Jump to content


  • Content count

  • Joined

  • Last visited

  • Days Won


Posts posted by DavidHertzberg

  1. @David Hertzberg: This is off-topic, but could you please stop with that childish "He Who May Not Be Named Lightly"  and "derek500 doesn't like me to use smilies in these forums" stuff?


    It maybe was funny once, but it is becoming annoying now.


    I'll discuss the point I'm making with "He Who May Not Be Named Lightly" in this post.


    I see that Mayoff liked Hofstede's post.  That's good, because it indicates that Mayoff is actually reading this thread—which I had come to doubt.  Mayoff is, of course, the poster I intend to needle as He Who May Not Be Named Lightly.


    For me, the problem became evident in May 2016.  In the "Retrospect 9 or higher for Macintosh" forum I made a perfectly legitimate post (I no longer remember what it was) about a deficiency in the Mac 13 User's Guide, and accompanied it with a gentle gibe about Mayoff having a hypothetical daughter with a speaking Barbie doll that says "writing is hard" (referring to a Barbie doll—quickly withdrawn after widespread criticism—some years ago that said "math class is hard").  Mayoff deleted my post, and immediately re-posted this—with evident emphasis on its final paragraph.  If Mayoff was motivated by memories of a real daughter to whom something tragic had happened, I'm truly sorry, but I think an alternative explanation for his actions is much more likely.  Given that, on 26 September, he posted "Technical Support does not write the Retrospect User's Guide. This is handled by a different team of people in the company which includes Product Management and Engineering. That same team currently writes most of the KB articles", I now think he was instead motivated by embarrassment at his own powerlessness to improve the UG.


    But that doesn't explain what happened over the last week on the Retrospect Forums when the spam tsunami started.  In the "Retrospect 9 or higher for Macintosh" forum I posted a report about a similar spam tsunami on the Ars Technica Macintoshian Achaia forum, and what the administrators and moderators there were doing to deal with it.  The whole thread containing my post was deleted "accidentally" along with a lot of spam posts, which I discovered when the link to it I had put into post #4 of this thread stopped working.  So I first put a direct link to the Ars Technica administrative thread on the spam tsunami into post #8 of this thread as a P.S..  Then, when spam threads kept reappearing in the "Retrospect 9 or higher for Macintosh" forum after Mayoff said he "adjusted the forum spam filters", I "spelled out" the Ars Technica forums solution as post #13 in this thread. Seeing that that didn't have any effect, I copied that post #13 as a Support Case.  Finally, realizing that continuing to implement the Ars Technica solution would put a strain on Mayoff, I suggested a labor-saving technique—which BTW Ars Technica can't use because they don't sell any software—as post #14 in this thread.  All in all I put a fair amount of effort into helping Mayoff deal with the spam tsunami; all I got for it was my Support Case being closed without any comment—not even "Thanks, but we already saw this on the forums."  Am I wrong to call Mayoff "He Who May Not Be Named Lightly"?  What epithet would you folks suggest?

  2. @David Hertzberg: This is off-topic, but could you please stop with that childish "He Who May Not Be Named Lightly"  and "derek500 doesn't like me to use smilies in these forums" stuff?


    It maybe was funny once, but it is becoming annoying now.



    I never intend that stuff to be funny;  I intend it as needling to make a couple of points.  


    I'll discuss the point I'm making with "derek500 doesn't like me to use smilies in these forums" in this post, and the point I'm making with "He Who May Not Be Named Lightly" in my next post.


    derek500 said "... there is a checkbox on the right for 'enable emoticons'. Uncheck that and B) works normally! Being a tech forum, I think these should be turned off by default!" in the last paragraph of this 25 March 2016 post.  I replied "Being a tech forum dealing with backup, and especially with Retrospect, I think we need all the nuance we can get from emoticons" in the next post in that same thread.  These remarks were part of a technical discussion related to what I said in a preceding post in the same thread, which was "So the answer to your Question 1) is B] (which I must write that way because—thanks to a 'peculiarity' in the Retrospect Forums software—if I write it as 'b' or 'B' followed by ')' it gets turned into this B) smiley—as happened in your post)."  The underlying technical problem IMHO is that what the Retrospect Forums software inserts for certain emoticons is too simplistic; the insertions are identical to certain other legitimate combinations of alphabetic and special characters.  The Ars Technica forums don't have that problem, because what their software inserts for emoticons is never identical to other legitimate character combinations; for instance they insert 'B-)' without the surrounding single-quotes instead of  'B)'.  As a matter of fact, I will belatedly file that problem as a Support Case.


    Now let's look at the second paragraph of post #16 in this thread.  In that paragraph I'm actually discussing a rather tricky problem, which is how to distinguish bad guys who can be expected to lie from good guys who can be expected to tell the truth.  If you think about that problem for 15 seconds, you'll realize it's the same problem as with the "extreme vetting" of Moslem refugees advocated by Mr. Trump.  In the Retrospect Forums case the bad guys merely want to subvert our forums, and the good guys merely want to discuss a legitimate concern with Retrospect.  In the "extreme vetting" case, on the other hand, the bad guys sympathize with ISIS and want to kill us, and the good guys want to get into the U.S. to avoid being killed.  Either way the problem of distinguishing bad guys from good guys is also rather funny; in fact I suspect there is a Monty Python skit about it someplace.  That's why I talked about "my years working as a KGB interrogator under Vladimir Putin", which was a joke (I actually make yearly contributions to Amnesty International) that would have been much clearer if I could have used an emoticon.

  3. If I read David's post correctly, that would effectively close out those interested in Retrospect, but have not purchased a license yet.



    If you read my post that way, then I didn't write it clearly enough.  In the situation you cite, the administrator should exchange e-mails with the person who has signed up for the new Forums account before allowing him/her to post.


    In spite of my years working as a KGB interrogator under Vladimir Putin (derek500 doesn't like me to use smilies in these forums), I'm not sure what I—if I were a Retrospect Inc. Forums administrator—would ask the person who has signed up for the new Forums account.  Asking him what his/her Retrospect question is sounds like a good start; if the person seems to have no familiarity whatsoever with Retrospect's facilities, that would sound more than a little suspicious.  Asking the person for three references who would testify under mild torture that the person wasn't doing any spamming in the last week seems a bit excessive, although we used to be able to do things like that in the KGB.


    In any case, a quick search of the Forums seems to indicate that such persons are few and far between.  AFAICT the only people who asked a question without already having a license showed up in "General Discussion—Retrospect" about 3+ years ago.  I'm sure He Who May Not Be Named Lightly would be overjoyed to encounter such a person, provided he/she were not a spammer.  In fact, that's why DovidBenAvraham greatly enhanced the Retrospect article in Wikipedia—but the readership statistics so far seem to indicate that the majority of people who are reading the article are already Retrospect users who are monitoring the article for any changes.

  4. Having read DovidBenAvraham's augmented Wikipedia article on Retrospect, I regret to inform brian163 that Retrospect can currently do cloud backup natively only to provider services that are either AWS-S3-compatible (as of Retrospect Mac 13) or WebDAV-compatible (as of Retrospect Mac 9).  According to the WP article on Google Drive, it is not compatible with either of these protocols.  Those incompatibilities are also likely to be true for any 15GB of cloud storage brian163's hypothetical cat—who may be named Larry Page or Sergey Brin—may have dragged in (insert appropriate smiley here).


    However lovable Larry Ellison may have come to brian163's rescue for a paltry US$1/TB/month for uploading onto Oracle Archive Cloud, as mentioned in an Ars Technica thread starting with this post.  Oracle Archive Cloud is apparently AWS-S3-compatible, but it is not certified by Retrospect Inc..  I'm not sure whether certification of a cloud backup provider is dependent on Retrospect Inc.'s having run actual tests, or whether it is dependent on Retrospect Inc.'s having reached some kind of deal with the provider, or whether it is dependent on both.  In any case, as pointed out by the poster of the linked-to post further down that thread, the expense of a Oracle Archive Cloud recovery is such that it is only appropriate in "Fireman, fireman, save my photos/videos/artwork!" situations.


    Another alternative would be, as brian163 says above, to try to use Mountain Duck to mount Google Drive as a WebDAV-compatible drive on the desktop.  In this case brian163 would not be paying attention to anything in the Retrospect Mac 13 User's Guide "What's New" chapter, but instead to mentions of "network shares" at various places in the UG.  The primary starting point would be "Adding network shares" on page 85, but brian163 should search for mentions in later pages of the UG.  Having done so, he might do well to adapt the procedure shown in this post, since he would have to cut down his Google Drive-stored Media Set to 15GB.  In step C3), brian163 would replace "setup your cloud account" with "use Mountain Duck to mount the Google Drive storage as a network share on the desktop".  In steps C4) through C7), he would replace "new Cloud Media Set" with "new WebDAV Media Set".  In steps C1) and C7), brian163 would replace the "months" mentions with whatever it would take to restrict his new WebDAV Media Set to 15GB—which in my experience is a laughably small amount of storage (my MacBook Pro's Macintosh HD started out at 29GB backed up in a Recycle Media Set script once a week, and has since grown to 50GB).

    P.S.: So that this post can serve as a reference for anyone asking about possible Retrospect compatibility with Google Drive and "other less than enterprise class cloud storage solutions", here is a blog post comparing Google Drive with Google Cloud Storage.  Google Cloud Storage is AWS-S3-compatible; Google Drive is not.  With the exception of Backblaze B2—which was cheaper than any other suitable cloud provider at the time Retrospect was enhanced to interface with it , every cloud provider Retrospect Inc. has since certified for Cloud Media Set backup is AWS-S3-compatible.  

  5. It occurred to me later this morning that there might be a way of making the manual exchange of e-mails requirement before posting can begin easier for the Retrospect Inc. administrator(s) of these forums than it is for Ars Technica forums administrators.  It depends on the fact that—IIRC—anybody purchasing a license for Retrospect must supply an e-mail address, and on the fact—again IIRC (its been 12 years since I signed up for a Forums account)—that anybody signing up for a Forums account must also supply an e-mail address.  It should be comparatively easy for He Who May Not Be Named Lightly to get Retrospect engineering to supply him with alphanumerically-sorted copies of those two e-mail address lists, plus simple programs to do binary searches of those sorted lists.


    Thus, for any person who holds a license for Retrospect and has not previously signed up for a Forums account with the e-mail address he/she now supplies, it should not be necessary to do a manual exchange of e-mails before allowing that person to make posts to the new Forum account; the administrator can safely approve posting semi-automatically.  On the other hand, if the person signing up for a new Forums account does not have a license for Retrospect using the supplied e-mail address, or if he/she has a Retrospect license using the supplied e-mail address but already has another Forums account using that same e-mail address, then the administrator should exchange e-mails with the person who has signed up for the new Forums account before allowing him/her to post.


    Being a retired applications programmer, it naturally occurred that this decision process could be semi-automated.  However I can see some potential problems with doing that.  For instance, what if the resourceful spammers get hold of a list of the e-mail addresses of Korean/other licensees of Retrospect?


    I would greatly appreciate anybody's thoughts on this subject. 

  6. >No updates during these five years?


    That isn't what I said. Using the same forum software, and it is kept up to date to the best of my knowledge.


    I have adjusted the forum spam filters to see if it helps.



    IMHO you (or your co-workers at Retrospect Inc.) are being obtuse, He Who May Not Be Named Lightly, so in the interest of all users of the Retrospect user forums I'll spell it out for you—based on  what I have read on an administrative thread of another set of forums that seems to have dealt permanently with this problem:


    The basic source of the problem is that there are some Korean-language (that's the Hangul script that they use) spammers who have discovered how to automate the account signup process  for forums that use the type of forum software Ars Technica and Retrospect Inc. use.  Among other things their automation (or possibly low-wage Bangladeshis they may have hired) can do is to cope with whatever CAPTCHAs the forum software may require for account signup.  Those Ars Technica forums posters who can either read Korean or use Google Translate have said that the spammers are creating accounts on popular forums and doing automated posting to them in order to promote some gambling sites—whether websites or physical sites I don't know.  The spammers' idea is to raise the online profile of these gambling sites by doing Search Engine Optimization, whose embodiment for Americans is gaming Google Search.  The spammers are incredibly persistent.


    The only way the Ars Technica administrators have found to cope with the spammers is to disable automatic posting capability for newly-created forum accounts.  Ars Technica evidently has introduced a manual exchange of e-mails requirement before posting can begin.  The Ars Technica administrators and volunteer moderators have followed this up by deleting any spammer accounts that had already been created before the manual-exchange-of-e-mails requirement was instituted.  BTW, the spammers have sent e-mails to the Ars Technica administrators asking why they can't post to their accounts—which in the opinion of those administrators is true chutzpah (that's a Yiddish term, not a Korean one, for those who don't know).


    The Ars Technica administrators are not happy with having had to institute the manual-exchange-of-e-mails requirement.  For one thing, it means more work for them.  For another thing, it inhibits the ability of new forums users to sign up for an account and immediately start posting about whatever software/hardware problems they are having.  But it seems to be the only way to permanently deal with these spammers.  I'm sure that the spam filters for the Ars Technica forums are at least as good as the spam filters for the Retrospect Inc. user forums, so if Ars Technica administrators couldn't deal with the spammers through filters alone then neither will you be able to do so.


    Please excuse me if I have unfairly belittled your efforts to date,  He Who May Not Be Named Lightly.  But, BTW, why have the Retrospect user forums suddenly started displaying an IPS Community banner instead of a Retrospect banner at the top of the forum pages?  How is that evident switch supposed to help?


    P.S.: As of 4:24 a.m. EST on 16 November the spammers are back on the "Retrospect 9 or higher for Macintosh" forum; I think that proves my point.


    P.P.S.: In second sentence of second paragraph, parenthetically allowed for the possibility that the spammers may be using low-wage humans to get around CAPTCHAs.


    P.P.P.S.: This morning (16 November) I copied this post as a Support Case, to make sure He Who May Not Be Named Lightly read it.  The case has now been closed by Support.

  7. Heavens to Betsy!  Retrospect Inc. seems to have been motivated by this problem to upgrade to non-free forum software from Invision Power Services Inc..  That's why the logo in the browser tab has changed from the red Retrospect cube to the 'V' in the blue box.  That will cost them a minimum of US$30/month for 40 simultaneous visitors.


    Couldn't He Who May Not Be Named Lightly just have done what I suggested, which is what Ars Technica has done?


    P.S.: Here's the Ars Technica Macintoshian Achaia thread that tells how they got rid of the spam in that forum.


    P.P.S.: When I try to post from Firefox 50.0, I still don't get any formatting buttons and I still get extra HTML character entities thrown in.  That would imply that the Retrospect forums are still using the same software they did.  So why are we getting the IPS Community banner at the top of the page?  Did Retrospect Inc. just pay IPS to get rid of the spam?


    P.P.P.S.: As of 22:54 on 15 November, there are 4 new spam threads in the "Retrospect 9 or higher for Macintosh" forum.  Based on what I have read on Ars Technica, that probably means there is one spammer account that is still registered and allowed to post.

  8. For reassurance jethro might first want to view the video linked to in this post, which was created by Mayoff.  Note that it is entitled "Staged Backup with Retrospect for Macintosh", which is exactly what jethro now wants to do—except that the video  shows disk-to-disk-to-tape and he wants disk-to-disk-to-disk.  You will note that Mayoff made the post in a thread that jethro should have some slight familiarity with (insert appropriate smiley here), since he started that thread and made several posts in it last spring.


    In other words, what jethro wants to do now is just a variant on the disk-to-disk-to-cloud that he wanted to do last spring.  Therefore I suggest that he consider basically following the steps in section C) of this post in the same thread.  All jethro would need to do is replace all mentions of "Cloud" with mentions of "Disk", eliminate mention of Grooming in steps C1) and C2) unless he still wants to cut his off-site backup down to the most recent year, eliminate steps C3) through C5), and not check No Verification in step C6)—he would be doing a Copy Backup to fallible Disk rather than ultra-reliable (thus sayeth Retrospect Inc.) Cloud.  If jethro wants to have both a Media Set A and a Media Set B on off-site Disks, he should repeat steps C1) and C2) to a second portable Disk and change the last sentence in step C6) to alternate the scheduled weekly Copy Backup between Media Set A and Media Set B.


    In return for all this additional help that he shouldn't really have needed, I'd like jethro to give us in this thread an answer to one question:  Why didn't he implement the disk-to-disk-to-cloud plan he wanted to do last March?  Was it because cloud storage turned out to be too expensive for his monthly budget?  If I point out that Retrospect Mac 13.5—released 14 September—would allow him to maintain a Cloud backup of up to 1TB on Dropbox Pro for only US$99/year, would that change his mind?  Or was it because he realized that, with "seeding" and "restore-to-door" eliminated by all U.S./Canada cloud backup providers except pricey Amazon and Google, initial population to and restores from the cloud would take an unreasonably long time?

  9. Here's the link to my post on the "Retrospect 9 or higher for Macintosh" forum discussing this, in relation to how it has been dealt with on the Ars Technica forums.


    P.S.: The reason the link doesn't work any more is that someone with administrator authority, presumably He Who May Not Be Named Lightly, deleted the entire thread containing the post.  I really don't understand why he did so.  The reports of spam in the early posts were correct.  The link in my post to the thread in the Ars Technica forums was surely helpful, along with my description of how Ars Technica administrators had dealt with their spam.  My characterization of the nationality of the spammers, reflecting statements by other posters to the Ars Technica thread who could read the spammers' distinctive script (which even I can recognize), was surely correct and referred to "some  ...s".   The suggestion in my post that this was an opportunity for He Who May Not Be Named Lightly to shine was not abusive, and—given the reference to a smiley after it—could surely be recognized by anyone without undue sensitivity as an expression of sympathy with the amount of work that He Who May Not Be Named Lightly would have to do to fix the problem.  All in all, I think He Who May Not Be Named Lightly owes all of us who posted in that thread an apology.


    P.P.S.: They're back.  Approximately 1.15 pages of Korean-script spam in the "Retrospect 9 or higher for Macintosh" forum as of 10:15 a.m. EDT on 13 November.

  10. OK, let's look at a related Mac User's Guide problem, from almost 9 months ago in the "Retrospect 9 or higher for Macintosh" forum. It's this post by JMcIntire in this thread . He is quoting the immediately- preceding post by me in the same thread.

    "Posted 18 March 2016 - 05:43 PM
        The question is: What is an 'active backup'?

    Your active backups depend on your current grooming settings. If you have grooming disabled then your active backups will be the most recent backup of each source. If you have grooming enabled, we will keep the number of snapshots (backups) in the catalog necessary to carry out your grooming policy. For example, if you have "Groom to keep this number of backups" set to 3, we will keep 3 copies of your recent backups in your catalog, and consider them active.

        But then, why does Copy Backup offer 'Copy all backups' as a choice in the drop-down?

    While I admit this is a bit confusing, and I actually had to test the behavior to be certain, the user's guide is indeed correct. Copy backup scripts will only copy active backups, meaning that with this drop-down option selected, we will copy all of your active backups. I'll preemptively agree with you that this is misleading, and have already logged a bug to correct the wording here."

    Note that, in his first paragraph, JMcIntire uses the term Snapshot, which has been an "un-concept" (in the Orwellian sense) since Retrospect Mac 8. JMcIntire—who as an "Escalation engineer" evidently works for Retrospect Inc. and probably normally for Engineering rather than Support—cannot explain the following quote from page 161 of the Retrospect Mac v.13 User's Guide without indirectly using the term Snapshot.  "Copy Backup scripts are different from Copy Media Sets scripts in a number of ways: • They copy only active backups; Copy Media Sets scripts copy all backups. • They provide different methods for selecting which backups get copied, such as the most recent backup for each source contained in the source Media Set; Copy Media Sets scripts always copy all backups."  JMcIntire's explanation of why Copy Backup scripts are different from Copy Media Sets scripts is much more significant than any explanation DovidBenAvraham might need to provide in the Wikipedia article about the Retrieve and Forget buttons in the List View Toolbar for the Console's Past Backups panel.  It caused me to alter step C2 in this post earlier in the same thread, and it is the updated version of that post which has made that thread wildly popular (1,018 views as of 9 November) in the forum.


