FD64-812-multisession USB -hosts file overwritten-Solved

[baldronicus]

Hi,

EDIT- please ignore this post. The problem was with me and not with Fatdog64. For explanation please see edit below. Thanks.

When attempting to modify a Fatdog64-812 set up, using multisession with a zero save interval, I have noticed that the /etc/hosts file seems to be automatically overwritten. Notes I had from Fatdog64-811 (I think) suggest that the first line in the file used to be overwritten on boot. However, it now seems to be the whole file.

The save seems to be working, as some adjustments that had been made to the eztables configuration file, at the same time, have been retained.

Of course, I have taken way too long to try this, and am as 'rusty as' with it. Hence, I have a suspicion that I might be just missing something simple that I had failed to note before. Indeed, I apologise for taking so long to try it, and for probably just wasting your time through my ignorance.

Thanks.

EDIT- Please ignore this post. As noted above I was the problem. Thinking on it, I have realised that I have posted a bit prematurely, as I haven't checked things properly, and I don't want to be creating a "Red Herring". I had added some complications in what I was attempting, and had made a few errors that I thought should have been corrected, but perhaps they weren't. I will try a simple test tomorrow to check on whether there does appear to be a problem (apart from me).
Thanks and apologies.

[Flash]

Do you mean multisession CD or DVD?

[baldronicus]

Hi @Flash . Please accept my apologies. I had tried to cover the info, but messed it up (as usual ).
The boot medium is a USB drive. With the multisession saves and the save interval set to zero, I can use it to boot and then remove it, only needing to insert it again if I decide to update something.

Thanks and, again, apologies.

[jamesbond]

@baldronicus, I don't think I changed anything on multisession code or hostname editing on 812; but it has been a while.

To confirm your problem: all the previous modifications you have done to /etc/hostname is cleaned out on the next boot, even after loading all the sessions? Can you manually look at all the session SFS files, and see if/when the hostname file gets cleaned out?

I will also check from here, but it will take a while for me do to so.

@Flash, Fatdog supports multisession both on DVD and on harddisk. In either case, the "session" data is written out as an SFS file; so every session is easily inspectable - as easy as looking into an SFS file.

[baldronicus]

Hi @jamesbond and @Flash . Please accept my apologies for wasting your time. The problem is one of operator error somewhere. I also should have edited my previous post, rather than just the first one, to note my suspicions regarding this.

Some checks done by changing the /etc/hosts file on some Fatdog64-812 setups including one on an SSD that is using a savefile, another USB drive with straight multisession (save interval not set to zero) and the same USB drive with multisession with save interval set to zero (after having deleted the multisession files from the earlier test), all worked fine.

As I had started to suspect, I have managed to do something wrong somewhere with respect to the USB drive that presents the problem. I was complicating things, and am not entirely sure of what I had done. So I think I will start again from scratch.

Again, I apologise for posting too quickly, for not editing the previous message, and for the inconvenience.

Jamesbond, I hope you haven't gone far with checking things. I had hoped I would intercept any action until I had checked further.

For info only, the new host name in /etc/hostname was as I had changed it when setting up the save. I had been attempting to use a modified /etc/hosts file. I think my poor wording regarding some notes I had made in the sample hosts file (from Fatdog64-811, I think) might have caused confusion. I was referring to a suspicion that the first line of the /etc/hosts file might be automatically overwritten with the hostname from the /etc/hostname file. However, I cannot recall where I might have got that idea from, so I could easily have it wrong.
In an earlier "episode", that I thought I had cleared, I had copied the sample hosts file that I use to /etc, but with the wrong owner and permissions (even though I had a note to check them, D'oh). I thought I had cleared that, but maybe I hadn't gone far enough. I will try checking the session sfs files as you suggest, just out of curiosity, but I think a start from scratch, and then try to stay alert (ha, ha), and keep things tidy, is the only way that I am personally going to be confident that I have done things right.

Thanks and apologies again from the Fool In The Hollow.

Further info- I checked the multisession sfs files, as suggested, and the updated (from sample) /etc/hosts file appeared to be present. However, I noticed that the names of the later sfs files were missing "fd64save" in the names.

Which makes me think that I have been trying to be too clever and too greedy for my own good, which, through my stupidity, has unfortunately impacted others.

Previously I had placed the multisession saves in partitions. This time I thought I would try putting them in directories in the one partition (I set up different sets of multisession saves for different internet purposes). To add further complication, I copied the (uncombined) save files from one "set" as a base for another. In other words I think I have been doing my best to set myself up for a fall, particularly since I wasn't really alert when trying to do this. However, since everything had seemed to be working until I came up against the /etc/hosts file problem, I thought it was all right, and reacted the wrong way by not checking further before posting.

Hence, being dull and stupid, I posted too early, without checking whether the problem might be related to the "experimental" nature of what I was attempting (which really irks me, since I should know better, but keep doing this sort of thing).

It doesn't help, but I cannot apologise enough. Thanks.

[jamesbond]

@baldronicus, it's all right. We're bound to do experiments from time to time, sometimes successful, and sometimes not.

And none of us in the Fatdog team are perfect either; as much as we want our creation to be bug-free, by testing and using changes before we publish it for others; we're not perfect too and mistakes can and do happen from time to time.

In fact, we're glad to hear any feedback, bug reports or otherwise; because more often than not we're blind to our own mistakes. If we can fix it, then that's great; if we can't fix it, then by reporting it here, perhaps there are other, smarter people who can fix it, and we are all the better for it.

No harm is done, no offense is taken, all is good with the world and always nice to hear from people who find Fatdog useful.