Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Good news all, the Tiger mini-browsers are almost done...

10.5Trim(ph) is still in development / refinement...

And TenFourTrim has a renovation in the works. :)

@wicknix, I strongly suggest changing TenFiveTube's homepage from youtube.com to tonvid.com. The performance difference is unbelievable. Watching videos in-browser is now actually possible on my 400mhz G3 (rocking its Rage 128!).
 
Last edited:
The Tiger mini-browsers are as fast as Leopard WebKit, and almost as lightweight.

Releasing later today...
 
Last edited:
  • Like
Reactions: G4fanboy
TenFourTrim has received a minor update alongside a rebranding as AquaTrim.

It has also been newly accompanied by AquaAccelerator, the Tiger equivalent to AuroraAccelerator. Unfortunately, all it can really do on Tiger is disable BeamSync...

This is TenFourTrim's final revision.
 
Last edited:
I know this is a bit of an old thread, but I was wondering how high the risk is of something breaking by running this and if there is any application that is known to break. I would love to speed up Tiger, but I want to know if there is anything that could possibly break due to this.
 
@RogerWilco6502 To my knowledge (and prior experience), the risk is very very low, outside of what the included Read Me details.

I've run it on every machine of mine with Tiger, and have never encountered any resulting problems. But if there ever is, Undo AquaTrim provides an escape hatch out back to the stock config.

The script file can be opened with TextEdit, and then thoroughly examined for every action it makes.

ADDENDUM: If the user disagrees with any one change made by the script, they can simply remove the subject line, save, and run it like that.
 
Last edited:
@RogerWilco6502 To my knowledge (and prior experience), the risk is very very low, outside of what the included Read Me details.

I've run it on every machine of mine with Tiger, and have never encountered any resulting problems. But if there ever is, Undo AquaTrim provides an escape hatch out back to the stock config.

The script file can be opened with TextEdit, and then thoroughly examined for every action it makes.
Ok, thank you! :D
 
This seems great! Thanks for sharing this "accelerator" ;) I wonder though:
Would it be better to update my PB running on Tiger 10.4.11 to 10.5 or rather use this?

I've thought about upgrading to 10.5 before, but the thing, my external hard drives won't mount on my Mac so I cant save what's on my HD. Maybe I could upload all in a cloud or something then download it later... But at the same time, I cant be bothered to look and find versions of the adobe apps I use (PS and ID) from time to time. It's not that I am lazy - I am quite busy at the moment.
 
@SnowBlush Upgrading to 10.5 will give you a second browser option, better Web performance, and slightly better security. Staying with Tiger will give you less modern applications, frameworks, and libraries, but the OS and applications will be faster on your PowerBook.

If running older offline applications are a priority, Tiger will probably give you a better experience. If Internet-related tasks are a higher priority, Leopard will likely serve you better.

Of course, even though Aqua / AuroraTrimcelerator are there to streamline whatever you choose, the decision mainly depends on the user's needs.

Hope I was helpful. :)
 
Yes, it was helpful - thank you @z970mp ! :)
I think I should upgrade to 10.5 then... (but like I said, I can't save everything that's on my hard drive as I cant connect my externarl drives on my mac...) I might try and find a shop in my town that allows me to buy a new computer whatsoever - because I only own this old mac atm - and the speed (or lack thereof) and incompabilities are a hassle.

Anyway , I installed the Aquatrim and it seems to accelerate a bit :) The only thing is, I can't read videos anymore, I get this message evrytime there's a youtube video embeded, or even directly on the YT website: "Your browser does not currently recognize any of the video formats available.
Click here to visit our frequently asked questions about HTML5 video."

I do have the QuickTime plugin installed in that TenFourFox suggest to use. (sorry I forgot the name of it)
And I can still watch YT videos in a TenfourFoxBox (but not in the main browser) Would you happen to know why? thanks!
 
@SnowBlush Would you have installed foxPEP, perchance? You will need to install the MP4 Enabler into TenFourFox after doing so if you wish to watch YouTube videos in-browser.

 
@z970mp Yes, i have installed foxPEP. The thing is, I had installed the mp4 enabler before and videos ran fine. And now, when I go the github link you gave, all I can download is a file named "readme.markdown" and I cannot open it. I'm feeling very dumb at the moment! :rolleyes:
I think it's not clear whether I also should install AquaWeb and AquaVid.
 
@Dronecatcher Heehe... Yes indeed - that worked much better this way :D :rolleyes:
@z970mp yes MSE is enabled. The videos works now ^^ I don't find Aquavid much different or faster than the TenFourFoxBox I created with youtube as a homepage. Is it, though? (also, it seems you used a android (or maybe safari-ipad) string agent for aquavid? I find it not as responsive as the TFF string agent, (whilst using youtube, I mean)
Could you explain why you you YT is better experienced with a "mobile" interface? Thanks!!
 
Could you explain why you you YT is better experienced with a "mobile" interface? Thanks!!

Traditionally, mobiles were less powerful than desktop machines and with limited screens, so content would be optimised for them - lower res video and less page elements and coding.
However, as mobile technology has advanced the gap is much smaller now and soon the mobile user agent will offer little advantage as sites will only accommodate newer (and more capable) user agents.
 
  • Like
Reactions: z970
@SnowBlush What mainly sets AquaWeb and AquaVid apart from regular FoxBoxes is that foxPEP is built right in, which is supposed to make a marked difference when navigating the Web.

(Strangely enough, I've also noticed that AquaVid benefits from this less than AquaWeb does...)

