News, tips, advice, support for Windows, Office, PCs & more
Home icon Home icon Home icon Email icon RSS icon

We're community supported and proud of it!

  • PaulK

    Forum Replies Created

    Viewing 15 posts - 1 through 15 (of 409 total)
    • Author
      Posts
    • in reply to: Wanting to find duplicate files #2363230
      PaulK
      AskWoody Lounger

      Finally, I changed the output file to write to the root of the C: drive as that should be out of the way.

      (emphasis added)

      Great idea!
      But there may be a problem here, by a Standard user, to write to C:\ .

      On Windows 7, and I presume on Windows 10 too, one can
      – Make a Directory, MD  –  (New folder)
      – Copy an existing file there, either
      — from elsewhere, or that was originally there
      but one cannot [ “Access denied” ]
      – directly create a file there
      – Save[As] there a file that was already there, even if unchanged

      Two workarounds that I’ve tested are:
      – Standard user
      — MD a directory
      — Save to, or Copy to, this directory
      — ‘Run as administrator’ the program that is to write to C:\
      – Administrative user – has no restriction

      So, to use Mr. Geek’s Script, I suggest that one
      – first, create a folder there (such as: MD C:\Comps )
      – modify part of the last line of the script:
      — old: C:\CompareFiles.txt
      — new: C:\Comps\CompareFiles.txt

      Edit addition:
      I’m always a bit nervous when working directly with C:-root.
      Crossed fingers for luck are acceptable, but crossed hands can lead to a disaster.
      It is far too easy to bypass the UAC caution when one is focusing on something else.

      in reply to: Misbehaving Mouse #2363063
      PaulK
      AskWoody Lounger

      Do the right-button and the scroll wheel both work OK?
      If so, perhaps there is contamination inside associated with the left button.

      But this doesn’t explain why tapping the Tab key seems to help.
      Is it only the Tab key taps, or other keys too? Only one tap doesn’t help?

      Try vigorously shaking the mouse, slapping it on the desk (not too hard).
      If these help, you’ll probably need to open the case and look carefully.

      in reply to: Wanting to find duplicate files #2363062
      PaulK
      AskWoody Lounger

      How about doing this in two ‘batches’?
      One of them is the original – deleting all the ‘space-dot’ names.

      The other run: Make this change: FIND ” .NET”
      NOTE: I just tested, and the content within the ” “s IS case respected.
      Thus: ” .NET” does not find ” .net”, and vice versa.
      Also, if there is mixed case in some names, one will have to be careful to match it.
      However, IF the word ‘Framework’ is always present: FIND “Framework” ?
      Again, verify the precise case of the spelling of the word.
      And if there are other folders with ‘Framework’ as part of the name … collateral damage.

      Note also that finds may be chained: FIND … | FIND … | FIND … . This is useful if one is not interested in saving the results of each individual stage of filtering.

      Now, as a bonus. If you DO have ‘intermediate’ files (one run w/o periods; one run for .NETs), you can combine them back into one file. Look at the syntax of COPY /? .
      COPY filein2 + filein1 fileout
      The spaces around the plus sign are not required.
      So [ COPY filein2+filein1 fileout ] is the same. User’s choice for readability.
      (I’ve shown filein2 before filein1, assuming that ‘2’ is the .NETs, which is where they would naturally sort to if they hadn’t been dropped in the first run.

      in reply to: Wanting to find duplicate files #2363023
      PaulK
      AskWoody Lounger

      Excellent observations.
      1 – Concerning the CaSe of the parts of a file-id.
      In old DOS, everything was upper case. (Historically this was because old (pre big/main frame) computers used a limited character set that consisted only of ‘capital’ letters, plus digits and some symbols. As time and technology and user-friendliness progressed, character sets expanded to include lower-case letters and more characters. And now, more character sets also support other ‘non-Latin-character’ characters. But we’re on the edge of the topic now.)

      When one keys a letter into a command that results in an output name, the case of that letter is honored in the result. But, Windows ignores the case when using it to locate a file. Note, however, that case IS respected when it is pertinent. (Other operating systems DO respect case in file ids, so one need to be fluent in them — and give them respect.)

      2. In the Find command, why is [ FIND /V ” .” ] included? A question which leads to another dissertation (or distraction?). Look at this sequence. Explanations follow.
      periods

      Perhaps you have noticed that the ‘raw’ result of a DIR always has two lines at the top: one has a name of just one period, the other has two periods. These have a special meaning when used in a command that includes a path specification.

      Two periods = back up one level in the path; that is, to ‘my’ parent directory.
      One period = this directory where I am now.

      For your screening you didn’t need to see hundreds of these redundant repeated lines, so you eliminated them with that Find.

      (The [ CD . ] above is only an illustration of what its effect is, and is not useful as-is. However it is used in some batch command applications.)

