Hacker News new | past | comments | ask | show | jobs | submit login
CudaText: Cross-platform code editor (cudatext.github.io)
196 points by nitinreddy88 on July 24, 2021 | hide | past | favorite | 69 comments



> Disclaimer: word "cuda" is taken from Serbian language, it means "miracles".

This has to do with a Serbian word and not the Nvidia GPU programming language.


Weird, the actual serbian word is "čuda", pronounced "chuda", and they're completely ignoring the diacritics.


In Polish it’s actually “cuda”, without diacritics (same meaning). Regardless, one often ignores diacritics in writing in English/Western context. My last name has “ń”, though all the airlines insist to put “n” there; same with Western European banks. While for some words they maybe ambiguities if you skip the diacritics they are usually resolved by the context. Polish ppl are def used to read a text without diacritics and not surprised by this. (Of course a larger body text is a bit harder/slower to read, but it happens often when text messaging or in a Western context). I don’t know for sure, but I suspect that Serbian is similar.


> Regardless, one often ignores diacritics in writing in English/Western context

For example I think most people will write the name of the car make Škoda as Skoda.


I think a lot of this is because, especially historically, people outside Czechia etc didn't know how to type Š. It's an effect of English-speaking countries defining global standards for computers, aviation and so on.

Hypothetically, and with a little exaggeration to make the point, if Sweden had somehow had that global influence in the mid-20th century, the American actress "Raquel Welch" might have been stuck with "Rakuel Velch" in her passport, and Swedish people would wonder why she was annoyed by this.


And that's how, (in my head,) we ended up with the international spelling of "Česko" using the Polish "cz".


The “international” spelling is actually the English one. E.g. France and Germany respectively use “tch” and “tsch”, which correspond to the pronunciation in their languages.


It hindi language, chuda means 'get f*ckd'.


Writing ČudaText rather than CudaText would help clarify it's nothing to do with Nvidia.


Yeah, I went in thinking I'd see some kind of wild text editor implemented entirely on the GPU


Or an editor that somehow makes Cuda programming easier.


As a croat, I don't think this works. It didn't even cross my mind that it would be miracles, a) because it's plural, b) it's combined with the english word "text", c) would be more correct to spell it with 'ch' and not just drop the diacritic d) it's a noun and not an adjective which would be more suitable.


Cuda is a trademark of Nvidia. I expect the outcome to be the same as if it was named NvidiaText.


Nvidia doesn’t have any connection to text editors. Can Apple sue people for selling fruit under the name “apple”?


They've essentially done that and won, so.... Yeah?

There was a smallish apple producer (the fruit) that used a red apple with a bite taken out that lost the right to their logo. More recently there was another recipee startup with a pear logo that apple sued


What a broken system. The new blasphemies of the modern age are corporations overstretching into ownership of words representing things older than the language itself.

At least Google had the decency to invent a word (itself a derivative of a word invented by a child).

In the UK the cloud IaaS vendor Skyscape was forced to change name by a certain media company for only part of their name, in an unrelated field.

How about if you want to protect a trademark you don't use a word with real world meaning? What benefit does it give society to give these protections to extant words?


Also Microsoft OneDrive used to be known as Microsoft Skydrive and the same company objected.

Wouldn't be so bad if Sky had contributed anything useful to the world in the last twenty years.


Less sport on TV? Actually I do follow F1 through the Channel 4 highlights, but last weekend it was all live from Silverstone. I realised my life is too short for hours and hours (and hours) of coverage. The highlights are perfect. Sorry Sky.


They also sued polish company a.pl a few years ago because they said it sounds like 'apple' when you read it in english (.pl being of course polish TLD and it is read very differently than 'apple'). Not sure how this was resolved but the site is still up


I clicked thinking it was some NLP project that ran on the GPU.


>CudaText: Alternative open source version of Sublime Text

Doesn't seem to have any relation to Sublime Text. "CudaText: Free Software alternative to Sublime Text" might be a better title.


Or just text editor


I'm really impressed with this so far! It's very fast and looks great. The lexer I was able to find within 30 seconds of opening it for the first time highlights Haskell code better than Sublime- I'm not seeing weird syntax errors like I do in Sublime and quasiquoting works correctly.

I immediately miss "Find in files" bound to Control P in Sublime- I didn't see similar functionality in a list of plugins https://github.com/halfbrained/cudatext_plugins_list but I may have missed it.

I've been looking for a go-to lightweight editor so I'll keep trying it out. I know some people will disagree but it doesn't quite sit right with me that Sublime charges $100 and not only relies on the community to maintain plugins, but charges those plugin maintainers as well (if they don't want to get hit with nag messages). I don't think that's really working out.

Edit: Title referred to this editor as "open source version of sublime" at the time of the writing of this comment


The FindInFiles plugin is mentioned in that plugin list. Version 4 is better than 3.


