/hydrus/ - Hydrus Network

I am very excited to hear about the IPFS plans, I look forward to retrieving and sharing content via IPFS, I much prefer the idea of that over a centralized file repo.

Thanks as always Hydrus.

>>1783

>>1787

Adding OR searching is definitely something I would like to do. It was one of the choices on the large poll I had up a couple of weeks ago and came fifth, so it is quite popular. Once I am finished with the current larger project priorities of IPFS, suggested tags control, and faster dupe searching, then please vote for OR as your favourite thing for me to next work on.

>>1798

>>1808

Sorry about the delay in response here. My laptop has ~280k files and took about 12 secs, so I expect your db updated in less than a minute but then the init hung for some reason.

If the client is working in v190, then you are now ok. Unless there is a huge amount of stuff going on, and I warn an update might take an hour or so, DB updates are always atomic, which means they either succeed completely or fail with no changes, which also means you can kill the program half way through an update and it shouldn't break the db. (In this case, the db will try again the next time it boots).

Since you are a special user, odd problems are unfortunately more likely for you. If you encounter something like this in future and the program doesn't seem to be doing anything for too long, force-killing the program and restarting is probably the prescription for now. In this case, I would be interested in:

Is there any odd stuff at the bottom of your install_dir/logs/client.log file? (For instance, you might have 30MB of 'tried to start gui but failed'-type errors that somehow looped)

How big is your install_dir/db/client.db-wal file, if it exists? (In rare cases, especially with complicated dbs, this file can grow to like 17GB before the transaction completes–I'll be turning this journalling mode off for db updates starting next week, at least until sqlite devs fix it)

In your current client.log, what does it say about the update boot attempt? Did it try to boot the gui, and what are the timestamps for those steps? I have this:

2016/01/13 15:28:21: hydrus client started
2016/01/13 15:28:21: booting controller...
2016/01/13 15:28:21: booting db...
2016/01/13 15:28:23: updating db to v190
2016/01/13 15:28:23: updating file tables
2016/01/13 15:28:25: updating db to v190
2016/01/13 15:28:35: booting gui...
2016/01/13 15:28:37: The client has updated to version 190!

>>1803

It seems the startup/shutdown code isn't causing it. I made a special build here:

http://www.mediafire.com/download/dxkvm4ny29z3ab7/Hydrus_Network_v190_with_no-daemons_-_Windows_-_Extract_only.zip

If you launch that with the switch '-no-daemons', i.e. by opening the command line (shift+right-click on an explorer window looking at your install_dir->open command window here) and typing 'client -no-daemons', do you get the BSOD? I've been talking about it at >>1782 .

>>1809

>Is there any odd stuff at the bottom of your install_dir/logs/client.log file

I can't find anything related to v190 update in the log file. I tried searching for "updating db to v190", but no matches are found, yet I found matches for v189,188, so on. Other than that, the log seems fine.

>How big is your install_dir/db/client.db-wal file

6mb, so I'm assuming this is fine too

>In your current client.log, what does it say about the update boot attempt?

Yeah I don't have any of this in my log, yet it shows I'm on v190. I'll try doing a clean update after I get back.

>>1813

Since you client is working now, I will for now assume this problem was something rare and unusual. If your client hangs like this on the v191 or any subsequent update step, let me know, and we can explore this further with debug builds or whatever.