic way of managing a resource that really should be treated more like a commons, and I'm not going to use them on my web servers anymore. It would have saved me many hours if I had never set a quota-limit in the first place.
-
Love it when a confirmation code takes forever to come in, so I go off and do something else while I'm waiting and by the time I check back and it has come in, it has also expired.
-
"luckyBackup" lets me select remote locations via sftp, but then refuses to do the command because (it says) the source doesn't exist.
At least it will compose the necessary rsync command for me.
...but, like, #WTF -- how could you write an entire GUI for rsync, one of whose big strengths is doing backups involving remote locations, and not work properly with sftp (while still fully supporting the necessary browsing operations)?
...and why is it including the "--delete-after" flag when I'm doing a backup?
Apparently it hasn't been maintained since 2018, so I can't be too upset. Add this to the list of projects that make me wish I had time to learn how to do C++ GUI stuff in Linux, so I could tweak it (and fork it to Git)...
-
...and this -- this , friends -- is why glob expansion is a Bad Idea:
rm *.pdf-bash: /usr/bin/rm: Argument list too longBackground: For some unknown reason, #Syncthing decided to make 1.9 TB of backups of four documents (2 PDFs, a PNG, and an ODG), none of which were over 10 MB in size.
(According to the settings, Syncthing is only supposed to keep 10 versions of files in that particular share. ...and I don't think we ever made more than 1 or two edits to any of the files in question; certainly not all of them.)
...and as you may or may not know, when you use filename-wildcards ("
*" and/or "?") in a command viabash, it automagically generates a list of the matching files and replaces that term with the entire list.This has always seemed to me like very bad praxis, but here we are.
I'm currently using Caja (GUI file-browser) to delete the files; it's got about an hour left to go.
-
When drag-dropping anything. a file manager should do one or both of the following:
- always ask to confirm the action
- provide a "what did I just do??" feature, with the option to undo whatever it was
I'm getting seriously tired of accidentally moving stuff and either not knowing what I moved and where it ended up or else not even realizing I did anything except things suddenly break and I don't know why.
Lookin' at you (yet again), #Caja.
(Also, WTF is it with the mode you get into where all the cut/copy options on the right-click menu are greyed out?? If that's a legit thing, I have no idea how to get out of it except to close that Caja window, and sometimes all of them.)
-
I'm sorry, but Markup's syntax for numbered lists is FECKING TERRIBLE.
It simply should not happen that I can type something like:
4. (item four)
5. (item five)...and get this:
- (item four)
- (item five)
-
CW: Apparently I don't actually understand .gitignore syntax.
(...and apparently CWs render HTML as plaintext, so: .gitignore syntax)
What I want is to exclude any folder that has "RETIRED" in the name -- even if it's prefixed (I sometimes stick a "0 " at the beginning so it will sort to the top rather than being in the middle of the other, active folders).
Any of these lines will exclude any folder that is named exactly "RETIRED":
/*RETIRED/*RETIRED//*RETIRED/**/**RETIRED//0?RETIRED/0?RETIRED/So far so good... except none of these will exclude folders named "0 RETIRED", which they all kind of seem like they ought to do.
I've also tried the following:
*RETIRED**/*RETIRED...but neither of them exclude anything.
What am I doing wrong? Is the documentation even correct?
-
#softwareGripe Video-recording apps that don't warn you when audio or video or both aren't actually producing any output to the file.
Like, if the file is 264 bytes and never changes, that should be a red flag that the user might want to know about. Also, if the output file is zero bytes and never changes. That's probably also bad.
Lookin' atchoo, #Webcamoid 8.6.1.
-
#softwareGripe
file-scanning software which somehow find a file which it then decides doesn't exist, causing the scan to fail on that file, at which point it quits scanning. No way to resume. #WTF(How about, you know, just skipping that one??)
Lookinatchu, #FileLight :kestraglow:
-
#softwareGripe disk-space-usage analyzers that can't tell the difference between local folders and folders that are actually links to folders on another disk
#FileLight, I'm lookin' at you :kestraglow:
Yes, you can list folders to exclude -- but that's imprecise and also I shouldn't have to.
(mumble mumble reliably work over sftp, too, dang kids get off my lawn)