v. 5.3

• leech emails, web links, numbers and alphanumericals (phone numbers, dates, ZIPs, program serials) and HTML page resources (SRC/BACKGROUND/HREF links) + optionally turn system files back into normal files (eg: folders' hidden album pictures, which cannot be un-hidden, and whenever they're copied you're asked a boring prompt of confirmation)
• report favorites, cookies, PC+internet history, recent docs and recent programs
• re-save previously sieved data, without having to re-sieve
• sieving of OpenOffice documents "vs." Microsoft Office's
• simpler "newbie-friendly" interface, with separate drawer for power users
• lots of bugfixes and improvements, plus two shortcuts (try doubleclicking the button-labels "help" and "settings" ;-)

v. 5.2

• off-line help now included in the installation .ZIP, plus the missing DLL for Vista and the latest releases of XP
  The VisualBasic runtime (msvbvm50.dll) is normally already available in previous versions of Windows, so it can be safely deleted.
• recognizes OpenOffice documents (ODT ODP ODB ODS ODG), plus ASHX files (HTML/Java)
• optional quick-erasing of Internet Explorer cache
• image recognition is now faster
• a few bugfixes

v. 5.1

• new shell-extension for extracting embeddings from any file
  eg: right-click on a powerpoint and get all the images/sounds it contains! (also recognizes filetype, in case of unknown/wrong extension)
• drag'n'drop extra functions: the enhanced version of an old Sieve/WinSieve tradition
  • trash duplicates across multiple files/folders
  • batch-extract file embeddings: why keep tons of PowerPoints, when all you care for is just an image contained there?
  • regroup files and/or folders: just drag'n'drop your scattered data, and WinSieve will be your shepherd dog :-)
  • sieve custom data: extract valuable data from anywhere on your PC, and have it organized WinSieve's way!
• recognize Office documents (DOC PPS PPT XLS MDB) and ASX
• sieve more Windows media videos (ASF and ASX)
• faster file scanning (on large volumes, goes by 20)
• faster sorting (required to find duplicates and process exclusions)
• automatically skip locked files (files in use by the system might crash any program)
• disable exclusions with one click (instead of erasing the paths)
• ..and a bunch of bugfixes (the disabling of shell-extensions, better parameters-passing on right-click, no more messy multiple instances or duplicate definitions on Thumbnails and Junk-pics, and no more stack-overflow when processing huge amounts of data)
• also, a stricter nag-system was implemented in order to encourage donations - yet always guaranteeing free sieving for everyone

v. 5 (what's new since the last "DOS shell" version)

• brand new algorythm since the first Sieve: by sieving in memory, and an "a la carte" selection of file types only, the process gets 4.3 times faster!
  The basic idea here is: "instead of sieving and then deleting things I don't care for, just get me what I want". You save time on sieving, and also when going over the found stuff.
• multi-browser: extract from the cache of Internet Explorer, FireFox, Opera, Safari (with LiteX plug-in) and Chrome (with ChromeCacheView plug-in)
• 100% self-configuring: finds the paths to the Internet cache, pre-sets the output folder and the exclusions
• a versatile progressbar, telling how much time is left before the current operation is over, a percentage of completion (morphing through traffic-lights colours, from red = busy to green = done), and some extra informations (the current path examined, how many duplicate files were found, etc.)
• seek duplicate files by right-clicking on any folder: they'll be safely moved to Recycle, so you can restore them "just in case"
• identifies lots more filetypes: CUR, ANI, RMI, CDR (coreldraw), ASF (windows media), LNK (Windows shortcut), MPG (MPEG video), HLP (file help), URL (internet link), 3DS (3D studio), PSD (photoshop image), AI (adobe illustrator), PST (Office Outlook personal folder), CAB (cabinet file), CHM (compiled help file), DWG (autocad drawing), TTF (TrueType font), DXR, WRI (writer), CLP, PSP (PaintShopPro), WPG, SID (Commodore 64 songfile), FLV (YouTube video)
• sieves many more file-formats (CLA, LOG, INI, PPT, PPS, CSV, ICL, OTF, FON, VBP, VBW, BAS, FRM, FRX), especially pictures (JIF, JFIF, PCX, TIFF, PIC, AI, BW, CLP, CT, CUT, DIB, EMF, EPS, FPX, GEM, IFF, IMG, JSL, LBM, MAC, MSP, PBM, PCT, PFR, PGM, PPM, PS, PSD, PSP, RAS, RAW, RGB, RGBA, RLE, SCT, SGI, TUB, WPG), and you can also enter your own custom formats to sieve for!
• new options: skip banners, and/or graphics and/or small images, and/or non-photoquality full-sized pictures
• banner detection also by name (ie: if the filename contains "banner")
• new and/or improved definitions of thumbnails (TNTHUMB* | *_THN) and of junk-pics (108x87 | 150x150 | 90x* | 180x*)
• name recognition of known junk (index.dat, goog-malware-shavar*, goog-phish-shavar*, v=*.PNG/JPG, vt*.PNG, mapdirect*.JPG, shinystat.PNG)
• custom renaming of bad filenames and custom renumbering of same-name files
• no more folder "over-crowding": once reached the preset file-number limit, a new folder is created
• no more "I'm stuck on a big file": large files (you choose how large) now trigger "quick processing" mode
• accurate to the very last byte: "trashbin" option for inspecting all non-sieved files, and a detailed sieving log
• after the sieving is done, automatically pop-up internet options and/or sieved data and/or exit WinSieve
• INI file is now optional (only needed when moving WinSieve to another PC)
• simpler small-pic definitions (width and height instead of thousands of pixels)
• some advanced options from previous versions were removed as non-essential, but may possibly be reinstated in future versions
  namely: the favourites' and cookies' indexes ("who needs them, really?"), the picture-size index (useful in Windows98 but no longer since XP's in-folder image preview), the leeching of URLs within HTML (nobody ever used that, not even me!), and the text/multimedia extraction (too slow, but I'm working on that ;-)

