This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2023-09-03
Channels
More “fun” on emacs-devel, where the conversation about Clojure support in Emacs continues to rage https://lists.gnu.org/archive/html/emacs-devel/2023-08/msg01287.html On days like these working on OSS feels like a very bad use of one’s time and energy. 😿
Joao IME is extremely taxing to engage with.
His solution isn't awful, just needs some naming adjustment. A general inf-lisp-mode
that can be called for clojure, racket, scheme, whatever sounds like a good way to appease everyone
But I suppose that's just lisp-mode
and a call to eglot, which doesn't really solve anything does it...
What I found absurd in the whole thread is how… NOBODY participating in it besides yourself is an active Clojure user
When I saw two Emacs maintainer ask each other “does Clojure support macros?” and even admit they never use the language, yet deciding to write a major mode for it, it was just one of the most disastrous things I saw last week
with that said, if they must add a barebones clojure mode, I do think Phillip offers a decent middle ground: name the mode anything but clojure-mode
… call it something like basic-clojure-mode
… in an ideal world, we could have clojure-mode
itself in emacs core, but copyright exists and most people will not be excited having clojure-mode
stuck to annual releases
its kinda frustrating seeing the maintainer of SLY (my favorite IDE plugin for Common Lisp) blatantly demonstrate ignorance of Clojure yet insist on hacking together something that nobody will seriously use
my BINGO card is almost filled out after that thread hit "Clojure isn't really a lisp"
Correct me if I'm wrong, but wouldn't adding a new clojure-mode
to core emacs break at least something in the current clojure-mode
+`cider` "ecosystem", which would have to be worked around in those existing libraries + individual users would have to fix their configs? And then there's all the existing documentation and tutorials, which refer to clojure-mode
. Seems so absurd how casually people on the list ignore the ramifications of taking over a well-established name.
@U0178V2SLAY Of course, there would be breakage if they do something like that, but it doesn’t seem they care much about the practicalities of their proposals. That’s part of the reason this conversation infuriated me.
@U043RSZ25HQ In the past clojure-mode
and inf-clojure
were just super thin wrappers around lisp-mode
, but this caused a lot of problems which caused us to change our strategy. Now they are planning to go full circle which I find this very funny. 🙂 Of course, it’s hard to make informed decisions if you have no idea what you’re talking about.
> What I found absurd in the whole thread is how… NOBODY participating in it besides yourself is an active Clojure user I'm the other active clojure user in this thread, fwiw.
In what I have been thinking of as "damage control efforts" I have gotten what I feel like are assurances from EliZ that 1. Emacs devs will not FORK clojure-mode or clojure-ts-mode (that was another large thread) 2. That he does not support taking the name clojure-mode
Eli being the lead developer has the ability to say "no" to such decisions. I think he is reasonable and wants whatever comes out of this to not majorly fuck users over, which forking and name hijacking would definitely do
Eli is extremely reasonable here I think, rms was the one being unhelpful but it looks like he later apologised
RMS I think does not read all the theads completely. Or maybe he does, but doesn't remember some things. He also comes off as abrasive but I don't think he can help himself. Even when I keep that in mind interacting with him can be frustrating.
Joao on the other hand has been a huge asshole. I wish someone like Eli would reign him in. This thread has been extremely disappointing to me because Joao was the person who convinced me to write my first emacs package.
Mmm, all I can say is I'm glad he no longer has much sway over the project 😄 Seems like positives are coming out of the thread though. > Joao on the other hand has been a huge asshole I'm surprised he goes so unchecked tbh, he comes across as quite hostile
If I try to check Joao or someone he is hyping up he just gaslights and insists I'm arguing against strawmen. Extremely frustrating and disappointing behavior from him.
Something not visible in that thread: https://github.com/clojure-emacs/clojure-ts-mode/issues/17#issuecomment-1698513481 I got some out of band condolences from Yuan, the developer of the tree-sitter integration in Emacs. He has always been very kind to me and Sogaiu. He's fixed numerous bugs that clojure-ts-mode uncovered in the core, even applied workaround for bugs we have found in tree-sitter itself.
Yuan is a pleasure 😄 The few interactions I've had directly have been great, whenever I've seen them around it's been positive
Indeed Joao is very hard to reason with. Used to have a lot of respect for him back in the day, now I can barely stand having to communicate with him. Every conversation is pain.
I highly doubt anything will come out of all these conversations, but I took the time today to write about 10 email responses to this thread. I hate venting out, but today I really needed to.
Danny has been a real trooper - he has nerves of steel compared to me. 😄 (I’ve got a pretty explosive personality after a point that Joao managed to trigger today)
You've said some things that needed to be said. I'm about to reply to one of your emails with a small anecdote about CA
On second thought, maybe I'll continue to let this stew. I'm quite exhausted.
FWIW, i spoke briefly with technomancy about the CA. He won't contribute to projects with one. Can't blame him after what we've seen how some companies wield them, wrt Hashicorp/terraform, elasticsearch, redis, etc.
I think I haven’t been in such full-on rage mode in years. I’ll ask them to put this convo on the backburner so we can discuss it properly internally in clojure-emacs.
