Aleksandra Fedorova :fedora:

CI/Devops/Infra Engineer, working on RHEL & CentOS Stream
Fedora Ambassador, former FESCo and Fedora Council member
PhD in Geometry and Topology
she/her

Matrix: @bookwar:fedora.im

I'll leave this account for FOSS debates.

For more personal stuff I am going to use @bookwar

Feel free to follow one or the other, both or none.

--
I am not a native speaker, and I appreciate private notes correcting my English or German.

Public Key

npub12u3h3dchshl60nl3qwqylucf323l87d5yxhd403gsgq8jfjafqcsscq0uk

NIP-05 Address

bookwar_at_fosstodon.org@mostr.pub

Profile Code

nprofile1qqs9wgmckutctla8elcs8qz07vyc4glnlx6zrtk6hc5gyqreyew5svgxz0sk5

Author Public Key

npub12u3h3dchshl60nl3qwqylucf323l87d5yxhd403gsgq8jfjafqcsscq0uk

Show more details

Last Notes

2024-08-27 11:00:40

Hot take:
"town square" is not the best model for open public discourse, and should not be treated as the Holy Grail of social networking.

2024-08-17 16:04:22

- reply

Hm, this doesn't uniquely define the workload.
So it should probably be:
Workload is a *minimal* set of applications, so that if you remove them from the server, you stop paying for it.

2024-08-17 15:54:49

- reply

In the context of a single server,
Application is a leaf service. Nothing else depends on it.
Workload is a bundle of applications serving the purpose of the server. If you remove these services from a server, you wouldn't pay for running it.
Monitoring service running on a server is an application but not a workload.

2024-08-17 14:53:42

Let's play my favourite game "Define the obvious" again:
What is a *workload*?
Is every service running on a server a workload? How many workloads can a server run? Is workload an application or a function?
What is the difference between an application and a workload?

2024-08-17 12:20:38

- reply

@npub1wx0…cyga
I don't see a reason to bring discussion of harassment and "cancelling" into this conversation.
Maybe consider not jumping into a random conversation all defensive, when there is no one attacking.

2024-08-17 10:17:23

We didn't organise the Fedora booth at FrOSCon this year. But you still can get some leftover Fedora stickers at the Deutsche Welle booth.
And of course you can ask me anything about Fedora and CentOS.
#FrOSCon2024 #froscon

2024-08-17 07:28:54

- reply

@npub1jfd…s8ga
In many cases "Where are you from" question should not really be considered a small talk.

2024-08-16 21:13:48

I wish our language lessons in school were focused less on grammar and more on communication.
Instead of training the declinations, they would ask something like: rewrite this phrase to sound angry, unsure, sad, excited or childish. Than give the result to the classmate and ask how does it sound and compare the results.
Or write a clickbait headline, and then the informative one to the same article.
Or write a support request to IT.

2024-08-16 18:03:39

- reply

Just so you know, there is no hiking route in Germany which doesn't end in a Biergarten. No matter how much you need to climb to get there.
https://cdn.fosstodon.org/media_attachments/files/112/973/014/215/616/232/original/3d1e9db1b657edde.jpg

2024-08-16 17:55:37

- reply

https://cdn.fosstodon.org/media_attachments/files/112/972/982/472/897/020/original/e183528d9bcf1a07.jpg
https://cdn.fosstodon.org/media_attachments/files/112/972/982/929/115/606/original/2107671b2b09c024.jpg
https://cdn.fosstodon.org/media_attachments/files/112/972/984/225/756/551/original/019b69cb51fe2ad6.jpg

2024-08-16 17:15:42

Starting a little bit later than planned.
#FrOSConFriday
https://cdn.fosstodon.org/media_attachments/files/112/972/808/784/358/304/original/62a00ecd4bfef711.jpg

2024-08-13 23:06:35

- reply

@npub1rgx…6y88
This definitely deserves a project-wide celebration.
Thank you for driving this work.

