There should be a concept called "engineering privilege".
I see some companies do things that are widely out of their core product and engineers there just having fun with a bunch of different stuff with seemingly no commitment to bring in revenue. They'll re-invent/re-write things constantly, write a new compiler/package manager for their chosen language, create new frameworks, new open source projects unrelated to their companies, etc.
It must be awesome to be able to work in these conditions and still get paid. I'm not hating on it, it's how we get a lot of the good stuff we use.
My favorite example was from Joel Spolsky's company, where bright engineers got bored with working on bug tracking software so they wrote their own C# compiler.
Because every company may have different needs that are not fulfilled by standards software. We have seen the large number of companies whose goal is to make custom software for other companies So is not always that reinventing is fun, but it may actually be needed.
The same logic also applies for software that is used internally. It's not always that developers want to rebuild stuff, but that they need to rebuild or alter software to fit their / company needs.
We have not even talked about cost or licensing or other issues.
Yes except a lot of it isn't even fit for their purpose and they don't rely on it. Facebook switched from Cassandra to HBase because "eventual consistency is a difficult pattern". No shit, really?
> Every change is linked to the agent conversation that produced it. From any line of code, find the conversation. From any message, jump to the code it touched.
This is how Zed leapfrogs the field and becomes a centicorn.
If you can train a model to operate at a higher level and with looser requirements, you've gotten rid of engineers in the loop and built a system for the automated synthesis of high quality software.
You're training systems to produce higher quality chain of thought across the entire software development lifecycle.
"There is the software. And then there are the databases beneath the software, two enormous databases, encyclopedic in their comprehensiveness. One is the history of the code itself -- with every line annotated, showing every time it was changed, why it was changed, when it was changed, what the purpose of the change was, what specifications documents detail the change. Everything that happens to the program is recorded in its master history. The genealogy of every line of code -- the reason it is the way it is -- is instantly available to everyone."
also from ^ article on Lockheed Martin Corp space shuttle software group ...
Importantly, the group avoids blaming people for errors.
The process assumes blame - and it's the process that is
analyzed to discover why and how an error got through.
At the same time, accountability is a team concept:
no one person is ever solely responsible for writing or
inspecting code. "You don't get punished for making errors,"
says Marjorie Seiter, a senior member of the technical staff.
"If I make a mistake, and others reviewed my work, then I'm
not alone. I'm not being blamed for this."
It's funny, zed started out really refreshing, and now it's degraded to the point where it's fairly unusable. Regularly it would tie up tons of resources on my fairly beefy laptop. I switched back to sublime and it is a breath of fresh air, I didn't realize how slow zed had gotten. It seems like maybe they lost their way a bit, but I recognize that also they need to make money presumably, and it's hard to do that with a large team and a focused product.
I would like Zed to have feature parity with VSCode from 6 years ago. Working well over SSH or WSL, which it currently doesn't, is a part of that, even if I don't use that everyday.
I am a paying user, and have no plans to stop supporting them financially, but I wish they locked in and went through GH Issues queue, starting from the oldest.
Of course vs code accommodates wsl. Same company. Personally I couldn’t care less how badly MS wants to make windows a viable development environment. It’s more irrelevant every day. I’m glad zed doesn’t waste too much time on it.
Their ACP integration is good (finally), at least with Codex. For when I want to be near the code, I prefer it now over TUIs or the Codex app. It feels much more immediate that way. Sticking to the basics will kill them IMO. They have been pitching themselves as agent-native for a while btw.
And the workflow they are describing in the post sounds amazing, I‘ve been wishing for that. It‘s orthogonal to CVS, no? You can still make PRs. But, crucially, you can loop people in early on, when it matters the most.
Unfortunately their Zed agent is completely broken when running on Windows connected to WSL/SSH remote because the terminal tool checks for paths using the local Path style (windows) which breaks when applied to paths on the remote (linux).
> Their ACP integration is good (finally), at least with Codex.
you must not use it much. try pasting in an image, quitting zed and restarting (ie: upgrading the app). it shows you the base64 encoded data instead of just showing the image. heck, it doesn't even show you the image when you paste it in. it just shows the filename.
ACP integration is very good, IMO. I'm using it with Claude Code. They've made so many improvements over time. I use image pasting myself, TBH I was excited it worked at all. Surprised that some rough edges around images are enough to ruin the whole thing for you.
images were just one glaring example... i didn't say anything about ruined, i just think a lot of those 'rough edges' are a bit more rough than they could be.
it is a code editor after all. how about integrating the LSP so that it doesn't have to resort to rg/grep/sed/scripts to edit code?
I agree with you that I wish they'd just nail the editor. I've been using it for years now and I really like it, but there are a few rough parts (settings and LSPs especially) that could probably be tuned up in no time if the focus was there.
Having said that, I love the idea of correlating the agent sessions to the changes. I'm glad someone is working on it.
You don't need a tool for that just configure your agent to write commit messages that link to a LLM log. And changes between commits. Maybe don't care why do we even care?
Of course you are right. I could write my own software to do what I want. Or, I could just do it all myself manually. This argument is valid in pretty much all cases, so I concede the point.
I've been hitting that or something recent recently; it's what I would consider a severe bug because it can result in lost work and related. Basically, if you `git pull`, edit a file elsewhere, or use an LLM, zed's tree will remain stale unless you re-close and open the file. (Restarting the application doesn't sort it). So you end up overwriting the file with whatever was in the buffer.
Oh, that sounds related to what I experienced. I recently had another look at Zed and tried to use it to full-text search through my workspace consisting of symlinked folders. Relevant things would just now show up, cmd+P also wouldn’t find the files. That’s where the evaluation stopped.
One could argue using WSL is an anti-pattern: use a real OS. I kid, I kid.
Yeah I’m an avid user of the editor. I don’t care for their AI stuff I bring my own but I just want to see the editor become awesome. A more detailed object plugin ecosystem would be nice.
Did you even read the post? The idea is to tie conversations to code. This is a more sophisticated version of something like doltdb, not a git replacement.
> This is my #1 annoyance: clicking microscopic icons at the very bottom of the screen to switch views.
Have you considered using a keybind? They even show which keybind to use when you hover them.
Amazing that someone would prefer a vertical bar that is 85% empty, wasted space instead of pressing a few keys on the keyboard - for a text editor of all things.
> An unprivileged userspace application, even if coded poorly, shouldn't crash the machine.