      To answer your question (thought I’d never get around to it?) – Yes, the dot-NET lines WERE filtered out because of this specification. Candidly, I didn’t think of them. You might then ask – why not change the find to [ “. ” ], that is, period+space? The problem here is that, for those lines, the last period is the end of the line: there is no trailing space IN THE RECORD. On the screen we see blanks, but the actual line has already ended. Try it.
      = = = = =
      The replies by others (concerning the limitations here) are quite correct. That is why

      ‘True duplicates’ likely will have identical file dates/times and sizes; these are good discriminatory indications.

      was stated in paragraph 5.
      = = = = =
      To answer

      One question, though: after the list of possible attributes, the information says “prefix meaning not”. Can you explain how that works with an example??

      You already used that. In your Find, the [ /v ] said: Find all lines that do-not have a [ space+immediate-period ]. This dropped all the lines discussed above. Were you to do a Find WITHOUT the /v, the result would eliminate everything except those (repeated) two lines for each folder. Try it. In any folder: [ dir /s | find ” .” ]. Hint: to abort the hundreds or thousands of lines scrolling down your monitor, press [ Ctrl+C ].

      Attachments:
      in reply to: Wanting to find duplicate files #2362757
      PaulK
      AskWoody Lounger

      Do you have one to recommend?

      in reply to: Wanting to find duplicate files #2362748
      PaulK
      AskWoody Lounger

      Congratulations, you are doing famously well.

      Background info:
      You have been using the ‘pipe’ operator [ | ]. In context, it means “take the output of the operation that precedes it, and pass it to the the operation that follows”.
      Another operator that is very useful is: “take that which precedes me, and put it in
      the location that I specify next”. This is the redirection operator: [ > ].
      And another command that is useful is TYPE. It “types out” (by default to the console, your display) a file. But the default destination can be changed by piping or redirecting.

      Will this work?
      FindDIR

      The output sequence is not changed; only a[nother] Sort does that.

      NB – For some reason in these AW postings I cannot insert a string with the word DIR surrounded by the double-quotation symbols and angle brackets (less-than and greater-than) symbols. The result is a broken mess. Hence I’ve had to insert a png file of the desired line.
      How did YOU manage to post it above: paragraph 2, line 5?

      Attachments:
      in reply to: Wanting to find duplicate files #2362706
      PaulK
      AskWoody Lounger

      Good questions.
      0 – 4K FILES? You’re probably pushing the limits of patience. But it is worth a try just for the educational experience.

      1 – Just plain [ /A ] means: include all entities, regardless of their attributes.
      If one specifies something, for instance /AR, then ONLY Read-only records are listed.
      Or, /AH, then ONLY Hidden files are listed.
      Omitting /A altogether will list only those items that are NOT: Hidden, Read-Only, etc.

      2. Column 40 is the fixed location where, in a Directory Listing, the names (of Files, Directories) start.

      3. It is not documented well. This appears to be an indication to the Sort program as to the maximum record length to be expected, in order to set up internal variables. The default is 4096. When I was experimenting for this, I had set 200 for one run. But the Sort aborted because one record (” Directory of C:\Users\Paul\AppData\etc.” ) was 221 characters long. You most likely can set your number much lower; I listed 400 just to be safe. The SORT command can sort any records-length file, up to 64K-1.

      This Sort command originated away back in the early DOS days, and apparently has not been enhanced much since then in flexibility. Records are sorted beginning at the /REC-number column, through the rest of the length of each record. (Other programs and applications support the specification of multiple sets of starting+ending column numbers upon which to sort.)

      4. The way that DOS and Windows Commands work is that the default location of a file is in the “current directory”. Hence, if I am in C:\Users\Paul\Documents, then – unless I otherwise specify a Path designation – any entity that I create will be there too.

      So, yes, when you are through with this exercise you’ll want to go back to the directory (or directories) where you were, and purge these working files.

      5. Correct. This will group all duplicate file names together. Eyeball+brain are needed to determine what to do with these. The CTRL+F applies within Explorer windows in order to make the determination that ‘these files are true duplicates, and one needs to go’ and ‘these files are valid duplicate names, but different valid folders, and need to be retained’. ‘True duplicates’ likely will have identical file dates/times and sizes; these are good discriminatory indications.

      6. No. ‘What you see (in a DIR listing) is what you get’. One advanced editing program that I have DOES include the full path information. And it supports multiple sort specifications.

      The method outlined here – DIR/FIND, SORT – is a brute force attack, best limited to relatively small sets of names to be reviewed. It is cheap and easy.

      If you want to experiment with sorting on other fields, here are the column numbers:
      1 – Month part of Date
      4 – Date (in month) of Date
      7 – Year part of Date
      – Note that the above assumes the format MM/DD/YYYY.
      10 – time
      19 – AM/PM
      25 – DIR, if present (There are angle brackets too, but they corrupt this entry if I show them.)
      30 – file size
      40 – file name
      But remember, a sort starts at the specified column, and continues all the way through the end of each record.

      1 user thanked author for this post.
      in reply to: Wanting to find duplicate files #2362586
      PaulK
      AskWoody Lounger

      How many folders and files do you have to check?

      My \Users\Paul has >43K Files and >7500 Folders.
      But the majority of these are in AppData: >24K Files and >6600 Folders.

      My \Users\Paul\Documents has about 2K Files and 195 Folders.
      (The \Users\Public\Documents is larger.)

      How comfortable are you with using the Command Prompt?

      The following is slightly cumbersome, but may be worth trying.

      Key in each of the following in a Command Prompt.
      (You may type either in CAPS or in lower-case; I use CAPS here for clarity.)

      To review the syntax for these commands:
      FIND /?
      SORT /?

      Navigate to the highest level directory (folder) that you need to check.
      e.g.: CD C:\Users\WCHS\Documents
      – Adjust as needed; substitute the correct folder name for WCHS.

      Extract the names of the folders and files. And sort them.
      DIR /A /S *.* | FIND /V ” .” > WCHS1.TXT
      SORT /+40 /REC 400 WCHS1.TXT /O WCHS2.TXT
      Notes:
      1) If you are doing a Copy/Paste from here, ensure that the quotation marks are ‘straight’ not smart/curly. This has been a problem in the past, but may not be now.
      2) Careful: the string following /V is exactly “spaceperiod”

      Now, using two copies of an editor – Notepad is fine – Open both WCHS1.TXT and WCHS2.TXT.
      In WCHS2, visually scan for duplicates. Folder names are intermixed with File names.

      For each duplicate in WCHS2, in WCHS1 do an Edit>Find (Ctrl+F) for the name of interest.
      Find all duplicates in WCHS1, and deal appropriately.
      You may wish also to have (an) Explorer window(s) open to handle the deletions.
      I suggest that Deletions be to the Recycle Bin, not immediate deletes, just-in-case.

      in reply to: Firefox vs 20H2 #2360106
      PaulK
      AskWoody Lounger

      As krism asks: what’s broke?

      If Firefox itself does initiate, but the default Profile fails, create a new profile to test with:
      "C:\Program Files\Mozilla Firefox\firefox.exe" -P

      Or start in Safe Mode:
      "C:\Program Files\Mozilla Firefox\firefox.exe" -safe-mode

      When you copy the lines above, be sure to change the ‘curly-quotes’ to ‘straight-quotes’. (Postings here corrupt the characters.)

      If Safe Mode works, suspect a problem with an Add-On to Firefox.

      Edit addition: If FF itself doesn’t come up, check that (the newest) Windows hasn’t futzed with the Default Programs and their authorized Accesses. (Courtesy of the new Edge, perhaps?)

      [Moderator edit] fixed your quotes, copy / paste will be fine.

      2 users thanked author for this post.
      in reply to: mail sent from my computer to my server #2360103
      PaulK
      AskWoody Lounger

      Another needle in the grass.

      In a Command Prompt: RESMON -or –
      Task Manager > (tab) Performance > Resource Monitor

      In Resmon:
      In tab Overview, see the Network panel -or-
      Tab Network.

      Perhaps you can spot something that flits in and out periodically.

      in reply to: mail sent from my computer to my server #2360064
      PaulK
      AskWoody Lounger

      Maybe help, maybe not –
      Take a stroll through the Task Scheduler forest? Is there something triggered every 5 minutes?

      1 user thanked author for this post.
      in reply to: My first OEM Linux PC (laptop)! #2356175
      PaulK
      AskWoody Lounger

      Rhetorical snipe from a Win7 user, to Ascaris and Oscar:
      Which is worse for those of us PC users who grew up on DOS and Windows < 8:
      – KUWW10 – Keeping Up With Windows 10; or
      – Wrestling with the vagaries of all the different generations and distributions of Linux?

      1 user thanked author for this post.
      in reply to: Testing if I could post a thread #2356171
      PaulK
      AskWoody Lounger

      As GoneToPlaid notes just above, it is recommended always to create your post material in the ‘Text’ mode; this ensures that any embedded control characters (e.g., HTML, but others exist) are visible to you and can be removed. Then click on ‘Visual’ to see (about) how the post will appear. ‘About’ because line breaks are not consistent. Then switching back to Text mode will show any BBCodes that have been inserted by the editor; these are OK and may be left alone.

      in reply to: Testing if I could post a thread #2356138
      PaulK
      AskWoody Lounger

      I’ve sometimes had a problem getting an entry actually to be posted when there is an included link. I don’t know what the filtering rationale are that trigger a rejection.

      So you may try a posting again, but break up a link into pieces: insert spaces, for instance. Human readers easily can reconstruct the string if necessary.

      in reply to: Does it exist? #2354522
      PaulK
      AskWoody Lounger

      I wonder if the reason that ‘it doesn’t exist’ has to do with USB signaling protocols? We know that ethernet cables cannot be just connected in series or parallel. Is there the same issue with USB – the Data connections? Comments from anyone who has knowledge of this?

    Viewing 15 posts - 1 through 15 (of 409 total)