« first day (2157 days earlier)      last day (1667 days later) » 

1:00 PM
[ SmokeDetector | MS ] Potentially bad keyword in answer (1): Using Qt To Draw the Graph of Sin(x) by raju kumarkhaniya on stackoverflow.com
 
@double-beep that was a test post I made myself to test the notification system
 
naa- feedback received
[ SmokeDetector | MS ] Url-only title, bad keyword in body, bad keyword in title, bad keyword in username, blacklisted website in body, +11 more (1180): supplementgear.com/balanced-max-keto/ by Balanced Max Keto on askubuntu.com
 
the vegas slots one
 
you were the author of it?
 
1:04 PM
yup
Spanish.SE allows answers from unregistered users
I just had to answer in a private tab
 
yes, but wasn't your IP blocked?
 
I really hope it wasn't 😅
Haven't really tried to ask or answer since then
let me check
I edited a CW answer after that post with my actual user, no problem
@double-beep I guess it wasn't
 
[ SmokeDetector | MS ] Pattern-matching website in body (97): The Premier Healthcare IT Software Vendors - HealthAsyst by HealthAsyst on apple.SE
 
was this answer deleted as spam?
 
1:09 PM
It was deleted by a mod. It currently appears as deleted with the usual "This answer was marked as spam or rude or abusive" but it was not an automatic deletion.
(the mod was aware of the test)
 
[ SmokeDetector | MS ] Username similar to website in answer (58): Why does my dog push his food out of his bowl? by dadogsyard on pets.SE
[ SmokeDetector | MS ] Link at end of answer (60): Firefox Download from Command Line by TheUnseen on superuser.com
 
@SmokeDetector Undisclosed affiliation, modflagged
 
@MichaelDodd That really wasn't undisclosed.
 
1:24 PM
Although you are quite right that the answer can't stand on its own.
 
@terdon I've made my comment. I felt the "my" referred to the dog food they had purchased and is now feeding to the store, not a link to the blog
I've re-read it several times and don't see how it can be interpreted as affiliation
 
[ SmokeDetector | MS ] Url in title, body starts with title and ends in url, potentially bad ns for domain in body, potentially bad ns for domain in title, potentially bad keyword in body, +1 more (198): www.allwoodbicycle.com/ by Wrive1 on academia.SE
tpu- feedback received
 
I didn't go red-flag, and I've withdrawn the modflag, but to me that's not clear cut at all.
 
@MichaelDodd Well, the my made it clear for me. But when I reacted, I had misread and thought the post was more than just a signpost to the blog. I thought the answer could stand alone.
 
[ SmokeDetector | MS ] Link at end of answer (60): Easily accessible snow in Nepal by Pasang Lama on travel.SE
 
1:39 PM
[ SmokeDetector | MS ] Bad keyword in answer, toxic answer detected (166): Why does Puck refer to Emilia as 'My Daughter'? by asd on anime.SE
tpu- feedback received
tpu- feedback received on Easily accessible snow in Nepal [MS]
tpu- feedback received on RPA Express SwitchWindowAction error [MS]
[ SmokeDetector | MS ] Potentially problematic ns configuration in answer (1): One page website menu too long by Amelia jones on ux.SE
ignore- feedback received on How do you change age settings on ROBLOX? [MS]
[ SmokeDetector | MS ] Potentially bad keyword in body (1): python, BeautifulSoupでgoogle検索のタイトル取得できない ✏️ by user9191983 on ja.stackoverflow.com
[ SmokeDetector | MS ] Potentially bad keyword in username (1): Laravel+React+Reduxを用いたJWT認証のリロード(f5)時の認証エラー by khkhkh on ja.stackoverflow.com
[ SmokeDetector | MS ] Potentially bad keyword in answer (1): Daily stock data request by James C on economics.SE
ignore- feedback received
[ SmokeDetector | MS ] Potentially bad keyword in body, potentially problematic ns configuration in body (2): fashionfabric online fabric store by fashion fabric on stackoverflow.com
tpu- feedback received
fp- feedback received on javascript shopping cart email checkout [MS]
[ SmokeDetector | MS ] Potentially bad keyword in answer, potentially problematic ns configuration in answer, username similar to website in answer, blacklisted user (131): Unable to change width and height on image by fashion fabric on stackoverflow.com
tpu- feedback received
 
!!/report stackoverflow.com/a/58378375 "Undisclosed affiliation; user is the author of the article"
 
thanks
 
I can't type today, it seems
 
[ SmokeDetector | MS ] Manually reported answer (94): Make flutter project swift compatible AFTER creation by Serge Shkurko on stackoverflow.com
tpu- feedback received
naa- feedback received
 
2:21 PM
!!/report stackoverflow.com/a/58378305 "Undisclosed affiliation; user is the author of the article"
 
The untagged domains review queue is (finally) empty!
 
