I love the editor, it's still better than its competitors.
But in terms of management I think they're making two mistakes:
1) I bet almost everyone that uses Sublime Text for production work are using the "Beta" version (3) and not the release version (2). However, version 3 is suppose to be a paid upgrade from version 2. How long can you leave money on the table like this and still have a business, especially when the "beta" status probably doesn't have much meaning with your customer base? Maybe these guys have day jobs and this is just funny money anyway...
2) By remaining in beta so long and now that many people are using it... when they do pull the release trigger on version 3, that will be a very sudden expense for people that maybe have grown to depend on it without having paid for the upgrade. While everyone on version 3 should know that the day will come when that's a new license to pay for, if it rolls out as a standard update and doesn't obviously leave room to stay on the last beta version while people digest the upgrade fee... you might see a lot of people upgrade to the alternative editors instead. (OK, OK... that's a potential management mistake, not a concrete one as yet...)
1) Yes, well over 90% of Package Control users are running Sublime Text 3. We haven't brought it out of beta yet because there were still a number of rough edges and improvements we wanted to make before then. Jon mentioned in the last beta blog post (https://www.sublimetext.com/blog/articles/sublime-text-3-bui...) that this dev cycle is leading up to the 3.0 launch. And no, we don't have day jobs. Sublime Text is my day job. :-)
2) In terms of the expense, the only people who will be required to pay an upgrade fee are users who purchased a license before the Sublime Text 3 beta was announced. So effectively users who have been using ST2 (or ST3) for more than four years will be paying. Everyone who has purchased since the ST3 beta announcement have an ST3 license already (http://www.sublimetext.com/blog/articles/upgrades).
We haven't brought it out of beta yet because there were still a number of rough edges and improvements we wanted to make before then.
Let me compare this to how I decide if something should be released: Is it stable (ie: it doesnt crash frequently)? Did I add some features so I can tell users its better than the last version? If you answered yes to those questions, then fix any major bugs and ship the thing as stable.
You would have been on 3.0 stable a long time ago.
I also would package each of these fairly minor releases as 3.x.. so this release would probably be 3.25 or something.
Why? You mark it as stable so new users (who aren't aware of your process) know they can use this (and you can close the sale). You want people to feel like the project is active, and they're getting a good deal for the license.. which feels like a better deal to you?
- I bought ST2 and will get ST3
- I bought ST 3.0, and received over 30 releases from 3.0 to 3.30.
Yes, well over 90% of Package Control users are running Sublime Text 3.
This is a warning sign. You were so conservative with marking it stable, that the users have made the decision for you. The beta is now your product, and if you break it, that will be your reputation. You no longer have a real beta.
I would agree, but I would also say that they should err on the side of caution. Don't see all of these posts and just call the next release 3.0. Do an absolute thorough round of testing. Be sure that it is stable. Then bring it out of "beta".
As bad as breaking the beta would be, releasing a product after such a long beta period that isn't completely stable would be worse.
All that said, Sublime is easily the most stable piece of software I use on a daily basis. I prefer JetBrains IDEs for daily coding, but still do plenty of work in ST3 and have customized the heck out of it to fit my needs almost perfectly. I'd be very much surprised if it wasn't ready to exit beta today
ST seems to call the entire period through active development "beta".. and then when they've finished developing the entire thing and hit what would normally be the extended support period (bug fixes only), they call it stable.
It's certainly costing them money. And reducing the value of licenses (since its "stable" for a shorter period of time)
If I were them, I would wait a couple of weeks (so that they get the entire benefit from their announcement today of this new build), and then re-release their software as stable -- even if no changes were made to it during that period.
This thing has been in beta since 2013. It's been tested enough. It's ready.
(And to ST: if you want free advice on your release process, you're more than welcome to email me. Or at least talk/consult with someone with product management experience.. your software is great, but this release process needs to be fixed.)
I paid for a license already, and I'm not sure where I fall in the timeline of being eligible for a free upgrade, or not. That said, I'm happy to pay for a license for the upgrade.
