Intype has new web!

Intype has it's own website now. You can find it at intype.info. You will soon find there more info about our progress, planned features... and about us.

Alpha release is comming closer every day, so update your RSS readers and stay tuned. See you at intype.info.

What's new?

We are currently (more than ever before) focusing on programming, so I was too busy to post some updates. Juraj is 400km far from me, so instant messaging and remote source code repository are our best friends now :) We are going to spend October at home, suspended from any of our job duties. We will be working fulltime whole month here in Prague on Intype. This will definitely be the greatest step in development.

SonyaEditor is doing very well and we are starting to be proud of it. Editor now fully supports Unicode, and is flawlessly rendering Hebrew, Thai or whatever you can imagine in Unicode. We are preparing some internal tests of rendering engine, encoding detection and encoding conversions. I'll let you know how it went. We also have new and slick text-run based styling engine, that will be soon joined with parser. In the background, the 'web team' is silently working on Intype website, that will be launched in first days of October.

There is also one big change to Intype. We decided to leave Lua, and replace it with JavaScript for scripting, that is more common for web-developers, and its implementation supports Unicode, so it saves us from a lot of trouble in the future.

C++ guru in da house

During this weekend, I've settled Intype brainstorming session with friend of mine, professional C++ programmer, Juraj Durech. We did some changes in Intype core and the result is 3 times faster parser (3,3 MB in 22.5 seconds and counting), new data model for document styling and AST, fast word-wrapping algorithm and much more.

Alpha release date term (middle of September) is too short for us to finish what we want to. None the less, we are working hard every day to release it as soon as possible. Official Intype website will be launched during September.

In-logo-type

Yesterday, I needed to relax, so I've designed logotype for Intype, that is to be final. What do you think about it?

Planned features for alpha

Yesterday, I've finished feature list for first alpha version.These are core features, that we will try to test and fix:

1. Editor with basic functionality (in-progress)
2. Syntax-highlighting (todo)
3. Scopes and scope-selectors (todo)
4. Parser (ready)
5. Snippets (ready)
6. Basic file operations (ready)
7. Unicode support, including UTF-8 without BOM (ready)

Intype Team and SonyaEditor

In spare time, I just want to focus on Intype development, but there are many other tasks to do before I release alpha. Therefore I've configured :) an Intype Team with my close friends to share the tasks of preparing web-site, maintaining it's content and to plan some marketing strategies. There are 3 other people: Tomas Lopasovsky (Marketing Specialist), Ivan Centes (Web Developer) and finally beloved Lenka Mazancova (Content Maintainer). We will introduce ourselves on new Intype's website.

So yesterday, we started work on support web for Intype with blogging system, wiki and forums. I decided to use PHP platform, althought my favorite one is Ruby on Rails :) and to use existing solutions as many as possible.

WordPress is choice for blogging system now, but we are considering also Swell Blog. For Wiki we are considering DokuWiki or WikkaWiki. And for forums, there was no other choice than beautiful Vanilla.

Now, topic number two: SonyaEditor. "SonyaEditor" is name for my own editing component, that is going to fully replace Scintilla editing component. I decided to finish it before first alpha due to need of much stronger document model for faster and more reliable parsing (coloring with Scintilla was very slow due to it's own styling invalidation mechanism). I've started yesterday, and today I have working document model with undo/redo actions. I hope that this decision will not postpone the alpha release date.

Simple parser prototypes benchmarking

Yesterday, I've tested speed of parser prototype. I've used full HTML grammar with 33,2 kb HTML file (with 700 lines of code). First parser prototype made it in 7.548 seconds, what is very poor result. After few consultations and modifications, I've used little different algorithm and the result is 0.618 second. Hooray! So good news are:

• parser is 100% compatible with TextMate grammars
• parser is now quick enough for alpha
• parser is running in separate thread, so you don't have to wait until it finishes its job.