    So I have suggested, in my Support Case that covers posts #2 through #7 in this thread, that Retrospect Inc. bite the bullet and "rehabilitate" the term Snapshot in the Mac User's Guide. Doing so will make both explanations much simpler. Besides, the Windows User's Guide still uses the term Snapshot, so Retrospect Inc. can copy passages from that document where necessary.  And, after all, we know that Snapshots still exist in Retrospect Mac; Retrospect Mac 12.5 shows in the Activity detail Status and the Log that Snapshots are being built and copied and compared every time I run a Backup script.

  11. As of today, the Retrieve button in the List View Toolbar for the Console's Past Backups pane is enabled (not grayed-out) no matter which list line I select.  Since I reboot my "backup server" each time I use it, that implies that experimenting with the Retrieve and the list line Browse buttons over the last three days has done something to the Catalog-associated Snapshots for my Media Sets.


    Note that Snapshot became an "un-concept" (in an Orwellian sense) for Retrospect Mac as of the version 8 User's Guide. Not even the Retrospect Windows 11 User's Guide provides any useful answer to this question.  The best answer I have found for the Retrieve button is contained in this 2014 post by Lennart Thelander.  The latest Mac User's Guide also contains no explanation of the Forget button In the same List View Toolbar.


    Therefore DovidBenAvraham will not attempt to provide an explanation for either of these buttons in the Wikipedia article.

  12. DovidBenAvraham now has an interesting question about Snapshots in relation to the behavior of the Retrieve button in the Past Backups panel of the Console.  Since he and I discovered the question in relation to Retrospect Macintosh 12.5, I have posted it here in the "Retrospect 9 or higher for Macintosh" forum.


    Retrospect Macintosh 8 officially did away with Snapshot as terminology, as noted in the last two sentences of the second paragraph in post #2 in the thread where I made my post.  However, since the question relates to the underlying mechanism of Retrospect—which is the same for Retrospect Windows as for Retrospect Macintosh, please feel welcome to contribute in that thread any knowledge you may have.




    P.S.: Corrected first sentence in second paragraph, in regards to precisely where in the linked-to thread I said that Retrospect Macintosh 8 officially did away with Snapshot as terminology.

  13. On 11/3/2016 at 5:26 PM, David Hertzberg said:



    Based on what you've said, I just tried it out again on my copy of Retrospect 12.5.  As described here, I rotate my Media Sets once a week.  I confirmed that, if I select a backup done this week from Past Backups, the Retrieve button is grayed-out.  However, if I select a backup done either of the preceding weeks—onto a different Media Set—the button is not grayed out.  It makes no difference if I mount the drive containing the two-weeks-ago backups; selecting a backup done during that period still leaves the Retrieve button un-grayed-out.


    I guess this relates to what JMcIntire was talking about in this post.  Since I don't have grooming enabled on my Media Sets, the only active snapshots in the catalog are the most recent ones for each Source.  It seems that if the selected backup from Past Backups is one covered by an active snapshot, Retrospect realizes that I don't need to do a Retrieve—I can just click the Browse button for the selected backup at the right-hand end of its list line and then do a Restore of whatever files I want.  Neat, if a little confusing.


    P.S.: Do you think the Mac 13 User's Guide mentions the Retrieve button?  Think again.  How about this Tutorial from Mayoff?  Ditto.



    DovidBenAvraham is trying to precisely understand the concept of the Snapshot, so he can explain it in the "Concepts prior to Retrospect Windows 7" section of the Wikipedia article. Page 38 of the Retrospect Windows 11 User's Guide says "For each volume, one Snapshot is stored in the Catalog File and a copy of the same Snapshot stored on the backup medium (tape, disk, cartridge, or CD).  Following each successful backup or archive operation, the old Catalog File Snapshot is replaced but old media Snapshots remain untouched and Retrospect adds new Snapshots to the medium."  Since the underlying mechanism of Retrospect Macintosh is the same as that of Retrospect Windows—per the last two sentences quoted in the second paragraph of post #2 in this thread, isn't the latest Snapshot copied into the Catalog File for each Media Set—not just the most-recently-updated Media Set?  If so, why is the Retrieve button enabled when I click a Past Backups list line for the latest backup of a volume that was done to a cataloged Media Set that was not the latest Media Set backed up to?  Shouldn't the Catalog File Snapshot for that volume in that Media Set be automatically accessed?


    Actually I now find I need to correct my observation in the quoted post.  This morning, being a Saturday, I ran a Recycle Media Set backup of all my drives to "Media Set White" (my Media Sets are named "... Red", "... White", and "... Blue" in the patriotic American sequence).  I now find that if I click any line in the Past Backups list that relates to "Media Set White", the Retrieve button remains grayed-out.  The Retrieve button also remains grayed-out if I click any line in the Past Backups list that relates to "Media Set Red".  However if I click any line in the Past Backups list that relates to "Media Set Blue", the Retrieve button becomes enabled.  Is there a limit of two Media Sets for which the Snapshot is automatically accessed from the Catalog File?  Or is the limit two weeks?


    Lennart?  Mayoff? JMcIntire?