How would you prevent that? It can write until root disk is full, can take all your memory, saturate your CPU and a whole bunch of other stuff we (users) like to be able to do for the things that need them.
Unless you're suggesting something like Qubes or similar UX?
Why develop this when the core editor is almost unusable? On Linux wayland, copypaste is broken, the file manager is laggy and snippets do not trigger before the heavy hint logic is done. This is just on top of my head, the list is long
"Almost unusable"? Slight hyperbole don't you think? I initially tried the Flatpak version on Fedora 44 and had a few minor problems, but since switching to the terra repos version (terrapkg.com) I've had zero issues, Have you tried that?
For me, its a massive improvement over IntelliJ for my daily workflow.
This is surprising to hear. I have Zed on my tablet as IntelliJ (PyCharm/RustRover) turn it into a space heater. Zed's introspection/refactoring/general language knowledge etc isn't on the same tier at JB; it's closer to VsCode in that regard. I.e., I'd classify Zed as a project-based editor, while JB is an IDE.
Im on nixos and tried all immediate flavours of zed: stable, unstable and nightly. Nightly seems to be most stable and every time i pull the flake, i prey for more bugs to be fixed. Sometimes it works
I have been using Zed for about a year now on Debian 12 / Wayland and it works flawlessly. Even with Toshy (macOS) keybindings. I have it installed now on my Framework 13 Pro (also Fedora 44, and also with Toshy) and it works great aside from a subtle (100ms?) lag when using ctrl+arrow-key. Otherwise it is great.
The only thing that doesn't work well is inertial scrolling[0]. But scrolling is a fucking shit show on Linux no matter what... so I don't necessarily fault Zed for this. The same issue occurs in other apps.
I can't use an editor where the file contents I'm being shown don't match the reality of the file. I guess this issue doesn't affect most people somehow, but in the age of CLI agents I'm not sure how more people aren't running into it.
I was really enjoying Zed for a short while before I ran into this fundamental issue.
All I can say is it's the only editor I've ever used with this type of issue and unfortunately it's a deal breaker for me. I can get by with any other editor even if they aren't my favorite, but only Zed made me realize that the most important thing an editor can do for me is accurately show me files.
Have you checked your fs.inotify.max_user_watches value? Maybe your value is too low and Zed can't watch the files that are changing. I know the JetBrains IDEs need the default to be increased on Linux because they detected it and prompted me to increase it.
Just installed it, works on my kde and niri setups (I logged out and switched to each), copy paste works as well. File manager laggi-ness I didn't notice I don't use it, I use it mostly to type stuff out.
I have never used snippets so unsure about that too, going from command bar seemed to work.
I think it might be bad, but could also be it's not correctly configured maybe try asking codex/claude or something to check if it can fix/fine tune the config, if that doesn't work file issues?
I do suspect it might be working for me cause I got lucky I do see an open issue[1], but given Zed is used a bunch the lack of comments does suggest maybe it's not that wide spread either way not a good look.
There are a lot of subtle bugs. I'm on macOS too, but perhaps my workflow is unusual? Every time I open it, I find small bugs. There are so many features that it's hard to maintain them, I guess.
For example, a lot of disable buttons do not actually disable stuff. They just hide it from the UI.
Agree, copy-paste feels like it works 80% of the time, which is maddening when 1 in 5 copy paste attempts fail. (mostly from zed to something else) Why??
At least Excel has the courtesy to telling you "Hey the copy-paste buffer said it didn't catch that, can you try again?"
It still doesn't actually display a window my end until I press a keyboard key or move the mouse. So very often I sit and wait for it to open, only to realize I have to press a random key for it to unhide.
That's something I've never encountered on Linux/KDE/Wayland/Flatpak across Debian Trixie, Arch, and Fedora. Might be worth nuking your local Zed files and trying again.
I started using Zed a couple weeks ago and immediately had issues with copy/paste/buffer on an M1 Mac. I tried a few suggested fixes and they didn't work.
I've been using this "almost unusable" editor daily on both Linux and Mac for the past year and now experience no issues. while I still have JetBrains I haven't used it at all in at least six months.
A lot of the copy on this page feels like the kind of copy an AI would produce. But I’m assuming, surely, Zed writes their own copy so I don’t want to accuse them of doing that.
Marketing aside, this is the kind of thing (advancements in VCSes) that I’d usually be excited about, but which I’m less excited about coming from Zed because it makes me start to worry that they’re going to continue focussing less on the core editor experience.
I love Zed and I hope that it doesn’t start to go flakey.
Anytime a company/group announces some random new project now, I tend to immediately suspect over-reliance on AI. From what I’ve seen, AI addiction frequently manifests as an explosion in scope, possibly because LLMs make so much initial “progress” on new projects which you can then demo and have resource allocators extrapolate to some glorious shipping version. But then you have to go back & clean up the slop, progress slows to a crawl, and it’s all way less exciting. That initial rush can be addicting. Ultimately a supercharged version of eternal human impulses I guess.
I'm always hesitant to assert something is definitely AI copy because everyone's tone of voice is different.
I'm also starting to wonder how much real, human-made copy is mimicking AI-created copy that has been filling marketing pages as of late. When people make marketing websites, they often riff and rip off other websites' (à la "Good artists copy; great artists steal"[0]) design ideas and language -- I wouldn't be surprised if there are humans now inadvertently adopting LLM's writing style like that.
In my opinion / experience, AI copy is a generally more cumbersome to read and process. So I would assume a company at Series B stage would have a human (with marketing guidelines) go over the language used on a one-pager to announce a new product.
> Every change is linked to the agent conversation that produced it
A total nightmare and imo every dev should push back using this stuff. You'd otherwise have management brag about you not being a good communicator with the ai agent, the same management that have no idea how to steer these tools.
I really like Zed the editor, I'm not sure I like the path they getting into
These sort of paper trails are for developers not management though, does your manager open all your commits and start commenting on them? If so, maybe apply somewhere else.
I really love Zed and I want them to really succeed. Unfortunately it seems that they are swamped with various things and the issue tracker is not a priority for the team, which fair enough. OSS is a gift and I take it as such. Wild proposal but if someone is willing to start a "user advocate" contributor, I'll be happy to donate monthly so user requests can be taken a look at :)
I disagree. JetBrains PyCharm has a similar "file history" feature that solves this very real and existing problem and has saved my ass more than once when accidentally overwriting changes before committing or fubar'ing a git reset.
