service scanning filesystem

[tajcg]

(http://img441.imageshack.us/img441/93/tag2findae7.png)

Good evening,
I find that scanning is a long task.
Can you say some words about this operation. How we see everything is OK? How long (in time) is this operation?
Thank you.

[martin]

Hi,

initial scanning enumerates all directories on the drive. This is necessary for how the NTFS Change Journal works.

Subsequent scans occure on any change that are written to the NTFS change journal, so we can keep track of files. The notification is displayed if the service has been active for more than 1.5 seconds following changes in the NTFS change journal.

After booting up the PC and starting tag2find, it will also show "scanning filesystem", which is a misleading text which we already fixed for the upcoming 0.8.2 release: to ensure your tagged information is safe, we use two sorts of databases: one for each volume, storing in all relevant data in the backend, and one for searching on the frontend. The latter is built from the individual volume databases. It is being rebuilt on every start of the service for the current user. Depending on the number of tagged files this can take some time.

For the next 0.10 version we will specially address issues in the file tracking performance and in the building of the database, which is currently sub-optimal.

At the moment, this notification should be seen more as a "we are tracking file changes". It was primarely meant for indicating activity. Please feel free to deactivate the notice in the options dialog -> "advanced settings" tab.

Best regards,
Martin

[tajcg]

Good evening

I find that tag2find is very hungry in memory

(http://img441.imageshack.us/img441/7161/tag200701201520144zd.jpg)

Is not it?
Thank you

[martin]

Hi,

you are posting a screenshot of Process Explorer (I suppose), yet I'm missing which memory columns you are using. We are trying to keep the working set as low as possible, yet as many .NET applications we need some "Private Bytes". Could you tell me which memory performance counter your screenshot refers to?

I totally agree that we need to do a better job in trying to minimize our memory usage. We will concentrate on this area in the next releases.

Best regards,
Martin

[tajcg]

(http://img404.imageshack.us/img404/2407/tag2finf200701211927134vh.jpg)
Good evening,
Now the image gives the names
I think it's good
Have a good week
Good bye

[martin]

Hi,

thanks, that's what I have expected. Private Bytes is the amount of virtual memory used by the process, it does not really count for the real physical RAM being allocated (that's what the Working Set Size is for).

Still, you are right, it is higher than of native C++ processes, but that's part of the nature of .NET. We will try to optimize memory usage in the next releases.

Best regards,
Martin

[tajcg]

Good morning,

One can consider that tag2find aims at:
1) to know the path of a file,
2) to give a name or several names to these ways (tag1, tag2, tag3…)
3) perhaps, to allow the user to put notes on these files.

One can consider that tag2find aims at doing on a PC what the service “myweb” of Yahoo is doing  for Web pages. The realization of tag2find has thus a model on the Web.
 In addition, all the elements for tagging being on the PC there is a question : why to be connected to the Net? Why not make a traditional software?

Thank you

On peut considérer que tag2find a pour objectif :
1)   de connaître le chemin d’un fichier
2)   de donner un nom ou plusieurs à ces chemins (tag1, tag2, tag3…)
3)   peut-être de permettre à l’utilisateur de mettre des notes sur ces fichiers.

On peut considérer que tag2find vise à faire sur un PC ce que permet le service myweb pour des pages web.
La réalisation de tag2find peut donc trouver un modèle sur le web.
Par ailleurs, tous les éléments étant sur le PC est-il nécessaire de se connecter au net ? Pourquoi ne pas faire un logiciel classique ?
Merci

[martin]

Hi,

I'm not completely sure what you mean exactly. tag2find is a classical application, it is not connected to the web in any way. Could you please elaborate on your point a little further?

Best regards,
Martin

[tajcg]

Good afternoon,

Indeed , I have checked up : I was wrong, so the actual questions are :

1) it would be perhaps good to be able to add notes to a tagged file with tag2find
That is done by yahoo on "myweb" for each bookmarked page on the web


(http://img321.imageshack.us/img321/5967/yah1bn.png)

2) A second question : where is the "event log" of tag2find?

3) do you know if it is easy to find the path of a file which is opened on the the desk, for instance the path to a mail in Outlook, the path to a mail in Thunderbird, the path to a page opened in httrack,  ?
So I shall be able to write for each file I want (designed by its path) all the notes I wish in a file written with software Word.
If it's possible, I shall try to open the file with the address toolbar of Internet explorer or Firefox


Thank you

[martin]

1) it would be perhaps good to be able to add notes to a tagged file with tag2find
That is done by yahoo on "myweb" for each bookmarked page on the web

Thanks for the suggestion. We will look what we can do about this.

2) A second question : where is the "event log" of tag2find?

At the moment, tag2find has no event log. The error log is stored under "C:\Documents and Settings\All Users\Application Data\tag2find" for English installations (sorry, I don't know the path for a French installation).

3) do you know if it is easy to find the path of a file which is opened on the the desk, for instance the path to a mail in Outlook, the path to a mail in Thunderbird, the path to a page opened in httrack, ?
So I shall be able to write for each file I want (designed by its path) all the notes I wish in a file written with software Word.
If it's possible, I shall try to open the file with the address toolbar of Internet explorer or Firefox

Mails in Outlook have a virtual URL which is only correctly interpreted by Outlook itself. Maybe Outlook registers itself with Internet Explorer to be able to resolve such an URL. Sorry, but I'm not aware of how to retrieve such URLs from Outlook or Firefox.

Best regards,
Martin

[tajcg]

Good morning

I am trying to optimize the Windows Xp operation because sometimes I find my PC slow
Thank you to say what rules I have to follow to have a good operation for Tag2find but as light as possible.
Is it  possible to modify tag2findservice which is put with the rules : “ started” and “automatic”.
I have linked three pictures :
A/ The tasks (2 tasks : taf2find et TF2service)
http://img167.imageshack.us/img167/7012/tasktagos4.png

B/ The service (tag2find)
(http://img101.imageshack.us/img101/2443/tag2findservicezj6.png)

I think that it is possible to choose “manual”?
Do you think this action is important for optimisation?
C/ I have put tag2find in delayed  startings with Winpatrolplus
http://img102.imageshack.us/img102/2410/delayedtagcz1.png

Thank you

[martin]

Hi,

You will have to keep it at automatic if you want your files to be tracked. Otherwise copy, rename and delete operations will be missed and the tag database will get out-of-sync with the data on the disks.

Please also note that while tag2find itself is running it depends on a running service.

If you delay the startup about 3 minutes it is most likely that no entries from the NTFS change journal are missed and it will work. However, it will try to catch up all changes since startup (which are still recorded in NTFS change journal), which could cause a spike in CPU usage for several seconds if you start it delayed.

I'd advice you to try it out, you will not lose tags as long as you do not move files around in the time (and even if you do, tag2find will most likely find the entry still after several minutes). Of course delaying the start of the service will also delay the use of the frontend.

Regarding performance of the service, we will work on this in the next time.

Best regards,
Martin