I'm just hoping that ST3 will be marked as final soon, so that Skinner may collect the upgrade fees before VSC has completed its take over of the world (of free editors). As a long time fan, I do get - and admire - the perfectionism, but hopefully there's room for dot releases after the 3.0 final.
You'd t think with their big team and resources, they would have been able to use a more performant cross platform toolkit, say Qt or maybe some of their .Net stuff.
Why did they choose to base their editor on Electron, of all things, especially considering the existing results of Atom (horrible) is beyond me.
It is an example that all the gazillion dollars and huge number of s/w engineers will not automatically do the right thing just because it is right. Most likely they will hop on the bandwagon, add cool words to their quickly concocted webapp masquerading as powerful editor.
In addition 1M+ plugins, open-source, cross-platform etc will keep cool kids not just happy but they will also aggressively challenge whoever is using old but efficient software products.
I think the development of VS Code was kind of an accident. They built Monaco (the editor component) for Visual Studio Online, then made it a standalone text editor as a side project and it ended taking off big time.
If you compare it to the likes of Brackets and Atom, it was significantly faster than the others at release... I'm not sure what can be done to make significant performance gains... I mean being able to even handle large files (disabling any highlighting/autocomplete, and only displaying portion of the file may be necessary)... as it is, for most of what I've used it for vsc is great.
Electron is essentially Node grafted to Chromium... There are a few things that could be done, but would depend on a lot of upstream cooperation to do so.
Sublime uses skia as its rendering library, which is what blink (chromium frontend) uses for rendering. May be electron can provide a nicer api to interface directly with skia.
By giving performant lower level access to render, may be VSC can move some of the perf sensitive code directly to compiled optimized binaries.
Sort of like what node native libraries do.
Sublime does some pretty great dark magic. I was impressed to see vscode approach similar speeds to sublime in search. I believe there is plenty that can be done to make rendering fast, reduce memory consumption and make it a bit light weight. But it would require access to metal APIs
VSCode's plugin model is pretty impressive and tends to isolate the side effects of poorly performing plugins (including language/autocomplete, etc)... so that the main editor tends to be responsive. There should be room to tweak, but I'm pretty happy overall. Way better experience than any full IDE I've tried... I only hope they keep it that way as features come in.
Also, as an Electron app, there is a node element, so compiled components are already an option.
I'm between VS17 with Resharper and Ncrunch, VSC, and Sublime Text myself. (VS17 for all my .net stuff, which picked up about 16 months ago). Even with a 4.2ghz 6 core CPU/32/NVME, the formers are slow compared to ST. Sublime is just so instantaneous it never gets old.
Yet VSC is like a polished machine with so much weight behind, it's almost a wonder ST3 can stand up to it. That's why I'd like the 3.0 release to be official.
(Fwiw I used Textpad for over 15 years before finding ST. Compared to Eclipse etc that was another speed demon).
I tried to switch to VSC the other day but its insane. I have a Surface Pro 3 and it was so slow to edit an 80 line Golang file (With the go plugin). I just cant use a text editor like that to get anything done. Like scrolling lagged, keystrokes lagged. Meanwhile with Sublime and GoSublime on the same laptop there is zero lag.
Using both is very good. Sublime has some text transformation features that VSC doesn't have (I use a ton of regex replace and multi-cursor typing -- VSC has these but there are some menu items in ST that are not in VSC), VSC has a debugger, source control out of the box, and can use symbol servers.
VSC is very much prepared for cross-platform .NET Core development. ST3 is more generally prepared for a lot of other languages and project types. If you're on Windows using a .NET language, Visual Studio should supersede VSC in almost every way though, and ST works better as your lightweight text editor.
And now, a moment of silence for those poor Notepad++ fans. :)
Sublime does keyboard macros that make my life easier which VSC and its add-on don't have. Now if Sublime had a option to delete trailing whitespace every time I hit return, I could return to my PFE days of macro productivity.
When I start doing coding and text manipulation again, it looks like it will be Sublime I purchase.
I use the editorconfig plugin at work. You have an .editorconfig file at the base of your projects and everyones text editor gets the same settings on a per-project basis.
One of the default settings is to trim excess whitespaces.
I actually want it to trigger on each line when hitting return. It really helps with macros you can write when text munging. I didn't spend as much time as I would have liked looking at each editor but a command-key that deletes all ending whitspace on a line would also be acceptable.
VS Code is one of the most organised/well-run open source projects that I know. Just take a look at their updates ie. (https://code.visualstudio.com/updates/v1_11). I'm just as excited to read the updates as I am to try each new release.
It's a bit unfair to compare a well-backed enterprise team to a small one (with 2-3 devs?), but it sure does make for an awesome user experience for developers.
Much better, more unique and somewhat memorable, says a little something about what it does instead of just communicating "orange S" like the last one.
I wonder how that delicate shadow holds up at small sizes and in monochrome.
The creator of Sublime Text tends to go silent for long periods between stretches of activity, which always makes me wonder about the future of the project. It's a shame, since ST is a well-thought-out editor with great attention to detail, and of course the speed argument (compared to Atom and VS Code), while beaten to death, still holds true.
Of course, it's up to the author himself to choose how he communicates, most likely the current method is working out just fine for him :). And the editor works very well, and the updates eventually return. Just saying that some occasional communication regarding the future of the project, or the current development model ("I spend half the year perpetually wasted in Ibiza, radio silence expected"), would be nice!
I've been very active on the forum, Discord, IRC and the Packages repo on GitHub. I don't think the stereotype of Sublime HQ being radio silent holds true any longer.
We don't, however, make blog posts or tweets saying "we're still working on the next build." However, whenever anyone asks on the forum (twitter, etc) I'll be happy to tell you we are still working on the next build! :-)
>I've been very active on the forum, Discord, IRC and the Packages repo on GitHub. I don't think the stereotype of Sublime HQ being radio silent holds true any longer.
Yeah, it's the blog posts/tweets stuff that might be needed.
Because most devs don't bother checking the forums (much less irc).
In fact, and this is 100% true, I just spoke to a guy in the next office asked him enthusiastically "did you get the new ST update?", and he looked at me quizzically: apparently he still uses ST2 (which has bought 2 years ago or so) and didn't even know there was a ST3 out at all.
The ST page _still_ says "Sublime Text 3 is currently in beta. ". This leads the cautious and uninitiated user to install ST2 instead.
I used ST2 for years after everyone else was using ST3, before I realized that ST3 was basically what everyone else is using -- which I mostly only realized as more and more packages were available only for ST3 and not ST2.
I have no idea why the author still calls ST3 beta, but I am curious.
Thanks for the reply! I'm sure you're right about the stereotype not holding true, I haven't really been following things closely.
I also understand that time spent convincing people all over the internet that you're still working on the next build may be better spent... well, working on the next build :)! What I had in mind was more along the lines of "what's happening to Sublime Text in the long term", though that may not even be relevant.
> We don't, however, make blog posts or tweets saying "we're still working on the next build."
Why not? I think it would definitely be in ST's best interest to be more engaged with the community. It's only natural for users to worry that ST will go the way of TM, and nipping those concerns in the bud seems to make eminent business sense.
Ignoring all the other rather skeptical replies, thanks for all the work you're doing! It's the most beautiful of all the editors I've used and works perfectly across all platforms.
I don't even really see the problem with the demand for new updates. If ST3 stopped now, it's still got an awesome set of plugins and it will remain an awesome editor.
I had one issue with a recent release, that was answered quickly via the Github and dealt with in a day or so. So I can verify that @wbond is true with what he says.
Total Commander released version 9 last dec, the first beta for version 9 was available a few months before in june. The previous version before that was 8.52 that was released in Sep 2015, so less them a year between the last two released. This hardly count as not updated for a long time.
Just saw that version 9 have automatic online check, its default is disabled, but when you check for updates the first time (in the help menu) it will offer to do it automatically.
I'm finding it hard to see where people are coming from with this complaint, though you're in good company so I assume there's something to it. But what features? What bugs?
Sublime seems to have a well-defined aim and to be resistant to feature creep. It's doing a good job of meeting that aim, so I'm happy with it only changing as needed to keep up with like, high DPI monitor support or other things that would actually make it unusable.
One reason for the 6 week update cadence that Firefox uses was to increase stability by reducing the pressure to include a feature in a release even if it's half baked. If not shipping at release time meant that the feature would be delayed for six months or a year while competing browsers pull ahead, there's a lot of pressure to just go ahead and release, even with known issues.
The approach, in my mind, has had mixed success, but has been a net positive overall. The biggest weakness is that it can be hard to maintain large architectural changes on a branch, which means there's still pressure to ship them in some form. (Though ideally with a configuration option to disable them by default.) Most things aren't like that, though, and for changes with a more controlled scope the six week cadence works very well.
Instead of writing blog posts as a vital sign, maybe it would help to offer some regularly updated pre-defined "bundles" of ST together with package sets for a certain audience (JS/TS/Python developer, etc.). Many packages have a higher update frequency, so even when there is no new ST build it would make sense to also update such bundles on a more regular basis (I know packages are updated automatically -- but this would give a better impression regarding the activity and the size of the ST community for new users).
I imagine the introduction of something like "Sublime Text distros" in package control, which would provide ST users something akin to the different Eclipse flavors (or the various .vimrc settings available online) could be quite useful. While the package control setup is already pretty straightforward, it is still rather confusing for new users that they have to manually install such a 'basic' feature (and Atom, for example, is much more polished in that regard).
> While the package control setup is already pretty straightforward, it is still rather confusing for new users that they have to manually install such a 'basic' feature
If that confuses you, why bother trying to code at all?
Yes, it's simple, but it's on almost every installation of Sublime Text, and moreover, package management is a baseline expectation in modern editors, so having it built-in makes a lot of sense.
There's a command to "install package control" with the newer versions of ST3 that doesn't require manually copying and pasting a script into the console. It's a big improvement and still allows users who don't want package control to not have it.
Some people just don't want to install packages and prefer to keep the editor as is. And i can't blame them, the only packages I install are my usual theme and Emmet.
That is exactly the reason why I haven't bought a license yet. I consider the license an investment in Sublime's future. Since I don't have to pay, buying a license is more of a donation in order to show support. But the communication seems a bit too sparse/erratic for me to have much trust in this project. The time spans between updates, blog posts, or even tweets are simply way too long.
I also consider the existence of http://docs.sublimetext.info a bad sign. The way I perceive this: At some point, an enthusiastic user was so fed up with the sub-par documentation that they decided to take matters into their own hands, and create a better documentation. But instead of working with the community, the Sublime project people/person just... did nothing? Was there any communication? I don't know. Why not provide the documentation on github and accept PRs? The Sublime project should push for docs.sublimetext.info to be merged into the official documentation.
Same with https://github.com/SublimeTextIssues/Core/issues. Why is the bug tracker "community run"? What does that even mean, considering the "community" can't fix any bugs? I absolutely cannot understand why the author of the software doesn't seem to participate at all in the bug tracker. Here's a quote from https://github.com/SublimeTextIssues/Core/issues/594 (from 2014, so I don't know whether it is still true)
> Unfortunately, the only thing I can say is it'll get fixed if Jon Skinner sees it as a big enough priority. Our intent in creating this unofficial Issues page was to have a curated place where known issues can be collected together and triaged, and he can use it as he sees fit to fix bugs in upcoming releases. Without a more direct feedback mechanism, this is the best we can do, at the moment.
What I'd also like to see is a road map for the future with regular (2-4 weeks) status updates.
> The time spans between updates, blog posts, or even tweets are simply way too long.
Is there a lot that you think ST still needs? As long as bugs are fixed, I'm happy with the functionality of the software today. The software feels done to me.
The community aspect of ST is a result of the product being somewhat unique in its positioning as commercial product, but also one that is 'free' for use as long as you can put up with the random nag window. The community aspect represents all the people who have never paid a cent for the product finding a different way to give back. Even paying for it costs a fraction of similar commercial software (although this is admittedly a harder argument to make).
Sounds like you want to reward PR over progress and development work. Writing PR and responding to the same questions over and over sucks time away from actual work.
If your point was that they don't respond to questions or are hard to reach, it would make more sense to me, but they seem very responsive.
The trend in software is definitely towards a lot of tooting one's own horn over substance. I don't think that's a good thing that we should be rewarding.
> I consider the license an investment in Sublime's future.
I'm exactly the opposite. I consider the license as compensation for work already done, I operate under the assumption that the vendor will have made enough to not care to work on it further.
Completely agree. What is there today is well worth the license fee even if there's never another update -- though this isn't one of my concerns as Sublime is so much more usable / lightweight / snappy vs Atom. [For it to be sunset, there would have to be a better alternative out there, and for the general case of multiple languages, I don't believe there is.]
Also packages are well separated from the Sublime releases so even in the case that there isn't another update, the ecosystem is still active. That said, I do wish the internal Python for plugins would receive a version bump though.
> The following pages contain the official documentation for Sublime Text 3. The Sublime Text Unofficial Documentation is an excellent supplementary resource, with a huge amount of information not covered in the official documentation.
Certainly true. My intent was to say that after a year or so of using Sublime the need to consult the unofficial docs was minimal. The unofficial docs are more like training wheels than a necessary reference in my experience.
I changed my mind and just bought a license. Bringing multi-cursor editing into the world is easily worth $70 (I think Sublime was the first?). And it's not just Sublime Users that benefited from this.
I just exported a text file that looked like this:
Col1 title|Col2 title|Col3 title|...
1||asdf|...
With multi-cursor editing it almost trivial to transform it into this:
Col1 title: 1
Col2 title:
Col3 title: asdf
...
That may be possible in VIM using macros, but would require way more fiddling.
I'm somewhat bummed that Limetext [1] seems to be dead. I like Sublime, but I don't want to rely on nonfree software for my work. I switched to vim for now.
It's not that we don't want to pay for software, and I don't think that's what OP meant when he used the word "nonfree". It's just that we need the four freedoms in case the software takes a sudden turn in a direction we don't like.
Now you're going to ask: How is it possible to support the four freedoms and still charge for the software?
And my answer is: I have no idea, and I don't care. It's not my job to come up with a business model for you, I'm just stating my requirements.
I don't think it's that big of a deal for a text editor, though.
One thing is if you're a designer and all your files are PSDs and you're locked in with Photoshop, but if Sublime Text dies you still have lots of other editors that will open your plain text files: Atom, Eclipse, etc.
You can still 'pay' for free open source software, many projects accept donations.
You can also donate your time by contributing to projects.
You can always help pay to keep an open source project running. The same thing is a lot more difficult with closed source software once its been abandoned.
Does Sublime let you print yet? (Hint: this is the part of the program where you tell me "why would you want to print! or don't print!" and I call you all granola-sucking hippie-daisy doofuses who don't understand the real world.)
If Sublime only had definite release! It's wonderful GUI editor! I stopped using it years ago when I switched to Vim and Emacs, but if it gets released I will buy it for sure, at least to have it on my work machine.
I like editing text in it, everything is so smooth and well looking, and it doesn't feel bloated. And I just hate all those Electron based editors, no matter how much incrementally they get better from month to month, they never even came close to Sublime's swiftness and polish. I tried VSCode last month, for few hours and uninstalled.
And people are often relaying on metrics and public opinion of usage of certain tool. Look, on latest StackOverflow Dev Survey Notepad++ was way ahead of everyone, it tells you something.
Release or not release is just a name. You shouldn't let that be a factor. In my experience it is absolutely stable and as bug free as software can be.
However, I'm not sure how well maintained it is. I fear all the time that the author, jps, has quit, but then every few months he has a development sprint and releases a new version.
I'd rather have it permanently called "beta" but a minor update every month, than one version declared finished, with rare or no updates afterwards.
That's true. Sublime 3 never felt like "beta" software. But as somebody said below, I am afraid of people quitting and leaving it to die slowly (like TextMate died) and that's the reason why I was hesitate to buy it for so long. Nobody guarantees if the developers drop the ball for some hypothetic reason thing will go open source, even that is not guaranteeing anything (again TextMate's case).
Textmate is still in development and has a good few releases a year. Last commit was a month ago but I think Allan likes to work a few months then have a few off.
If you like it, and you're afraid that it will be abandoned then don't hesitate to pay. I'm sure money is an incentive to keep working on it. I paid for S2 and S3 and I'm happy to have done so.
I am student, and not from rich enough country. We don't even have Starbucks. But I don't want it to be excuse. With this new update, Sublime feels fresh again, and I think it will be worth investing into.
That's true. I've been using Sublime for years and I've never had a bug or a crash. It pretty much perfect. And it's blazing fast.
I actually even use its fuzzy search to search my harddrive sometimes. Because I haven't found an easy app for Linux that works like Search Everything on Windows.
> Release or not release is just a name. You shouldn't let that be a factor.
I disagree. The name is the developer's way of telling you whether the software should be considered ready for general use. As long as it's called "beta", I'm not going to second guess the developer's judgement; I'll wait until the release version of ST3 before upgrading from ST2.
On the contrary, I still miss the speed of ST but the price-vs-feature ratio on VSC is too high for me to ignore anymore. No way the boss will shell out $70 per license if/when ST3 is ever released (now that half our team has dropped it anyway). Guess what you said about not relying on "public opinion" includes your own.
Well price-vs-feature ratio is debatable. The thing is that VSC has many IDE-like features, that I don't want in text editor. So everything comes down to a personal preference. And I don't know if somebody will finance piece of software for you, I mean they can, again depending on situation, but I don't turn off the possibility of me personally paying for piece of software that I like.
How so? VSC is approaching 1:1 feature parity with ST at a $70 price reduction. Some features you need extensions for in ST are included with VSC OOTB.
I am a big vscode user, have a couple of contributions too. But because of perf, sublime feels light weight. Opening up minified and large packaged js files larger than 100 MB, sublime shines.
I still have a licence and want to support the excellent work. There's a ton of features in VSC that were inspired from sublime. Minimap being one.
Well it depends what you define as editor features. Debugger isn't part of editor, at least for me, nor is git integration like VSC has. But speed of opening and editing big files, swiftness in searching and replacing even chars in source code and so satisfying scrolling and good minimap for code navigation are defining factors of text editor.
I don't want Sublime to be IDE, I have programable environment called Emacs for that. Integrating everything into editor is cool, to the point where you ask yourself "why am I doing all this?" and erase everything and just start writing code, not code for the editor. Again, this is my _personal_ experience, so take everything with grain of salt. I know your case is different, but I think you can see the reasoning behind my words.
Yikes, if $70 too high for developer productivity it seems someone isn't doing the right equation when you consider developer salaries cost in lost hours.
I feel the issue isn't the $70 which isn't a big deal (or shouldn't be) but more the point that other editors like VSCode are available now and have the backing of a big company like Microsoft. The biggest criticism of Sublime is that it often feels like it has been abandoned. It goes 6+ months without a peep from the developer. This can make people uneasy.
I prefer Sublime over VSCode, Atom, etc. but at the end of the day VSCode is free and still very good. Perhaps not quite as fast as Sublime but it also offers other features out of the box which Sublime doesn't so there is a trade off. Is Sublime "better" than VSCode to justify $70? That is something that can only be answered on a case by case basis I think.
I find this weird. I've been using sublime text for maybe 5 years and it has only crashed once. And it was fault of a plugin, not the editor itself.
Sublime Text has been feature complete (for my liking) for a lot of years now. The only thing I wish it had is an integrated debugger but it's not a deal breaker for me.
I've tried to jump ship to open source alternatives (atom and vscode) but they feel sluggish to me. Most people will be happy with their performance but I've been spoiled by Sublime Text and the slightest delay feels sluggish to me.
> No way the boss will shell out $70 per license if/when ST3 is ever released
That's crazy. My computer was waiting with a license for ST3 when I arrived at my first day of work at my job. Everyone's is, and if you prefer Atom or VSCode, they don't care if you switch.
> No way the boss will shell out $70 per license if/when ST3 is ever released
Yeah this is how it always goes. The 'almost as good but totally free' alternative almost always wins in the long run. Even when the cost isn't an issue it's just inconvenient to buy anything in most companies. Most people don't have expense accounts, company credit cards or simply don't want to deal with the hassle of paying out of pocket and getting reimbursed.
Notepad++ just works and is great. It's lean and has a common UI, has syntax highlighting and auto-complition. The Texteditor that comes with Gnome is similar. Same with Nano, it just works and has a good enough UI.
If you're using Vim / Emacs, people should consider giving an equal amount that you'd pay for a ST licence to Vim / Emacs instead. Even better, if you're a company then pay the equivalent of what Visual Studio costs.
It's already $70. I'm ok paying for my tools (I have a policy of no-piracy at my company, which is really weird in my country – We've bought IntelliJ, Affinity, etc), but paying $70 for a text editor is just above my threshold. I took out my credit card and went to the website thinking I had bought my first license at $50, but when I saw it asked for $70, for a closed-source license, I didn't complete the purchase.
I donate 1% of my revenue to open-source, by the way.
SublimeText no open source. If SublimeText open source, me agree pay $70 for my 2nd and 3rd employee, because open-source = greater good.
I reckon paying for open-source could be the other thing that doesn't make sense to you (although you didn't explain much). I reckon it's surprising, but it's entirely ok to charge for open-source and/or add an "Enter your activation key here" locker. You just have to know that your customers can remove the keylock and redistribute your software; but most businesses won't, because they're happy paying for someone's work, and they want support. The GPL even has a full page on this: https://www.gnu.org/philosophy/selling.en.html
Ligatures would be amazing and JPS has said previously that he'd love to implement it, but it's a huge amount of work at the moment when there are other pressing issues.
Why people cry about sublime not having a "release" yet. I have faced 0 bugs, seriously 0 and I'm using it for more than 3 years. Its way more performant than the rest "cool" editors and honestly...it just works! Maybe you need a little bit more configuration to make it perfect for you but thats just it.
On Mac, one bug that bites me about once per week is the editor becomes unresponsive. Not the whole application, I can still focus on windows, use the menus, and scroll the text, but the text editor just goes vegetable. The caret disappears, and I am unable to select text, type, paste, etc. Occasionally it returns after 20-30 seconds, but usually I have to restart Sublime.
It's still a great application in spite of this bug, but it is annoying when it happens.
Restarting is indeed quick, but sadly sublime text windows don't re-open on the same desktop they were on when I quit. They all pop up on whatever desktop is up when I re-launch, and I have to arrange 4-5 windows again. It's not a huge deal, but it is an inconvenience.
It's a testament to the quality of the app that the gripes are so minor.
Actually this dev build (well, the one released a couple of hours earlier than 3128) had a pretty major bug with gutter icons which broke text editing, scrolling etc.
Thankfully it was fixed very quickly in 3128, and my experience with ST dev and beta builds is much the same as you, that's it's been very stable in general.
I tried the new beta: it kinda breaks fonts in old themes (I use spacegray) but the new Adaptive theme looks pretty (I wish I could reduce tab bar height, though).
Also enjoyed switching UI and syntax themes via palette.
Overall looks great but for now I'll stick to ST3 stable release.
My #1 feature request is Sidebar API, you know, for version control integration.
OK, I read that and didn't understand the "registered" (should be "users that have a License").
I understood as "registered on the forum" and figured I could try it. It ended up erasing my non-dev sublime text :'(
I love Sublime. It is cross platform and has a great array of plugins from the community. I have used it for years. I know plenty of keyboard shortcuts and I am productive using it.
It never crashes on me, but maybe I am just not an edge case user of it.
Whenever version 3 is released for purchase, I do not anticipate having any problems getting my employer to pay for a license and I don't mind paying out of pocket if necessary.
Sublime Text is a great piece of software. My only gripe is that the license I've bought (which was kinda steep IMHO) doesn't carry over from ST2 to ST3.
For vim users, there's also the ActualVim[1] plugin, which uses neovim's headless mode for processing. I've always been somewhat disappointed by vintage mode / vintageous in ST, and while ActualVim still has some issues, it's definitely promising. (It even works on Windows now! Though I couldn't quite get my plugins running yet)
Shameless advertising: since the developer of Vintageous stopped maintaining it, @gerardroche and I have started a fork that we will review PRs for and occasionally work on. We've already merged a bunch of outstanding PRs, added some new features and fixed some bugs.
Awesome! Vintageous is nice, but there are outstanding bugs that are annoying and it has had that "it doesn't matter if you write a nice PR I'm going to ignore it" feel about it for some time.
240 comments
[ 2.6 ms ] story [ 228 ms ] threadWhy did they choose to base their editor on Electron, of all things, especially considering the existing results of Atom (horrible) is beyond me.
In addition 1M+ plugins, open-source, cross-platform etc will keep cool kids not just happy but they will also aggressively challenge whoever is using old but efficient software products.
Electron is essentially Node grafted to Chromium... There are a few things that could be done, but would depend on a lot of upstream cooperation to do so.
By giving performant lower level access to render, may be VSC can move some of the perf sensitive code directly to compiled optimized binaries.
Sort of like what node native libraries do.
Sublime does some pretty great dark magic. I was impressed to see vscode approach similar speeds to sublime in search. I believe there is plenty that can be done to make rendering fast, reduce memory consumption and make it a bit light weight. But it would require access to metal APIs
Also, as an Electron app, there is a node element, so compiled components are already an option.
Yet VSC is like a polished machine with so much weight behind, it's almost a wonder ST3 can stand up to it. That's why I'd like the 3.0 release to be official.
(Fwiw I used Textpad for over 15 years before finding ST. Compared to Eclipse etc that was another speed demon).
VSC is very much prepared for cross-platform .NET Core development. ST3 is more generally prepared for a lot of other languages and project types. If you're on Windows using a .NET language, Visual Studio should supersede VSC in almost every way though, and ST works better as your lightweight text editor.
And now, a moment of silence for those poor Notepad++ fans. :)
When I start doing coding and text manipulation again, it looks like it will be Sublime I purchase.
One of the default settings is to trim excess whitespaces.
But there's almost certainly a ST extension for your particular taste.
I usually have ST and vim open at the same time myself.
It's a bit unfair to compare a well-backed enterprise team to a small one (with 2-3 devs?), but it sure does make for an awesome user experience for developers.
Would be curious to know what "fans" are you talking about.
The slow is gonna take over the world? Ha.
But I'm happy for the update!
I think it's much better than the previous one, but I've always replaced the icon for that with this anyway: https://dribbble.com/shots/1836652-Final-Round-Sublime-Icon
I think I'll leave the new one in place for a while and see whether it grows on me.
I wonder how that delicate shadow holds up at small sizes and in monochrome.
Of course, it's up to the author himself to choose how he communicates, most likely the current method is working out just fine for him :). And the editor works very well, and the updates eventually return. Just saying that some occasional communication regarding the future of the project, or the current development model ("I spend half the year perpetually wasted in Ibiza, radio silence expected"), would be nice!
We don't, however, make blog posts or tweets saying "we're still working on the next build." However, whenever anyone asks on the forum (twitter, etc) I'll be happy to tell you we are still working on the next build! :-)
Yeah, it's the blog posts/tweets stuff that might be needed.
Because most devs don't bother checking the forums (much less irc).
In fact, and this is 100% true, I just spoke to a guy in the next office asked him enthusiastically "did you get the new ST update?", and he looked at me quizzically: apparently he still uses ST2 (which has bought 2 years ago or so) and didn't even know there was a ST3 out at all.
(which motivated me to post this thread on HN)
It is however "major" in features.
I used ST2 for years after everyone else was using ST3, before I realized that ST3 was basically what everyone else is using -- which I mostly only realized as more and more packages were available only for ST3 and not ST2.
I have no idea why the author still calls ST3 beta, but I am curious.
I also understand that time spent convincing people all over the internet that you're still working on the next build may be better spent... well, working on the next build :)! What I had in mind was more along the lines of "what's happening to Sublime Text in the long term", though that may not even be relevant.
Why not? I think it would definitely be in ST's best interest to be more engaged with the community. It's only natural for users to worry that ST will go the way of TM, and nipping those concerns in the bud seems to make eminent business sense.
I don't even really see the problem with the demand for new updates. If ST3 stopped now, it's still got an awesome set of plugins and it will remain an awesome editor.
I had one issue with a recent release, that was answered quickly via the Github and dealt with in a day or so. So I can verify that @wbond is true with what he says.
With ST we see slow but constant development, personally I don't care if the author is not a Twitter addict.
I think Chrome and Firefox has the right idea, with updates every 6 weeks.
Most commercial desktop software, and even web services like Gmail, go on for years on end without an update.
Browsers are always adding crap (like Firefox's nth cosmetic update), they should be stabler too. Once every 6 months or so at worst.
Yes, and I'm saying that's too long. I don't like waiting months or years for fixes or new features.
I also stated that this is my personal opinion. Not sure why people are downvoting.
Sublime seems to have a well-defined aim and to be resistant to feature creep. It's doing a good job of meeting that aim, so I'm happy with it only changing as needed to keep up with like, high DPI monitor support or other things that would actually make it unusable.
The approach, in my mind, has had mixed success, but has been a net positive overall. The biggest weakness is that it can be hard to maintain large architectural changes on a branch, which means there's still pressure to ship them in some form. (Though ideally with a configuration option to disable them by default.) Most things aren't like that, though, and for changes with a more controlled scope the six week cadence works very well.
I imagine the introduction of something like "Sublime Text distros" in package control, which would provide ST users something akin to the different Eclipse flavors (or the various .vimrc settings available online) could be quite useful. While the package control setup is already pretty straightforward, it is still rather confusing for new users that they have to manually install such a 'basic' feature (and Atom, for example, is much more polished in that regard).
If that confuses you, why bother trying to code at all?
I also consider the existence of http://docs.sublimetext.info a bad sign. The way I perceive this: At some point, an enthusiastic user was so fed up with the sub-par documentation that they decided to take matters into their own hands, and create a better documentation. But instead of working with the community, the Sublime project people/person just... did nothing? Was there any communication? I don't know. Why not provide the documentation on github and accept PRs? The Sublime project should push for docs.sublimetext.info to be merged into the official documentation.
Same with https://github.com/SublimeTextIssues/Core/issues. Why is the bug tracker "community run"? What does that even mean, considering the "community" can't fix any bugs? I absolutely cannot understand why the author of the software doesn't seem to participate at all in the bug tracker. Here's a quote from https://github.com/SublimeTextIssues/Core/issues/594 (from 2014, so I don't know whether it is still true) > Unfortunately, the only thing I can say is it'll get fixed if Jon Skinner sees it as a big enough priority. Our intent in creating this unofficial Issues page was to have a curated place where known issues can be collected together and triaged, and he can use it as he sees fit to fix bugs in upcoming releases. Without a more direct feedback mechanism, this is the best we can do, at the moment.
What I'd also like to see is a road map for the future with regular (2-4 weeks) status updates.
Is there a lot that you think ST still needs? As long as bugs are fixed, I'm happy with the functionality of the software today. The software feels done to me.
IMHO, Sublime Text should fight feature creep.
Sounds like you want to reward PR over progress and development work. Writing PR and responding to the same questions over and over sucks time away from actual work.
If your point was that they don't respond to questions or are hard to reach, it would make more sense to me, but they seem very responsive.
The trend in software is definitely towards a lot of tooting one's own horn over substance. I don't think that's a good thing that we should be rewarding.
I'm exactly the opposite. I consider the license as compensation for work already done, I operate under the assumption that the vendor will have made enough to not care to work on it further.
Also packages are well separated from the Sublime releases so even in the case that there isn't another update, the ecosystem is still active. That said, I do wish the internal Python for plugins would receive a version bump though.
I've already bought my ST2 and ST3 licences and honestly just looking for a way to pay more if it would ensure continuity.
There's an official docs site as well now that's actually pretty good. I haven't referenced the unofficial docs in some time.
https://www.sublimetext.com/docs/3/
> The following pages contain the official documentation for Sublime Text 3. The Sublime Text Unofficial Documentation is an excellent supplementary resource, with a huge amount of information not covered in the official documentation.
I just exported a text file that looked like this:
With multi-cursor editing it almost trivial to transform it into this: That may be possible in VIM using macros, but would require way more fiddling.[1] http://limetext.org/
Usually paying for something helps it not to die, like limetext, for example.
For my work, I'd rather pay for something I rely on to increase the likelihood of it staying alive and keeping up with the work I need to do.
It's amazing how people still don't want to pay for software.
Now you're going to ask: How is it possible to support the four freedoms and still charge for the software?
And my answer is: I have no idea, and I don't care. It's not my job to come up with a business model for you, I'm just stating my requirements.
I don't think it's that big of a deal for a text editor, though.
One thing is if you're a designer and all your files are PSDs and you're locked in with Photoshop, but if Sublime Text dies you still have lots of other editors that will open your plain text files: Atom, Eclipse, etc.
You can always help pay to keep an open source project running. The same thing is a lot more difficult with closed source software once its been abandoned.
1, 2, 3. The violation of Freedom 3 is somewhat ameliorated by the plugin feature.
I like editing text in it, everything is so smooth and well looking, and it doesn't feel bloated. And I just hate all those Electron based editors, no matter how much incrementally they get better from month to month, they never even came close to Sublime's swiftness and polish. I tried VSCode last month, for few hours and uninstalled.
And people are often relaying on metrics and public opinion of usage of certain tool. Look, on latest StackOverflow Dev Survey Notepad++ was way ahead of everyone, it tells you something.
However, I'm not sure how well maintained it is. I fear all the time that the author, jps, has quit, but then every few months he has a development sprint and releases a new version.
I'd rather have it permanently called "beta" but a minor update every month, than one version declared finished, with rare or no updates afterwards.
wbond is the author of Package Control for ST and I think he lurks here sometime.
Textmate is still in development and has a good few releases a year. Last commit was a month ago but I think Allan likes to work a few months then have a few off.
I actually even use its fuzzy search to search my harddrive sometimes. Because I haven't found an easy app for Linux that works like Search Everything on Windows.
https://github.com/albertlauncher/albert
In that vein though, see also Synapse: https://launchpad.net/synapse-project
I disagree. The name is the developer's way of telling you whether the software should be considered ready for general use. As long as it's called "beta", I'm not going to second guess the developer's judgement; I'll wait until the release version of ST3 before upgrading from ST2.
I still have a licence and want to support the excellent work. There's a ton of features in VSC that were inspired from sublime. Minimap being one.
Speed is a major feature.
And battery/memory efficiency is too.
And, on principle, I'd pay to NOT have to use an Electron app.
That said, I've used VSCode from time to time.
I don't want Sublime to be IDE, I have programable environment called Emacs for that. Integrating everything into editor is cool, to the point where you ask yourself "why am I doing all this?" and erase everything and just start writing code, not code for the editor. Again, this is my _personal_ experience, so take everything with grain of salt. I know your case is different, but I think you can see the reasoning behind my words.
I prefer Sublime over VSCode, Atom, etc. but at the end of the day VSCode is free and still very good. Perhaps not quite as fast as Sublime but it also offers other features out of the box which Sublime doesn't so there is a trade off. Is Sublime "better" than VSCode to justify $70? That is something that can only be answered on a case by case basis I think.
Sublime Text has been feature complete (for my liking) for a lot of years now. The only thing I wish it had is an integrated debugger but it's not a deal breaker for me.
I've tried to jump ship to open source alternatives (atom and vscode) but they feel sluggish to me. Most people will be happy with their performance but I've been spoiled by Sublime Text and the slightest delay feels sluggish to me.
There was a time when that was the case, but for quite a while now we've been seeing frequent releases.
The way I think about it is: How many hours of time does Sublime save me per year relative to another editor. Even just 1 hour makes it worth it.
That's crazy. My computer was waiting with a license for ST3 when I arrived at my first day of work at my job. Everyone's is, and if you prefer Atom or VSCode, they don't care if you switch.
Yeah this is how it always goes. The 'almost as good but totally free' alternative almost always wins in the long run. Even when the cost isn't an issue it's just inconvenient to buy anything in most companies. Most people don't have expense accounts, company credit cards or simply don't want to deal with the hassle of paying out of pocket and getting reimbursed.
So I have to use Kate for non-Latin text. But it still miles better than VSCode/Atom.
I use a US keyboard now so I had forgotten :)
Is there any way to donate to SublimeText? We already have licenses at work, but would like to donate $10 or so to show support.
I donate 1% of my revenue to open-source, by the way.
$70 too much.
$50 = ok price.
SublimeText no open source. If SublimeText open source, me agree pay $70 for my 2nd and 3rd employee, because open-source = greater good.
I reckon paying for open-source could be the other thing that doesn't make sense to you (although you didn't explain much). I reckon it's surprising, but it's entirely ok to charge for open-source and/or add an "Enter your activation key here" locker. You just have to know that your customers can remove the keylock and redistribute your software; but most businesses won't, because they're happy paying for someone's work, and they want support. The GPL even has a full page on this: https://www.gnu.org/philosophy/selling.en.html
So I might add:
$70 for open-source is ok.
$50 for closed source is ok.
edit: manual download is very slow. Maybe things are just a bit bogged down at the moment.
I hold out hope for it happening one day though!
It's still a great application in spite of this bug, but it is annoying when it happens.
It's a testament to the quality of the app that the gripes are so minor.
The only time I've ever experienced something like you're describing, it was due to some bad math in a 3rd party theme.
Thankfully it was fixed very quickly in 3128, and my experience with ST dev and beta builds is much the same as you, that's it's been very stable in general.
I tried the new beta: it kinda breaks fonts in old themes (I use spacegray) but the new Adaptive theme looks pretty (I wish I could reduce tab bar height, though).
Also enjoyed switching UI and syntax themes via palette.
Overall looks great but for now I'll stick to ST3 stable release.
My #1 feature request is Sidebar API, you know, for version control integration.
> Sublime Text 3 dev builds are currently available to registered users only.
https://www.sublimetext.com/3dev
Anyone knows what this means in practice? What rendering is used instead and does this mean Consolas/Courier have a disadvantage or an advantage?
It never crashes on me, but maybe I am just not an edge case user of it.
Whenever version 3 is released for purchase, I do not anticipate having any problems getting my employer to pay for a license and I don't mind paying out of pocket if necessary.
[1]: https://github.com/lunixbochs/ActualVim
https://github.com/NeoVintageous/NeoVintageous