    P.S.: A bit over an hour ago, following up on my corrected observation in the second paragraph, I bravely selected a Past Backups list line relating to "Media Set Blue" and then clicked the Retrieve button.  I got a file browser window and a message saying I might need to wait several minutes for it to be populated.  I took the opportunity for a short bathroom break, and when I returned to my Mac Pro "backup server" the file browser had been populated with a single disclosure-triangle-equipped line for the backed-up volume.  Retrospect did not call for mounting the HDD for "Media Set Blue" (which I could have done if necessary, since I brought it back from the bank safe deposit box yesterday), which means it used the Snapshot in the Catalog File to populate the file browser.  I therefore conclude that the requirement for clicking the Retrieve button for an old-enough Media Set has to do with some kind of RAM-saving strategy, and is not an methodological constraint of Retrospect.  DBA will therefore not stress it in the Wikipedia article. 


    P.P.S.: Oops, left out "post #2 in" in the fourth sentence of the first paragraph in this post; its absence really made that sentence confusing.

  14. Are you referring to this button?


    attachicon.gifPast backups.png


    That button is not grayed out in my copy of Retrospect 13.5. It pulls up the past backups associated with the selected media set, and enables retrieving the actual snapshot.



    Yes, that's the button I'm referring to.  Thanks, twickland.


    Based on what you've said, I just tried it out again on my copy of Retrospect 12.5.  As described here, I rotate my Media Sets once a week.  I confirmed that, if I select a backup done this week from Past Backups, the Retrieve button is grayed-out.  However, if I select a backup done either of the preceding weeks—onto a different Media Set—the button is not grayed out.  It makes no difference if I mount the drive containing the two-weeks-ago backups; selecting a backup done during that period still leaves the Retrieve button un-grayed-out.


    I guess this relates to what JMcIntire was talking about in this post.  Since I don't have grooming enabled on my Media Sets, the only active snapshots in the catalog are the most recent ones for each Source.  It seems that if the selected backup from Past Backups is one covered by an active snapshot, Retrospect realizes that I don't need to do a Retrieve—I can just click the Browse button for the selected backup at the right-hand end of its list line and then do a Restore of whatever files I want.  Neat, if a little confusing.


    P.S.: Do you think the Mac 13 User's Guide mentions the Retrieve button?  Think again.  How about this Tutorial from Mayoff?  Ditto.

  15. Here's a DovidBenAvraham discovery of a probable "blast from the past" that's not just a problem with the Mac 13 User's Guide; it appears to be a problem with the Mac 13 program itself.  In the "Retrospect Macintosh 8" section of the old Wikipedia article, the bulleted item for the "All-new, customizable [administrator] interface" mentions a Retrieve button in the List View Toolbar for the Past Backups panel.  The Retrieve button in that toolbar is grayed-out in DBA's copy of the program, as it is in my copy.  Under what circumstances is that Retrieve button not grayed-out?


    Here's a clue as to what's going on:  Page 122 of the Retrospect Windows 11 User's Guide, in the sub-section "Restoring in Advanced Mode" of the "Immediate Operations" chapter, says "When you select a Snapshot and click Retrieve Retrospect obtains the older Snapshot from the Backup Set media (which may require you to insert media) and adds it to the list in the restore source window.  The Retrieve button is disabled when you select a Snapshot that is already available."  As DBA understands it, Retrospect Mac 8 essentially did away with Immediate Operations as anything other than a fast way of creating one-shot scripts.  Also, page 267 of the Mac 13 User's Guide has the Glossary of Terms definition "Snapshot— In pervious [sic] versions of Retrospect, a Snapshot refers to the point-in-time file and folder listing that is captured during a backup operation to depict a volume’s state (that is, all its files and their paths).  Makes it easy to restore a hard disk to its exact state as of a given backup. Retrospect now uses the term backup to include both session and Snapshot data."


    So it looks like the Retrieve button in the List View Toolbar for the Past Backups panel may be either an always-disabled holdover from pre-Mac-8 days, or is only enabled if you somehow mount a Media Set so old that it doesn't appear in the Past Backups list pane.


    Lennart?  Mayoff?

  16. Yesterday DovidBenAvraham posted a reply to's Wikipedia talk page, giving a three-sentence explanation of administrators' use of Retrospect that assumed was an employee at the Finnish telco Sonera. replied that s/he is not an employee of Sonera, but only uses them as the ISP to his/her residence.  Since the IP address is therefore probably a DHCP one that could theoretically be re-assigned to another Sonera customer, that in DBA's opinion is why has not officially signed up for a WP account.


    In the same reply DBA thanked for suggesting that he use the Wikipedia-style paragraph-leading-asterisk format for bulleted lists. thereupon removed the Manual-of-Style-violation tag from the Retrospect article, but added a "This article contains embedded lists that may be better presented using prose. (October 2016)" tag.  DBA doesn't think this is a good idea; his bulleted-list items are prose descriptions of Retrospect features, and are presented as part of bulleted lists because there were multiple feature additions to each major release of Retrospect.


    *Indented bulleted list item doesn't work for Retrospect forum—must be a WP-specific feature.