2024-08-06 16:49:31

- reply

CC @npub1rnj…vtmy
Lesson in effective marketing strategies :)

2024-08-06 16:48:32

- reply

"..Only when they saw this film did they realise that it was about Rosetta, at which point we all took our jackets off to reveal ESA polo shirts, and then gave some talks on the mission, the film, and the connection between science fiction and fact. Quite effective: got us lots of press ;-)"

2024-08-06 16:48:06

- reply

From the comments 1/2
"...this was a kind of stealth project. We released a teaser trailer for the film in 2014 without any reference to ESA or Rosetta at all, but showing Aidan and getting a lot of viral interest. That then allowed us to invite a bunch of journalists to the British Film Institute in London for the premiere of "Ambition", again without any mention of ESA..."

2024-08-06 15:45:16

Thanks to #Rosetta mentions in my timeline, I re-watched the Ambition
https://www.youtube.com/watch?v=H08tGjXNHO4
Still good.

2024-07-22 14:03:23

- reply

@npub147r…rhvl
It seems that Crowdstrike was designed for the first usecase (gated updates), but applied for the second (direct consumption)

2024-07-22 13:21:44

- reply

I am really puzzled by the fact that the company with so many customers had not implemented some sort of sharding for its customers and didn't introduce a staged rollout.
How did it happen that everyone got hit by the update at the same time?

2024-07-22 13:19:19

- reply

Then next week the update would be pushed to Siberia (where you may need a day of travel to reach the support).
And only week after it would be rolled out to the Moscow region.
2/

2024-07-22 13:18:43

Years ago I worked in the IPTV company. We built software for those small boxes which you would connect to your TV before TVs became "smart".
We were updating those boxes remotely over Internet. And our biggest fear was that we will brick them.
So we had tests, but also a rollout strategy where update would go first into a central region, which is well connected and well-covered by the offices, but not as densely populated.
1/

2024-07-21 16:05:34

Is it the time yet, when we should talk about integration testing which needs to happen on the delivery side of the pipeline, rather than on the production side?
I guess I should give it a week or two..

2024-07-09 16:39:13

CentOS Stream 10 compose tests - now with TMT tool.
https://sig.centos.org/integration/compose-tests/
#Centos #CentosStream

2024-07-09 15:11:31

Irregular reminder for junior engineers:
You absolutely SHOULD argue with random people on the internet.
People who tell you that flamewars are useless, do it after they spent significant time participating in them.
And yes, generic flamewars indeed *become* useless over time once you learn the main arguments and the rules of the game.
You absolutely should learn those rules though. And no one will teach you that in the university.
Do remember that it is a sparring, not a fight. And enjoy.

2024-07-08 12:20:01

In both #markdown and #asciidoc we have a possibility to create a nicely highlighted code block:
``` python
def test(input=zero):
pass
```
I don't write code snippets though. Usually i write:
```
1. Install the app:
# dnf install app
output
1. Run the app:
$ app --version
output
```
And here # means root and $ means user.
Can someone more capable than me introduce the new style like:
```cli
```
with a proper highlight for such cases?

2024-07-04 13:35:29

- reply

@npub15s3…8j2z
It is different.
After CentOS 7 EOL, you can not get CentOS 7 content without changing the paths in the repo files on the installed system.

2024-07-04 13:26:23

- reply

@npub1hlt…ck4e
It doesn't break the running system.
It breaks the attempt to update it. Because there is no update.

2024-07-04 12:53:20

On one hand, seeing how many people are impacted by taking down #CentOS 7 repositories and moving them to vault, I wonder whether we should have a different procedure for the EOL...
On the other, seeing how many people are impacted by taking down centos 7 repositories, looks like it was a right move. If people are still surprised and haven't checked what does EOL means for their systems in advance, it means we need this wake-up call.
Yes, it has impacted many people. It is a good thing.

