The Rust compiler always produces quite large binaries compared to other programming language. I notice there's a (closed) issue on the Zed github [https://github.com/zed-industries/zed/issues/34376],
> At this time, we prioritize performance and out-of-the-box functionality over minimal binary size. As is, this issue isn't very actionable, but if you have concrete optimization ideas that don't compromise these priorities, we'd be happy to consider them in a new issue.
I think there should be a best-of-both-worlds type of linking - during compilation, the linker places a statically compiled library at a certain address, but doesn't include it in the binary. Then, during startup, the OS maps the same library to the given address (sharing the data between processes). This would improve memory use and startup time both and performance, avoiding dynamic linking.
Of course you need to match the exact versions between the compiled executable and the dependency, but this should be a best practice anyways.
Static linkers generally don't compile a "full copy" of the library. Just the code paths the compiled application uses. The compiler may have also made optimizations based on the apppication's usage patterns.
I say "strangely" because honestly it just seems large for any application. I thought they might not be doing LTO or something but they do thin LTO. It's just really that much code.
> The world moved into dynamic linking in the 1980's for a reason.
Reasons that no longer exist. Storage is cheap, update distribution is free, time spent debugging various shared lib versions across OSes is expensive.
Tbh, the rights and wrongs aside, I suspect "everyone" is complaining about it because it's the easiest thing to talk about. Much like how feature discussions tend towards bikeshedding.
My /usr is 15G already, and /var/lib/docker isn't that far off despite people's obsession with alpine images. If more people would dismiss storage as cheap it'll quickly become expensive, just not per GiB.
> update distribution is free
I wouldn't be surprised if at one point Github would start restricting asset downloads for very popular projects simply because of how much traffic they'd generate.
Also, there's still plenty of places on the planet with relatively slow internet connectivity.
Storage doesn't really feel cheap. I'm considering buying a new laptop, and Apple charges $600 per TB. Sure, it's cheaper than it was in the '80s, but wasting a few gigabytes here and a few gigabytes there is quickly enough to at least force you to go from a 500GB drive to a 1TB drive, which costs $300.
It's the reality of storage pricing. The general statement "storage is cheap" is incorrect. For some practically relevant purposes, such as Apple laptops, it's $600/TB. For other purposes, it's significantly below $50/TB.
You could say "just don't buy Apple products". And sure, that might be a solution for some. But the question of what laptop to buy is an extremely complicated one, where storage pricing is just one of many, many, many different factors. I personally have landed on Apple laptops, for a whole host of reasons which have nothing to do with storage. That means that if I have to bump my storage from 1TB to 2TB, it directly costs me $600.
If you're buying Apple then you should expect inflated prices.
I got a 4TB NVMe SSD for like 350€, a 2TB one goes from 122 - 220 € depending on read/write speeds.
I don't check the installation size of applications anymore.
I'm just saying that $600/TB is a real storage price that lots of people deal with. Storage isn't universally cheap.
This feels especially relevant since we're discussing Zed here, the Mac-focused developer tool, and developers working on Mac are the exact people who pay $600/TB.
A 2TB SSD for the Framework 13 cost me 200 euros. But I agree that it's not cheap, files are getting bigger, games are big, apps are huge, and then you need backups and external storage and always some free space as temp storage so you can move files around.
I don't need to "get far in the Apple universe", I need a laptop. My current MacBook Pro cost about the same as the Dell XPS I was using before it, I like nice laptops
RAM isn't cheap (it may be for your tasks and wallet depth, but generally it isn't, especially since DDR5). Shared objects also get "deduplicated" in RAM, not just on disk.
What objects is the Zed process using that would even be shared with any other process on my system? Language support is mostly via external language servers. It uses its own graphics framework, so the UI code wouldn't be shared. A huge amount of the executable size is tree-sitter related.
I 100% agree.
As soon as you step outside of the comfort of your Linux distributions' package manager, dynamic linking turns into dependency hell.
And the magic solution to that problem our industry has come up with is packaging half an OS inside of a container...
OSes don't load the full executable into physical RAM, only the pages in the working set. Most of the Zed executable's size is tree-sitter code for all the supported languages, and only needs to page in if those languages are being used in a project.
Even if it’s delta, it cannot patch itself when running on Windows. So it runs the updater, creates a new exec and switches to it after relaunch. Same as Chrome or Firefox.
Is it? On Linux, you can overwrite the file, but the underlying inode will still be open, and the 'invisble' old version will linger around - you don't have any easy way short of restarting everything to make sure the new versions are being used.
And with Chromium this directly leads to crashes - when you update the browser as its open, the new tabs will open with the new version of the binary, with the old ones still using the old binary - which usually leads to crash.
I prefer 'you cannot do X' instead of 'we allow you to do it, but it might misbehave in unpredictable ways'.
I don't use Chromium. I never had issues with Apache, MySQLd, Firefox, Thunderbird, ... . You can even swap out the Linux kernel under userspace it still keeps all running.
> maybe programs shouldn't be allowed to update themselves.