  Read-on to learn about the two "missing" versions: the never-released WinSieve 3 and 4.

v. 4.0 lite

Let's keep things simple: this was the guiding concept I decided to attain to, upon fulfilling my promise to complete the new WinSieve "right after I've finished my Book". That's how I turned a Swiss-army knife with a blunt blade (v4) into sharp switchblade (v5).

• non-essential options (including all drag'n'drop functions, but the finding of duplicate files) were plainly stripped: "let's focus on sieving alone!"
• just three forms: the main (shrinked down to a button bar with a pull-down "drawer"), the settings and only one minimalistic but versatile progressbar
• multi-browser support: Internet Explorer, Firefox, Opera and Safari (and later also Chrome)
  According to the latest statistics available (March 2009), FireFox was the most used browser (46.5%) followed by Internet Explorer (43.3% of IE6+7+8) and an amazingly quickly growing Chrome (4.2%) which had already surpassed Safari (3.1%) and Opera (2.3%). That's why it was imperative to make WS4L multi-browser capable.
• sieving in memory: files are only copied/moved at the end of the process (later the low-diskspace moving option was discontinued)

  And, man, it DID work! A benchmark comparison with WinSieve 2.9 on a cache of 15302 files indicated that v4.0 lite was 18% faster when keeping the same restrains of v2.9 ("sieve all") and an amazing over 4 times faster when using the new sieving algorythm!
    After 25 days of hard work, I was about to release it as WinSieve 4.1 on the 5^th anniversary of v2.9 - when suddenly the recurrence of number 5 struck my attention: 5*5 days of work, version 4.1 that is 4+1=5, 5^th anniversary on the 5^th day of the 5^th month (May).. I had already skipped one version, after all, so why not make it two? And besides the gone-multibrowsing option alone called for that, so here's how we jumped from v2.9 to v5!

v. 4

The revolution! Featuring a totally brand new concept since the first (DOS) Sieve: only examines an "a la carte" selection of file types instead of wasting time on sieving everything and skipping just a few exceptions.


• it quickly grew up to consist of seven forms: the menu, the settings, the drag'n'drop functions, two separate progress-bars, an "about" with usage statistics and even a teaser to sponsor OpenOffice
• it kept v3.03 new options, and introduced the indexing of the computer's recent files and the removal of permanent history (a copy of internet chronology + recent files that is withheld by the system even if you purge them)
• it kept and improved v3.03 new drag'n'drop functions, and introduced a disk-lister (making a CSV database of files, with ID3 tag infos on MP3s)
• Opera-browser ready (not really implemented, but was on the to-do list)

  In a few words, its ambition to be the perfect computer's Swiss-army knife.. had turned into gigantism. While adding more and more extra-features (including an extractor of email-addresses / telephone-numbers!), I was somewhat losing sight of the core.
    Then I had to focus on my novel for many months, and by the time I got back to WinSieve I could hardly see my way through all the code I had written before. Which is why I finally decided to make a bonsai out of the sequoia, and WinSieve 4 lite was born.

v. 3.0

The early Windows-GUI porting, rewritten from (Free)Pascal into VisualBasic5.

