Suggestion/Fix for HRZR entries

buggage

While using the lastes version I had an idea for the HRZR entries that (if possible) should make this a more powerful/safer feature for everyone to use:

For Rot13 entries where the entry names start with HRZR_EHACNGU (UEME_RUNPATH) and HRZR_EHACVQY (UEME_RUNPIDL) why not build in a Rot13 decryptor that can check whether the data actually points to a valid location or not, and then either skip or delete the entry based on this information. I noticed that some of these entries which are deemed useless actually point to real files/locations, while others do not. MRU data or not, Whether or not these are used, the app shouldn't delete entries that are actually valid. I believe this would also solve the issue of wiping the frequently used apps list fromt the start menu. Thanks.

A simple online Rot13 decryptor can be found here: http://www.arantius.com/tools/rot13.php

EDIT: I see that this was already breifly brought up in the topic here: http://www.macecraft.com/phpBB2/viewtopic.php?t=1188 and also states that the app is only supposed to delete those entries that are invalid, though doesn't seem as that is the case yet.

tullik

buggage wrote
... that can check whether the data actually points to a valid location or not, ...

PT (1.5.2.344) already does the decryption and check for existence even if you don't check the MRU scan (just tried it on this Norwegian system which has never seen PT before ...)

... I noticed that some of these entries which are deemed useless actually point to real files/locations, while others do not. MRU data or not, Whether or not these are used, the app shouldn't delete entries that are actually valid. ...

Trouble is a system stable for 2 or 3 years with turnover of many applications can have 10,000s of these entries ...

... I believe this would also solve the issue of wiping the frequently used apps list fromt the start menu. ...

With 344, it is now a user choice, so the Personalized Menus in the Classic Start Menu is also less of an issue.

buggage

tullik wrote
buggage wrote ... that can check whether the data actually points to a valid location or not, ...

PT (1.5.2.344) already does the decryption and check for existence even if you don't check the MRU scan (just tried it on this Norwegian system which has never seen PT before ...)

Sorry, but I have to disagree with this statement. Tried it here using .334 and it does indeed list valid entries for removal - Actually it flags ALL of these entries as bad, so it would appear that no decrypting/checking is actually going on. Mind you it's not a huge deal, but why remove valid items at all? Shouldn't the registry cleaner only touch non-valid items?

tullik wrote
buggage wrote ... I noticed that some of these entries which are deemed useless actually point to real files/locations, while others do not. MRU data or not, Whether or not these are used, the app shouldn't delete entries that are actually valid. ...

Trouble is a system stable for 2 or 3 years with turnover of many applications can have 10,000s of these entries ...

Right, so again, why not only delte the non-valid items? Surely the majority of them will turn out to be non-valid anyway. What's the trouble with keeping the left over valid entries?

tullik wrote
buggage wrote ... I believe this would also solve the issue of wiping the frequently used apps list fromt the start menu. ...

With 344, it is now a user choice, so the Personalized Menus in the Classic Start Menu is also less of an issue.

I understand it's a user choice - and that's fine, but there's no need to disable it by default, if it only cleans the bad entries, right?

buggage

Well I suppose I see that this is by design, in the fact that ALL MRU items are flagged invalid, even the "regular" entries. I guess what may be something to think about might be a third option to the MRU cleaning, so where you now have this:

Clean all items

Do not clean

You might have:

Clean all items

Clean only invalid/nonexistant items

Do not clean

tullik

buggage wrote
tullik wrotePT (1.5.2.344) already does the decryption and check for existence even if you don't check the MRU scan (just tried it on this Norwegian system which has never seen PT before ...)

Sorry, but I have to disagree with this statement. Tried it here using .334 and it does indeed list valid entries for removal - Actually it flags ALL of these entries as bad, so it would appear that no decrypting/checking is actually going on. Mind you it's not a huge deal, but why remove valid items at all? Shouldn't the registry cleaner only touch non-valid items?

That was .334 ... jv16 changed the behavior in .342 (& .344 of course) ... I don't mind having an option to remove what (in my usage) is useless debris.

With .344, I got 22 "File or directory "path" does not exist" type HRZR_ errors with MRU scan not checked. With it checked, I got 369 additional items called "Useless registry entry."

buggage wrote
tullik wrote
buggage wrote ... I noticed that some of these entries which are deemed useless actually point to real files/locations, while others do not. MRU data or not, Whether or not these are used, the app shouldn't delete entries that are actually valid. ...

Trouble is a system stable for 2 or 3 years with turnover of many applications can have 10,000s of these entries ...

Right, so again, why not only delte the non-valid items? Surely the majority of them will turn out to be non-valid anyway. What's the trouble with keeping the left over valid entries?

I ran for about a month with UserAssist encryption turned off. Wound up with about 400 entries clear text readable -- "valid" but useless to me. Those who reported 1000s & 10000s of entries did not do a frequency analysis by type of "error" -- would have been interesting if they had ...

buggage wrote
tullik wrote
buggage wrote ... I believe this would also solve the issue of wiping the frequently used apps list fromt the start menu. ...

With 344, it is now a user choice, so the Personalized Menus in the Classic Start Menu is also less of an issue.

I understand it's a user choice - and that's fine, but there's no need to disable it by default, if it only cleans the bad entries, right?

PT Reg Cleaner picks out the "bad" MRU and UserAssist entries whether MRU scan is checked or not. Having it off by default causes the "valid" entries not to show as "errors" -- if a user wants to see the "valid" but probably useless items, the choice is there. If these items were not shown as removal choices, the entire MRU scan option could be removed.

tullik

Sorry, we encountered an error while displaying this content. If you're a user, please try again later. If you're an administrator, take a look in your Flarum log files for more information.

erronius

tullik wrote
Trouble is a system stable for 2 or 3 years with turnover of many applications can have 10,000s of these entries ...

I recently upgraded to PT2006 (1.5.2.352) and went from about 100 entries a week to 6000(!) with the first scan of the 2006 version. Almost all of the new ones were the HRZR entries which I had never before seen or realized even existed. I use Classic without personalized menus so deleting them wasn't a problem. I'm glad I got rid of them - that's a good chunk.

My point is that without the registry cleaner setup the way it is I may never have found out about them.

Regards,

Erronius