If it's really so for some people, a case could be made that their problems can already be half-solved by changing the granularity of their commits: just commit more often.
Once I've tried one SCM to track every single change to my files, and the other for bulk commits. The result was quite satisfying.
It was some 15 years ago, we just switched from CVS to SVN and everyone thought it was the end of the history, at least for SCMs. Main branch was called master, and we were locking files by simply telling our teammates not to touch them.
At one time I decided I really missed the Eclipse feature which was tracking all local changes to the files, even without an active SCM. Making tiny commits was not an option, because I want it to track every file save, and we haven't yet figured out working with branches.
So I decided to install mercurial and made a hook which made a dummy commit every time I saved a file. Everything mercurial related was ignored by SVN.
Suddenly I was able to teleport myself to a version from before the lunch while SVN history was crystal clear.
Somehow I still miss integrated, per file, timeline in IDE.
This feels very useful for training agents but almost completely useless to humans? Do I really want to go back to every change I made? Whatuse would that be?
Jetbrains has had a "local history" feature for many years that had this, and I only use it occasionally but it feels pretty essential now. I once accidentally wiped uncomimted work in the terminal, and brought up Local History to restore it in a click. Or more casually, I'm struggling with a CSS design: what did it look like at 3pm? It's an effortless escape or experimentation valve between commits.
(It's also why you never need to manually save a file in jetbrains. It just assumes you want to save, because you can always safely go back if needed. Editors with manual saving feel cumbersome now)
I kind of do the conversation tracing thing already. Every change is attributed by Claude code session id (and timestamped obviously), then I have a jq based tool over cc session jsons that I save indefinitely. I've found it very important to be loud about what I actually wrote vs what the llm wrote.
Personal use solo repo. I don't know id want people seeing my cc sessions lol.
Looks like a great feature/product/functionality. I hope that Zed editor will make it work with LLMs/agents that run in the terminal, not just the Zen Agent itself. Is it even possible? I use OMP.sh agent running in Zed terminal.
This is the killer Zed feature, IMO, that they support almost any agent via extensions. I deleted Cursor but still use the agent via Zed because it’s a lot nicer experience, Zed never had popups telling me Grok is half off. I also use OpenRouter and Zed.dev with it, that’s three providers out of dozens that it supports.
Exactly. The more competitive the space is, the more important it is that products purposefully support 3rd-party providers. Now that people are increasingly switching from Anthropic/OpenAI for inference reseellers and on-prem compute, transparency and portability will be the new values, not being slightly better than the next agent or model.
Even more than supporting any agent that support almost anything period. They seem to take the default approach of using open community standards first when adding new features. By investing very heavily in LSP, ACP, DAP, tree-sitter etc it makes it easy for Zed to integrate with tons of different workflows, language, ecosystems etc quickly and easily. Its just good engineering and I love to see it.
I guess an interesting use case for this would be to pinpoint exactly which change was made by who in cases where multiple people / agents work on the same code at the same time. The Zed team is working on two features that are probably going to be built on top of DeltaDB: async collaboration and instant sharing. Instant sharing used to be talked about more in the past and it's still the feature I am most eager to see.
i think a system that uses git and file system artifacts (txt/md/xml files whatever preference) is much better. eng teams need to define their artifacts, eg plan file, reqs etc whatever is needed and important to that team. the agent convo itself doesn’t add much value IMO and adds unnecessary bloat.
Maybe I’m pessimistic, but I think this just gives micro-managers the data they need to micro-manage you. It invites people to scrutinize the process, instead of the results.
Not once have I wished to review a colleague’s conversation with an AI. I want to review finished work.
I look at this and I can only imagine layoffs being justified with “bad prompt quality”. An incident retrospective broadcasting your agent conversations to the entire company, with people speculating where you should have put the “make no mistakes”.
It’s a shame, really, because I like Zed as an editor, but I can’t bring myself to support them if this is the direction they’re choosing.
The sort of company that would do those things is going to be mandating you to use their enterprise AI plan anyway, which already lets them read all your conversations if they want to.
Letting them read my conversations is okay. Systematically reconstructing a timeline of every keystroke that went into a project, then immortalizing this timeline into a VCS, that’s the creepy part.
I suppose companies could do this anyway just by installing spyware in your work computer, but I would very much prefer if that spyware was not my editor of choice.
The way I’m seeing AI used by organizations is to produce process slop: tickets, summaries, docs, etc. rather than drive process change and optimization, or enable individuals to be more effective within the organization.
This landing page is not good. It's mired in out-of-context jargon and implementation detail. (What does "virtualize the worktree" mean? Git already has "effectively free" branching. What does branching "mid-run" mean? How does a teammate "join while the work is happening"? What/where are they joining?).
What actually is it and what value do I get out of it?
Reading the linked blog posts: It's a layer on top of git for CRDT-based collaborative editing that captures addressable code deltas + agent conversations. I could have real-time collaborative humans + agents development and can resume any agent conversation from any point in history, because it is all stored (powered by DeltaDB). I guess?
I actually might want this, but wouldn't know it from the landing page.
Jujutsu already does this. I had agent screwing up changeset over couple hour long session and it took one shell command to split on pieces with longer pause.
The only downside is that because of Jujutsu snapshots it can't be used inside sandbox (since all jj commands are mutating)
You still get snapshots of the working tree every time an agent calls a jj command, including reading ones. This is regular for me that I'm not worried to lose any intermediate state.
I had the great pleasure/misfortune of being forced to work with LaminDB, and I can tell you folks; data provenance isn't going away -- it will be used to train the very AI models snapping at our heels
This local history has saved my bacon on occasion. E.g. when I went too long without committing and something silly overwrote the file I was working on. Or because I did something silly with git. Then it is nice that there is still an independent backup elsewhere with different failure modes.
I really do like Zed, it's snappier than VS code and with pretty good vim emulation. Even through SSH, the native UI allows for no network latency, compared to something like vim+terminal. Like many other posters, I do feel like the core editor is getting worse.
Some issues I've encountered recently at work:
- Routinely encounter bugs with the python language server, it's particularly slow when compared to using VSCode. All that snappiest disappears when finding symbols after the '.' is considerably slower than my typing speed.
- I found similar issues with C++ code bases.
- It's also so hard to debug when something goes wrong. I was trying to find out why pyright keep spinning on startup, but the "LSP logs" never show anything.