2024-06-30 12:15:26

RB VIA 10 from Frankfurt to Koblenz is getting a bit crowded, but we all fit in.
I am slowly getting better with the gear, for example I did bring the cord this time, which holds the construction together.

2024-06-28 19:48:17

Double achievement:
1) managed a smalltalk with a random person in the hotel restaurant in German!
2) advertised the "Ada & Zangemann" to a non-FOSS person.
...
Wait, does it still count as a smalltalk when you inevitably start to advertise FOSS?

2024-06-26 14:03:21

- reply

@npub14ze…xqw8 Channel your frustration towards the US patent system which allows patents on software and algorithms.

2024-06-20 11:58:33

- reply

@npub1duy…e0e8
Hm, I don't have such a fancy label on mine.
I am now even more curios, what data is on my tape, as I got it directly from the IT department and not via the souvenir shop.

2024-06-19 12:31:47

I like how if I write
"Red Hat supports RHEL 7 (and systems converted to RHEL 7 from CentOS 7) till June 30, 2028"
it is not a newsworthy announcement.
While at the same time my timeline is full of news about how about everyone else is going to "support" CentOS 7 by fetching RHEL 7 patches into it, as if this is a self-contained achievement.

2024-06-12 18:19:52

It has been three years, as I am trying to communicate this message to people:
----------------
#RHEL 8.5 is a branch.
#AlmaLinux 8.5 (and #CentOS 8.5, Rocky 8.5, Oracle 8.5...) is a tag.
----------------
I am glad Alma wrote it explicitly on https://wiki.almalinux.org/release-notes/ that
"Release of new 8.x version ends support for the previous version."
I am sad that people don't read and internalize it.

2024-06-12 13:27:26

- reply

@npub1dtd…vkyt
Hooray for #CentOS 10!

2024-06-12 10:50:59

https://marc.info/?l=openbsd-tech&m=171810878113010&w=2
"I added 10000+ checks so far, in about 4 or 5 hs. Final count will
likely be close to a million.
It's true that many are useless, perhaps up to 50% of them. Most
stack protections put into place by the compiler are also useless.
But the question is, how many are not useless?"
No-no-no, this isn't the question.
I always knew framing programming as coding and not reviewing was wrong, but I didn't expect it to bite us this hard.

2024-06-10 15:50:50

Honestly, one of the most exciting things I saw in CERN was the report about the state of the OpenStack infra which stated that all hypervisors (~1600 of them) have been updated to EL 9 and they have only two remaining #CentOS 7 machines left.
This is so refreshing to see the infra team doing its job correctly and not waiting for the EOL date to start planning their migration efforts.
Unlike some of the enterprise customers, which are going to come on June 30 saying "EOL, what EOL?"

2024-06-07 20:03:42

<rant>
I think O2 wins the top place on my personal list of design failures.
They "upgraded" the app so that it can not load without internet. At all.
I mean, what do they think I use their app for?
Casually open it to check new "special offers"?
I open it _only_ when the internet stops working to see if it might be caused by the negative balance. And the only question I have then is how do I get my internet back. There is no other reason to open this app ever.
Now that reason disappeared.

2024-06-06 12:57:32

Got my piece of CERN.
https://cdn.fosstodon.org/media_attachments/files/112/569/781/576/274/539/original/d2aededd9c1442b8.jpg

2024-06-05 12:16:15

How to see the feed of CentOS Stream activity without Gitlab account:
https://blog.centos.org/2024/06/how-to-watch-for-centos-stream-activity-on-fedora-message-bus/
#centos

2024-06-04 13:33:50

Well, on one hand, it is nice to see the that we respect the `ln` tradition, no matter how high in the cloud we go..
https://cdn.fosstodon.org/media_attachments/files/112/558/593/958/687/131/original/8bc61e5c67fcee9f.png

2024-06-01 08:47:55

How does your PhD help in your daily life?
I use it to hit people in the face, when they tell me that my lack of fundamental education doesn't allow me to see how programming is superior to all other IT and non-IT jobs.