• it consisted of single-form with 4 tabs: the menu, the settings, a built-in HTML help, and the drag'n'drop functions.
• the interface was still obviously too similar to the DOS/Windows-console versions, and just as "esoteric" for the average user. However..
• portions of this interface are still recognizable in today's versions. (eg: the settings and the buttons)
• it kept the old v2.9 options (including the indexing of favourites and cookies) and introduced the indexing of chronology (Internet history)
• it kept the old v2.9 drag'n'drop functions (find duplicates, index URLs, images, leech HTML, extract text and multimedia) and introduced data-reorganizing (ie: sieving any folder like it does with the Internet cache), directory listing and a smarter "file search & replace" than Windows98's
• ..but little code was actually written, for it soon became clear to me that there was a much smarter way to sieve

WINSIEVE v. 1 - v. 2.9
Complete chronology of Windows-console versions (March 28^th,2003 - May 5^th,2004 - 4921 downloads)

v1.0 beta @ March 28th, 2003 v1.0 @ March 31st, 2003 v1.2 @ April 3rd, 2003 v1.3 @ April 8th, 2003 v1.4 @ April 9th, 2003 v1.5 @ April 13th, 2003 v1.5 @ April 15th, 2003 v1.6 @ April 17th, 2003 v1.7 @ April 17th, 2003 v1.8 @ April 19th, 2003 v1.8 @ May 1st, 2003 v1.9 @ June 3rd, 2003 v1.10 @ June 17th, 2003 v1.11 @ June 25th, 2003 v1.12 @ June 28th, 2003 v1.13 @ July 1st, 2003 v1.14 @ July 3rd, 2003 v1.14 @ July 7th, 2003 v1.15 @ July 19th, 2003 v1.16 @ October 19th, 2003

v2 beta @ November 11th, 2003 v2 @ November 11th, 2003 v2.1++ @ February 11th, 2004 v2.2 @ March 4th, 2004 v2.2 @ March 5th, 2004 v2.3 @ March 7th, 2004 v2.3 @ March 9th, 2004 v2.4 @ March 27th, 2004 v2.4 @ March 28th, 2004 v2.4 @ March 29th, 2004 v2.5 @ March 30th, 2004 v2.5 @ March 31th, 2004 v2.5+ @ April 6th, 2004 v2.5+ @ April 7th, 2004 v2.5+ @ April 8th, 2004 v2.6 @ April 9th, 2004 v2.6 @ April 10th, 2004 v2.6 @ April 11th, 2004 v2.6 @ April 13th, 2004 v2.6+ @ April 13th, 2004 v2.6+ @ April 15th, 2004 v2.7 @ April 15th, 2004 v2.8 @ April 29th, 2004 v2.8+ @ May 4th, 2004 v2.9 @ May 5th, 2004 SieveGUI @ May 6th, 2004

v. 2.9

This was the last Windows-console ("DOS shell") version, written in FreePascal.

• diagnostic mode skips files that would cause a disk read error (=no more crashes :-)
• recognizes ANSI-codified XML
• drag'n'drop function to seek duplicate files: now also erases read-only duplicates
• recognizes new thumbnail: thumbphoto*

v. 2.8

• extracts embedded multimedia from PowerPoint™ documents!

