I’ve been using Cygwin at work (at CrossLoop) basically since I started, and as far as just getting around the filesystem it works great. Adding a good number of utilities, gcc/gnu products, all that works fine, but any time you have to make changes after the fact it’s a little nightmarish, as anyone who’s been actually reading what I have to say on Twitter could tell you. Not that there are any folks matching that description. Basically, though, symlinks can get hosed, and you frequently have to perform a “rebase all” before a number of things work correctly. Not only that, but you have to perform it from a crippled shell and sometimes it doesn’t even want to run from there. Some tools and tricks I have taken to using in Cygwin are:

• Mintty – a replacement terminal with middle mouse button pasting, select-to-copy, and a few other improvements over the vanilla cmd.exe terminal.
• Charade – a bridge utility that fakes out all the ssh-* utilities (like ssh-add, etc) into thinking ssh-agent is running, when in fact it talks to an instance of Pageant, the Putty key agent, so you can use keys inside and out of Cygwin with ease. The only trick here is sometimes Pageant goes out of sync with Cygwin, so you have to kill it and start it with cygstart, a utility that gives ownership of processes to Cygwin.
• Symlinks – lots of em. Link all your frequently-accessed folders under ~, because having to jump through /cygdrive/c/ is just annoying.
• I’d like to say I’ve gotten z working, but I still haven’t. I’ve got python installed, I get no errors, but it doesn’t jump to directories like I’d expect. The author, rupa, has been kindly helping me jump through hoops to figure what’s wrong, but to no avail as yet.

I’ve also ditched E as an editor on Windows in favor of Gvim, so I’m now using vim across the board on Mac/Win/Linux. (more on Linux in a moment) On Mac I’m using the lovely MacVim, whose author has kindly implemented a few fixes since I whined about not being able to add a simple set fu in .gvimrc to enter fullscreen at startup. (previously you had to set up an autocommand). Yesterday morning I finally adopted pathogen on my Mac and committed my setup to github (it’s at http://github.com/thynctank/thyncdotvim) but am currently having issues getting it working under Windows, which is no surprise given all the hell I generally have to go through to get anything working on Windows. I’ve installed a copy of AutoHotKey, but I have yet to do much productive with it beyond getting a script (slightly out of date, will provide a copy of my tweaked version if anyone wants it) to convert my TextExpander abbreviations converted for AHK. Pretty cool utility, though a little clunky in the syntax department. Another great Windows until is VirtuaWin, a virtual desktop/workspace program that gives you a completely separate taskbar per workspace, and lets you assign apps to load into specific workspaces, very similar to Spaces under Mac. (I’ve been having issues with Spaces lately, too… )

I’ve gotten a little Asus EEEPC netbook, which came preinstalled with a trimmed-down version of Windows 7. Well, it works, but it’s pretty damned slow. I didn’t want to completely trash it, as Windows does have its uses, like testing and running Win-only software. Instead of trashing it, I installed a copy of EasyPeasy Linux beside it, which I highly recommend. The install was silky smooth, I didn’t even have to quit running software in the livedisc-like environment it loaded with off my Pre, which I loaded the ISO onto for installation. (remember, netbooks have no optical drive) EasyPeasy has a lot of great general purpose software preinstalled. The only things missing were developer tools and those were pretty easy to get rolling, for the most part. I’m still a little confused as to where Apache root is, but I haven’t had to fool with that much yet. Under EP I’m running Chromium as my main browser (with a fully-loaded Firefox for good measure), and a copy of Turpial as a Twitter client. It works pretty well, it’s definitely the nicest simple interface I’ve seen for a Linux client. Most software is easy enough to get installed, just find it in Synaptic package manager or add the repo to Synaptic if it’s not already there. A few items, like Chromium, Dropbox (yes! It works great!) and maybe one or two other apps, I had to download as .deb files and let the system install for me. I’ve even got the Windows Kindle app running under the latest WINE beta, though it’s a little uggo. I’m also running AutoKey, a simple AutoHotKey-like utility that works well for the most part. It’s only crashed a little bit. Mostly I’ve just used the netbook for reading and browsing the web, but I do plan on doing some work on it. (I bought it because I can’t do any dev work on the iPad, after all…)

I’ve been reading The Definitive Guide to Dojo ebook, which is a couple years old now, but still a good intro to the framework. Other than the crazy number of typos, the book is great. Dojo’s a lot more robust than I had realized before. I saw Pete Higgins say that Base alone is useless the other day, but I’ve gotta disagree. Being able to dojo.connect to any event or arbitrary function is awesome, and built-in pub-sub and deferred objects is real cool, not to mention the module loading and build tools, which is what attracted me in the first place. I’m reading about dijits right now. I plan on using several parts of Dojo (not the least of which is dojo.gfx) for a mindmapping app I’m building called MindSnap.

