• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

* HFM.NET - Client Monitoring Application for Folding@Home *

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Counting pauses as a fail is not correct, I'm aware of it. ;) The reason this happening goes back to v6, where core returns of INTERRUPTED are considered a failure.

The log parsing code for v7 is the log parsing code for v6, with a few overrides. The fix for this issue is to simply change the way the completed and failed counts are derived. In fact, the solution I want to implement removes taking these values from the log file all together and instead rely on the values in the WU History Database. This way you could see a must more historical count or a count from the last time you start the client.

I was working on database upgrades and need to get back to it. Once those are done fixing this is one of the next things on my list.

hey harlam, awesome :attn:
 
Thanks ZL1! Worked on things a bit tonight. I've got LIKE and NOT LIKE query functionality working. Also, I solved the problem with dates not showing correctly when viewing the history entries from a v7 client. The problem stemmed from the upgrade to SQLite in the last version. UTC date/time values are now written with a "Z" (zulu) ack tacked onto the end of the date/time.

Previously this didn't happen and I read the values back in as is... and my code knew they were UTC. However, the newer entries written with the updated version of SQLite knew the entries were UTC also... but then assumed that I wanted the entries to be translated to local time... which I don't. So, there is a simple SQLite connection string switch that can be added to force SQLite to assume all dates are to be returned as UTC, whether they have the "Z" (zulu) designation or not.

The nice thing about this fix is that all the data is in tact as is. It's mostly just a presentation issue. Also, this will fix the issue with duplicate entries being written to the database... and I'll clean those duplicates up when the new version is run and the database is upgraded. So everyone should be back to having and seeing a reliable history list.
 
Been watching your updates on the google code page and I too am looking forward to your next release. Thanks for this awesome app and the tons of time that I know you continue to put into it! :salute:
 
Hey taz... thanks so much! I'm trying to work on it every chance I get. Trying to minimize to the changes that will give you guys the most bang for the buck. I hope I can give you guys something soon. An update is well overdue.
 
hey harlam any way we can contribute ? :) other than folding that is :)

btw my botnet :D
(sadly a single i7 can do more or same, but oh well maybe someday)
 

Attachments

  • Capture12.JPG
    Capture12.JPG
    113.6 KB · Views: 229
Update or not I'm still glad to be using the program on my comps. :comp:

Don't mind us folding addicts lurking in the shadows (tightening tourniquets while pumping for veins) for waiting for the new fix, I know I can wait for it to be released when all is good. ;)
 
hey harlam any way we can contribute ? :) other than folding that is :)

Download the source code and have at it! I gladly accept patches. :) Even if it's as simple as adding a detection for a new core or return code... that's one less thing I have to do.

I know svn isn't the most friendly source control for those that want to contribute. I've considered moving HFM to a git repo... and lookie there, google supports that transition basically out of the box. Maybe this is something I'll tackle after I get v0.9.2 out the door. Should make it a lot easier for folks to play with the innards of HFM and contribute. :)

Update or not I'm still glad to be using the program on my comps. :comp:

Don't mind us folding addicts lurking in the shadows (tightening tourniquets while pumping for veins) for waiting for the new fix, I know I can wait for it to be released when all is good. ;)

Thanks taz! I certainly appreciate all the support HFM gets from you guys in the community! :salute:
 
HFM.NET is great. Thank you for developing this.

Capture.PNG

Is there any way you could add a lower client refresh rate? I would like the option of updating my clients every few seconds, if possible.
 
New Version: 0.9.9 - Revision 789 - Project Site - Download

Change: Log viewer and message viewer font changed to Consolas 10pt. The mono-space font makes reading the log much better on the eyes.

Change: Auto updates source location. HFM will auto update from v0.9.8 for a short time. But it will begin to fail soon. I've maintained some paid web hosting over the years that I intended to do more with. At this point I've decided to cut my losses there as there are many more options now. The new update location should do well for the foreseeable future.

Change: Lots of things behind the scenes. I won't bore you with the details. These are all things I've been working on for the last few years, as time has allowed, to make the application leaner and meaner. Again, users won't really care, unless I created a bug.

Notes: I hope those of you using the project enjoy. If you have an issue please post in the Google Group or add a new Issue here on GitHub. I'll do my best to respond as quickly as possible. I'm working my way through the Issues here on GitHub to clean them up, some of them are very old and were imported from Google Code.

---

Need help with HFM? Have a feature request? Please join the discussions. Post in the HFM Google Group.


http://groups.google.com/group/hfm-net
 
New Version: 0.9.10 - Revision 811 - Project Site - Download

- Fix: Don't allow config.xml read errors to stop HFM from loading. Load default values if this situation occurs.
- Fix (Issue 300): Web Generation - include ClientVersion, TotalRunCompletedUnits, TotalCompletedUnits, TotalRunFailedUnits, and TotalFailedUnits in SlotData structure in web xml data.
- Fix: Web Generation - format production (PPD, UPD, Credit) numbers that were not being formatted according to the user preferences.

- Change: UI - failed unit count now toggles between "current run" and "total" just like completed unit count.
- Change: v7 client interface - allowing CPU and OS values from v7 client to fully flow into the queue viewer. Was previously doing some checks and parsing on this data which required code updates for each new CPU or OS. This could result in some failed detections when the application isn't updated to detect a new CPU or OS. That's too much maintenance, so the values reported by the client are now used verbatim.

---

Need help with HFM? Have a feature request? Please join the discussions. Post in the HFM Google Group.


http://groups.google.com/group/hfm-net
 
Back