Anyway, AquaVid was made before PPCMC 7 existed, when at the time there were no functioning YouTube clients for Tiger. As far as I see it, it can at the very least now provide a tailored and convenient method of copying URLs to stream, download, and process videos with the latter.

I hope that more or less answers your question...
 
Okay, taking a second look back at this project, I can finally understand why my connected network servers have been dropping out after so long on Tiger client. Not only that, I'm getting sick of seeing the missing font sign (for Dingbats) across the Internet.

Plus, I don't like the fact that this script has only provided marginal improvements for some, and killed AltiVec for others. This is unacceptable.

Looks like development hasn't finished after all...

*rolls up sleeves*
 
Last edited:
Hi @z970mp firstlt thanks for your hard work on this - I was just messing about with my iBook G4 1.33 Ghz, and after doing some reading here - I didn't use the Trimcelerator, but I did use the Accelerator, and it made a marked difference. I will also say that I added the entire iBook's hard drive to be excluded from Spotlight in the preferences pane. And both these made a lot of difference to my performance - it is maxed out with 1.25GB of RAM though, so not sure how much more I would have gotten out of this.

When I installed Tiger on this I already elected for a nice and trim install - do the fonts make so much of an impact? I am tempted to run the script with just sections of stuff that's really not needed. I wanted to ask, do we know why AltiVec was hampered with this patch? I know nothing compared to yourself, but I was looking at the script and couldn't see anything that would explicitly remove it
 
@Toyface19 The fonts do make an impact when launching certain applications, namely TenFourFox. More specifically, TFF parses through system fonts while initializing, which can compound startup times. So, the less fonts there are, the quicker it loads.

That's what I'm trying to figure out now. Generally, I tried to keep to a "don't remove if it's an unknown" development model because there was a fair amount of components in the system which I couldn't find any documentation for, so they were left in, but there were a couple of exceptions to this.

Unless my memory has already failed me through the course of only a year, I don't know what the VSP framework relates to, but it sounds suspiciously similar to "VMX", otherwise commonly known as AltiVec, so that's definitely subject to removal.

Otherwise, the only other things I could imagine might have anything to remotely do with hardware acceleration / technology leveraging would probably be XgridInterface and XgridFoundation, however I recall them to relate to resource clustering and are only actually useful for data center usages, for instance with hardware like the XServe.

If it strikes your interest, here's one of the resources I used during development:


So far, I'm getting the feeling that it might be best to go with a split product availability, with one being meant to strip the system to the minimum for old and unsupported machines to run better, and another to simply optimize it for newer ones without removing any functionality. That way, there is no compromise for any group.
 
Last edited:
Thanks for your reply - I will definitely take a look at the link you've provided. All of this reminds me of my teens running OpenSUSE SLICK on a Pentium III, and with all the CK kernel optimisations and packages it was one hell of a distro!

Am I to take it that your approach has been to remove what you can in order to put less strain on the system upon boot and to prevent any of this from ever entering RAM? It seems like sound logic, for sure. I wonder if the 'X' in those two frameworks you mentioned relate to X11 or anything UI based? I know very little of the OS X of old though, so I really wouldn't know. I would be more than happy to help you test any changes to see if your progress is keeping altivec intact. I have an iBook G4 1.33 GHz 2004 with 1.25GB of RAM with Tiger, my PowerBook Al 2005 1.67 Ghz 17-inch is running Leopard, where I have tried AuroraAccelerator to great effect - but I will post there so as not to go off topic here.

Back to the point about fonts - I am a printer/typesetter/graphic designer/whatever by trade, so it is ingrained in me that more fonts is more good, clearly not on performance, but for giving a client a choice, that is. I wonder, if TFF and its brethren are to look for these fonts on startup, it will speed up any load up times as there are less fonts for TFF to index or whatever, but what happens if for whatever reason something needs these fonts and they're not there, would that cause any issues or whatever?
 
Yes, that is precisely the approach, and has been demonstrated to hold water in the past.

Referring back to Apple's framework resource, XgridFoundation.framework, introduced in OS X 10.4, "contains interfaces for connecting to and managing computing cluster software", proving my memory correct. I assume XgridInterface.framework is used for identical purposes, so that is included too. As I'm sure you can imagine, nobody still using PowerPC machines in 2020 tends to cluster them together, so seeing as how this is an actively-loaded framework (according to System Profiler), removal ought to be not only harmless, but beneficial, as that's two less items to load on startup.

That's great to hear; I appreciate your generosity of time. If you're willing, you can run AquaTrim in its current state, and download XBench to run a benchmark. Then, move VSP.framework back to its original location, repair permissions, reboot, and then run the benchmark again. Afterward, compare the benchmark results when the system lacks VSP.framework compared to when VSP.framework has been restored. This may confirm or disprove my initial suspicions regarding the component, saving us time.

I've seen that happen before. If I recall, if the fonts are missing when, say, Photoshop has opened a document where those fonts are required, I believe the system will just use a different font instead. Obviously, this is unideal because we want the performance enhancements provided by the script without compromising on the ability to use the wide range of built-in fonts.

...I just got an idea. Instead of removing the fonts outright, perhaps the script can simply rename them so that when TFF or other applications parse the database upon startup, it does not find the fonts it is looking for, and therefore does not load them. Then, when the user wishes to use them for a professional instance, like Office or Creative Suite, they can simply run another script, and the fonts' names are restored, thus allowing them recognition.

... Then again, I don't know if TFF parses through the database for a specific list of fonts, or just for inhabiting items, in which case that route won't work.

Anyway, we'll figure it out...
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.