Messages - jity2 [ switch to compact view ]

Pages: prev1 2 3 4 5 [6] 7 8 9 10 11 ... 26next
26
Hi Netrunner,

After trying to understand your WSW requests I am not familiar with :
Maybe use a third party service for canary : https://ifttt.com/makers/canaryhome ? or ask Microsoft flow ?...

>download page with PhantomJS or similar if the page is not compatible with the built-in IE (as it seems IE won't be updated anymore after Edge lunched and the developer said there is no plans to use another engine with JavaScript support).
Can you give a page example so I can check with WSW ?

Thanks ;)


27
Dear all,
I am a WSW customer for a little more than 10 years now! I don't know why Martin (WSW's developer) stopped his forum. Maybe lack of time ?
For my part I have always preferred to ask him many questions directly. He has always provided fast and helpful answers.
Alas I am not a coder so I can't help much with WSW scripts. But this is a great software. So thank you Martin anyway. ;)

>Netrunner :
For your RSS Full text script, I advise you to test this http://fivefilters.org/content-only/ in conjunction with WSW. ;)

See ya ;)

28
Dear all,

First some background for the idea :
I have some PDF files which are damaged. My goal is to OCR what can be repaired (I recently tested "Nuance Power PDF Advanced2". IMHO it can OCR many pdf that have problems that other OCR softwares can't even open. But alas it has still problem with some pdf files.)

I have tried several tools and techniques. The best ones so far being :
- 3-Heights™ PDF Analysis & Repair (they also sell a shell version).
https://www.pdf-tools.com/pdf20/en/products/pdf-converter-validation/pdf-analysis-repair/
(The free version can be used here : https://www.pdf-online.com/osa/repair.aspx )
The problem is that it doesn't repair all defects properly. ;(

- and a batch script using SumatraPDF and the printer Bullzip ( see https://www.donationcoder.com/forum/index.php?topic=42713.msg399623 ).
The problem here is that it takes a lot of time,CPU and memory. For instance a pdf of 100MB uses 16GB of temporary SSD space in order to produce ("print") finally, after 10 minutes, a 300MB pdf !
Also for several pdf files, the process is done and at the end no pdf file is created ! ;(

So I got this idea :
I realize that the nice thing is that I can open most of the pdf (that have errors) with SumatraPDF.  ;)
So it would be great if some software could once the pdf openned in SumatraPDF, take a screenshot of each pages in burst mode (one screenshot then turn to the next page, then repeat). Then I could probably make a pdf from the image files and OCR them very fast ?
I did test SCREENSHOT CAPTOR VERSION 4 https://www.donationcoder.com/Software/Mouser/screenshotcaptor/index.html but I wasn't able to do it (the automatic page "down/up" did not work - win8.1 64) !


Thanks in advance ;)

29
update: A few months later now, I stopped using Otixo (too expensive now for me).
For uploading my data I use : SynckbackPro and Syncdocs (for converting some of my files into (free) Google Documents).
And now, once I have uploaded my data into one service (Google Drive, Amazon Drive...), I now use rclone.org with a cheap VPS (I have used OVH. Similar are: Digital Ocean, Scaleway...) About $3.5/month. Not only this is cheap but it is far much faster than using my DSL line to move data between online services. ;)

Hope this helps ;)

30
2016 edition : How Backblaze Personal Backup lost me 2TB on purpose !

tl;dr : Backblaze uses file checksum to their advantage, so their customers can loose easily most of their previously uploaded data when changing hard drive (or moving a big folder inside a new partition) !


Dear all,

Instead of using the 30 days data retention policy to the customer advantage (by transferring automatically ALL the files already uploaded previously), Backblaze software is designed to lose most of its customer updated data when the customer remove an old drive and add a new bigger one (with the same data). Logic would be that they first find all the already uploaded files whatever their size and then starts uploading the missing ones ! After all, their software does a file checksum for something. But, in fact, it ranks the files by size and then starts to upload the smallest files first. So if a customer has uploaded a lot of big files before (max 2GB zip files in my case), they are lost IF all the smaller files are not uploaded during the 30 days lapse.(*)

For years, their client has also some on purpose bugs :

    1) It does not tell you that when you add a new drive in your computer you have to manually add it in their client settings.
    2) And when you do so, it again silently removes all the exclusion list that you ave added before ! And you have no choice but to again add manually all the folders you need to exclude !
    3) Caution: their new client version (v 4.20) added a "stupid feature : " exclusions now work across all attached drives. " Too bad for those that have the same folder name on different drives. Now you can't exclude only one in all drives or in none ! I just can't believe you did that ! ;("

It is why :

    - Their support team always answer: "Backblaze client needs time to find already uploaded data" ! They acknowledge this but it is hidden in their help file : "Backblaze prioritizes smaller files, and uploads larger files later." https://help.backblaze.com/hc/en-us/articles/217666728-How-does-Backblaze-handle-large-files-

    - They sell it as unlimited but recommend in initial backup in 30 days : "(...)3. Ideally, Backblaze should be able to complete your initial backup in 30 days. If your initial backup is estimated to take longer due to a lot of data or slow internet connection, then Backblaze is not the best solution for you.(...)" https://help.backblaze.com/hc/en-us/articles/217664608-Best-Practices

I like them a lot (see their blog) and I understand that they need to stay profitable (from memory 250GB threshold in 2012? 1TB threshold in 2016) but their customers must know that they should upload their files also elsewhere (Amazon Drive, ..etc..) where, contrarily to Backblaze Personal Backup, they can move the uploaded files very easily thanks to APIs (with for instance cloudhq.net or a cheap VPS and rclone).

(*) in my case: Win 8.1 64bits running 24/24 7/7 - 16GB RAM - DSL (upload max speed 10GB/day). Customer since 2011. About 3TB uploaded in July 2016. About 2TB lost as of today and only 1TB recognized (30 days period ended). ;( I have tried to add many folder exclusions (note: you can't just remove C:\ !) so backblaze could find easily my previously big files. Alas for me they were in 2 big folders previously. And I have added them into one big folder in my new drive. And could not remember or check easily within the 30 day retention period. Note2: I also lost about 2 weeks in order to realize that the new drive was not added automatically by Backblaze client) which exact many parts were previously uploaded. ;( Hopefully for me I use other storage services (Amazon Drive...). See here : https://www.donationcoder.com/forum/index.php?topic=41873


update: 2016, Mid December : Backblaze lost me all my backup again! This time I did not add any new hard drive to my computer. I just put it on "run once every day at 10pm" for about one month. Then, I changed that again back to "continuous" and the 1.2 TB of data that I had there disappeared ! I just recovered 300GB since that day. ;(

Pages: prev1 2 3 4 5 [6] 7 8 9 10 11 ... 26next
Go to full version