Oh, wow, this email thread really went of the rails. I remember some Github and reddit "conversations" between Yyoncho (lsp-mode maintainer) and Joao about eglot vs lsp-mode and Joao came also across as a bit of a d*ck like in this thread. Imho it's probably best to just ignore some people there. It's a bit of a shame as both the Emacs and Clojure community is not really big and we should get along instead of this 😕
> Can't blame him after what we've seen how some companies wield them, wrt Hashicorp/terraform, elasticsearch, redis, etc But there's a big difference between assigning copyright to the FSF vs a big corporation. The FSF needs copyrights so they can defend you and your freedom in court vs others so the software stays free. Corporations want copyright so they can take your code and make more money and less free.
> The FSF needs copyrights so they can defend you and your freedom in court vs others so the software stays free. Indeed, that's the only reason I signed. Not everyone will see it that way though, or want to go through having their employer sign as well. Overall people are very skeptical of CAs, no matter what the other party's intentions might be
In the future, since they don't seem to care about being hostile and non collegial, the problem could be phrased in terms of being user hostile - there are many current users who whose setup would break by a wring decision. Making this decision for them is infringing on their core freedoms
heh, I remember that LSP flamewar too. again, its a shame, because eglot and SLY are great packages
> In the future, since they don't seem to care about being hostile and non collegial, the problem could be phrased in terms of being user hostile Myself and others (Philip, Dimitry) have argued heavily from this perspective. https://lists.gnu.org/archive/html/emacs-devel/2023-09/msg00059.html I took wading through more dismissals and gaslighting from Joao but they eventually recognized that it will be bad to screw over users this way.
Even RMS conceded that taking the name is probably not great. Like you suggested Ben, I think the argument from a users perspective was the most persuasive > RMS: >
I'm ok with making small adaptations to cooperate with the developers
> of the external Clojure language mode, when that does not cause a
> problem for Emacs. There is no reason to be ornery. But we should
> not do that in ways that mess up our general conventions.
Thanks lol. This has definitely been trying on my patience. I've not replied to it in a couple of days though. That helps
I learned some time ago to not argue with people online. You get worked up over what might as well be voices in your head. Better to leave them to it. Unless you're having fun
Indeed. It would be fun to send that image to the mailing list
There are certainly winning moves here, but what would playing any of them serve besides stroking one's ego?
It would be fun to vent a lot of annoyances on emacs-devel but alas no good will come 😛
Some smooth word smith could probably run circles around them, goad them, make them lose their temper all while remaining squeaky clean. But for what purpose
Honestly they mostly ignore anything like a good "gotchya" argument. Probably for the same reasons I try to ignore Joao's taunting
Well, it seems @UDVJE9RE3 is a lot more emotionally restrained than me - I simply can’t tolerate people making shit up, being passive-aggressive all the time and making personal attacks on me. I’ve lost my entire respect for Joao and frankly speaking - I don’t want to interact with him ever again.
I acknowledge he has done a lot of work on Emacs, but I also remember all too well how he attacked flycheck
in the past and was part of the reason it’s awesome maintainer decided he doesn’t won’t to work on Emacs anymore. And everyone knows how the situation with eglot vs lsp-mode unfolded.
I recall at some point he was also explaining to me how projectile was a lot worse than project.el, but at least that conversation was slightly more constructive.
Seems he has some real issue with people who don’t believe that the GNU/FSF way is the only way.
Think in his mind, everything he writes is flawless and all other software is just a waste of time. Seeing him talk about lsp-mode, flycheck & projectile, at least
Yeah, yeah. He has a very high opinion of himself, that’s for sure. And infuriatingly he was constantly accusing me of having some ego problem myself and demanding some special treatment by him and the Emacs developers.
The same on the eglot "vs" lsp thread (which started out as an objective difference comparison, since the table was outdated), I think ignoring him unless necessary is a good idea
That’s my plan. I don’t think I’ll ever engage in another conversation with him. It’s a pity that very few people on emacs-devel care about the quality of the conversations there.
I’ve been on this list for 20 years now and very little has changed. I’m not holding my breath…
Perhaps back then it was even more violent, so there might be some small improvement.
I read most of that thread (of emacs-devel). I discovered it separate from this discussion and the thread from Sunday 27th. I lost my mind several times just reading it and was so shocked as to what was being discussed & what was the reasoning. Kudos on not starting a flame war. Having endured most of the emacs-devel discussion, I somehow missed this crucial agreement: https://lists.gnu.org/archive/html/emacs-devel/2023-09/msg00311.html
I only got it, since I contacted Phillip K directly… Otherwise I would’ve missed that and would be fearful as to what Emacs devs are going to do, to clojure-mode.
The emacs thread has derailed much further than I thought was possible. As a mock-lisp developer using closure[sic], I just want to send some virtual :hugging_face: and 🍻 to Bozhidar and Danny. Kudos and props for being so patient and civil. Funnily enough, CIDER, Doom, and evil-mode are the only reasons I'm still using Emacs (a trifecta of heresy for emacs-devel I would guess).


Sigh... I got tired skimming that thread. Add one more to "political problems that exist because proper global namespaces don't". As a grateful user of clojure-mode, CIDER and all sorts of other tooling, I hope emacs developers decide to avoid clobbering clojure-mode and all its dependencies. Just do their own thing in their own way in their own namespace.
Here's the final episode of this - now that I'm home I took the time to summarize all of this from my perspective and created a nice rant to vent out in style https://metaredux.com/posts/2023/09/09/clojure-support-in-emacs-one-unnecessary-drama.html 😄