- Connecting via SSH always opens the repo in restricted mode, even though I check the "Trust all projects in" checkbox. Never happened before.
- Debugger UI is a mess. Make a run, stop the run, edit the task.json, try running again, and 2 tasks show up. One for the old and one for the new.
--
I stuck with it because I really do like the UI, the performance, and initially it seemed to be an improvement from VS Code for my workflows, but lately I've been moving back to code as it seems more stable. I don't really use agent features, so I don't think I'm the target customer for this or other priority Zed features.
I do see the usefulness of something like this, I'm interested to see if the new agentic style of programming will result in the "code" part of editors becoming second class (for the better) and will allow for more rapid development. Although I'm not looking forward to read my coworker's "threads" of how a feature was made, I have a feeling that it will be more difficult to read than the raw code, as from my experience, LLMs don't really employ terse and effective communication (might not be entirely fair, I found ChatGPT /study mode pretty good.)
--
Anyone know of a good replacement editor that might meet my usecase? I'm really a batteries included type of developer.
Yeah apparently Zed was engineered for speed but I find it very slow in general, especially on startup.
VS Code can slow down as you add functionality, but Zed often just doesn’t have that functionality at all. So the proper comparison is “VS Code minus all extension-provided features” which is actually pretty fast.
this looks like jetbrains "local changes" but globally synced. "for the agents" mights scare people who jump to think about micromanagers, but these people mostly hide or fake work. even micromanagers will eventually give you an agent to use so it won't matter what is in the change log, what matters is speed+quality. and since you can iterate from any branch you have inginite try and retry possibilities
I have been using zed for almost 2ish years now, great overall, but last month when I tried ACP with opencode and even omp, it gave me many issues that it was easier just to use the terminal directly, oh, and I couldn’t select the model that time not sure if it’s fixed now.
Snapshots have an intention behind them as a chunk that the programmer thought of. It has infinite flexibility depending on the programmer/team. IMO they are a necessary feature, but the article considers it a bug. It’s an explicit and deliberate chunking of code writing process which is what you want in a chaotic “collaborative” environment. The system doesn’t get to decide but you decide when to commit.
Yes, it is open source, and I am not saying they should not do it. Chromium is also open source, but if they added too much nonsensical AI functionality, I would be upset. I am free to improve it on my own, but that does not mean I cannot express how I feel.
I have not gone to the Zed repository and created an issue telling them to stop developing DeltaDB because it takes away the maintainers’ time.
Zed should focus on basics. When it focuses on basics Zed is good.
https://github.com/zed-industries/zed/discussions/54150 failing to show newly created files and declining to provide a refresh button, instead adding a polling backend, breaks Zed on WSL.
Why a new version control system? Why not git, jj, or another existing system?
There should be a concept called "engineering privilege".
I see some companies do things that are widely out of their core product and engineers there just having fun with a bunch of different stuff with seemingly no commitment to bring in revenue. They'll re-invent/re-write things constantly, write a new compiler/package manager for their chosen language, create new frameworks, new open source projects unrelated to their companies, etc.
It must be awesome to be able to work in these conditions and still get paid. I'm not hating on it, it's how we get a lot of the good stuff we use.
My favorite example was from Joel Spolsky's company, where bright engineers got bored with working on bug tracking software so they wrote their own C# compiler.
Because every company may have different needs that are not fulfilled by standards software. We have seen the large number of companies whose goal is to make custom software for other companies So is not always that reinventing is fun, but it may actually be needed.
The same logic also applies for software that is used internally. It's not always that developers want to rebuild stuff, but that they need to rebuild or alter software to fit their / company needs.
We have not even talked about cost or licensing or other issues.
Yes except a lot of it isn't even fit for their purpose and they don't rely on it. Facebook switched from Cassandra to HBase because "eventual consistency is a difficult pattern". No shit, really?
> Every change is linked to the agent conversation that produced it. From any line of code, find the conversation. From any message, jump to the code it touched.
This is how Zed leapfrogs the field and becomes a centicorn.
If you can train a model to operate at a higher level and with looser requirements, you've gotten rid of engineers in the loop and built a system for the automated synthesis of high quality software.
You're training systems to produce higher quality chain of thought across the entire software development lifecycle.
This is training data for the next stage.
It's not just a step, it's a leap towards real productivity that your agents will actually use!
From 1996:
https://www.eng.auburn.edu/~kchang/comp6710/readings/They%20...
"There is the software. And then there are the databases beneath the software, two enormous databases, encyclopedic in their comprehensiveness. One is the history of the code itself -- with every line annotated, showing every time it was changed, why it was changed, when it was changed, what the purpose of the change was, what specifications documents detail the change. Everything that happens to the program is recorded in its master history. The genealogy of every line of code -- the reason it is the way it is -- is instantly available to everyone."
also from ^ article on Lockheed Martin Corp space shuttle software group ...
It's funny, zed started out really refreshing, and now it's degraded to the point where it's fairly unusable. Regularly it would tie up tons of resources on my fairly beefy laptop. I switched back to sublime and it is a breath of fresh air, I didn't realize how slow zed had gotten. It seems like maybe they lost their way a bit, but I recognize that also they need to make money presumably, and it's hard to do that with a large team and a focused product.
I don't want Zed to focus on accommodating WSL's quirks.
I would like Zed to have feature parity with VSCode from 6 years ago. Working well over SSH or WSL, which it currently doesn't, is a part of that, even if I don't use that everyday.
I am a paying user, and have no plans to stop supporting them financially, but I wish they locked in and went through GH Issues queue, starting from the oldest.
Of course vs code accommodates wsl. Same company. Personally I couldn’t care less how badly MS wants to make windows a viable development environment. It’s more irrelevant every day. I’m glad zed doesn’t waste too much time on it.
Their ACP integration is good (finally), at least with Codex. For when I want to be near the code, I prefer it now over TUIs or the Codex app. It feels much more immediate that way. Sticking to the basics will kill them IMO. They have been pitching themselves as agent-native for a while btw.
And the workflow they are describing in the post sounds amazing, I‘ve been wishing for that. It‘s orthogonal to CVS, no? You can still make PRs. But, crucially, you can loop people in early on, when it matters the most.
Unfortunately their Zed agent is completely broken when running on Windows connected to WSL/SSH remote because the terminal tool checks for paths using the local Path style (windows) which breaks when applied to paths on the remote (linux).
