Sun 18 May 2008
PithHelmet 2.8.3a Beta (for Safari 3.x.x) # 5:08PM EDT

This fixes the installer for Safari 3.1 on Tiger.

PithHelmet-2.8.3a.tbz

Sat 26 Apr 2008
PithHelmet 2.8.3 Beta (for Safari 3.1) # 9:20PM EDT

This fixes PithHelmet for Safari 3.1.

PithHelmet-2.8.3.tbz

Sun 02 Dec 2007
PithHelmet 2.8.2 Beta (for Safari 3.0.x) # 9:34PM EST

This build works with Safari 3.0.4 on either Tiger or Leopard but fixes the menu conflicts properly.

I also changed some of the installer error messages to be more meaningful. Hopefully that will cause less confusion.

PithHelmet-2.8.2.tbz

Thu 29 Nov 2007
PithHelmet 2.8.1 Beta (for Safari 3.0.x) # 3:47PM EST

This build works with Safari 3.0.4 on either Tiger or Leopard.

I changed a few of the command keys so that they no longer conflict with new Safari shortcuts.

Fri 16 Nov 2007
Safari 3.0.4 on Tiger 10.4.11 # 1:21AM EST

This doesn't work right now. Unfortunately, I'm out of the country at the moment and I won't have a chance to work on this until I return home.

Wed 14 Nov 2007
TerminalColors in Leopard # 12:41PM EST

I took a quick look, but bascially I have to rewrite the thing from scratch. In the meantime, someone else wrote a plugin called "TerminalColours" which seems to be about what I'd do. I can't get it to work exactly (saving seems busted), but I don't have the time to fix my version right now.

Sun 11 Nov 2007
PithHelmet 2.8 Beta (for Leopard) # 9:59PM EST

This build works with Mac OS X 10.5 and Safari 3.0.4. By far, I spent more time figuring out the new installer and it's charming quirks than I did actually writing code.

Anyway, I've used it for a few days without incident, so I think it should be relatively bug free. Enjoy.

Update: There a number of reports about the plugin just not loading. It's releated to the embedded SIMBL installer.

Update 2: I think I've fixed the installer problem, reinstalling should correct everything.

Update 3: Ok, still hearing about problems.

Update 4: 3rd time's the charm? I changed the installer once again - hopefully this should fix all variants of the problem. I also just noticed that Command-Shift-K and Command-Shift-R have been hijacked by Safari, so I need to remap these in a new version.

Tue 23 Oct 2007
Reports of My Death... # 1:49AM EDT

I haven't been doing much Mac development for a while. Still, I managed to release a new version of SIMBL for Leopard after some urging by a friend. When Leopard finally releases, I hope to take a few days and finish updating PithHelmet.

Sun 05 Aug 2007
PithHelmet 2.7 Beta # 9:17PM EDT

This is a quick build to get PithHelmet working with Safari 3.0.x beta.

The GIF animation filter is broken - I'm not sure when I'll have a chance to sit down and fix that. If you have any troubles, send me an email.

PithHelmet-2.7-78.tbz

Mon 10 Jul 2006
"PithHelmet" Bugs # 11:58PM EDT

It seems like there have been more instances of bug reports with the new version of PithHelmet, despite the fact that the number of code changes was small. The core problem is not in fact PithHelmet - it is actually WebKit/WebCore. Most bugs reported are simply the consequences of making uncommon circumstances more common.

There are two main bugs - one crash in the DOM code and one server response handler.

I had previously submitted a reduction of the 1st bug and believed it to be fixed. However, I don't think that code is in the Apple official Safari release.

I actually spent most of my Saturday trying to figure out why WNYC.org crashes Safari. After 4-5hrs I narrowed it down to the XMLHttpRequest call. This call tries to enforce that the caller and target share the same domain name. When PithHelmet changes the path to a local empty file, WebCore treats this as a violation of the security model and decides to bail on the request. This is all fine and well - unfortunately, it does not clean up the request properly and when data flows into the response, the application state treats the appearence of this data as a critical error and crashes the application.

Sadly, the fix is not as simple as "ignore the data" - that would like cause trouble elsewhere. No, this is a more sinister problem that I haven't had time to wrap my head around. The worse news is that even if I were to fix the bug, submit a patch and have it accepted, it would still be months before it would make it to production. Unfortunatley, I can't see how to work around the crash either - the information I would need to avoid the situation is not readily available to PithHelmt at the time I would need to make the decision.

Kind of frustrating. :-/