    It does not seem that has actually proposed the Retrospect article for deletion from Wikipedia.  DBA therefore concludes that is simply a carping kibbitzer who either doesn't like articles longer than the WP one on Time Machine, doesn't like DBA's prose style (I don't much like it either), or doesn't like Retrospect.

  17. Early this morning DovidBenAvraham  posted the following reply on's Wikipedia talk page : "Come on, you can't walk away from your criticisms this easily! You're the administrator who talked about proposing the article for deletion. I've just finished revising the article so that a full 50% (13 out of 26) of references are to third-party secondary sources, and—as I explained previously on my Talk page—the remaining references to primary sources are in the article only because there were no reviews of the Retrospect software for several major releases. I think my latest revisions would justify your deleting the tag that reads "This article relies too much on references to primary sources." As far as the "Original research" questions I myself raised for the "Documentation" section, I haven't had an answer yet from the Help Desk, but I think the two issues are simple enough that any WP administrator—including you—should be at least able to express an opinion. As for the "MOS" criticism, I'll try to deal with it this afternoon. DovidBenAvraham (talk) 05:12, 27 October 2016 (UTC)".  Note that the times in the signatures in this and the next paragraph are UTC (Universal Time Coordinated), which means non-DST Greenwich time (4 hours ahead of NYC DST). apparently posts from Finland; Helsinki is 2 hours ahead of UTC.


    A few minutes later posted the following reply to the reply: "@DovidBenAvraham: I am not an administrator. I am an editor like you, but unlike you I don't have any knowledge of Retrospect. Any contributor can propose an article for deletion, where depending on community consensus a proposed article may be deleted. You're doing a good work. (talk) 05:15, 27 October 2016 (UTC)".


    P.S.: The reason DBA was able to revise the article "so that a full 50% (13 out of 26) of references are to third-party secondary sources" is that he belatedly found 3 reviews of Retrospect on the venerable and respected TitBITS.com.  Those reviews, although not as detailed as the corresponding Retrospect Mac User's Guides, contained information that was detailed enough to serve as substitutes for those UGs as references in the WP article.

  18. Last weekend DovidBenAvraham ran into additional flak from a senior WP editor who goes by the catchy name (s/he apparently works at  the formerly-state-owned Finnish telco, and—DBA thinks—doesn't want to register a handle with WP because that would restrict anyone else from using the same IP address for WP work).  His/her main complaints are "This article may contain an excessive amount of intricate detail that may only interest a specific audience" and "This article relies too much on references to primary sources".


    DBA justified his reliance on primary sources (meaning the Retrospect User's Guides) by the fact that—because of the debacle of Retrospect Macintosh 8—there has apparently only been one review of Retrospect on an reputable site since then.  DBA accepted in part the "excessive amount of intricate detail" complaint, and deleted about 14% of the article that really just re-hashes User's Guide material rather than provides the intended informative overview of Retrospect—plus a few short paragraphs that tout speed improvements rather than describe new features.  The UG-level deletions have been copied into the Talk page for the article, where you can still read them if you care.

  19. The Mac KB article for Block Level Incremental Backup, reachable from the Release Notes for Version 11.0.0, is here.  It has now been "disappeared" from the directory of KB articles shown under Support, but is still hiding on the Retrospect Inc. website.  The article, as is typical for KB articles, omits a step—that you have to click the Pencil icon at the bottom of the window to get the dialog.  I found that trick out by watching this Tutorial; Heavens to Betsy, Mayoff, the video is over 3 minutes long—think of the children's attention span!   (derek500 doesn't like me to use smilies in these forums)


    I only have Mac version 12.5, but the dialog—or anything else in the Version 13 UG—doesn't have the security settings option Scillonian describes.  I guess the option is Windows-only.

  20. In the Retrospect Mac 13 User's Guide, the section "Using the Dashboard" on page 208—in the chapter entitled "Managing Retrospect"—still describes the Dashboard prior to Mac version 11 and Windows version 9, when it was simply "a [sic] overview of some of the reports that come with the program, as well as the ones that you have created yourself and chosen to display in the dashboard."  However the section "High-level Dashboard" on page 35—in the chapter entitled "Introducing Retrospect"—correctly describes the Dashboard as it was enhanced in Mac version 11 and Windows version 9. 

  21. As noted in this thread in the Windows Professional Forum, DovidBenAvraham has just done a major enhancement of the Wikipedia article on Retrospect, bringing it up from around 2006 to the present.  He uses Retrospect Mac—although he tried to give adequate coverage to Retrospect Windows, and therefore recently noticed an interesting documentation problem within the Mac Version 13 User's Guide: "Because the Retrospect Windows User's Guide is now substantially different from the Retrospect Macintosh User's Guide, there seem to be some problems keeping the updates in sync. For example, both User's Guides have since v12/v10 had a section in an introductory chapter entitled 'High-level Dashboard', which describes the feature added in v11/v9. However the latest Macintosh User's Guide also has a section entitled 'Using the Dashboard' in a chapter towards the back of the UG entitled 'Managing Retrospect'; this section still describes the Dashboard prior to v11/v9, when it was simply an overview of some 'reports that come with the program' as well as some 'ones that you have created yourself'."


    Just sayin', on DovidBenAvraham's behalf.