2024-05-31 22:27:30

- reply

@npub1uje…yxv9
FESCo is a challenge. I hope you'll like it.

2024-05-30 14:46:19

- reply

Looks like I can indeed rent a bicycle from one train station to the other
https://www.rentabike.ch/en/rent-and-tours/reservation-daily-rental

2024-05-30 13:36:51

- reply

@npub1amd…qlf5
There is a reason why they let visitors in only when everything is switched off for maintenance :)

2024-05-30 13:35:51

> We're contacting you because you registered for the event OpenInfra User Group meeting that will take place at CERN on June 6th...
Hooray, I am going to #CERN :)
I am thinking if on the way back from Geneva I should make a couple of stops in nice places.
Is renting a bike to get from Geneva to Lausanne a feasible idea?
It is tempting to try and go there all the way with my own bike, but four train transits make it quite a challenge.

2024-05-30 09:28:29

- reply

Jokes aside, publically available Jupyter notebook with read-only access to key Fedora datasources(koji api, message bus, pagure, pkgdb) which you can ask questions about all things Fedora could be a nice thing.

2024-05-30 09:25:10

Ok, my second* royalty-free AI innovation idea:
We do a website, chatPT-7o.ai, which looks exactly like ChatGPT, but internally is a python prompt (IPython?).
Basically, ChatGPT skin for Jupyter Notebook. Where all error messages are ignored and replaced with a random fortune quote.
And then we publish a draft of a research article on how this novel approach to add occasional brackets in the prompt string can reduce LLM hallucinations by 93,6%.
* The first one was: https://fosstodon.org/@bookwar/112235835417984724

2024-05-29 19:15:39

- reply

@npub1p6p…vv94 I had a colleague who had a habit of shortening "test" to "tst".
Yes, I hated that.

2024-05-28 13:54:07

- reply

@npub1djr…tf7y
But I wonder if LLM efforts can be refocused somehow from trying to build the owner and provider of the answers to every possible question, to building the ultimate proxy.
The API rather than the backend.

2024-05-28 13:05:01

I wonder how many people like LLM not because it is an LLM but because it is a cli.
The CLI they were afraid of for so long, but then found how powerful it can be and what a variety for tasks it can do.
If only it were slightly more reliable, and knew how to do the exact thing you want it to do..

2024-05-25 08:24:15

- reply

@npub1pz3…dldm CentOS Stream 10 could be a good target for this, imho.
Keeps the excitement at the beginning, then, at RHEL 10 GA, stabilizes and runs smoothly until you are feeling adventurous again. :)

2024-05-24 09:02:21

- reply

@npub1d7l…546q
Looks like it also wants to have some fun
https://github.com/bookwar/fm-metrics/blob/main/fm-metrics/conf.py#L20-L22
But at least I have something to blame now :)

2024-05-24 08:25:01

- reply

@npub15nw…s0l7
Well, with numbers we find other ways how to make it harder for ourselves. :)
Looking at `06/08`...

2024-05-24 08:10:49

- reply

@npub15pn…4un8
I usually try to remember that Tuesday is a nice word and Thursday is ugly (pronunciation-wise). And so are the days.
But then I forget which day of the week I should like more :)

2024-05-24 07:59:42

The Calendar UX issue only non-native speakers can understand:
Almost every time I refer to Tuesday or Thursday I do a quick double check on which one is which.
I usually try to do it by a quick look at the calendar. But if you shorten the names to a single letter, they look the same.
Thus in the current Gnome calendar, there is no way to see what is the full name of the weekday for May 28.
#gnome
https://cdn.fosstodon.org/media_attachments/files/112/494/985/257/002/821/original/7e7794022795139b.png

2024-05-23 22:33:48

- reply

@npub1ql7…f9xl Better late than never :)
We could have called it Mid-release party, but there wouldn't be that much difference.