I fixed this and a couple related issues in a PR but it hasn't gotten any attention yet. I guess they are a bit swamped. https://github.com/zed-industries/zed/pull/59937
> Their ACP integration is good (finally), at least with Codex.
you must not use it much. try pasting in an image, quitting zed and restarting (ie: upgrading the app). it shows you the base64 encoded data instead of just showing the image. heck, it doesn't even show you the image when you paste it in. it just shows the filename.
ACP integration is very good, IMO. I'm using it with Claude Code. They've made so many improvements over time. I use image pasting myself, TBH I was excited it worked at all. Surprised that some rough edges around images are enough to ruin the whole thing for you.
images were just one glaring example... i didn't say anything about ruined, i just think a lot of those 'rough edges' are a bit more rough than they could be.
it is a code editor after all. how about integrating the LSP so that it doesn't have to resort to rg/grep/sed/scripts to edit code?
I use it constantly but haven’t felt the need to paste an image yet. I guess I‘m using pi and the Codex app more for that.
sorry, i didn't mean a quantity of time, i meant depth of usage.
I think conceptually, the deltas are exactly the type of “basics” Zed is built on.
I agree with you that I wish they'd just nail the editor. I've been using it for years now and I really like it, but there are a few rough parts (settings and LSPs especially) that could probably be tuned up in no time if the focus was there.
Having said that, I love the idea of correlating the agent sessions to the changes. I'm glad someone is working on it.
You don't need a tool for that just configure your agent to write commit messages that link to a LLM log. And changes between commits. Maybe don't care why do we even care?
Of course you are right. I could write my own software to do what I want. Or, I could just do it all myself manually. This argument is valid in pretty much all cases, so I concede the point.
Claude, for example, already does write a git commit trailer with a session ID.
I've been hitting that or something recent recently; it's what I would consider a severe bug because it can result in lost work and related. Basically, if you `git pull`, edit a file elsewhere, or use an LLM, zed's tree will remain stale unless you re-close and open the file. (Restarting the application doesn't sort it). So you end up overwriting the file with whatever was in the buffer.
Oh, that sounds related to what I experienced. I recently had another look at Zed and tried to use it to full-text search through my workspace consisting of symlinked folders. Relevant things would just now show up, cmd+P also wouldn’t find the files. That’s where the evaluation stopped.
One could argue using WSL is an anti-pattern: use a real OS. I kid, I kid.
Yeah I’m an avid user of the editor. I don’t care for their AI stuff I bring my own but I just want to see the editor become awesome. A more detailed object plugin ecosystem would be nice.
This specific issue drove me mad for a couple of hours, I thought my LLM wasn't creating files and lying about it
Did you even read the post? The idea is to tie conversations to code. This is a more sophisticated version of something like doltdb, not a git replacement.
So like Fossil then.
I like Fossil be because it can be all in one. WIki, a place to store architectural diagrams, tickets, basic webgui, etc.
Hmm, yeah before "DeltaDB" I have a big list of things I would rather they develop/fix:
- Vertical activity bar (https://github.com/zed-industries/zed/pull/58868#issuecommen... - rejected because it "doesn’t match our design vision for Zed")
This is my #1 annoyance: clicking microscopic icons at the very bottom of the screen to switch views.
- Don't crash my entire PC when there's a large (100+ MB) JSON file in the working directory
- Don't freeze when resuming from suspension (I suspect this is GPU-driver related on linux) forcing me to `killall zed-editor`
> This is my #1 annoyance: clicking microscopic icons at the very bottom of the screen to switch views.
Have you considered using a keybind? They even show which keybind to use when you hover them.
Amazing that someone would prefer a vertical bar that is 85% empty, wasted space instead of pressing a few keys on the keyboard - for a text editor of all things.
> Don't crash my entire PC when there's a large (100+ MB) JSON file in the working directory
I don't think you can put this on Zed. An unprivileged userspace application, even if coded poorly, shouldn't crash the machine.
> An unprivileged userspace application, even if coded poorly, shouldn't crash the machine.
How would you prevent that? It can write until root disk is full, can take all your memory, saturate your CPU and a whole bunch of other stuff we (users) like to be able to do for the things that need them.
Unless you're suggesting something like Qubes or similar UX?
Why develop this when the core editor is almost unusable? On Linux wayland, copypaste is broken, the file manager is laggy and snippets do not trigger before the heavy hint logic is done. This is just on top of my head, the list is long
"Almost unusable"? Slight hyperbole don't you think? I initially tried the Flatpak version on Fedora 44 and had a few minor problems, but since switching to the terra repos version (terrapkg.com) I've had zero issues, Have you tried that?
For me, its a massive improvement over IntelliJ for my daily workflow.
This is surprising to hear. I have Zed on my tablet as IntelliJ (PyCharm/RustRover) turn it into a space heater. Zed's introspection/refactoring/general language knowledge etc isn't on the same tier at JB; it's closer to VsCode in that regard. I.e., I'd classify Zed as a project-based editor, while JB is an IDE.
> massive improvement over IntelliJ for my daily workflow.
Why?
I used Zed for about 6 months but I lost motivation.
I have the full jetbrains subscription and use it for some debugging etc. I got vim for quick edits and cli tooling for ai. I got no use case for Zed.
Im on nixos and tried all immediate flavours of zed: stable, unstable and nightly. Nightly seems to be most stable and every time i pull the flake, i prey for more bugs to be fixed. Sometimes it works
I have been using Zed for about a year now on Debian 12 / Wayland and it works flawlessly. Even with Toshy (macOS) keybindings. I have it installed now on my Framework 13 Pro (also Fedora 44, and also with Toshy) and it works great aside from a subtle (100ms?) lag when using ctrl+arrow-key. Otherwise it is great.
The only thing that doesn't work well is inertial scrolling[0]. But scrolling is a fucking shit show on Linux no matter what... so I don't necessarily fault Zed for this. The same issue occurs in other apps.
- [0] https://github.com/zed-industries/zed/issues/14473
>> For me, its a massive improvement over IntelliJ for my daily workflow.
That's a bit of a low bar. Perhaps that's the disconnect?
I had to stop using Zed because of this issue (which persists despite being closed): https://github.com/zed-industries/zed/issues/38109
I can't use an editor where the file contents I'm being shown don't match the reality of the file. I guess this issue doesn't affect most people somehow, but in the age of CLI agents I'm not sure how more people aren't running into it.
