Jump to content
519 posts in this topic

Recommended Posts

Revised version (v2.4.6) attached to the opening post.

The changes are as following changes:

 

Changes from 2.4.2 -> 2.4.3

 

- Updated cmosDumperForOsx with latest v0.1 from STLVNUB.

- Corrected typo in HTML report RTC sub table header for cmosDumperForOsx title.

- Changed HTML report script to dump BootLog.txt if it does not contain 'CLOVER' string, so should work when using UEFI BIOS.

- Changed folder name to use initial cap. So if CLOVER, use Clover.

- Hopefully fixed outstanding bug where gTheLoader string for AMERICAN_MEGATRENDS should be renamed to AMI.

 

Changes from 2.4.3 -> 2.4.4

 

- Updated cmosDumperForOsx v0.1 with Leo compatible version (Thanks Slice).

- Updated dmidecode to v2.11b (and Leo compatible version) (Thanks Slice).

- Removed underscore from American Megatrends string.

- Revised/tried to improve the routine for detecting gTheLoader.

- Included updated version of pci-ids.gz.

 

Changes from 2.4.4 -> 2.4.5

 

- Fix rEFIt rev detection to work with UEFI boot (thanks for hex Kynnder).

- Amend generateHTMLreport script to work with Clover bootlog without 'Clover' in the name (in the case of UEFI boot).

 

Changes from 2.4.5 -> 2.4.6

 

- Dump routines are now run in three stages of background processes. Dumps can complete nearly 2 x faster (depending on the system hardware and current running processes / free resources at the time of execution).

- App now uses a text window, instead of a progress bar display with collapsed window.

- App text window stays open when script completes to allow viewing/saving log.

- Amended verbose output log, and added time taken to complete.

- Removed system version check for RTC dump as latest version now works for OS X 10.5.

- Disk partition table info dump now produces a separate file for each disk and the corresponding HTML report now shows separate sub-tables to match.

- PCI Utils driver is only loaded if not already loaded.

- HTML Kext dump is now built from dumped text file rather than running kextload multiple times.

- Revised CSS spacing for nav links to allow a gap under collapsed sub tables.

- HTML anchors re-positioned for better view when clicking nav links.

- HTML Diskutil & Loader dump sub tables are now collapsable.

Link to comment
Share on other sites

  • 2 weeks later...

Revised version (v2.4.9) attached to the opening post.

 

Changes from 2.4.6 -> 2.4.7

 

- When an MBR and PBR boot code is found but not known, it's now marked unknown in dump files.

- PBR disk sector dump file now shows unknown code rather then just ignoring it if it couldn't detect what's there.

- Removed an unnecessary for loop when checking to see if a volume is mounted.

- Re-enabled application progress bar.

- App closes when finished.

- Log file now written to reports folder, and most verbose output directed there.

- Ownership/permissions of reports folder fixed.

- Added nvram.plist dump.

- Pashua window resized to accommodate new option.

- Existing Pashua window selectable dump options re-ordered alphabetically.

- Added option 'None' for archive.

- Shortened some HTML nav link titles.

- Moved some nav links to keep alphabetical order.

- Fixed HTML anchor for OpenCL.

- Added option to select collapsed or expanded HTML tables from Pashua window.

 

Changes from 2.4.7 -> 2.4.8

 

- Adjust HTML nav links length so it doesn't overflow on to two lines.

- Fix check for 'Collapse Tables' checkbox so it works.

- Fix HTML table arrows so they point the right way.

- Correct spelling of LSPCI in Pashua window.

 

* Note: I'm aware of an issue where expanded tables initially require clicking twice on the arrow to collapse it.

 

Changes from 2.4.8 -> 2.4.9

 

- Use the OS X 10.5 compatible version of dmidecode as standard.

- Moved SMBIOS HTML table to appear after the RTC table to keep it sorted alphabetically.

Link to comment
Share on other sites

Revised version (v2.4.9) attached to the opening post.

 

Changes from 2.4.6 -> 2.4.7

 

- When an MBR and PBR boot code is found but not known, it's now marked unknown in dump files.

- PBR disk sector dump file now shows unknown code rather then just ignoring it if it couldn't detect what's there.

- Removed an unnecessary for loop when checking to see if a volume is mounted.

- Re-enabled application progress bar.

- App closes when finished.

- Log file now written to reports folder, and most verbose output directed there.

- Ownership/permissions of reports folder fixed.

- Added nvram.plist dump.

- Pashua window resized to accommodate new option.

- Existing Pashua window selectable dump options re-ordered alphabetically.

- Added option 'None' for archive.

- Shortened some HTML nav link titles.

- Moved some nav links to keep alphabetical order.

- Fixed HTML anchor for OpenCL.

- Added option to select collapsed or expanded HTML tables from Pashua window.

 

Changes from 2.4.7 -> 2.4.8

 

- Adjust HTML nav links length so it doesn't overflow on to two lines.

- Fix check for 'Collapse Tables' checkbox so it works.

- Fix HTML table arrows so they point the right way.

- Correct spelling of LSPCI in Pashua window.

 

* Note: I'm aware of an issue where expanded tables initially require clicking twice on the arrow to collapse it.

 

Changes from 2.4.8 -> 2.4.9

 

- Use the OS X 10.5 compatible version of dmidecode as standard.

- Moved SMBIOS HTML table to appear after the RTC table to keep it sorted alphabetically.

 

:D

maybe it's better to put those words in UI when choosing which categories to perform gathering infos.

All options are selected by default

Link to comment
Share on other sites

I'll make a note to do this for the next release.

Thanks for the suggestion p.H.

:D

Thinking of the first time when I used DarwinDumper, I don't even know how to make a choice.

Maybe besides the note, marking all the options by default is another choice :)

Link to comment
Share on other sites

Yes. Logically, that's how it should be done and that's what I wanted to do. However, if for example the user only wanted to run the 'Diskutil & Bootloaders' dump, then they would have to untick all the others which would be a chore.

 

One way to overcome that would be to have two buttons, one reading 'Select All' and the other reading 'Deselect All'. However, the design and execution of DarwinDumper doesn't allow that. The problem is because DarwinDumper is essentially a shell script which uses an app named Pashua to draw it's UI for taking user input. While this is good, the UI is not reactive in the sense that clicking a 'Select All' button will not refresh the UI to give visual feedback to the user.

 

I appreciate the feedback and DarwinDumper is in continuous development. I am quietly looking at other methods of presenting a UI for the script(s) and have a few routes to explore, for example CocoaDialog and JQuery. Unfortunately, I just don't have as much spare time as I used to which means this get's worked on only as and when.

 

Having said that, DarwinDumper is an open project so if anybody else has the skills, time and the desire to take it to the next level then they're welcome to. All that's asked is the source remains open and all changes and improvements are shared at the original projectosx thread.

Link to comment
Share on other sites

Yes. Logically, that's how it should be done and that's what I wanted to do. However, if for example the user only wanted to run the 'Diskutil & Bootloaders' dump, then they would have to untick all the others which would be a chore.

 

One way to overcome that would be to have two buttons, one reading 'Select All' and the other reading 'Deselect All'. However, the design and execution of DarwinDumper doesn't allow that. The problem is because DarwinDumper is essentially a shell script which uses an app named Pashua to draw it's UI for taking user input. While this is good, the UI is not reactive in the sense that clicking a 'Select All' button will not refresh the UI to give visual feedback to the user.

 

I appreciate the feedback and DarwinDumper is in continuous development. I am quietly looking at other methods of presenting a UI for the script(s) and have a few routes to explore, for example CocoaDialog and JQuery. Unfortunately, I just don't have as much spare time as I used to which means this get's worked on only as and when.

 

Having said that, DarwinDumper is an open project so if anybody else has the skills, time and the desire to take it to the next level then they're welcome to. All that's asked is the source remains open and all changes and improvements are shared at the original projectosx thread.

:D

If it's really of great work, a kind notice would just be fine.

Once again, great appreciation.

Link to comment
Share on other sites

Revised version (v2.5.0) attached to the opening post.

 

Changes from 2.4.9 -> 2.5.0

 

- Re-designed the Pashua UI.

- Updated pci.ids.gz

 

 

:D

I saw the notice in the latest version.

Plus, after running safari automatically opens the result page while my default program for web is chrome.

Does DarwinDumper contain code to launch safari to open the result page ?

Link to comment
Share on other sites

Plus, after running safari automatically opens the result page while my default program for web is chrome.

Does DarwinDumper contain code to launch safari to open the result page ?

The DarwinDump.htm file is a standard file and should open in your OS the same way as any other .htm file.

If you select 'Get Info' for the file in the Finder, what do you see under the 'Open with:' section?

If it reads Safari and you want to open that and all other .htm files in Chrome then select Chrome from the list and click the 'Change All...' button.

Link to comment
Share on other sites

The DarwinDump.htm file is a standard file and should open in your OS the same way as any other .htm file.

If you select 'Get Info' for the file in the Finder, what do you see under the 'Open with:' section?

If it reads Safari and you want to open that and all other .htm files in Chrome then select Chrome from the list and click the 'Change All...' button.

It's already chrome there.

Along with DarwinDumper first came out, I've been using it. At that time, it just open safari for the result page. So I've already modified the default program for htm file. At first glance, the htm file in my hack has chrome's icon on it. I also double check the "get info" tab. It also showing chrome as the default program. Dunno what's wrong here. :(

  • Confused 1
Link to comment
Share on other sites

hmm... I've just run a test here and I see the same result you are describing.

 



Chrome is set to be the default application for .htm files and the file opens in Chrome:


1 - when I double-click it in the Finder


2 - running open /path/to/file/DarwinDump.htm from the Terminal



 

So the above two work as they should and behave according to the man page:

The open command opens a file (or a directory or URL), just as if you had double-clicked the file's icon. If no application name is specified, the default application as determined via LaunchServices is used to open the specified files.

 



So to echo your question, why is the .htm file load up in Safari when the open command is issued from the DarwinDumper script?

Link to comment
Share on other sites

hmm... I've just run a test here and I see the same result you are describing.

 



Chrome is set to be the default application for .htm files and the file opens in Chrome:


1 - when I double-click it in the Finder


2 - running open /path/to/file/DarwinDump.htm from the Terminal



 

So the above two work as they should and behave according to the man page:

The open command opens a file (or a directory or URL), just as if you had double-clicked the file's icon. If no application name is specified, the default application as determined via LaunchServices is used to open the specified files.

 



So to echo your question, why is the .htm file load up in Safari when the open command is issued from the DarwinDumper script?

I don't know either. Since you already experienced the problem, maybe you can find related stuff in source.

Link to comment
Share on other sites

Lol.. yeah. My last statement was really a rhetorical question rather then me asking you :)

But I think the issue is because the script is running as the superuser so the default app for .htm files in LaunchServices for the superuser would be Safari. When you/I change our default app associations then they're only set applicable for our user accounts. I'll see if I can find a suitable fix.

  • Sad 1
Link to comment
Share on other sites

Lol.. yeah. My last statement was really a rhetorical question rather then me asking you :)

But I think the issue is because the script is running as the superuser so the default app for .htm files in LaunchServices for the superuser would be Safari. When you/I change our default app associations then they're only set applicable for our user accounts. I'll see if I can find a suitable fix.

Yep, nice guess. It may be the real cause.

Link to comment
Share on other sites

Hey blackosx

Try this, could be handy :wink2:

Hi STLVNUB

 

Thanks. I've tried it and it running the following works great:

./DefaultApplication -ext htm

However, it doesn't let me read the default application for a different user.

 

I've since found that after enabling Chrome as default app for .htm files.

~/Library/Preferences/com.apple.LaunchServices.plist gains the following:

<dict>
<key>LSHandlerContentType</key>
<string>public.html</string>
<key>LSHandlerRoleViewer</key>
<string>com.google.chrome</string>
</dict>

So I'm thinking I can use this as a reference to see if the user has changed their default web app.

Link to comment
Share on other sites

As a user, I do not care in which browser it opens. I need DarwinDumper's output. Just my hint.

:D

It's just an imperfection. :thumbsup_anim:

It doesn't bother too much just as you said.

Link to comment
Share on other sites

Revised version (v2.5.1) attached to the opening post.

 

Changes from 2.5.0 -> 2.5.1

 

- exFAT format partitions should no longer be unknown to the gatherDiskUtilLoaderinfo script (thanks dmazar). I've only worked from blank and NTLDR examples so I'll need to run more tests in the future to make sure everything is identified correctly. Please tell me if what I've done doesn't work.

- HTML report nav links now expand (by design they do not collapse) relevant table as well as jumping to it.

- The collapse / expand arrows have now been replaced with clickable table header bars. Way better!

- Re-ordered nav links to match alphabetical order of tables, and removed 'Top' nav link.

- Adjusted HTML report for the Diskutil & Bootloader table to better fit longer names. The Active column is now gone, though a single space remains for an asterisk.

- Renamed 'Windows NTFS Loader' to 'Windows NTFS NTLDR'.

- Tweaked the look of the Pashua window.

 

As a user, I do not care in which browser it opens. I need DarwinDumper's output. Just my hint.

You're right. Thanks for helping me keep focused.

 

@p.H - I've left the code as it was. It would be easier if you changed the .htm file association to Chrome for your root account than for me to work out an algo to find your local user browser association setting and open the file in that from the script. Thanks for the feedback though.

Link to comment
Share on other sites

Revised version (v2.5.2) attached to the opening post.

 

Changes from 2.5.1 -> 2.5.2

 

- Added EDID dump using edid-decode by Adam Jackson. (Link to source in the opening post).

- Appended some info to the header of generateHTMLreport.sh script.

- Log now terminates after HTML report is complete and saved in actual dump folder.

Link to comment
Share on other sites

 Share

×
×
  • Create New...