2024-05-23 19:24:25

Sometimes I write code...
https://github.com/bookwar/fm-metrics/blob/main/fm-metrics/count.py#L61-L70
https://cdn.fosstodon.org/media_attachments/files/112/492/031/744/891/484/original/07145abbe1c686da.png

2024-05-22 13:41:23

#Fedora 40 Release Party in Frankfurt on June 13th
https://discussion.fedoraproject.org/t/fedora-release-party-f40-in-frankfurt-germany-june-13-2024/117045
Come visit, if you are around!
---
It is free, but requires a (very simple) registration.

2024-05-02 15:39:32

- reply

Apparently, some puzzles become much easier once you know English well enough that you can read the text and listen to the story, not just look for pictures trying to guess the meaning in them.

2024-05-02 15:36:08

I am binge-replaying #Myst series this week.
Some years ago I got the full set on GOG, but at that time wine was not ready and lutris was not available, so I couldn't get them running.
But it seems we finally have reached the state of Wine and #Fedora where Myst I - IV work flawlessly.
The controls feel a bit flaky but overall it is still as immersive and epic as I remembered.

2024-05-02 12:02:33

- reply

@npub1y3q…wmll
Funny how it would all make sense if "About Us" were just "About".

2024-05-01 10:03:48

- reply

@npub19la…hm4g Block them all, but please keep on writing.
You are one of the best tech writers in the area of system programming, and your posts, toots and blogs are always interesting, insightful and easy to understand.

2024-04-27 16:58:12

- reply

It was of course purely random, but it worked like a trap forcing me to look for common properties trying to find some pattern again and again and again..
If you ever watched Beautiful Mind, you get the idea.
I am kind of afraid of generative AI stuff, because of that effect. You get a pile of random nonsense, and you try and try and try to make sense out of it...
It is not just that AI is hallucinating. It is that you start to hallucinate with it.
2/2

2024-04-27 16:52:52

When I worked in Moscow I spent 1.5 hours on a subway daily.
And there was one ad in the subway train which literally tortured me for a year.
It was a schema of subway lines, where all the stations were replaced with some generic activities like vacation, family, walking in a park..
I don't remember the point of the ad, but I spent crazy amount of time trying to find logic in it: Why family-park line intersects with vacation-shopping line at the "museum" station?
1/

2024-04-09 14:12:14

All the talk about eclipse makes me go re-read the Asimov's "Nightfall" novel.
I don't like the word 'awe' because it is hard to pronounce, but this book for me is the perfect representation of it. How you feel curios, amazed and scared at the same time.
An Ode to Science, if you like.

2024-04-08 14:13:22

- reply

@npub1tny…z6px
We should organize a separate online event for people "who are not there".

2024-04-08 14:08:00

- reply

@npub1tny…z6px
Do you *really* need to make us more jealous than we are already? :)

2024-04-08 13:28:57

Let me introduce my new product called *Distributed Crowd Sourced AI*
It is completely free for all commercial and non-commercial use.
To add an AI generated photo to an article, add a line with a prompt directly to the text of the article:
AI Figure 1.: Generic rainy day in a small neighborhood.
and post it.
On the client side parser will read the prompt and translate it into a picture.
Parser is already included in every browser and reading device!
Supports HTML, SVG, doc and docx.

2024-04-04 10:56:06

As usual a good Fediverse conversation is a best motivation to write an article.
So here is the new, slightly provocative one, featuring #DevOps fever, Professional Users and why developers are always doing things wrong :)
https://quantum-integration.org/integration-engineer-in-a-java-shop

2024-04-04 10:02:47

- reply

@npub1kxe…69je
The worst one for me is "deshalb". Because it sounds like 'despite' but means exactly the opposite.
Unlike "also" which just sounds slightly weird, it can lead to huge misunderstandings.