2:37 PM
[ SmokeDetector | MS ] Potentially bad keyword in answer (1): Add Visa or master card Payment integration to eCommerce android app by mastodon on stackoverflow.com
 
@AndrewT. Post 1: Already recently reported [ MS ]
 
of course...
 
@double-beep I suspect the lack of autoflagging is a result of the issue that MS is having with creating a record of new domains. All of the posts which I've checked which weren't autoflagged had at least one new domain in them. I suspect that the error in creating a new domain resulted in halting the rest of the processing of an incoming report. I pinged @ArtOfCode about the domain creation issue earlier, but wasn't aware of the autoflagging problem at that time.
I can take a more detailed look at the problem, but I'll have to learn some more Ruby to figure out why what's being done in that section of code is being done, so I can determine an appropriate solution that doesn't cause more harm.
 
2:49 PM
That... that makes this a more severe issue. I'll be home in the next 20 or so, and I'll take a look.
 
@double-beep Thanks. The process was locked up, consuming all of a CPU. I'm not sure what the problem was, but I saved the console output to look at later.
@geisterfurz007 Yes, SD should catch those. Reporting them is OK.
@ArtOfCode Thanks. I was hoping you'd be around. While I like learning new languages, and I intend to learn more Ruby/Ruby on Rails and how the MS code is interacting, I wasn't looking forward to doing so under pressure of this bug. :-( I expect I know a couple of ways to fix it, but to determine the potential side-effects of each method of resolution appeared to require a better understanding of the surrounding code than I currently have.
 
[ SmokeDetector | MS ] Potentially bad keyword in answer (1): Python excersise gives wrong answers by CountryTk on stackoverflow.com
tpu- feedback received
 
3:06 PM
@Makyen it's also not obvious what's causing the bug (or I'd've fixed it on the move this morning), so it's going to need a development environment to diagnose
 
@ArtOfCode Then, I'm glad I didn't try to go for it. :-) Thanks for taking a look at it.
 
Running Regexp.new(nil) in IRB gives me the same error, which does make me wonder if that's it, but I'm not seeing any records that have a nil regex field
 
fp feedback received on Cannot download youtube-dl [MS]
 
Are there any deleted/completely nil domain_groups records?
 
3:17 PM
[ SmokeDetector | MS ] Manually reported answer (batch report: post 1 out of 3) (94): Are holidays included by default in SharePoint 2013 calendar? by Max on stackoverflow.com
[ SmokeDetector | MS ] Manually reported answer (batch report: post 2 out of 3) (94): SharePoint Online: Webpart that opens links in the same tab in modern pages by Max on stackoverflow.com
[ SmokeDetector | MS ] Manually reported answer (batch report: post 3 out of 3) (94): Creating a top navigation menu by Max on stackoverflow.com
 
@Makyen That's not possible - at least the id field will always have a value, because it's the primary key and cannot be null. If there's a record, even if everything else is null, it would show up in the table.
I've just added a migration to add domain groups to the data exploder, so that should help diagnose.
 
[ metasmoke ] ci/circleci success on 29ff8b5: Your tests passed on CircleCI!
 
...really?
 
[ metasmoke-deploy ] deploy started by ArtOfCode
 
@ArtOfCode Sorry, that's basically what I meant.
 
3:22 PM
[ SmokeDetector | MS ] Potentially bad ns for domain in answer (1): Continue debugging after SegFault in GDB? by MicStretch on stackoverflow.com
 
@ArtOfCode I fixed the JS issue. It's been nice about the random rails tests issue.
... for the last couple of commits.
 
sweet, cheers
 
np.
 
[ SmokeDetector | MS ] Potentially bad asn for hostname in answer (1): ShellExecuteEx error 299 when launching installer created with inno setup by Vas on stackoverflow.com
 
BTW: My current theory on the random rails test issue is that it's having a problem loading the database when switching between test workers. Adding local_infile: true to database.yml gives permission for loading (current error), but just changes the error to a file not found error. I suspect (hope) that changing the configuration so that it's only using one worker for the tests may resolve the problem, but I haven't gotten the config changed to just one worker.
I tried a few ways, but wasn't successful and had to switch to doing other things.
 
