←back to thread

“I'm basically giving myself a permanent vacation from being BDFL”

(mail.python.org)
2024 points randlet | 8 comments | 12 Jul 18 15:22 UTC | HN request time: 0.022s | source | bottom
Show context
bla2 ◴[12 Jul 18 16:02 UTC] No.17515883[source]▶
> I don't ever want to have to fight so hard for a PEP and find that so many people despise my decisions.

Leading a large open source project must be terrible in this age of constant outrage :-(

replies(9): >>17515955 #>>17515972 #>>17516193 #>>17516427 #>>17516776 #>>17516884 #>>17517282 #>>17517716 #>>17517821 #
sjm-lbm ◴[12 Jul 18 16:09 UTC] No.17515955[source]▶
It's PHP and not Python, but every time I read something like this from a major open source figure, I always think of this old PHP mailing list thread:

https://bugs.php.net/bug.php?id=50696

replies(8): >>17516108 #>>17516130 #>>17516216 #>>17516240 #>>17516461 #>>17516708 #>>17516836 #>>17517666 #
Y_Y ◴[12 Jul 18 16:24 UTC] No.17516108[source]▶
That's a good read. I feel like the "customer is always right" mentality does quite a bit of harm to OSS support.

Also reminds me of that dev (who I can't seem to search up) who had their email printed as part of a open-source software license in a car manual and would get ridiculous email from people who had car trouble.

replies(6): >>17516199 #>>17516206 #>>17516230 #>>17516371 #>>17516964 #>>17517308 #
1. krylon ◴[12 Jul 18 16:45 UTC] No.17516371{3}[source]▶
OTOH, way back, when I had a TV receiver card by Hauppage, and a new Linux kernel broke something, I posted a description of my troubles on Usenet, and within 24 hours, the person who had written the Video4Linux subsystem replied asking for more details (which I gladly provided), and a few days later, the bug was fixed.

That, I think, was the most awesome "customer support" experience of my life. I did make a point of being polite about it, however, which I consider a ground rule for dealing with people, especially if I want something from them.

But it was so awesome to post to a random usenet group about a driver problem and have the person who wrote the driver personally approach you for details. You don't get that with Windows, for sure. ;-)

replies(4): >>17517050 #>>17517497 #>>17517581 #>>17519508 #
2. kwhitefoot ◴[12 Jul 18 17:57 UTC] No.17517050[source]▶
I had the same experience. Suddenly Emacs Gnus reader stopped working, about 11:00 on a Friday morning. I posted the bug and had a reply from Lars Ingebrigtsen with an explanation and a workaround by 13:00. I think it was fixed by the beginning of the following week.

My experience of trying to get help from Microsoft on the other hand is, well let's just say not quite so impressive; they kept me on hold for 45 minutes once and never did solve my problem.

replies(1): >>17517426 #
3. undseg ◴[12 Jul 18 18:41 UTC] No.17517426[source]▶
Back in the days Microsoft would have amazing customer support. And I'm talking about "API customers", many times I've seen them go out of their way to fix third party programs not working correctly by adapting their platform, it was a very efficient process. I guess that doesn't work as well at their current scale, or they care less because they're not building up market share.
replies(2): >>17517534 #>>17517999 #
4. eithe ◴[12 Jul 18 18:48 UTC] No.17517497[source]▶
I used to have similar experience with most of the free source (that, or silence when project got abandoned), until Laravel, where I was yelled at because somebody apparently needed an ego boost that day. I still love the framework, but I don't think I'll be trying to reason there anymore. Still - maintaining projects most of the time is unappreciated job, so kudos if you do it
5. marris ◴[12 Jul 18 18:55 UTC] No.17517534{3}[source]▶
> go out of their way to fix third party programs not working correctly by adapting their platform, it was a very efficient process

This kind of thing probably doesn't even scale with itself, since there are only so many acrobatics your code can go through before no one can even understand why it's doing something, let alone add new workarounds. So the first K broken programs get special treatment, and the others face a much higher bar to get the same treatment.

6. Kadin ◴[12 Jul 18 19:00 UTC] No.17517581[source]▶
Yep. This attitude isn't dead, assuming you approach the developer via the right/preferred venue, and are polite, and they are still actively maintaining the project. And it really helps if you give a thorough bug report!

I recently submitted a bug report to a fairly niche OSS package that I use, and within a few hours the author replied (on GitHub) something like "oh wow, yeah that's an edge case but I definitely want to fix it, can you send me the test data you used..." and once I gave him the test data, he had it fixed in two hours and now anybody who grabs the source won't have to deal with that bug.

It was great, and even though I didn't do anything except write up a bug ticket properly (the same way I'd expect anyone on my team at work to do it), the software is a little better now.

7. TheDauthi ◴[12 Jul 18 19:45 UTC] No.17517999{3}[source]▶
They once shipped my company a custom MSVCRT.DLL to fix a crash bug in a third-party application under heavy load. True, it was a bug in the C runtime itself, but they got us a fix for it about a day after we (along with the third party) got in touch with them.

Just a few years later my team had to contact them for a BSOD that kept happening after one of their patches. We were put off for about a week before throwing our hands up on it.

8. bigger_cheese ◴[12 Jul 18 23:39 UTC] No.17519508[source]▶
Years ago I purchased a motherboard with a built in Raid controller. There were no drivers for it in mainline Linux kernel but there was code for a kernel module on the chipset vendor (Via's) website. The provided code wouldn't compile because there was a bug.

People from Fedora IRC channel helped me live debug the code over irc and we found the bug and were able to get code to compile and my raid controller working. Was an awesome experience.