2024-04-02 21:29:12

- reply

@npub1rnj…vtmy
What if every year we choose the default for next year by a lottery?
I am joking.
But still.. :)

2024-04-02 15:09:46

- reply

@npub12lc…e4yu
Than everyone brings you a cake :)

2024-04-02 14:44:11

There is a lot of talk about how lucky we were. And that somehow this is a bad thing.
I want to bring up the Birthday Paradox:
https://betterexplained.com/articles/understanding-the-birthday-paradox/
It is extremely lucky for you to have a person in the room with the same birthday date as yours. It is though quite reliably true that in a room of more than 70 people there will be at least one pair of people with a same birthday date.
Opening the code, and encouraging tinkering with it is how we bring more people in the room.
#xz

2024-03-27 20:58:52

- reply

And no, don't come close to me with your #TOGAF books or certificates. You are already broken.

2024-03-27 20:49:53

- reply

And no, don't come close to me with your #TOGAF books or certificates. You are already broken.

2024-03-27 20:44:57

Just invented new criteria:
You can not be an Enterprise Architecture expert if you don't know what the term "Upstream" means.
Yes, it puts 99,99% EAs out of their job. For a good reason.

2024-03-27 16:16:13

- reply

@npub1cgg…0n34 If they are starting planning, I can recommend going for CentOS Stream 10.
It is right about time you can reasonable begin looking into it and prepare for adoption within a next year or two :)

2024-03-14 12:34:33

- reply

@npub1x4p…lgyx Thanks, fixed.

2024-03-14 11:53:07

To celebrate the #PiDay today I wrote this article about my favorite mathematical abstraction:
https://quantum-integration.org/pi-day-special-manifold
Mathematics is much bigger than Calculus. And #Topology (and Differential Geometry) is the best part of it :)

2024-03-13 14:16:41

- reply

If there is a collective aha moment here: I wish it is not about how to choose a vendor that allows you to postpone your problems for 10 more years.
But rather how to design systems so that they evolve rather than explode after everyone with a knowledge about them is retired.

2024-03-13 14:16:26

#CentOS 7 EOL is such a huge event for the industry, that everyone is trying to get a share of it with good and bad intent and marketing.
There are so many unresolved questions, postponed decisions and just poor planning, which all are exploding in that very last moment, because 10 years ago everyone just decided not to think about "complex stuff".
I don't even want to argue on CentOS or RHEL side. Make your own choices.
I want to say - piling up problems for 10 years is bad.

2024-03-05 21:56:11

- reply

@npub1djr…tf7y Classics!

2024-03-05 19:12:56

- reply

@npub1csu…qna3
The second is that "data-driven" people put so much effort into collecting and measuring data, that they forget they need to develop a theory to use it. And that developing a theory is a different skill set.
No, you can not develop a theory by piling up data.
While yes, I know, if you apply AI to a pile of data you can do cluster analysis and pattern matching. But developing of a theory is none of those things.

2024-03-05 19:11:26

- reply

@npub1csu…qna3 The problem is with the perception.
And there are two points:
First is that when people say data-driven, they usually imply that it is objective.
They say: "You can argue with me, but it is not me, it is data, which says it!"
But data doesn't speak. It can not make conclusions.
And you absolutely can and should argue with people who hide behind that "data speaks for itself" position.

2024-03-05 17:25:09

- reply

@npub1p0y…jtpw All these assumptions are not included in that number. You have them before you look at the display.
It is your theory of how thermometer works, which allows you to make conclusions about room temperature based on a single value on the screen.
You can not derive this theory from the data. It is just not there.

2024-03-05 17:22:48

- reply

@npub1p0y…jtpw
Consider the example with a room thermometer.
You look at it, it shows the number on a display. You think it is a room temperature.
But room temperature is already an abstraction. It is the interpretation of the number you see, which you make based on a ton of assumptions included.
You assume that that the room has a more or less stable temp over a period of time, and that that display is a thermometer, and it checks the air, and the air freely circulates in the room...