  22. On 10/9/2016 at 4:00 PM, David Hertzberg said:

    The friendly soul, who goes by the Wikipedia handle DovidBenAvraham, yesterday ran into problems with a senior WP editor who goes by the handle Diannaa.  She wrote "The reason I removed the content is because the quotations were excessive, comprising 30 percent of the article."  He wrote back "The reason I put so many quotations into the 'Retrospect Macintosh ...' sections of the enhanced WP article is that I would find it difficult to succinctly paraphrase descriptions of software features, particularly when those feature descriptions were originally written by expert technical writers."





    DovidBenAvraham has finished rewriting the sections of the Wikipedia article that had "excessive" quotations from scratch in his own words.  The latest version of the (old) "Retrospect (software)" article is here .  Diannaa said she was happy with the first section of the rewrite, and DovidBenAvraham hasn't heard a peep from her since.


    Please let me know of any changes you think should be made in the article, or make them yourself—but remember that DovidBenAvraham reserves the right to revert them if they seem unjustified.

  23. >cloud backup providers have limits ≦ 30GB per day per user


    I have never heard of Amazon S3 or Google Cloud having a limit like this. I don't think this is a true statement.



    Mayoff, I said cloud backup providers.  I hate—on this forum—to provide a link to the competition, but here's a link to a section in a generally rather favorable review of CP.  It says "The backup started rather fast at around 300KB/s but later plummeted to only about 30-50KB/s which is only a fraction which our connection is capable of. One reason might be our location: we’re based in Germany so files need to travel over the Atlantic all the way to California. We tried to change the network and CPU settings but couldn’t get an improved throughput, unfortunately."  Please check my arithmetic: 50KB/sec * 3600secs/hour * 24hours/day * 1000bytes/KB = approximately 4.3 billion bytes/day = 4.3GB/day.  So 30GB/day/user doesn't sound unreasonably slow as a practical limit for an cloud backup provider.


    You're correct that, since Amazon S3 and Google Cloud are not cloud backup providers, they wouldn't necessarily have practical limits like that.  So let's look at an example of reported upload speeds for Amazon S3.  In that example, the reported worse-case upload speed is approximately 10MB/sec.  At 86,400 secs/day, that's 864,000 MB/day = 864GB/day.  So uploading wouldn't be so bad for ggirao, but it would still take him/her 5 days to do the initial upload at that pessimistic speed.  But then there's downloading for a restore, which might run at the same speed.  And this just in: See the 2013 comment by Christian C. Berclaz in the review linked to in the second sentence of my first paragraph; he says it took 3 months for him to upload "3.4Tb" (I hope he meant 3.4TB) to Amazon Glacier.


    I think jethro must have done equivalent analysis, which is one reason why he decided to only store a cut-down copy of his Media Set in the cloud.  And when we start talking about cutting down a Media Set to most-recent backups , which I did in the post linked to as the second link in the first paragraph of post #3 in this thread, we have to start thinking about the problem I suggested two solutions for in a Product Suggestions thread


    P.S.: Moved last sentence in second paragraph to a new third paragraph, to which I added another sentence linking to the suggestions I started making 8 months ago about  enhancing Copy Media Set so a user doesn't have to first copy a Media Set to a destination with equal capacity, and then Groom it down.


    P.P.S: Added sentence to second paragraph, saying commenter reported upload speed in 2013 on Amazon Glacier equivalent to my CP optimistic estimate in post #3.

  24. IMHO the first thing you need to think about is how to upload that 4.5TB of data to Amazon.  I can't find the Web page early this morning, but IIRC leading cloud backup providers have limits ≦ 30GB per day per user—because they have to accommodate thousands of users simultaneously.  At that rate it would take you 150 days—almost 5 months—to upload 4.5GB of data.  What you need to do is to rent an Amazon Snowball appliance; it will cost you US$200 plus shipping (both ways).  You then need to follow the procedures in steps C1 through C5 of this post, replacing the first sentence in step C3 with "Copy the new local Media Set to the Snowball appliance with the Snowball software, and ship the Snowball appliance containing a copy of the new Media Set member back to Amazon."  Note that you will need 5TB of extra local disk drive capacity for steps  C1 through C3, but—since you won't be shipping the drive(s) containing that local Media Set anywhere—maybe you can temporarily re-purpose part or all of a 5TB drive(s) you have sitting around your local installation (you must have a lot of money at your disposal to be even considering this, so I assume you have local disk capacity "coming out of your ears").


    If you really intend to store that entire 4.5TB of data on Amazon, you can ignore anything about Grooming in steps C1 and C2.  However, once you notice that the price you can get for oil from your on-the-premises oil well has dropped (insert appropriate smiiey here), you will probably want to revise your plans and do something like what jethro has stated he wants to do in the thread linked to in the previous paragraph, which is to store the most recent 1TB in the cloud.  Personally I think cloud storage for large copies of Retrospect backups is a snare and a delusion, and that people who are thinking about it should think instead about storing off-site copies in a safe-deposit box at their local bank branch (or at a manager's home)—which is what I do—instead.  OK, Mayoff and the Retrospect Inc. engineers, go ahead and hate me for saying that.

    • Like 1