As a longtime customer, I wish there was more communication about the state of things, and plans for features, things like that.
Basically, more communication, and less radio silence.
ST3 is not beta anymore. You just call it that but as you said over 90% of user uses ST3 on daily basis. I highly doubt that all of those 90% are beta testers. I would even risk saying that all of those 90% users are using ST3 in production and are relying on its stability. It's Gmail Beta story all over again http://gizmodo.com/5309225/gmail-finally-comes-out-of-beta
Thanks for your response and hard work. I want you guys to make more money to have incentive to make Sublime Text better/faster since that helps me to be better/faster in my work! I will be in the paid upgrade class. I'm happy to do it, and the price is trivial compared to the return on that investment.
I'll leave you with this... My wife worked in Hollywood for a long time and we knew some wanna-be script writers... some of them wouldn't ever try to get their scripts sold or read because... for years on end... the scripts "weren't ready yet"... those guys usually just ended up doing other work and never selling a script since they couldn't live without cash flow. Don't be those guys :-)
I bought mine pre beta 3, and you know what, I'll be happy to send more money your way when 3 is released considering this is my main code punching tool. Thank you.
> We haven't brought it out of beta yet because there were still a number of rough edges and improvements we wanted to make before then.
Your standards of quality are admirable! ST3 is the most immaculate text editor I've ever used (never used ST2). If ST3 is beta, then in comparison Atom or even Vim are still in alpha.
Good call! That's changed then since I last looked at the license (probably about four years ago or so when I purchased version 2 and version 3 was going to be the paid upgrade).
So they dodged my point two. I hope they don't amplify point 1 with that move.
[Update] Actually, this is not fully correct in regard to my point. I re-read your link and this is the key bits.
"Until such time, upgrading is not required, as Sublime Text 3 will accept Sublime Text 2 license keys during the beta period. [...] For customers who purchased in the 90 days prior to the announcement of Sublime Text 3 [Jan 2013], we are reducing the cost of upgrading from $15 down to $11. Customers who purchased Sublime Text 2 before this time period are still subject to a $30 upgrade fee when 3.0 is released."
So version 2 subscribers are not automatically grandfathered into version 3; there is a class of version 2 users that are, but not all. Apparently there is a time at which they considered licensing to be version 3. http://www.sublimetext.com/blog/articles/upgrades.
So for example, I bought a version 2 license in 2012, have been using version 3 since 2013... an automatic update to a full release version would force the update issue.
Using web.archive.org, it looks like licenses have counted as version 3 licenses since April of 2013[1]. So I guess licenses purchased before that are for version <=2 like you said. At least it isn't as bad as you'd thought.
I used to publish my own shareware. So maybe I see these things differently:
Today's customers are paying for continued development, for future releases. If a product is dead, there's near zero chance I'd pay for it, or use it (willingly).
If this dev is happy, I'm happy. When I was in the game, I was thrilled to have steady income, enough to stay solo and continue to do what I loved.
(I bought a personal license and make each of my employers buy me a work license. I love SublimeText and am happy to pay.)
There was a time I couldn't live without Sublime text but with both Atom and Visual Studio Code I have no longer any use for it.
I understand people who need/want it because it's snappier and handles large files great. I am however not really one of those people. But I am happy to see development on it.
For me, ST is just so much faster that it's worth it.
Plus, the Latexing plugin has mendeley support and really, that's _so good_ that I am willing to pay for it and stay and be happy. You have not known latex joy until you have created a bibfile using intellisense-like reference searching and nothing else..
Latex Workshop for vscode [1] is catching up fast. It doesn't have Mendeley api support yet, but if you use the desktop client for syncing, it works perfectly and is under very active development.
Longtime Sublime user here as well, who switched to vim a couple of years ago. I still use Sublime occasionally for text manipulations though - it's support for transforms with multiple cursors is second to none and I absolutely love how snappy it is compared to other desktop text editors.
>I understand people who need/want it because it's snappier and handles large files great. I am however not really one of those people.
If ST3 had the same features (basically, more support for friendlier plugin integration) it would totally devastate those, apart from in price (which should be irrelevant for professional developers anyway).
But even as it is now, it's way handier (due to the snappiness) for everyday coding than VS Code (Atom is so slow it doesn't even qualify).
I read about atom's github integration that they released last week and I've been playing with Atom at work this week.
How big is your project that you find atom slow? I have gone back and forth between atom and sublime 10 times in the last 10 minutes and I have not found a single reason why anyone would notice a lag in a project of just a shade over 10,000 files. The difference in speed is completely imperceptible.
Not saying I love atom more than any other editor (I typically use terminal vim) but in my experience over the last week, any complaints about atom's sluggishness seems either outdated or wrong to the point of absurdity.
I haven't given Atom a good chance since the beginning of the year, but startup was noticeably slower for me. I often use it to read/modify extremely large JSON files (tens of thousands of lines) and Atom struggled to render that. I also frequently have dozens of tabs open and Atom seemed to slow down with that many open. I've had 50+ open in Sublime before with no issue.
All that said, Sublime is not my daily driver. I use JetBrains products for that. I mostly use it to jot down notes that will eventually be transferred somewhere more permanent, temp access tokens for APIs, small snippets, JSON files, Mardown files, etc. If my main use with ST3 wasn't frequently opening and closing it, relying on its great performance when opened with my actual IDE(s), and abusing its second-to-none ability to keep files saved without actually saving them to disk, I may feel differently.
Different guy here, but I don't thing project size has much to do with it. For me it's UI for the most part. Most of the time I can see, how the moment text goes from "simple text" to the "highlighed text" (and all of the formatting too). I simply can't stand this. Same happens when searching a large file (I do work with logs pretty often). But this problem is solved in Atom for the most part - you still can't open files larger than 10Mb...
And as a separate issue - the latest update made it impossible to update Packages for some reason. All I'm getting is "undefined" error.
I'd say Visual Studio Code is not only not an IDE, it's also lighter weight than Atom, though obviously both are Electron. VS Code lacks even basic project management, having less tooling than even Sublime for project layout/folder structure - there's no equivalent for .sublime-project or .sublime-workspace.
For the most part, the Electron UIs use a ton more RAM and are much laggier. They do not tend to handle large files well. However, in exchange for that, they offer rich features. Plugins are way better - since obviously they have HTML/CSS at their disposal.
Also, VS Code's community is very active - same for Atom - and it leads to quite high quality plugins. The plugin for Go, for example, is quite great. I've been hoping for first-class Go support in Sublime just so I'd have a direct comparison, but in comparison to the VS Code plugins for Go, Sublime feels outright clunky and hard to configure.
I'd say give it a shot. If you can get past the fact that it's undoutably slower, and some of the semantics are different, it's really nice having full code intelligence and rich plugins.
Also the feedback from the Microsoft team (they genuinely seem to put effort into VSCode), the frequent updates and new features. I love(d) ST and it was always one of the first tools I installed after a fresh OS install. My latest install does not include ST, only VSCode (Windows and Linux). I do not notice the speed in the type of work I do.
Still glad that there's an alternative available to VSCode also. :)
I'll list some pros for each which I have experienced on top of my brain. Note that for Sublime I mostly wrote in php, while the others javascript or typescript has been the main target.
Sublime Text:
* Faster, performs better which is especially visible on budget computers or when handling larger files. Sublime is really, really good in handling large files while Atom or VS Code really stinks when it comes to that.
* A lot of plugins because it was "first" of these new, great text editors. Many plugins varies in quality but they exist.
* Very easy to configure and make the editor a perfect fit for you. Atom and VS Code has the same configurability, but they simply copied Sublime on this part.
Atom:
* A LOT of plugins and extensions. Seriously. It feels like there is an extension for everything already.
* Snappy enough, easy to search and find stuff in your code.
* Great, simply great UI and font rendering. With some theming this editor is by far the most beautiful one.
VS Code:
* Snappier than Atom, but with deeper features in the box. Got stuff like git tools, intellisense and other goodies that is just missing in Atom.
* Insane support for Typescript (and perhaps other languages as well). There is simply no other editor you want to develop Typescript in. Also good for compiling .NET Core (these are the ones I've tried).
* A lot of great quality extensions from Microsoft and others. While Atom might have more extensions, I find those for VS Code often to be more quality ones and I have yet to uninstall a extension because it doesn't work or is buggy.
Personally, I prefer VS Code because it feels more snappy than Atom, has intellisense and good support for compilers like .NET and Typescript.
For me the biggest advantage of Visual Studio Code is support for Language Server Protocol [0] that makes the editor a lightweight IDE. If I'm using statically typed languages the level of support from the editor easily outweighs longer startup time.
Sublime is better when working with structureless files (plain text logs etc.).
You got it backwards: VSCode is like Sublime or Vim with plugins, not an IDE. Don't let the name fool you, it's not a version of Visual Studio (the IDE), but a new editor project.
Atom, while also not an IDE, is way heavier, to the point of being slower than actual full fledged IDEs.
I like VSCode a lot but a few plugins for Sublime swayed me back. Plus, it's not a Microsoft product - I really don't like supporting them, I hate their approach to privacy etc and I'm "voting with my feet".
I had started with Atom, then VSCode and moved to Sublime. With a little Vim here and there as well. VSCode is great and has a lot of great features. Just something about Sublime gets me. I tried Atom again recently and man it doesn't even compare to Sublime or VSCode. No longer in the same ballpark in my opinion.
Sublime is perfect example of extremely well (almost perfectly, for me personally it is) executed native desktop multi-platform app! I payed even if I use it today mostly for reading the code, since Vim and Emacs took over writing code.
2. See definition of a function - copy name of function
3. Open new folder -> say 2 folders above the current file
4. Right click -> Find in Folder, paste name of function, enter
5. Get an overview with:
- a list of files where the function name is found
- for each line the surrounding code is also shown
- double clicking on a filename opens said file in a new tab
I have yet to find any vim/emacs tool that can do this so easily and effortlessly. Even with Emacs Projectile, you have to define a "project"? And then grep in that (but the user interface is just subpar).
Maybe I'm missing something, but for me, this is effectively stopping me using Vim/Emacs most of the time. I use Vim for editing individual files, but anything more than that, it is either IDEA (ok, I do Java dev a lot, so that's a must), or any Commander clone (which has a Find functionality).
>I have yet to find any vim/emacs tool that can do this so easily and effortlessly.
You want cscope integration. Nothing compares to this in terms of ease of navigation - and I say that as a dedicated ST3 user. How to integrate cscope with vim:
So I don't want to tell you to change your workflow. Like most things that evolve over time, they are the way they are for good reason. No one way is better than another. That said, here is how I do what you described above in Vim.
There are a few components: 1) Some fancy keybindings, 2) A plugin, and 3) an optional wrapper script for launching commands inside a git project.
First, keybindings. You know about '#' and '*' in vim that search for the next occurrence of token under your carat? I have an additional binding that does the same, but doesnt move the carat. I bind it to <CR> (Enter), but you could bind it to whatever. If you bind it to <CR>, it will interfere with the minibuffer, which requires some shenanigans to work around (see the vimrc)
This is nice for a few use cases. One, it is nice to see where a token is used in the current file, and two, it puts the highlighted word into a register, that you can later paste.
I use mhinz/vim-grepper to execute searches, with my vimrc I can move the carat to the word I am interested in, press "enter", then type :Gr<TAB><ENTER> which will put me into grepper. Then I can press ^R/ (that is ctrl-r then /) which pastes the content of the '/' register. Grepper will then launch the selected search tool (ag, grep, git grep) and create a mini buffer of matches that you can click on, or highlight and press enter. Here is my vimrc if you want to give it a go. BE WARNED the first line of the vimrc calls git clone, if this is bad on your setup for some reason, dont use my vimrc.
The last piece is executing the search from the correct directory, to do that, I use two methods. First, vim has a built in way to change its working directory (:chdir), so you could do a :chdir ../../. Second, if you are using git it is nice to run commands at the top level of the git directory sometimes. For that I have a script that translates commands from the CWD to the git top level. Some things you can do with it:
Rebuild cmake style project:
g make -C build && g ./build/my_program # you can execute this from anywhere inside the repo
Open a file inside vim:
cd src/my/nested/folder/
g vim my_source_code.cpp # open vim at git top level directory, and translates the file to src/my/nested/folder/my_source_code.cpp
Thanks for the efforts for describing this workflow, but please note my point above about being "effortless" ;)
To be a bit more specific: I need to see the structure of what I'm searching in, so at minimum, that means neotree/nerdtree. I don't have photographic memory, or alphabetic memory, I always forget the names of folders, files, everything. I just couldn't care less. "../.." means nothing to me. What is that directory? Show me, and I will point to what I want.
Also, I don't want to search in a project. I want to search in a project OR in multiple projects OR in a subdirectory within a project, ad hoc. OR just select some files and search in them.
And in my opinion, for this specific use-case there is nothing wrong with using the mouse. When I'm searching for something in the way I described above I am usually browsing the code, reading it, trying to understand what is going on. This is different to when I'm actually coding something, then vi/emacs suddenly gets much more attractive, depending on the language used :)
This is why I prefaced my response :) every one places different importance on different things. Glad you have something that works for you. Vim does show the age of its design from time to time. The worst has got to be Java, where you end up having to run eclipse in the background to make it work. Talk about a hack...
Well I use Vim more for navigation and skimming around, or reading when I know where I am going. But for opening for example Redis source just wanting to read the source code and learn, I still like Sublime the best for that, it is so smooth and nice looking. Reading source code in Sublime really is a pleasure.
Emacs on the other hand provide very nice integration with various REPLs and functional languages have best tooling in it. Vim I use as terminal editor, and when I do work inside repo, reading diffs and programming in some languages (like C for example).
I really want to switch to only one editor, but whenever I try, I come to the point where I miss the feature of some other one of these 3. : ) I'm bit too weird when it comes to editors and configs...
IMHO: Sublime was once leading the way as a light weight, easily customizable editor. Since then, a slow release cycle and various compelling alternatives like Atom and VS Code, has led many users away. So in this thread you see people not discussing this release so much as the present day dynamic of Sublime and its place among these (arguably faster, better, more customizable) editors.
People are happy Sublime is still alive. It's still one of the best editors out there, and with recent updates it's been getting a lot better at a pretty fast pace.
I don't know if you've been following the sublime situation or not but...
it's not the contents of the release that's the story here, but more the fact there is a release after a long period of Inactivity from the dev side of it. And also the fact it's still in Beta
I think not necessarily in better place. The idea of 1000s of Javascript/electron developers hacking on a text editor, does not seem very useful to me. Atom/VScode already exists and extremely popular and rapidly released.
Some editors which are released slowly and liked by few is ok. I am using Sublime 3 in beta for very long time. Sometimes it gives popup to upgrade which I do then. Else it is fine, doing the job at hand without latest and greatest feature which I am not using.
I love sublime text and bought the ST3 license a couple years ago. For the last couple of months I've been holding off using VSCode as I felt the same way that it was an unfair comparison between the work of one or few developers vs a big team at Microsoft. But just last week I finally caved in and started using VSCode on a regular basis and I seem to like it a lot more than ST3 despite the issues about large files. I am just worried though if more people switch over and ST development stops because of funding and if Microsoft abandons VSCode development, then what are we left with? At least VSCode is on GitHub so the community might take over but I feel bad for indie developers who have to compete with large teams from big companies that have no direct revenue from the product.
If you look closely you can see that the VScode guys took a very, very good look at Sublime. They use a .json config with kind-of the same style, they have per folder settings, they've implemented multiple cursors and indent guides. I switched over a month or two back and for me VScode feels like a superset of Sublime, minus some of the snappiness. Hell, if you look at VScode GitHub issues, there's a fair amount that boil down to 'please implement this Sublime feature'.
Glad to hear that sublime is still alive as I find it very intuitive (I grew up using Windows shortcuts). The thing that makes me still search for alternatives is the closed source condition. If I can find another editor with similar speed and capabilities (but software libre) I'd switch without thinking. Atom it's great but as a front end developer dealing with large code bases in a text editor that can't handle large files, fast search, and keep many files open it's a no-go.
I have donated to organizations such as Wikipedia, FSF, EFF, etc. so in my case software libre it's not about saving money. I acknowledge that making money with a software libre project it's still a challenge.
I use UltraEdit and find it much better than ST. I have genuinely asked to users of both what all the fuzz about ST is, in comparison to UE, but have never gotten any response.
I personally also think that the VSCode vim plugin is even better than the atom one, and it's currently in a very active state of development. We recently added neovim integration for Ex-commands, which means that practically every Ex-command is now supported.
The guilt of never paying for Sublime Text finally overtook my conscience so I ended up switching to Atom. Atom would chug memory on Windows 10 so now I curiously find myself using Caret[1] which is much lighter on resources despite using the same web technologies in the background.
After years of using Sublime without licence I've looked at my paycheck and I realised that probably 80% of it was earned while being in Sublime Text. I don't understand why developers earning tens (sometimes hundreds) of thousands of dollars refuse to pay $70 for life time licence for an application that helps them earn those money.
I paid for ST a while ago and love it. But I'm finding more and more that the specialized IDEs make me so much more productive.
For example, I just started using Gogland and it is much better than ST with whatever plugins I could find. And it's unfortunate because I really like the convenience of using ST for all my different languages I use at work, but I'm just so much more productive with Gogland that I'm going to make a hard switch.
Even with IntelliJ and IdeaVim, if you ever need Sublime Text, you can easily switch to it temporarily using IntelliJ's External Editors feature. This will let you edit the file neatly in the external editor and return easily to the IntelliJ window.
Go to Preferences > Tools > External Editors
Add a new entry called Sublime Text
Turn off ‘Open Console’
Set program to /Applications/Sublime Text.app/Contents/SharedSupport/bin/subl
The only major flaw this editor has is that sometimes by mistake I enter only one character (or even just a space) into the 'Find in Files' input and run search. And it just grinds to a halt, so I have to force quit or wait 10 minutes for the search to end. All the while the UI is unresponsive, beachballing. Still my go-to editor.
I really want to like ST but its search/replace UI has always been completely counter intuitive to me.
I'm used to a lot of different ways of doing search/replace (IDEA, emacs, vim to name a few) but somehow, my brain doesn't seem to be able to become fluent with ST's approach.
I almost didn't buy an ST2 license because ST3 was already in beta and while I was happy ST3 beta was available to ST2 licensees I was worried I might have to re-buy fairly soon after.
Looking back 4 years later I'm certainly glad I didn't hold off until St3 was officially out.
But in terms of management I think they're making two mistakes:
1) I bet almost everyone that uses Sublime Text for production work are using the "Beta" version (3) and not the release version (2). However, version 3 is suppose to be a paid upgrade from version 2. How long can you leave money on the table like this and still have a business, especially when the "beta" status probably doesn't have much meaning with your customer base? Maybe these guys have day jobs and this is just funny money anyway...
2) By remaining in beta so long and now that many people are using it... when they do pull the release trigger on version 3, that will be a very sudden expense for people that maybe have grown to depend on it without having paid for the upgrade. While everyone on version 3 should know that the day will come when that's a new license to pay for, if it rolls out as a standard update and doesn't obviously leave room to stay on the last beta version while people digest the upgrade fee... you might see a lot of people upgrade to the alternative editors instead. (OK, OK... that's a potential management mistake, not a concrete one as yet...)