@Alec Mihailovs 

Not sure to understand what you mean, but for sure, the link in Will's first post now points to the wrong place  because of his branching of that conversation.

@Will 

Certainly, time of one, or both, of these posts was wrongly recorded. In my backup I observe them in the correct order, though not in "stamped" time order:

Maple Advisor Database                        
Comment on 2008-06-12 04:34 from alec

Individual vs collaborative                   
Comment on 2008-06-12 03:33 from jakubi

So, an hour time zone shift or a similar problem has occured that recorded some post time wrong making appear my post one hour earlier when it was an answer to Alec's post, hence later.

This all means that you cannot rely blindly on time stamp order, and some reordering will be needed in some cases as this one, to restore logical order within the thread. If you have access to a runnig version of Primes 1, you can check this fact easily. 

I have to add that removal of most of the thread tree structure, makes following the logic of the threads even harder. So, once again, I request you restoring this structural information. If not by indenting, by some other marking.

@Will 

I hope that you could recover all the posts that did not get imported. When done, probably, the next issue to be addressed is their sorting. As they stand now, some of them are not placed in their correct order, breaking the logic of the thread. For instance, right now it appears:

Individual vs collaborative
Commented: June 12 2008 by jakubi

Maple Advisor Database
Commented: June 12 2008 by Alec Mihailovs

while my post is an answer to Alec's post, hence it should go later (they appear in the correct order in my local backup copy). 

But just note that if somebody is to blame for pointing out missing posts in that thread its me not Joe. 

@Christopher2222 

In my opinion, the responsible thing to do would have been to keep a backup of the whole Primes 1 system, its database in particular, so that it could be run again e.g. over an auxiliary system in case of (likely) problems.

In the past, I have done system administration in my department, and occasionally I did some stupid things, but always kept the means at hand to be able to undo whatever changes I did.

Edit: I have just met this post by Bryon. In his words the backup exist, so now I cannot understand why it is not used for recovering the missing information. Clearly, Will can do something.

The main block of 11 missing posts is located between Jacques post Whenever possible and mine post Case type/anyindex. This is a preliminary list of their headings (copy&pasted from my backup copy):

patch problem

Comment on 2008-03-29 21:05 from Joe Riel

Here is the trick

Comment on 2008-04-02 10:09 from JacquesC

thanks

Comment on 2008-04-02 12:15 from Joe Riel

typematch

Comment on 2008-04-02 12:21 from JacquesC

nasty drawback

Comment on 2008-04-02 15:40 from jakubi

Maybe the drawback has been fixed?

Comment on 2008-04-02 16:34 from JacquesC

So, the issues are

Comment on 2008-04-02 17:09 from jakubi

And the answers are

Comment on 2008-04-06 10:44 from JacquesC

orphaned features

Comment on 2008-04-06 15:08 from jakubi

pattern matching

Comment on 2008-04-06 22:54 from JacquesC

continuation

Comment on 2008-04-07 02:25 from jakubi

@bthur 

Time may tell you, but in what concerns to me it works all the other way round.

The blog articles that I have posted in Primes 1 were on subjects that I needed to investigate and think thoroughly. And writing those results in a blog article, when I did it, was just a final publication step of that research process (not intended to be updated regularly). So, I would never have posted those articles in casual fashion. And, your downgrading of user blogs to ordinary posts, makes me avoid posting in Primes 2 on similar subjects.

Missing a Maple journal and a Maple wiki, the blog/book articles in Primes 1, have played an almost unique role as a basic reference material on current Maple. It seems to me that you have not realized fully of this fact.

So, yes, the term is loaded for me, and for very good reasons. And I find that the Maplesoft blog articles are well thought out and original. So, it seems to me that it is loaded too for their authors.

@Will

It is not true that additions to primes 1 threads are unlikely. From past experience, they may occur months or even years after start, as you could check. Please, give my primes 1 subscriptions back.

About IDE issues, you may get some information here.

About IDE issues, you may get some information here.

@Joe Riel 

A method that I use to get some information about undocumented, builtin routines is producing errors. Moreover, I have found that error messages provide, sometimes, even better information than the documentation itself. In this case:

bind(a=b);
Error, bindings of the form a=b only allowed in `use` statement

bind(a);
Error, `a` is not a module or member

bind(a,b);
Error, `a` is not a module or member

bind("a");
Error, no bindings were specified or implied

Anyways, the obsession for undocumentation is beyond my comprehension.

@Christopher2222 

I agree with old primes easiness of navigation, though the problem there was that the indenting method made deep level replies being displayed so narrow that they became too hard to read. A better balance between navigation and reading convenience is needed.

The numeric problem seems to be the failure of the implemented Levin's u-transform algorithm. Tracing shows that it gets a result, but not good enough, and gives up after many iterations:

trace(`evalf/Sum/LevinU_Symb/1_inf`):
evalf(Sum(ln(n)/n/(n-1),n=2..infinity));
...
                                    N := 43

    u_new := 1.257884847516482586720354280388654480463785060997874308221517

 prec_new := 0.0001583494697352504123705658970687817873585232635970462080245253

                                 progress := 0

                                   fl_pr := 0

                               inf_cond := false

                                  fl_inf := 0

                                  fl_fail := 2

                                fl_fail_max := 2

    u_old := 1.257884847516482586720354280388654480463785060997874308221517

 prec_old := 0.0001583494697352504123705658970687817873585232635970462080245253

  _Sum_guess := 1.257884847516482586720354280388654480463785060997874308221517

<-- exit evalf/Sum/LevinU_Symb/1_inf (now in evalf/Sum1) = FAIL}
                               infinity
                                -----
                                 \        ln(n)
                                  )     ---------
                                 /      n (n - 1)
                                -----
                                n = 2

Yes, I have these files:

Download 143_Surfplot.lib
View file details

Download 143_Surfplot.ind
View file details

The links that I have posted in the other thread worked at that time. I do not know why they fail now.

Yes, I have these files:

Download 143_Surfplot.lib
View file details

Download 143_Surfplot.ind
View file details

The links that I have posted in the other thread worked at that time. I do not know why they fail now.