I was really enjoying Zed for a short while before I ran into this fundamental issue.
I stopped using it due to this issue as well; too dangerous/risky.
I do not have this issue. (Yes, I am on Linux/Wayland)
All I can say is it's the only editor I've ever used with this type of issue and unfortunately it's a deal breaker for me. I can get by with any other editor even if they aren't my favorite, but only Zed made me realize that the most important thing an editor can do for me is accurately show me files.
Have you checked your fs.inotify.max_user_watches value? Maybe your value is too low and Zed can't watch the files that are changing. I know the JetBrains IDEs need the default to be increased on Linux because they detected it and prompted me to increase it.
Just installed it, works on my kde and niri setups (I logged out and switched to each), copy paste works as well. File manager laggi-ness I didn't notice I don't use it, I use it mostly to type stuff out.
I have never used snippets so unsure about that too, going from command bar seemed to work.
I think it might be bad, but could also be it's not correctly configured maybe try asking codex/claude or something to check if it can fix/fine tune the config, if that doesn't work file issues?
I do suspect it might be working for me cause I got lucky I do see an open issue[1], but given Zed is used a bunch the lack of comments does suggest maybe it's not that wide spread either way not a good look.
[1]: https://github.com/zed-industries/zed/issues/47004
Maybe I got lucky?? Does it not auto update on linux maybe I am on an old build... Very suspicious.
The copy-paste bug manifests itself after some time and the bug is silent.
I regularly leave it open for days and never run into any copy/paste bugs in Wayland/Flatpak.
I use it as a daily driver with few if any problems, but I'm on macOS. Maybe their Linux support is particularly crappy?
There are a lot of subtle bugs. I'm on macOS too, but perhaps my workflow is unusual? Every time I open it, I find small bugs. There are so many features that it's hard to maintain them, I guess.
For example, a lot of disable buttons do not actually disable stuff. They just hide it from the UI.
Yep. Many issues are Linux only. And even with the issues it’s still more pleasant to use than vscode/ium.
Agree, copy-paste feels like it works 80% of the time, which is maddening when 1 in 5 copy paste attempts fail. (mostly from zed to something else) Why??
At least Excel has the courtesy to telling you "Hey the copy-paste buffer said it didn't catch that, can you try again?"
Linux Wayland (Ubuntu 26.04)
It still doesn't actually display a window my end until I press a keyboard key or move the mouse. So very often I sit and wait for it to open, only to realize I have to press a random key for it to unhide.
Terrible Linux support.
That's something I've never encountered on Linux/KDE/Wayland/Flatpak across Debian Trixie, Arch, and Fedora. Might be worth nuking your local Zed files and trying again.
I started using Zed a couple weeks ago and immediately had issues with copy/paste/buffer on an M1 Mac. I tried a few suggested fixes and they didn't work.
I've been using this "almost unusable" editor daily on both Linux and Mac for the past year and now experience no issues. while I still have JetBrains I haven't used it at all in at least six months.
Ha if you thought it was laggy try it on software accelerated graphics.
why develop this when it seems jetbrains had a local history feature which recorded every change for the past decade(s)
> when the core editor is almost unusable
Shit, as someone who uses Zed everyday, I didn't realize what I am using is unusable.
I guess we have slightly different standards and/or use cases
Because it's not broken, the issues you're describing are far from universal. Also, fixing them won't bring any money.
A lot of the copy on this page feels like the kind of copy an AI would produce. But I’m assuming, surely, Zed writes their own copy so I don’t want to accuse them of doing that.
Marketing aside, this is the kind of thing (advancements in VCSes) that I’d usually be excited about, but which I’m less excited about coming from Zed because it makes me start to worry that they’re going to continue focussing less on the core editor experience.
I love Zed and I hope that it doesn’t start to go flakey.
Anytime a company/group announces some random new project now, I tend to immediately suspect over-reliance on AI. From what I’ve seen, AI addiction frequently manifests as an explosion in scope, possibly because LLMs make so much initial “progress” on new projects which you can then demo and have resource allocators extrapolate to some glorious shipping version. But then you have to go back & clean up the slop, progress slows to a crawl, and it’s all way less exciting. That initial rush can be addicting. Ultimately a supercharged version of eternal human impulses I guess.
You're too kind. It's obviously AI copy, and a huge negative signal (to me, anyway).
I'm always hesitant to assert something is definitely AI copy because everyone's tone of voice is different.
I'm also starting to wonder how much real, human-made copy is mimicking AI-created copy that has been filling marketing pages as of late. When people make marketing websites, they often riff and rip off other websites' (à la "Good artists copy; great artists steal"[0]) design ideas and language -- I wouldn't be surprised if there are humans now inadvertently adopting LLM's writing style like that.
[0] - https://en.wikiquote.org/wiki/Steve_Jobs#:~:text=Good%20arti...
Why do you think an AI-focused company would hesitate to put AI-generated copy on their website?
Hah :)
In my opinion / experience, AI copy is a generally more cumbersome to read and process. So I would assume a company at Series B stage would have a human (with marketing guidelines) go over the language used on a one-pager to announce a new product.
But -- evidently -- anyone's guess!
Sounds like a side-effect of being VC-backed. Like what the ex GitHub CEO went on to build (Entire). Unnecessary.
It also feels like his former company could implement his new company in a couple of weeks if they choose to.
> Every change is linked to the agent conversation that produced it
A total nightmare and imo every dev should push back using this stuff. You'd otherwise have management brag about you not being a good communicator with the ai agent, the same management that have no idea how to steer these tools.
I really like Zed the editor, I'm not sure I like the path they getting into
These sort of paper trails are for developers not management though, does your manager open all your commits and start commenting on them? If so, maybe apply somewhere else.
I really love Zed and I want them to really succeed. Unfortunately it seems that they are swamped with various things and the issue tracker is not a priority for the team, which fair enough. OSS is a gift and I take it as such. Wild proposal but if someone is willing to start a "user advocate" contributor, I'll be happy to donate monthly so user requests can be taken a look at :)
A non-existent problem that nobody asked to solve...
I disagree. JetBrains PyCharm has a similar "file history" feature that solves this very real and existing problem and has saved my ass more than once when accidentally overwriting changes before committing or fubar'ing a git reset.
They say that "history between commits" matters.
If it's really so for some people, a case could be made that their problems can already be half-solved by changing the granularity of their commits: just commit more often.