3:34 PM
[ metasmoke-deploy ] deploy started by ArtOfCode
[ metasmoke ] ci/circleci: rubocop failure on ab074d6: Your tests failed on CircleCI
[ SmokeDetector | MS ] Potentially bad keyword in answer (1): How to get ERD diagram for an existing database? by feliciano popov on stackoverflow.com
 
3:50 PM
[ SmokeDetector ] SmokeDetector started at rev f2d48b7 (Makyen: Add 46061: # La Tertulia to rooms.yml -autopull) (running on Makyen/EC2-linux, Python 3.6.8)
Restart: API quota is 16290.
[ SmokeDetector | MS ] Potentially bad ns for domain in answer (1): Continue debugging after SegFault in GDB? by MicStretch on stackoverflow.com
 
@Makyen looks like we're going to give RHEL a test drive now lol
 
[ SmokeDetector | MS ] Potentially bad keyword in answer (1): Travis and Firebase: deploy only changed functions by Basti Humann on stackoverflow.com
 
@ThomasWard Cool. Hopefully it will be fun. :-)
 
well I already have RHEL setup
i just have to set up the environment for SE heh
 
[ SmokeDetector | MS ] Potentially bad asn for hostname in answer (1): Samsung Easy document Scanner won't communicate with scanner on OS X Catalina 10.15 by dub on apple.SE
 
3:58 PM
UNFORTUNATELY it won't be a nice venv but meh
it'll be in userspace xD
 
Not the best thing, but it should work. :-)
 
[ metasmoke-deploy ] deploy started by ArtOfCode
[ metasmoke ] ci/circleci: rails-tests failure on 72f6400: Your tests failed on CircleCI
[ metasmoke ] ci/circleci: rubocop failure on 72f6400: Your tests failed on CircleCI
 
@walen Yes, the !!/standby was what caused your original !!/notify to no longer be effective. Settings for notifications are per-SmokeDetector instance. Unfortunately, the active instance changed between when you ran your first !!/notify and when you checked that time.
@walen I've added room 46061, La Tertulia, to SD's config for posting reports for Spanish Language.
 
... 2fa is set up on the GitHub...
 
Currently, commands (and chat based feedback) are disabled in La Tertulia. Are you wanting to be able to give commands to SmokeDetector in that room? That includes being able to give feedback via the chat interface. If so, do you want the people authorized to run such commands to include anyone other than moderators? This chat message has a brief description of the more common configurations which are used for command permissions in various rooms.
 
4:07 PM
this means I can't easily set up my SSH keys :?
 
Regardless of being able to run commands in the room, userscripts (i.e. FIRE), once changed to run in that room, will be able to be used for providing feedback. Personally, I use FIRE, or directly on metasmoke, to give feedback, 99%+ of the time.
@ThomasWard You can set up SSH keys, but you shouldn't need to. SD now preferentially uses the credentials stored in the config file for all GitHub interactions.
 
@Makyen ... for git
git clone pulls from SSH
are yo usaying this is bypassed?
 
[ metasmoke-deploy ] deploy started by ArtOfCode
 
@ThomasWard For the git commands that SD runs, it uses the username and password credentials in the config file, if they exist. Having SSH configured is no longer a requirement.
 
@Makyen Appreciated, thank you very much :)
 
4:10 PM
[ metasmoke ] ci/circleci: rubocop failure on 0728617: Your tests failed on CircleCI
[ metasmoke ] ci/circleci: rails-tests failure on 0728617: Your tests failed on CircleCI
 
@Makyen hm, interesting. But until we have a working deploy script I still have to clone it - SSH is the only way that has SO FAR worked because of the proxy setups here at this environment
meh, i just recovered the SSH keys from the 'template' container I've been using to deploy my Smokeys and grabbed it
seems to be working
 
@walen np. While it might not be clear, the next couple of messages from me after that one are also about setting up SD in La Tertulia. Basically, if you want SD to listen for commands in that room, I need more information.
 