v. 2.7

• diagnostic mode: help me to prevent program-crashes!
• automatic upgrading of the definitions (of thumbnails / junk-pics / unmasking)
  Each time you install a new version of WinSieve, you're prompted to merge the new definitions with the pre-existing ones.
• makes extracted text more accurate, cutting-off undesirable characters
• recognizes new junk-pics: 75x*, *x120
• unmasks more .JPG and more encrypted pages
• bullet-proof (before, there was a slight chance that the program might crash in an attempt to identify fake images :-p)

v. 2.6

• extracts text from unknown files (available also as drag'n'drop function!)
• unmasks more .SWF, plus the new filetypes: .XML .THEME and the MS-Office™ documents from Access™ (.MBD), Publisher™ (.PUB), Excel™ (.XLS), Word™ (.DOC) and Publisher™ (.PUB)
• unmasks textfiles with uncommon characters too (eg: open/closed quotes, fractions, etc.)
• recognizes new thumbnail: *_t
• creates a link to the Internet Options depending wether it's running on Windows 95/98 or NT/XP
• the automatic configuration sets the default documents' folder as the one that must be excluded
  (Because normally the results of previous sievings are stored there.)
• the logging was improved further (also, small pics are counted separately now)
• a few bugs were removed
• the updated helpfile was translated to English
  You don't have to use the 5-months-old help of v2.0 anymore, at last! :-)

  PS: the old subfolder "Text" has been renamed into "Documents", and "JAVA" into "Script"

v. 2.5

• a whole lot faster, thanks to the new algorythm of unknown-file unmasking and to the optimized progress bar (which speeds up the sieving of the 20% alone!)
• early file-analysis: WinSieve verifies that the file is what it looks like to be
• option to avoid junk-files: ShinyStat page-counters and encrypted pages (more might follow in the future)
• unified report, which includes the brand-new..
• ..cookies' index, to reveal what's being hidden on your PC
• recognizes .COM files which actually are HTML or text pages
• improved logging, so to offer more (and more detailed) infos
• spotted and removed a terribly well-hidden bug
  (it only occourred when creating the image-log, and only with images larger or taller than 999 pixels :_)

v. 2.4

• displays how long it's gonna take before finishing the whole work, not only the current cache sub-folder
• calculates the time left! (depending on the current sieving speed)
• small-pictures: now also only-JPG
• quick unmasking of unknown files implemented
• recognizes new thumbnails: small-*, *_th*
• recognizes new junk-pics: 96x96, 35x*, 65x*, 110x*, *x110
• recognizes new filetype: VBS (visual basic script)
• skips ShinyStat page-counters
• when pausing, displays the last file examined and what percentage of the current cache sub-folder has been sieved so far
• you can now stop the autoconfiguration (by hitting ESC)
• some bugs fixed
• tested on an internet cache of more than 19,000 files (145 MB)!

v. 2.3

• configurazione automatica (anche multi-utente) per Windows 95/98/XP!
• nuovo filtro: junk-pics (esclude immagini di dimensioni predefinite), anche solo JPG
• miniature (thumbnails): ora anche solo JPG
• riconosce nuovi tipi: MOV (quicktime movie) e INS (internet settings)
• riconosce nuove miniature (thumbnails): tn2_*,sm*,thums*,tM-*
• č ora possibile escludere i contenuti di pių cartelle
• crea il link alle opzioni internet (per cancellare cache e/o cronologia)
• opzione per mostrare il file correntemente processato (per eventuale debug)
• copia il log sul desktop (o comunque nella stessa cartella che contiene NewCache)
• definizioni di thumbnails e junk-pics editabili dall'interno del programma (=senza aprire Sieve.INI)
• migliorata la procedura di input da tastiera (ora č uguale ai dialog di Windows: scrolla le righe lunghe, e inizia reinserendo - per aggiungere/editare, premere FINE o spostare a destra il cursore)
• migliorato il calcolo del lavoro rimanente (ora per cartelle con pių di 2000 files vengono indicati anche i centesimi di percentuale)
• rimossi alcuni bug

v. 2.0

• files are immediately copied (or moved), without creating the batch file (unless you ask for it): this speeds-up the process (=one step less has to be performed) and also makes WinSieve work under the newer versions of Windows (that no-longer support DOS-prompt windows)
• the logging has been improved even further, and now creates a separate file (Sieve.LOG)

v. 1.16

• recognizes fonts (.TTF) and icons (.ICO .CUR .ANI), and places them in separate folders
• recognizes new file-types: WMF, TIF, PDF, TGZ, DLL, DRV, HLP, CHM, PHTML, HTT, HTA
• improved unmasking of WAV, AVI, BMP and JPEG (FSPA type)
• fixed a bug that didn't allow to create a folder in root-directories
• fixed a bug about hidden files