VCs count as people unfortunately.
tangent
Once I've tried one SCM to track every single change to my files, and the other for bulk commits. The result was quite satisfying.
It was some 15 years ago, we just switched from CVS to SVN and everyone thought it was the end of the history, at least for SCMs. Main branch was called master, and we were locking files by simply telling our teammates not to touch them.
At one time I decided I really missed the Eclipse feature which was tracking all local changes to the files, even without an active SCM. Making tiny commits was not an option, because I want it to track every file save, and we haven't yet figured out working with branches.
So I decided to install mercurial and made a hook which made a dummy commit every time I saved a file. Everything mercurial related was ignored by SVN.
Suddenly I was able to teleport myself to a version from before the lunch while SVN history was crystal clear.
Somehow I still miss integrated, per file, timeline in IDE.
A "smooth" editor with uncountable bugs and without smooth scrolling... I agree they should focus on basics.
This feels very useful for training agents but almost completely useless to humans? Do I really want to go back to every change I made? Whatuse would that be?
Jetbrains has had a "local history" feature for many years that had this, and I only use it occasionally but it feels pretty essential now. I once accidentally wiped uncomimted work in the terminal, and brought up Local History to restore it in a click. Or more casually, I'm struggling with a CSS design: what did it look like at 3pm? It's an effortless escape or experimentation valve between commits.
(It's also why you never need to manually save a file in jetbrains. It just assumes you want to save, because you can always safely go back if needed. Editors with manual saving feel cumbersome now)
I kind of do the conversation tracing thing already. Every change is attributed by Claude code session id (and timestamped obviously), then I have a jq based tool over cc session jsons that I save indefinitely. I've found it very important to be loud about what I actually wrote vs what the llm wrote.
Personal use solo repo. I don't know id want people seeing my cc sessions lol.
Looks like a great feature/product/functionality. I hope that Zed editor will make it work with LLMs/agents that run in the terminal, not just the Zen Agent itself. Is it even possible? I use OMP.sh agent running in Zed terminal.
It supports anything with ACP. So it can actually run Codex and Claude Code, not just the Zed Agent.
Looks like omp supports ACP, so all you have to do is specify a Custom Agent in Zed and it should just work.
https://omp.sh/docs/acp
https://zed.dev/docs/ai/external-agents
This is the killer Zed feature, IMO, that they support almost any agent via extensions. I deleted Cursor but still use the agent via Zed because it’s a lot nicer experience, Zed never had popups telling me Grok is half off. I also use OpenRouter and Zed.dev with it, that’s three providers out of dozens that it supports.
Exactly. The more competitive the space is, the more important it is that products purposefully support 3rd-party providers. Now that people are increasingly switching from Anthropic/OpenAI for inference reseellers and on-prem compute, transparency and portability will be the new values, not being slightly better than the next agent or model.
Even more than supporting any agent that support almost anything period. They seem to take the default approach of using open community standards first when adding new features. By investing very heavily in LSP, ACP, DAP, tree-sitter etc it makes it easy for Zed to integrate with tons of different workflows, language, ecosystems etc quickly and easily. Its just good engineering and I love to see it.
I use this every day — https://zed.dev/blog/terminal-threads
I guess an interesting use case for this would be to pinpoint exactly which change was made by who in cases where multiple people / agents work on the same code at the same time. The Zed team is working on two features that are probably going to be built on top of DeltaDB: async collaboration and instant sharing. Instant sharing used to be talked about more in the past and it's still the feature I am most eager to see.
i think a system that uses git and file system artifacts (txt/md/xml files whatever preference) is much better. eng teams need to define their artifacts, eg plan file, reqs etc whatever is needed and important to that team. the agent convo itself doesn’t add much value IMO and adds unnecessary bloat.
This makes me uncomfortable.
Maybe I’m pessimistic, but I think this just gives micro-managers the data they need to micro-manage you. It invites people to scrutinize the process, instead of the results.
Not once have I wished to review a colleague’s conversation with an AI. I want to review finished work.
I look at this and I can only imagine layoffs being justified with “bad prompt quality”. An incident retrospective broadcasting your agent conversations to the entire company, with people speculating where you should have put the “make no mistakes”.
It’s a shame, really, because I like Zed as an editor, but I can’t bring myself to support them if this is the direction they’re choosing.
Not only this. I'm also shuddering at the thought of how much more noise this adds which can only be filtered through LLMs.
Don’t worry, Zed (or another company in their VC backers’ portfolios) will helpfully sell you inference to filter through that noise!
I guess this data is not intended to be used all the time. But occasionally you may find it useful in blame or bisect or similar things.
The sort of company that would do those things is going to be mandating you to use their enterprise AI plan anyway, which already lets them read all your conversations if they want to.
Letting them read my conversations is okay. Systematically reconstructing a timeline of every keystroke that went into a project, then immortalizing this timeline into a VCS, that’s the creepy part.
I suppose companies could do this anyway just by installing spyware in your work computer, but I would very much prefer if that spyware was not my editor of choice.
It's also delusional to think you can find that one edit that was really useful but you forgot to commit but then deleted in another giant commit.
As anyone who's ever tried to split a days worth of work apart with `git add -p` is well aware.
It may be feasible with a million agents, but they could remember what they typed three hours ago anyhow.
The way I’m seeing AI used by organizations is to produce process slop: tickets, summaries, docs, etc. rather than drive process change and optimization, or enable individuals to be more effective within the organization.
this is the perfect description of how I've seen AI used in most organizations
I don't know. I've been giving Zed a chance for almost half a year now and it just doesn't seem to be going the way I would want an editor to go.
I really don't want management critiquing when I swear at Claude, thanks.
This landing page is not good. It's mired in out-of-context jargon and implementation detail. (What does "virtualize the worktree" mean? Git already has "effectively free" branching. What does branching "mid-run" mean? How does a teammate "join while the work is happening"? What/where are they joining?).
What actually is it and what value do I get out of it?
Reading the linked blog posts: It's a layer on top of git for CRDT-based collaborative editing that captures addressable code deltas + agent conversations. I could have real-time collaborative humans + agents development and can resume any agent conversation from any point in history, because it is all stored (powered by DeltaDB). I guess?
I actually might want this, but wouldn't know it from the landing page.
You can simulate this by having a jj repo with file system watches.
Please just make a good editor and stop getting distracted by AI nonsense.