@Makyen yup, I was just thinking about it (i.e. do we want commands or not), sorry
This whole thing is rather new for us. By pure luck I came across a SD message in Tavern about a spam answer in es.stackoverflow, I checked and it was indeed spam and I though "hey, this would be cool if it worked for Spanish.SE too"
 
[ metasmoke-deploy ] deploy started by ArtOfCode
 
I think most people on Spanish.SE didn't know about it (maybe fedorqui but he seems to be offline today)
 
4:16 PM
[ metasmoke ] ci/circleci: rails-tests failure on 9fff942: Your tests failed on CircleCI
 
@Makyen bluh I broke it, it's not letting me login on the new env.
i guess ratelimiting, so I'll wait :)
 
So... I'm leaning towards just getting used to having SD leave reports in La Tertulia and, if some day we decide we need it to listen for commands, we can reach out to you again. @Makyen Would that be OK?
 
[ metasmoke ] ci/circleci: rubocop failure on 9fff942: Your tests failed on CircleCI
 
@ThomasWard For git stuff outside of SD, that's going to depend on your config, although I wouldn't have expected SSH to be required, just a non-user-authenticated HTTPS connection. I'm glad you were able to use your old SSH key. That helps/makes life easier.
@ThomasWard ouch.
 
(I say that having in mind that we can use FIRE in the mean time, if I understood correctly.)
 
4:18 PM
@walen no problem. It's not something that I need a response for immediately. Just something to think about as to what you want.
 
@Makyen yeah i have an SSH key that's been there since early on so xD
@Makyen yep. i'll just wait 20 minutes before starting it manually in tmux
 
[ SmokeDetector | MS ] Few unique characters in answer, offensive answer detected, potentially bad keyword in answer, toxic answer detected (246): How to make turtle move continuously without multiple key presses? by user12215712 on stackoverflow.com
tpu- feedback received
 
@Makyen I see, thanks again.
 
ayyyyyy there you are
 
4:22 PM
@Makyen Sending commands from CHQ would still work anyways, right? I mean, if we get a report in La Tertulia and some mod wants to give feedback, could they just get in here and issue any necessary commands?
 
@walen Yes, you can use FIRE. FIRE, and AIM, will need to be updated to include La Tertulia as a room in which it runs, but that's easy to do. I'll see if I can do so later today (I have some other updates, which I'll try to bundle at the same time). Meanwhile, if you want, you can just add the room number to the // @include /^https?://chat\.stackexchange\.com/(?:rooms/|search.*[?&]room=)(?: line near the top of the file in your local copy.
@walen Yes.
 
@Makyen if you want to failover to Chaos you can, right now i'm running it on my residential connection so the IP isn't static, it might work well, but i'll let you decide
it might flap a little which is why i'm hesitant to drop right into Full mode (it's standby now)
 
@ThomasWard Sure. Let's try it.
@ThomasWard Hmmmm... Seems like it would be good to try it.
 
@Makyen 👍
 
@Makyen cool we can do a MS failover then
hang on a moment
!!/location
 
4:25 PM
@ThomasWard Makyen/EC2-linux
teward/Chaos_RHEL received failover signal.
 
!!/standby Makyen
 
Makyen/EC2-linux is switching to standby
[ SmokeDetector ] SmokeDetector started in standby mode at rev f2d48b7 (Makyen: Add 46061: # La Tertulia to rooms.yml -autopull) (running on Makyen/EC2-linux)
Restart: API quota is 19999.
 
hmm well lets see how this behaves
looks stable for now
 
That's a plus. :-)
 
4:29 PM
oops
 
@ThomasWard No such command 'locati0on'.
 
!!/location
 
@ThomasWard teward/Chaos_RHEL
 
i'll rename it later to just Chaos
 
tpu- feedback received
 
4:29 PM
but lets see how stable it is now :)
if it seems stable I'll tweak the config and kick it into reboot to get the new name
but at least this is running stable (RHEL8 running in a pyenv 3.7.4 python env!)
 
@ThomasWard Nice!
 
Let's see if git commands work. :-)
!!/watch- dillionharperblog\.com
 
Looks good.
 
