The community Taskforce initiative has now come to a close.
Thanks to everyone who made thoughtful and genuine contributions to the website.
All submissions will be kept publically available for the forseeable future for reference purposes.

This website is part of the community Taskforce initiative

Comments by user "linuxforever"

Registered since: May 25, 2009

iTunes: improve performance

Wrote on May 25, 2009, 9:39pm

In my experience (Windows 2000, iTunes v5, some AMD 900MHz) the CPU load always rises from a few per cent to the full 100 per cent when viewing the active downloads! When viewing something else (music playlist, iTunes store etc.) it dropps again. I do not have the slightest idea what sort of strange programming mistake is causing this unintuitive behaviour (I mean, I could imagine, but how can you not reconize that, Apple Devs, please??)
So I agree with you, kelchm: Rewriting the app in Cocoa is the only sensible way of getting rid of all the performance issues most iTunes versions obviously seem to suffer from!

Inconsistency in OSX Leopard Mac-Native Aqua Apps

Wrote on May 25, 2009, 9:26pm

Unification seems to have huge advantages:
- You only need to change the look at one place to modify all of them (for customization by the user or system upgrade by the vendor); nonetheless, it should be up to you whether you use system standard or the individual theme provided by application which makes sense for photo viewers for instance where usually a dark background and unobtrusively colored icons make more sense (the icon shape does not necessarily have to change, maintaining a bit unification)
- You find the standard functions more easily since you know how the buttons look like
- You can trust the symbols meaning they represent the same very function in each application (e.g. different symbols if the action is delete permanently or move to trash folder of the application)

Unfortunately, the latter advantage is really one of the biggest flaws security wise because it enables programmers to copy the standard theme in order to trick users to invoke actions they di not see coming... Nothing new in the Windows World!

Moving files in Finder displays a Copy dialog

Wrote on May 25, 2009, 7:42pm

I conclude that we all agree: The title does not tell the average consumer the whole truth. Therefore, I do not understand why there is no help button/icon of any sort explaining what is actually going on and what would happen to the original files and the "new" ones on a user initiated abort of the running procedure. For this reason, I recommend a short explanation pressing an info-button and additionally showing a table listing of the paths of the "original" and the "new" files also indicating the progress by color encoding the data already processed.
I definitely disagree with polycat33 tending to generalize that such action is short enough to neglect the presentation of extensive information because I consider it mandatory, especially to debug operations involving network connections or two different storage devices. For example pulling out a flash drive while moving files onto it should make the OS show up a notification suggesting to re-plug the drive to continue and informing the user that the flash drive currently contains a corrupted/incomplete copy of the original files which are still untouched and not even partly (re)moved.

Pause / Resume File Copy

Wrote on May 25, 2009, 7:18pm

Additionally, the restart could be delayed by a timer or signal sent from the OS or a qualifying application (e.g. internet download manager using the whole network bandwidth if it is a copy process over the network...). Even better would be a customizable Quality of Service routine managing the storage bandwidth (e.g. to give priority to another computer accessing the very same storage device through the network connection).

Finder: Queued file copy/move operations

Wrote on May 25, 2009, 7:08pm

Thinking of SSDs that most future Apple products are likely to be equipped with, it would also make sense for the OS to communicate to the SSD what transferred information belongs to a continuous data stream (one single file). So the OS would order to reserve the right amount of storage capacity for each file reducing fragmentation and its consequences and at the same time delivering additional information for the dynamic wear-levelling algorithm built into the SSD controller. Unfortunately, current Operating Systems do not pass enough information to the SSD (e.g. via NCQ employed in todays HDDs that is primarily going to be used in SSDs to alleviate write amplification increasing the overall lifetime).
The write speed would be comparable to common SSDs since they can already write multiple data streams in parallel as they are writing to many physical cells at once. But the read speed and lifetime are potentially be increased thanks to the reduced fragmentation.
These features could be implemented by storage manufacturers like it is going to be done with the ATA-Trim command for instance.

Page: « Back 1 2 (5 of 20 results)