I start to sound like a Jujutsu fanboy but...
Jujutsu already does this. I had agent screwing up changeset over couple hour long session and it took one shell command to split on pieces with longer pause.
The only downside is that because of Jujutsu snapshots it can't be used inside sandbox (since all jj commands are mutating)
> The only downside is that because of Jujutsu snapshots it can't be used inside sandbox (since all jj commands are mutating)
I guess it depends? You could certainly start with another repo clone that's separate from the one you work on yourself?
Jujutsu is great, but it does not do this. It has no knowledge of agent sessions.
You still get snapshots of the working tree every time an agent calls a jj command, including reading ones. This is regular for me that I'm not worried to lose any intermediate state.
Sounds a lot like JJ.. could just make JJ a first class citizen in zed and skip the hard part.. plus it’s git compatible
I’m curious, are many people using Zed as their daily driver in their professional work? (Not just in side projects)
It'll take more than a mouse jiggler to fool this new bossware.
Previous discussion: https://news.ycombinator.com/item?id=48492533 (June 11th, 225 comments)
Is it available now? Still seems to be a waitlist?
I had the great pleasure/misfortune of being forced to work with LaminDB, and I can tell you folks; data provenance isn't going away -- it will be used to train the very AI models snapping at our heels
I have fallen in love with Zed.
So is this basically Beads + Dolt? This isn’t to diminish it, it’d be great if this ends up being a less “vibe code-y” alternative to those.
If not, beads + dolt seems like it would be pretty great.
that was my first question too. Here's hoping
This local history has saved my bacon on occasion. E.g. when I went too long without committing and something silly overwrote the file I was working on. Or because I did something silly with git. Then it is nice that there is still an independent backup elsewhere with different failure modes.
Like a branch on the server?
That would still be in git. The local history in IntelliJ works independently of any VCS.
Also I’ve had formatters botch files including changes I hadn’t committed yet. For that the local edit history is also useful.
Or add it to git? What's wrong with git?
I really do like Zed, it's snappier than VS code and with pretty good vim emulation. Even through SSH, the native UI allows for no network latency, compared to something like vim+terminal. Like many other posters, I do feel like the core editor is getting worse.
Some issues I've encountered recently at work:
- Routinely encounter bugs with the python language server, it's particularly slow when compared to using VSCode. All that snappiest disappears when finding symbols after the '.' is considerably slower than my typing speed.
- I found similar issues with C++ code bases.
- It's also so hard to debug when something goes wrong. I was trying to find out why pyright keep spinning on startup, but the "LSP logs" never show anything.
- Connecting via SSH always opens the repo in restricted mode, even though I check the "Trust all projects in" checkbox. Never happened before.
- Debugger UI is a mess. Make a run, stop the run, edit the task.json, try running again, and 2 tasks show up. One for the old and one for the new.
--
I stuck with it because I really do like the UI, the performance, and initially it seemed to be an improvement from VS Code for my workflows, but lately I've been moving back to code as it seems more stable. I don't really use agent features, so I don't think I'm the target customer for this or other priority Zed features.
I do see the usefulness of something like this, I'm interested to see if the new agentic style of programming will result in the "code" part of editors becoming second class (for the better) and will allow for more rapid development. Although I'm not looking forward to read my coworker's "threads" of how a feature was made, I have a feeling that it will be more difficult to read than the raw code, as from my experience, LLMs don't really employ terse and effective communication (might not be entirely fair, I found ChatGPT /study mode pretty good.)
--
Anyone know of a good replacement editor that might meet my usecase? I'm really a batteries included type of developer.
Yeah apparently Zed was engineered for speed but I find it very slow in general, especially on startup.
VS Code can slow down as you add functionality, but Zed often just doesn’t have that functionality at all. So the proper comparison is “VS Code minus all extension-provided features” which is actually pretty fast.
Too bad VCs won't give you money for just building a fast editor.
Nice, how about fix bugs?
Basically what IntelliJ calls local history. It is a feature that has saved my bacon. https://www.jetbrains.com/help/idea/local-history.html
Similar, but not as polished, also exists in VSCode and derivatives, named the same way. Definitely saved me a few times as well.
This saved me from accidental resets in the past. I learned it existed because the files were showing up in mdfind
Yes, I concur. Local History is wonderful; nice to see Zed getting its own version.
Ooooh, this is cool. I was unaware of this. Thanks for the link.
Can maybe think of one time in the last 10 years that I would have wanted this. Its only purpose is for training LLMs
this looks like jetbrains "local changes" but globally synced. "for the agents" mights scare people who jump to think about micromanagers, but these people mostly hide or fake work. even micromanagers will eventually give you an agent to use so it won't matter what is in the change log, what matters is speed+quality. and since you can iterate from any branch you have inginite try and retry possibilities
What problem does this product solve?
Collaborative (as in simultaneous multiplayer) editing I think, which has been Zed's mission since before LLMs were a thing...
See https://zed.dev/blog/introducing-deltadb but also https://zed.dev/blog/crdts from years ago.
Is this closed source?
I have been using zed for almost 2ish years now, great overall, but last month when I tried ACP with opencode and even omp, it gave me many issues that it was easier just to use the terminal directly, oh, and I couldn’t select the model that time not sure if it’s fixed now.
Snapshots have an intention behind them as a chunk that the programmer thought of. It has infinite flexibility depending on the programmer/team. IMO they are a necessary feature, but the article considers it a bug. It’s an explicit and deliberate chunking of code writing process which is what you want in a chaotic “collaborative” environment. The system doesn’t get to decide but you decide when to commit.
It should work with github inbetween commits, that i may use. Right now I'm using staging as an in bewteen in a dangerous way.
This is a big unsolved dev tooling problem. Glad someone competent is working on it.
Yes, but it saps away the time that could be used to improve zed.
I wonder if a large enough user community could fork zed and actually make progress on the bugs that matter?
Or it’s actively used to improve Zed faster (or more efficiently).
zed is open source, you are free to improve it on your own
Yes, it is open source, and I am not saying they should not do it. Chromium is also open source, but if they added too much nonsensical AI functionality, I would be upset. I am free to improve it on my own, but that does not mean I cannot express how I feel.
I have not gone to the Zed repository and created an issue telling them to stop developing DeltaDB because it takes away the maintainers’ time.
Rust cult stikes again.
I don't understand, what do you mean by this?