I’ve also been reading Learning Vi and Vim ebook, also by O’Reilly, which is an awesomely helpful tome. Between David’s pestering and this book, I was convinced vim was worth the trouble of the initial learning curve. There’s just so much power packed in modal editing, and so many useful commands available even out of the box. I’m using a few plugins, including nerdtree/nerdcommenter, command-T (haven’t gotten this working on Windows yet, go figure) javascriptlint (dumps errors into quickfix window, is more configurable than jslint) vcscommand and fugitive. Pathogen makes all of this very easy to maintain, or that’s the theory! Hopefully I get it working on Windows in the next couple days… I’ve also had a number of issues getting Gvim on Windows to respect the default gvim commands, such as Ctrl-V for entering visual block mode. I think it may be working now, but I’m not sure what I changed.

On Mac I’ve been using a utility called Optimal Layout for the past couple months for moving/resizing windows via keyboard. I like it a lot, especially since the author added custom keyboard shortcuts, which lets me keep it running while shelled into my work machine (the default keys for bringing up Optimal Layout are Opt-Tab, which just happens to coincide with a certain shortcut in Windows… I’ve also been using iTerm2, which adds split panes and more to the existing iTerm awesomeness.

I’m using SimpleNote to keep notes synced across all platforms (Mac/Win/Web/iOS/webOS) which is nice for maintaining work notes, shopping lists, etc.

Have been doing a lot of embeddable JS work lately, and plan to write some tutorials and maybe put together a presentation or two if I can drum up some interest on the subject. I need to get some actual metrics together, but once I have real data I’m hoping to put together a service for compiling multiple scripts into an optimized, 3rd-party-safe embeddable widget or bookmarklet. I’ve also been building some interesting on-site widgets, like reusable dialogs/lightboxes and image croppers. Most recently I’ve built a Frankenstein monster that encapsulates both classes for reuse across pages and multiple instances per page. To avoid managing dependencies manually and to avoid the possibility of accidental multiple inclusion of a script (which aside from causing unexpected/hard to trace behavior is also a waste of HTTP requests) I’ve used RequireJS for the first time. Aside from a few hiccups learning the syntax, it’s working great. I plan on learning the optimizer component next so I can build concatenated/minified scripts that include templating strings and so on within them. It’s almost like I’m using Dojo…

So this may have been a longish post, but it was mostly a mishmash of various “what I’ve been doing” bits. Hopefully someone finds some of it helpful. Please ask questions if you’re trying out any of the same software or ideas!

As far as Mac apps go…

I’ve been doubling my efforts to use QuickSilver, especially in terms of avoiding guis like iTunes and Finder. The book talks about the importance of not disrupting flow and train of thought. While I’ve read about how damaging this can be in Pragmatic Thinking & Learning, I never stopped to examine just how much they affect me in particular. I seem to have attention deficit like a madman. I start and stop tasks several times a minute at my worst. I’ve flirted with using apps like Think and WriteRoom (or the free JDarkRoom) to try avoiding distraction, neither have stuck very well. It seems I’m more inclined to manage my own time, but I still need to figure the best way to do so.

What has stuck as far as productivity already is Spaces, OS X’s built-in virtual desktop; Exposé, the likewise built-in window management tool; the aforementioned QuickSilver; JumpCut; j2 and assorted custom aliases in-shell. I’m a biggish user of TextMate’s macros, snippets and columnar/multiline editing, as well as the Git/SVN bundles.

Now I’m adding Text Expander to my list of must-have apps. I’d played with Typinator (a similar app) before, but it didn’t suit me. Text Expander absolutely does. Being able to customize not only your snippets but how and when they’ll trigger is great. I find myself adding new words while talking to friends on AIM. I’m up to almost 300 words now, hopefully I’ll plateau around 400 or so. Haven’t even used it for development work yet but it’s cutting back on time wasted typing in chats! Also check Doodim, an easier version of Think.app.

Can’t wait to see what apps, commands and ideas I discover going forward, reading the book and on my own!

Another app I’m interested in but haven’t used much yet is Path Finder, a Finder alternative with built-in terminal and lots more.

Anyone with ideas, app reviews or advice, please do contribute!

Like WordPress, Firefox became the powerhouse in market share it is at this point because of the plugins extensions (and getting an early jump on the tabbed browsing trend didn’t hurt).