4:33 PM
!!/watch lanarhoadesblog\.com
 
looks like its working
yay!
lets it run
 
wait why did that happen twice
:|
@Makyen let me know if it dies off, glad you've got backup instances. Looks like Supergate ran out of cloud credits on the hosting provider so I'mma have to reup that next paycheck, so we'll stick with Chaos for now
(this is aptly named - compared to Ubuntu, RHEL is a little chaotic with where things're stored xD)
 
4:36 PM
@ThomasWard Yeah, there's a bug that sometimes results in the message, if not the reboot/reload, happening twice, or more. I haven't taken a close look at it yet, but it's not just your instance.
 
ah OK
 
yeah i've been boycotting chat for a bit
because 'reasons'
but meh
 
@ThomasWard Ouch. At least you've got Chaos available.
 
good thing I have remote access to my environments at home xD
@Makyen yeah, that and the unused 'stargate' LXD on the business class network
which I think is just 'slower' on the upstream - slower than it wants
which is odd, a 20Mbps should be MORE than sufficient for what it's doing
(down is 150)
meh i haven't ruled out my firewall appliance, which I have to replace because it doesn't have IPv6 DHCP-PD for prefix delegations
 
4:39 PM
@ThomasWard Yeah, that helps a lot. I've definitely used access to home a few times myself.
 
@Makyen heheh, indeed. I made it better though I have remote access to both my residental personal and by at-apartment business class network from the residentail network's VPN
there's a IPSec tunnel between the two networks that lets my management nets through xD
 
Sounds like a good way to configure it.
 
well i did one better lol
both firewalls exist on the same virtual cluster
so there's a private network between the two firewalls that is ONLY for the inter-network IPSec tunnel and is FAST because it's internal to the VMware cluster
 
Nice!
 
yup. the internal interconnection between the boxes is all gig, so it's not ULTRA FAST but it's good.
so i've got access to both sides. Gonna be a pain to config this though in pfSense to replicate the existing configs but >.>
... kinda need DHCP-PD :P
 
4:45 PM
:-)
 
tp feedback received on Can you help me to improve my article? [MS]
[ SmokeDetector | MS ] Bad keyword with email in answer (85): Travel by ship from India to Canada by Richa Rishabh on travel.SE
tp feedback received on Chat bug? Pings and replies time out [MS]
fp- feedback received on Android Youtube Download mp3 [MS]
[ SmokeDetector ] SmokeDetector started in standby mode at rev f2d48b7 (Makyen: Add 46061: # La Tertulia to rooms.yml -autopull) (running on Makyen/EC2-linux)
 
@Makyen fixed. There was a record with a nil regex, but it wasn't showing up anywhere. I'm going to have a look at why.
 
5:02 PM
@ArtOfCode Great! Thanks for taking a look at that. It's unfortunate that it was such a pain in the rear.
 
Ah, gotcha. It was an INNER JOIN, not a LEFT JOIN - groups with no domains wouldn't show up
 
[ metasmoke-deploy ] deploy started by ArtOfCode
 
Ahhh... good. You found it.
 
[ metasmoke ] ci/circleci: rubocop failure on ac0203b: Your tests failed on CircleCI
[ metasmoke ] ci/circleci: rails-tests failure on ac0203b: Your tests failed on CircleCI
naa feedback received
[ SmokeDetector | MS ] Potentially bad asn for hostname in body (1): Publicar pagina web wordpress by Felipe Correa Kusjanovic on es.stackoverflow.com
[ SmokeDetector | MS ] Potentially problematic ns configuration in body (1): Php page not working when the SQL tables are populated by Paul Rodríguez on stackoverflow.com
 
5:12 PM
tp- feedback received
 
@Makyen well... looks like Chaos is stable :)
 
@ThomasWard Yeah, it looks like it's running well.
 
[ SmokeDetector ] SmokeDetector started at rev 7cecdd9 (SmokeDetector: Auto watch of lanarhoadesblog\.com by double-beep) (running on teward/Chaos, Python 3.7.4)
 
^ that was my doing, so we know it didn't crash :P
 
Restart: API quota is 19440.
 
5:16 PM
That's good to know :-)
 
tp- feedback received
 
!!/watch- dillionharperblog\.com
 
@Makyen That pattern looks like it's already caught by Potentially bad keyword in body and Potentially bad keyword in answer; append -force if you really want to do that.
 
Yeah dillon harper is an actress..of sorts
don't ask me how i know
 
@SmokeDetector OK. That's what I thought, but you only reported it as a blacklisted user.
 
5:18 PM
!!/test-a dillionharperblog.com
 
> Potentially bad keyword in answer
----------
Potentially bad keyword in answer - Position 0-21: dillionharperblog.com
 
wonder what the keyword is lol
 
Ahhh... OK FIRE was showing the first report, not the second, which had the content edited out, so no detecting the domains.
 