v. 1.15

• allows the (re)definition of the thumbnails!
  Just edit Sieve.INI at "THUMB=.." (eg: "THUMB=th_*", "THUMB=*_tn", "THUMB=*info*")
  If the file Sieve.INI does not exist, or does not contain "THUMB=.." definitions, the default definitions will be used.
• allows to (re)define the criteria used to identify the unknown files!
  Just edit Sieve.INI at "BIND=.." (eg: "BIND=docinfo|.HTM")
  If the file Sieve.INI does not exist, or does not contain "BIND=.." definitions, the default definitions will be used.
• option to create an image log (PicInfos.TXT)
  Creates a log with all the details of each sieved image (bytes, width x height, etc.)
• option to exclude the files from a folder
• unmasks fake textfiles and fake images (!)
  If a .TXT is actually an image, or vice-versa: if an image is actually a textfile (or an HTML page, or a JavaScript), it now gets recognized and moved to \PIX\CHECKOUT with the correct file-extension.
• banner-recognition improved
  Now panoramic pictures (much wider than tall) are no longer recognized as banners.
• new memory management
  Due to a program bug, previous versions used to fill the phisical memory and then skip (after about 5000 files): now there's a true full virtual-memory support, and the only limit is your hard-disk space (beta-test performed on 28.720 files, 383MB)
• added automatic configuration
• added front-end mode
• logging improved
• recognizes new types of thumbnails (mini*.*) and of multimedia files (*.mpga)
• very long paths are now shortened (the progressbar no longer "overflows" into the next line)

v. 1.9

• repairs corrupted images (!)
  Keeps the original file (the damaged one) in \Pix\Checkout and, if it was successfully repaired, adds the suffix _repaired. The repaired file can be found in the proper subfolder.
• recognizes the following types: PNG, WAV, AU, RAR, ZIP, EXE/COM, MOV, RM, PPS, SWF plus generic text-only files
• sends the recognized files to their proper subfolder
  adding the suffix [ext] (ext is the old extension)
• automatically corrects images with wrong file-extension
• improved un-masking of HTML & JavaScript pages
• improved un-masking of bitmap and jpeg
• now you can put on hold or abort the sieving (by pressing the ESC key)
• bugs removed from the drag'n'drop function of duplicates-seeking

v. 1 (what's new since the last DOS version)

• almost 5 TIMES faster: 4.5 quicker ¹ thanks to the 32bit code and the fact that..
• it searches for duplicate files WHILE sieving, without limitations ² (before it could only handle up to 18.120 files)
• recognizes ALL the images ³, thanks to the new algorythm
• duplicate-files precision doubled ⁴
• incresed the precision of thumbnail-recognition ⁴
• separates small images from bigger ones ⁵
• recognizes vertical banners also ⁵ (=the typical on-page-side advertisement of search-engines)
• unlimited filename length: before they were truncated to 40 chars, now that's become an option
• the drag'n'drop function for seeking duplicate files was enhanced: it now reads all files in all subfolders (before there was a limit of 18.120 files and 50 subfolders), and with long paths (DOS only supports paths of 64 chars); but mainly it's quite faster thanks to the new optimized algorythm, minimizing disk-access
• new drag'n'drop function, to extract URL of images and links from an HTML page ⁵
• you can keep more sievings in the same <NewCache> folder ⁶
• no need to only use DOS names: now also accepts Windows long names
• runs also on Windows 2000/ME/XP (or at least it should, I'm yet to test that ;-)
• and finally, I've corrected some subtle bugs (such as the one about MOVE)
   ¹ Complete statistics from the beta-test: 4 times faster to create the "favourites" index. As for the sieving, it is: 4.52 times faster (using all functions, as default), 2.38 times faster (disabling the search for duplicate files) or 1.15 times slower (sieving only, without image-subdivision, duplicate files search and recognition of unknown files)
   ² The only limit is the virtual memory that the computer can provide. (Depends on RAM and hard-disk free space.)
   ³ \Pix\CheckOut now only keeps corrupted files (=partially downloaded, or with errors of transmission) or those with a wrong extension (but they're likely to be identified by the option to identify unknown files)
   ⁴ This explains the (little) difference in the sieving-results provided by WinSieve and SIEVE 3.5 (DOS).
   ⁵ New feature! (=not included in the DOS version)
   ⁶ In order to avoid to add files that have already been previously sieved, the duplicate-files searching mode is automatically forced even if the user has disabled it.