I’m currently writing this post from ScribeFire, a badass little plugin extension that allows me to post directly from the browser in a rich text editor without worrying whether I’m online or not. I use several other plugins extensions every day. Most notably of these, as a developer, is of course Firebug. While the source of some of the aforementioned memory leaks and random crashes/misbehavior, it’s well worth the cost for the benefits of a true debugging environment for my JavaScript, an unmatched HTML inspector and the best analysis-and-live-tweaking CSS toolset on the planet. Nothing else matches it, on any platform. Webkit has Web Inspector and Drosera, which are fine, I guess. IE has various JS consoles (including Firebug Lite and several bookmarklets you need not include in your pages), DebugBar and maybe one or two other commercial tools, but nothing with the integrated firepower of Firebug. I also rely heavily on Web Developer toolbar for quickly enabling/disabling JS/CSS and launching validation checks when I’m bug hunting.

More bitching and moaning after the jump!

There are a number of other plugs extensions I use (Google Browser Sync, Server Switcher, YSlow, Colorzilla, MeasureIt get the majority of use, though I play with OneClick for WordPress now and then) which at this point do not yet work in Firefox 3, which is currently at Beta 5. Web Dev Toolbar, ScribeFire, Firebug all do. Or do they…?

Firebug has recently seen some leaps forward in version numbers after a long period of inactivity (save for a scarily bad-looking little site for FireClipse). With several versions now vying for my attention, which am I to choose? Currently, I’m running Firebug 1.1.0b11, but I’ve tried 1.2 alpha. Not going there again till it’s in beta. Complete fail. Won’t need to play with ChromeBug unless I ever get round to developing some extensions myself, which I don’t see happening any time soon.

What drove me to write this post, however, wasn’t the terrible experience with Firebug 1.2a. No, it was my burgeoning crush on Firefox 3b that triggered this rant. That’s what bit me tonight.

Firefox 3 runs immensely faster than FF2. Startup time, page load time, JS execution time, all seem much faster. I credit much of this to Mozilla handing off much of the work of the browser to native API for drawing form controls and for similar tasks. FF3 will fill the whole screen of my MacBook when I click the little green button, something FF2 never did, the bastard. FF3 respects my wishes when I ask it to go away, instead of struggling to stay afloat like a drowning cat who just won’t stop mewling no matter how many times I dunk the bag in the frigid water of Quit. Pretty soon I’m just going to have to let the cat out of the bag entirely (except for testing backwards compatibility) and start playing with a new pet.

Anyhow, with all of the good things to say about FF3b, there are a few bad ones. First being that I’m lonely for plugins extensions. I’ve read about an about:config hack for running FF2 plugs extensions untested on FF3 and may try that, we’ll see. Next up, they keep breaking things. I tried upgrading to FF3b5 which came out… I dunno yesterday maybe? and suddenly Firebug ceased doing a damn thing. Retro’d my ass back to beta 4 real quick, tell you what. Tonight I tried connecting my MacBook to my secondary monitor and Firefox 3 just wouldn’t let me drag it to the other screen. What gives?! The mouse cursor simply slid off the top of the app no matter how hard I leaned into the mouse button. FF2 worked just fine.

Enough getting worked up over bull shiz. Time for me to sleep.

Cake’s bake utility is vaguely useful if only for the fact that I don’t have to memorize 3 lines of code. I’m much more interested in learning how to use the Cake console, as that’s one of my most-frequently accessed parts of Rails. Any time I need to learn something, I dig through mysql to find out field names, and dig through console to quickly see exactly what format data will return in, etc.

I’m glad Cake uses associative arrays rather than simple argument lists, that’s for sure. However, the 1.2 validation, though more complete than pre-1.2 validation, seems to be one of the more unnecessarily complex pieces of Cake functionality. It can be treated more simply, though, which is great.

Next up on the Cake learning plate: caching, testing, and ACL.

The first makes sense as I’ll be serving up images from there for some time at this point. The second is a temporary concession; I need to make a local copy of the link image eventually and stop retrieving images from there, and the third is more interesting.

Another topic discussed in High Performance Web Sites is the maximum # of parallel downloads per domain. By the HTTP 1.1 standard this is 2. Browsers can choose to ignore this but by default they obey like Andre. However, a way around the 2 simultaneous download limit (without resorting to a true CDN) is accomplished through mirroring. The browser believes the source files (in this case a few scripts and a .swf) to have come from another site, though the files are located in the same structure. Brilliant, and as cheap a solution as any that can potentially double my # of downloaded components at a time.

Grade attained, I can finally sleep. Well. Soon.