@Makyen which is their moderator doing a stupid
because their own moderator removed the content BEFORE deleting
which fudges up spam detectors, etc.
like ours
... perhaps we should confront about it xD
 
I agree that it would be better to delete, then edit, if they are going to edit. However, it was already spam/R/A flagged, so the content would have been hidden anyway (i.e. I don't see the point of editing it just to delete it).
 
5:41 PM
@ThomasWard Hmmm.... that reminds me. We can probably drop Python 3.5 support/testing at this point. All of the recently active instances (last 3 months) are on at least 3.6.7. Well, actually, I can't tell for iBug/SandyShores, but I'd be surprised if it wasn't at least 3.6. @iBug What's your feeling on dropping Python 3.5 support/testing for SD?
 
@Makyen I'd say we add a platform check at the beginning of runtime then though
to validate we're on proper Python supported versions
if we drop testing we have to enforce minimumPythonVersion
not... sure if we do that already been a while since i glared down at the code :P
 
With this ^ if you hasn't already, I say get rid of Windows support too. I made a big deal of it a couple years ago. That was...stupid.
*you = we
 
@ThomasWard True, although we currently only support 3.5+. I'm not sure if we check for it, but we should.
 
well that's not hard
@Andy Oh, so I can smack you now for this xD
pulls the code
 
5:52 PM
@ThomasWard Yes. I'm sorry :(
>>> import sys
>>> sys.version_info
sys.version_info(major=3, minor=6, micro=8, releaselevel='final', serial=0)
@Makyen ^ Checking version
 
we can do that or we can use the platform which returns a straight tuple
ahhhh wait a minute
we use platform in nocrash
that makes things easy
 
>>> import platform
>>> platform.python_version()
'3.6.8'
 
fp- feedback received
 
@Andy I'd prefer to leave Windows support in. In fact, we would have been in significant hurt over the last couple days if we didn't have Windows support. We wouldn't have had an SD instance to cover all of the weekend. There really isn't all that much that needed to be changed for Windows support, at least by the time I started looking at it. As far as I know, it's currently running fine on Windows.
 
>>> platform.python_version_tuple()
('3', '6', '8')
 
5:54 PM
fp- feedback received on Conversion of PDF to EPub [MS]
 
@Andy or my way...
>>> import platform
>>> tuple(int(x) for x in platform.python_version_tuple()) > (3, 5, 0)
True
because the tuple values will actually matter
and this returns (3, 7, 4) in the LHS there for my tests
if you invert the test
>>> tuple(int(x) for x in platform.python_version_tuple()) < (3, 5, 0)
False
so we could add that to the nocrash.py where we already import platform to test if Windows is in use
 
[ SmokeDetector | MS ] Bad keyword in answer (86): How to green and thicken my grass by sriv1210 on gardening.SE
 
Yup, that works too. I was giving a couple options. :)
Plus, I like one liners
 
fp feedback received
 
Given that we don't support <3.5, we should at least test for >=3.5 and error-out if it's not >=3.5.
I'm fine with requiring >=3.6.8. The semi-recent 3.6.7 was Lunar Eclipse. Other than the unknown iBug/SandyShores, everything else was >= 3.6.8.
 
6:18 PM
[ SmokeDetector | MS ] Potentially bad asn for hostname in body (1): Woocommerce login not working on the first try by t_csba on wordpress.SE
 
6:32 PM
[ SmokeDetector | MS ] Link at end of answer (60): image compression by Fernando Palacios on stackoverflow.com
 
I'm looking forward to dropping support for P 3.5, but I'd rather be cautious. Debian Stretch and Ubuntu Xenial both ship with 3.5, and both are still supported now. Unless there's something preventing us from writing better code in 3.5, I see no benefit dropping support for it. The last time we dropped 3.4 was something like async (only available in 3.5+).
@Makyen Sandy Shores is in Docker with at least 3.7.3, not a problem.
 
fstrings
 
!!/watch squeezer\.tools
 
6:44 PM
BTW, I've always thought RHEL/CentOS are the most difficult-to-handle systems, and very likely Fedora as well. Why not make your life easy by using Ubuntu for desktop and Debian for server?
@Makyen I'm also fine with requiring 3.6 as I no longer have servers running older systems. All my machines are on Ubuntu Bionic/Disco and Debian Buster now, so I provide Python at at least 3.6.8.
 

« first day (2157 days earlier)      last day (1667 days later) »