SIEVE v. 1 - v. 3.5
Complete chronology of DOS versions (Late 2001 - March 29^th,2003 - 1287 downloads)
v. 1c Late 2001, on Windows 95 v. 2.1c (aka SIEVE98) November 21st, 2002, on Windows 98 v. 2.2beta March 10th, 2003 v. 3 March 11th-17th, 2003 v. 3.3 March 21st-20th, 2003 v. 3.5 March 29th, 2003

v. 3.5

The last DOS version, written on Borland TurboPascal 6.

• Windows-friendly interface, making the program easier to setup (you don't have to use DOS parameters nor edit the configuration file anymore)
• Amazingly faster, thanks to the new algorythm
• New drag'n'drop functions: creates an .URL index, and/or removes duplicate files
• Deals with (potentially unlimited) large cache, holding an average of 6,000 files in memory (or more, and never less than 3,631) then using the hard-disk as "RAMdisk"
  And besides we're speaking of already-sieved files, so the cache can be a whole lot bigger than that.
• Option to skip normally-uninteresting files, such as HTML pages, JAVA stuff, thumbnails and files smaller than N bytes
• Fixes-up bad filenames, by turning %20 back into a space (and other %n into their proper character), removing illegal chars ("." "?" "=" ecc.) and shortening too-long filenames
  (=no more "getgraph?f=c12345&u=www%2Atry%2Ait.gif" !)
• Creates unique filenames (if files are all moved into 1 single folder, there's no longer the chance that some would overlap)
• Some bugs fixed (now also works with files without extension)

v. 3.4

v. 3.3

March 21^st-20^th, 2003
• improved duplicate-seeking (faster, and up to 18120 files in 50 folders)
• now only requires 160K DOS RAM free

v. 3.2

• 21-28 files per second (on a cache of 1373 files)

v. 3

March 11^th-17^th, 2003
• amazingly faster, thanks to the new algorythm
• skip html & thumbs
• large-cache capable (first 5000 entries in memory, then on hard-disk)
• renames same-name files instead of over-writing previous
• rename %n and cuts illegal chars ("." "?" "=" etc.)
• trims filenames longer than 35 characters
• duplicate seeking (up to 9000 in 30 folders)
• bug-fixed (does not crash with files without extension)
• new windows-friendly interface
• note: unless in safemode, some files may fail to be moved (possibly because they're still in use by the system)

v. 2.2beta

March 10^th, 2003
• sieve pictures (GIF,JPG,JPEG,PNG,BMP,XBM,ICO), plus media (MPG,MPE,MID,WAV,MP3,RA,RAM,AU,KAR,SWF,WMV,SPL), text (TXT,DOC,RTF,NFO,DIZ), programs (ZIP,RAR,CAB,JAR,TAR,LZH,EXE,COM), HTML (HTM,HTML,SHTML,XTML,XML,HTC) and Java (JS,CSS,CLASS,PHP,CGI,VBS)
• subdivides JPG, GIF, BMP, PNG into checkout + banners + graph + thumbs
• thumbs: *_SMALL*, *TN_*, TH_*, THUMB*
• recognize unknown files if they're GIF, MID, BMP, HTM or JS
• removes duplicate files (quickmode: sequences of 10 bytes)

v. 2.1c (aka SIEVE98)