> It starts quite fast: ~0.3 sec with ~30 plugins, on Linux on CPU Intel Core i3 3Hz.

That is impressive!


Yes, especially the 3Hz. Basically, that means the editor is loaded in 1 cycle.


CISC vs. RISC, bah! My text editor is implemented as a single CPU opcode! It's KiloMegaGigaTeraCISC.

You should see the silicon that implements the web browser instruction. KMGTCISC may not dispatch many ops per second, but what ops they are!


Implementing the LoadCudaText opcode has proven surprisingly useful.


Fixed the homepage text. Thanks.


Just gave this a quick try out. It's really pretty great!

It's become my new default "occasional/quick" editor. I use VSCode for work & as my "main" editor. But when I just want to edit a file, quickly - using a GUI editor, but without all the baggage/projects/etc... this is it.


A little disappointed it is not massively parallel text-processing on GPUs...


The page says it: Disclaimer: word "cuda" is taken from Serbian language, it means "miracles".


Yeah, a bit odd to use cuda in the name since it's such a specific thing.


They explain: "Disclaimer: word "cuda" is taken from Serbian language, it means "miracles"."

Right or wrong, I wouldn't be too surprised if they got a nastygram from NVIDIA's legal department at some point.


When this will happen, I will rename the app to CudeText. I am the developer.


I'm glad you have a plan - but I hope they leave you in peace!


Written in Free Pascal? Fun we don't see that often.


The main reason I use sublime is it’s super fast and uses very little battery. How is this in those respects?


I just tried it, did some light testing - and it's definitely comparable to Sublime in speed.


In my experience, CudaText starts up slightly faster.


Unfortunately does not have the search capabilities of Sublime Text. You need a plugin for searching through files, which is a bit awkward to control (bulky key-bindings, strange result panel, etc...). That's the reason why I finally bought Sublime Text.


The plugin can be improved, it is open source. FindInFiles does it’s job good, though.


A past thread from 2017:

CudaText: A lightweight, cross-platform code editor - https://news.ycombinator.com/item?id=13851189 - March 2017 (44 comments)



Wow, that animated cheetah is really giving me those late 90s vibes



Neat. From the features it looked like it could've finally replaced wxMedit as my general purpose editor.

But alas there's a couple of things not quite working for me.

* re-interpreting a files encoding as UTF-16/UTF-8 doesn't seem to switch it for some reason

* no hex mode search and replace (sure you can use regex escape codes, but that's more a workaround than a solution)

* re-opening pngs/jpg other images in hex mode require re-opening the file, just dragging it into the window

* couldn't find the "search/replace in files"


I cannot understand what your items 1, 2, 3 mean (item 4 means you didn't find the plugin FindInFiles). Please file an issue in the Github, so we can talk.


FindInFiles plugin exists in addons. About other issues: I will check them, but you can also file a bug report on GitHub.


If you want all of that, and more, there is always Emacs :)


But why install a whole OS if all you need is a text editor?


Tangentially related: does Object Pascal (the language CudaText is written in) have comparable memory safety to say Rust?


Freepascal is similar to C++ in that it requires the programmer to allocate and deallocate memory if you are using pointer objects.


What is the relation to

    https://aur.archlinux.org/packages/cudatext-qt5-bin/
    https://aur.archlinux.org/packages/cudatext-gtk2-bin/
that is sold as “Cross-platform text editor, written in Lazarus”, but links to an entirely different URL?


That's the same software with different UI backends.

If you click the "entirely different URL", it redirects to the HN post's URL. The AUR URL is apparently the URL of the author's company, and I guess that the editor moved to its own URL now.


Thanks!


Interestingly enough there is no arm version for mac, but there is a version for Haiku.


That's open source for you. I once worked on a project that had RiscOS as a first-class platform, and I was pretty sure that the maintainer was the only user of RiscOS in the world.


Good to see a light weight editor. Another bloat-free editor I've my eye on is lite [1].

[1]: https://github.com/rxi/lite


A bit tangential.. but can anyone recommend some good resources on writing a (emacs-like) text editor? I'm curious how they look architecturally


You want this:

The Craft of Text Editing

—or—

Emacs for the Modern World

–by–

Craig A. Finseth

https://www.finseth.com/craft/


looks awesome. thank you


Maybe uEmacs source[0] will help? Torvalds still uses it according to his last interview..

[0] https://github.com/torvalds/uemacs


Is there any good guide on how to set up LSP for autocomplete for react/jsx or plain js in CudaText? I really want to give this a try.



Anyone known if CudaText has features to edit Lisp or Scheme family of languages? Will be checking out the project.


Yes, both lexers are present in the Addons Manager.


There a Lisp lexer, and that's it, it seems.


Yes




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: