AmigaOS 68k development - components, critics, bugs, work-arounds, tips&tricks

[Thomas Richter]

So, random bug - when selecting multiple devices devices on Workbench, and then format them from the icon menu, once the first one is formatted with "quick format", the "quick format" option for the remaining devices is grayed out, leaving only regular "format" and "cancel" as options.

Actually, not reproducable here. The logic between greying out and not greying out the QUICK option is quite simple: If there is a volume in the DosList that corresponds to the device to be formatted, it is not greyed out. Otherwise it is. Thus, the usual game: Remove all hacks and patches, and try again.

[kolla]

Actually, not reproducable here. The logic between greying out and not greying out the QUICK option is quite simple: If there is a volume in the DosList that corresponds to the device to be formatted, it is not greyed out. Otherwise it is. Thus, the usual game: Remove all hacks and patches, and try again.

No hacks, no patches. Booting the OS 3.1.4 Install floppy and plain 3.1.4 kickstart. Fresh hard drive - set up with 4 partitions in HDToolBox, installed "long filenames" (..307), maxtransfer, buffers... save, reboot with OS 3.1.4 Install floppy, select the 4 new partitions, icon menu, format... four format windows show up, all with "quick format" as option. Click "quick format" on either of them, and the remaining three windows refresh and their respective "quick format" buttons are grayed out. In genereal happens to the format window of all "Uninitialized" partitions. Happens on all systems I have tested on, be it "real", emulated or FPGA

But I am not at all surprised that you cannot reproduce - it is not the first time.

[Thomas Richter]

No hacks, no patches. Booting the OS 3.1.4 Install floppy and plain 3.1.4 kickstart. Fresh hard drive - set up with 4 partitions in HDToolBox, installed "long filenames" (..307), maxtransfer, buffers... save, reboot with OS 3.1.4 Install floppy, select the 4 new partitions, icon menu, format... four format windows show up, all with "quick format" as option. Click "quick format" on either of them, and the remaining three windows refresh and their respective "quick format" buttons are grayed out. In genereal happens to the format window of all "Uninitialized" partitions. Happens on all systems I have tested on, be it "real", emulated or FPGA

Oh, you mean you run multiple Formats in parallel. Yes, of course. This is expected behaviour. Once a disk is formatted, its volume node will vanish from the dos list. Hence, it looks unformatted at this time to all the other format instances. The "Quick" option comes back once the format is complete as it will generate a disk-inserted IDCMP, and the volume node will then become available again.

[kolla]

The "Quick" option comes back once the format is complete as it will generate a disk-inserted IDCMP, and the volume node will then become available again.

No, the "quick" option does _not_ come back once the format is complete - if it had come back, I would not have called this a bug. With 3.1.4, for "quick format" to come back, you have to close and reopen the Format program.

With OS 3.1, "quick format" never grays out, one can even do quickformat on many partitions at the same time.

[Thomas Richter]

With OS 3.1, "quick format" never grays out, one can even do quickformat on many partitions at the same time.

Which is curious since I have not changed the logic of Format.

[kolla]

With OS 3.1, "quick format" never grays out, one can even do quickformat on many partitions at the same time.

Which is curious since I have not changed the logic of Format.

Clearly something has changed.

And besides - why gray out only "quick format", and leave the rather more damaging "format" available?

[Thomas Richter]

And besides - why gray out only "quick format", and leave the rather more damaging "format" available?

The reasoning here is that if the volume is not present, but the device is, then the volume is not readable and requires a full format. Consider an MS-Dos formatted disk inserted, with a FFS mounted on top of the trackdisk.device. Quick format is not an option.

Besides, I cannot reproduce that "Quick" does not reappear after a successful format.I neither quite get why you want to open multiple format windows on the same device.

[kolla]

Good lord, why do you have such a hard time understanding??

One Format per partition - I want to format them, one after the other.

[kolla]

Partition up a disk with HDToolBox - two partitions.

Reboot.

They both show up as "unititialized".

Select both.

Select "Format" from the icon menu.

Notice how both Format windows, one for each partition, has "Quick format" as options.

Select "Quick format" on one of them.

Answer all confirmation requesters so the partition is formatted.

Let it finish - it is quick.

The partition is now formatted, volume icon shows up on Workbench.

Notice how _THE OTHER_ Format window, for _THE OTHER_ partition now has "Quick format" grayed out - EVEN AFTER FIRST FORMATTING IS FINISHED.

[my_pc_is_amiga]

Separate thing I've noticed in 3.1.4:

I have a rather large file: 2,287,031,227 bytes.  The size shows up correctly in the Information window and shell.  But in Workbench view by size, it shows up as ,007,936,069.

[Thomas Richter]

Note how hard it is to write good bug reports? You are still incomplete on this one.

If the other partition is invalid (i.e. has no volume name on it), then this is expected. This is, again, due to the rather simple logic by which Format checks whether quick formatting is possible or not. If there is no volume name, it assumes that quick format is not possible. This check is made when changing disks, and the disk change is triggered by the disk-inserted message of the first format finishing.

[kolla]

So the bug is there, “quick format” should be reappear as available option in the Format window of the second partition once formatting of first partition is done. Like you wrote earlier. Do you acknowledge that this is a bug, or do you write it off as “expected behavior”, even though it makes zero sense from a user’s point of view?

In general, I don’t have a hard time writing bug reports, but with you, and just you, it’s a damn fight every effin time.

[kolla]

If the other partition is invalid (i.e. has no volume name on it), then this is expected. This is, again, due to the rather simple logic by which Format checks whether quick formatting is possible or not. If there is no volume name, it assumes that quick format is not possible. This check is made when changing disks, and the disk change is triggered by the disk-inserted message of the first format finishing.

Why is Format doing this check _only_ when disks are changed, and not when Format is launched?
Should "uninitialized" be treated as a special case, a case where "quick format" should be made possible?
Why did this change from OS 3.1?

[kolla]

I have a rather large file: 2,287,031,227 bytes.  The size shows up correctly in the Information window and shell.  But in Workbench view by size, it shows up as ,007,936,069.

Yes.
And I would argue that Workbench windows can show file sizes as "human readable", for example "2GB" in you case.

[Thomas Richter]

Why is Format doing this check _only_ when disks are changed, and not when Format is launched?

I do not know exactly. The code has the gadget disable switch commented out, with a rather cryptic comment on top saying that it has been disabled, that the check through the presence of the volume is not ideal and that the code should instead check to access the disk directly. I *believe* the reason is that you should be given the option to quick-format a non-iniaitlized harddisk, but if you change disks, you are using a removable device where this logic does not apply and a non-present volume is instead the matter of a non-(hard-)formatted medium.

Should "uninitialized" be treated as a special case, a case where "quick format" should be made possible?
Why did this change from OS 3.1?

I do not see that this changed, the comment and this disabling is taken from the 3.1 release.