Honestly I'd be all for this if the OS had a good autoupdate mechanism for 3rd party applications. But that's not the world we live in. Certainly not on windows - which is too busy adding antivax conspiracy articles to the start menu.
Will it though?
I mean it's a lot for a "text editor", but much less than a classical IDE.
And 400M is pretty negligible if you're on Windows, where your OS takes up dozens of GB for no reason.
From first hand experience, every British millionaire I grew up with or know, have left the UK for a country that treats them better.
The only ones that haven’t moved, are those who are considering it, can’t move cause their business is dependant on the UK, or their family/kids need them in the UK.
If your income doesn’t require you to stay in the UK, why would you stay there?
How do I know the LLM isn't lying to me? AIs lie all the time, it's impossible for me to trust them. I'd rather just go to the actual source and decide whether to trust it. Odds are pretty good that a programming language's homepage is not lying to me about the language; and I have my trust level for various news sites already calibrated. AIs are garbage-in garbage-out, and a whole boatload of garbage goes into them.
They do this already, but the problem is it takes me more time to verify if what they're saying is correct than to just use a search engine. All the LLMs constantly make stuff up & have extremely low precision & recall of information
I don't understand how that's an improvement over a link to a project homepage or a news article. I also don't trust the "verbatim snippet" to actually be verbatim. These things lie a lot.
>Odds are pretty good that a programming language's homepage is not lying to me about the language
Odds are pretty good that, at least for not very popular projects, the homepage's themselves would soon be produced by some LLM, and left at that, warts and all...
None of the LLMs (not even Grok) are "continuously trained" on news. A lot of them can run searches for questions that aren't handled by their training data. Here's Grok's page explaining that: https://help.x.com/en/using-x/about-grok
> In responding to user queries, Grok has a unique feature that allows it to decide whether or not to search X public posts and conduct a real-time web search on the Internet. Grok’s access to real-time public X posts allows Grok to respond to user queries with up-to-date information and insights on a wide range of topics.
i can also use my human brain to read a webpage from the source, as the authors intended. not EVERY question on this planet needs to be answered by a high resource intensive LLM. Energy isn’t free you know. :)
Other considerations:
- Visiting the actual website, you’ll see the programming languages logo. That may be a useful memory aide when learning.
- The real website may have diagrams and other things that may not be available in your LLM tool of choice (grok).
- The ACT of browsing to a different web page may help some learners better “compartmentalize” their new knowledge. The human brain works in funny ways.
- i have 0 concerns of a hallucination when readings docs directly from the author/source. Unless they also jumped on the LLM bandwagon lol.
Just because you have a hammer in your hand doesn’t mean you should start trying to hammer everything around you friend. Every tool has its place.
It's just a different kind of data. Even without LLMs, sometimes I want a tutorial, sometimes I want the raw API specification.
For some cases I absolutely prefer an LLM, like discoverability of certain language features or toolkits. But for the details, I'll just google the documentation site (for the new terms that the LLM just taught me about) and then read the actual docs.
Search is best viewed as a black box to transform {user intention} into {desired information}.
I'm hard pressed to construction an argument where, with widely-accessible LLM/LAM technology, that still looks like:
1. User types in query
2. Search returns hits
3. User selects a hit
4. User looks for information in hit
5. User has information
Summarization and deep-indexing are too powerful and remove the necessity of steps 2-4.
F.ex. with the API example, why doesn't your future IDE directly surface the API (from its documentation)? Or your future search directly summarize exactly the part of the API spec you need?
I don't know the exact word for this case, but sometimes you want the surrounding information to what you're looking for. Often I skim documentations, books, articles,... not in search for a specific answer but to get the overview of what it discusses. I don't need a summary of a table of contents. But it's a very good tool for quickly locating some specific information. Something like
Language Implementation Patterns (the book) |> Analyzing Languages (the part) |> Tracking and Identifying Program Symbols (the chapter) |> Resolving Symbols (the section)
or
Unit Testing: Principles, Practices,and Patterns (the book) |> Making your tests work for you (the part) |> Mocks and test fragility (the chapter) |> The relationship between mocks and test fragility (the section) |> Intra-system vs. inter-system communications
or
Python 3.13.3 Documentation (docs.python.org) |> The Python Standard Library |> Text Processing Services |> string
Could never understand that obsession with summmarization. Sure, it may be be useful for long-form articles or for low-quality content farms, but most of the time you are not reading those.
And if you are reading technical docs, especially good ones, each word is there for a reason. LLM throw some that information away, but they don't have your context to know if the stuff they throw away is useful or not. The text the summary omitted may likely contain an important caveat or detail you really should have known before starting to use that API.
Yes, you can use grok but you could also use a search engine. Their point is that grok would be less convenient than a search engine for the use case of finding Frob's website's homepage.
Perplexity solves this problem perfectly for me. It does the web search, reads the pages, and summarizes the content it found related to my question. Or if it didn't find it, it says that.
I recently configured Chrome to only use google if I prefix my search with a "g ".