2024-03-05 16:19:04

- reply

Adding numbers/data points to your theory does not turn the theory in the objective universal truth.
And rather then hiding behind the neutral "data-driven" terminology, we should be embracing it.

2024-03-05 16:16:29

Hot take:
There is no such thing as "data-driven". Data does not drive, it exists. There is "agenda-driven, supported by data".
And I don't mean it as a negative thing.
To make sense out of data you have to have a theory, the hypothesis which you are trying to support. This theory gives you the way to interpret the data points you have collected.
And while the data is objective, the interpretation is not. It can be wrong. It will require adjustments the more you develop your theory.

2024-03-04 17:45:48

"According to industry standards" is not a praise, it is a curse.

2024-03-01 14:06:13

- reply

@npub1cm7…32y8
And to clarify by "development process API" I do not mean ci.yaml in the main project repo. I think having these .yaml files mixed into project sources is an antipattern.

2024-03-01 14:03:26

- reply

@npub1cm7…32y8 Looks like a terminology problem again.
When people say CI as a system running tests, it totally makes sense to NOT have one global CI system for a project like kernel.
But when we discuss CI as a framework so that various CI systems can be hooked into it, and provide test feedback in a way that project developers can act on it - then you do need a more generic approach to it.
Which of course goes back to that kernel development process needs better APIs than unstructured mails.

2024-02-27 20:23:13

Well, I am usually the first one to rant about Jira and cloud services, but I did try Jira integration with Miro board today, and I liked it.
There is something oddly satisfying in moving the cards on the board.
Unlike Trello it is a free-form movement, and you can group, arrange and order tasks in a non-prescribed way.
So if any one of you is looking into development of a lightweight issue tracker, consider this to be a feature request.

2024-02-27 17:40:19

- reply

IMHO this is the most undervalued quote from Azimov's "Foundation".
I can not count all the times I had this reference in mind but couldn't use it, because I couldn't rely that the audience will get it.

2024-02-27 17:38:05

"And where is the analysis?"
"That," replied Hardin, "is the interesting thing. The analysis was the most difficult of the three by all odds. When Holk, after two days of steady work, succeeded in eliminating meaningless statements, vague gibberish, useless qualifications – in short, all the goo and dribble – he found he had nothing left. Everything canceled out."
"Lord Dorwin, gentlemen, in five days of discussion didn't say one damned thing, and said it so you never noticed."

2024-02-27 13:48:00

- reply

@npub1csu…qna3
This reminds me: Over last 15 years I made at least three attempts to start reading the ITIL book.
The problem is - I can not get past the first paragraph. It is like you read words, and they look clever and then you ask - what each of those words supposed to mean? Nothing.
What this whole sentence can be summarized to? Nothing again.
My brain gets overloaded from trying to dig into the content and falls asleep by the end of the second sentence.
And it was even before the AI times.

2024-02-27 12:16:00

Using AI to generate a wall of text out of two bullet points should be considered a crime against humanity.

2024-02-25 01:12:52

- reply

@npub1s2v…u50h
Kernel folks declared their LTS to be 2-3 years.
Should we call them Short Long Term Support? :)

2024-02-22 13:42:48

- reply

@npub1el5…gt9m
Check any change on review.opendev.org
For example https://review.opendev.org/c/openstack/requirements/+/907647?tab=comments

2024-02-22 12:42:09

- reply

@npub1hjv…938a I have a relatively small project with the same issue, just didn't want to put it on the spot.
GitHub doesn't scale to anything beyond the single developer working on his own sideproject, imho.

2024-02-22 12:09:49

I really really don't get how people can complain about Gerrit Code Review interface, if they ever used GitHub Code Review UI.
Just go to a completely random PR and try to find a comment from the reviewer on the page:
https://github.com/kubernetes/ingress-nginx/pull/10984
And this is our "industry standard" which everyone tries to replicate?