November 21^st, 2002, on Windows 98.
Howdy everybody! =-) I'm proud to introduce you "Sieve", the first "cache manager" to deserve this title: differently from other programs (at least, the ones I know), SIEVE does not only VIEW the Internet cache - but does much more than that: 1) re-orders the cache, sub-dividing the files according to their type (all HTML, SHTML, DHTML, etc. pages in the folder \HTML; all the Java, CGI scripts etc. in \HTML\JAVA; all the images in \PIX; WAV, MP3, MID, SWF, etc. in \MEDIA; and so on) 2) sub-divides images according to their kind (separates banners from pictures, thumbnails from graphic elements - such as bullets, buttons, and all those things that can be "stolen" in order to use them on your own site.. ;-) 3) restores the original filenames (ie: without the [1] suffix which is automatically added by Internet Explorer), and whenever it's needed it renames keeping the long filenames (eg: "hallotherebaby.jpg" becomes "hallotherebaby_2.jpg") 4) unmasks "cheating" files (those with a different extension than the one that would be due, such as the JPG masked as GIF, or HTML pages without extension, or an .XYZ extension different from .HTM) 5) removes duplicate files (or simply lists them for you, if you don't trust SIEVE to do the whole work for you.. ye unfaithful! :-) 6) turns all the .URL files you've stacked-up while browsing ("Add to favourites") into a single, convenient HTML page with a list of all the links 7) goes shopping for you, prepares your dinner, drives your kids back home from school, and then wishes you "Have a nice day :-)" It's all true. All, but point 7: I'm still working on that! (Right now, I'm teaching Sieve to prompt you to have a nice day. As for the remaining parts, it's just a matter of time! haha :-D) Especially if you're an "old-fashioned programmer" as I am, you won't miss to fall in love with SIEVE - which finally allows you to know EXACTLY what your computer is hiding in its pockets, until the last byte. And it only takes a few kilobytes of DOS program. (But, with a clever trick, Windows' long filenames are supported too. Check out the attached Pascal source-code to learn how! ;-) That's it. I'm sending SIEVE to you, so you can make the classic beta-test of it. (So I'll know who to blame, should anyone complain that it doesn't work right! ;-) If you find any bug, or mis-functioning (in case you tested it with Netscape, Opera or any browser different from Internet Explorer v3 and v5), or should you simply have any idea on how to improve it.. just let me know, okay? [The only known bug so far, is that SIEVE seemingly ignores the files without extension.. I'll take care of this minor problem as soon as possible, but right now, after a whole week of programming, I need some rest! ;-) ] For any further information, just write to this same email address. JB ^u^ ;---------------------------------------------------------------------- ; SIEVE - the advanced Internet-cache manager ; (c) 2001 JB/The One Phoenix (the1phoenix@the1phoenix.net) ;---------------------------------------------------------------------- ;IMPORTANT! ;--------- ;* DO NOT CHANGE THE HIGHER-CASE FIELDS ;* ONLY USE DOS 8.3 FILENAMES AND PATHS ; (eg. "tempor~1" instead of "Temporary Internet Files") ;The path to the folder where all the cache will be re-organized into. OUTPUT FOLDER=c:\windows\desktop\newcache ;The path to the cache folder of your browser CACHE=c:\windows\tempor~1 ;The path to the bookmarks' folder of your browser (.URL files) BOOKMARKS=c:\windows\prefer~1 ;The name that will be given to the list of duplicate files LOG FOR DUPLICATE FILES=duplicat.log ;If it's smaller than this (width x height), it's a graphic element of ;a page (such as a bullet, a button, or a ruler..) and not a thumbnail SMALLEST THUMBNAIL=20 x 50 ;If it's larger than this (width x height), then it's not a thumbnail - ;but probably a full-sized picture LARGEST THUMBNAIL=100 x 80 ;If it's shorter than this, it's not a banner, so maybe it's a button MINIMUM BANNER HEIGHT=50 ;If the image width is N times its height, then this wide rectangle is ;a banner BANNER PROPORTIONS=4 ;Reads the file-header to un-mask files (some GIFs are really JPG, or ;some files with an unknown extension are really textfiles/HTML pages) TRY TO RECOGNIZE UNKNOWN FILES=yes ;Choose NO if you don't trust SIEVE to identify true duplicates.. ;this way, you can check out the list before actually deleting them REMOVE DUPLICATE FILES=yes ;Internet Explorer tags files with the same name with an annoying [n] ;suffix (eg. image[1].gif): choose YES to remove the suffix ;(No file will be overwritten: just in case, automatically re-renamed). REMOVE IE5 [1] SUFFIXES=yes ;Choose YES if you don't trust SIEVE to work properly, and want to ;verify it before actually clearing your precious cache ;-) CREATE COPY OF CACHE INSTEAD OF MOVING FILES=no ;Choose YES if you wish to clear the bookmarks after they've been ;converted into HTML index-page CLEAR FAVOURITES AFTER CREATING THE LIST=no
• subdivide thumbnails (20x50 - 100x80) and banners (Y>= 50 and 4:1 proportions)
• recognize unknown files
• skip duplicates
• remove IE5's "[1]" suffix from filenames
• list favorites

v. 1c

Late 2001, on Windows 95.