mORMot Open Source

TexShop on a Mac.
Latex as my engine

My first few lines are:
%\documentclass{tufte-book}
\documentclass{book}

I didn't have luck with tufts-book, so I went with classic {book}

For conversion to Ebook I've been trying calibre.

Good news, it's available, neater, better, more topics, cheaper. 
I've started a new thread about the 2nd edition.

Great timing Albert, I just signed in tonight to say

Good News: HOLD THE PRESSES

I've spent many hours yesterday and today converting the text to LaTex.  And there are some new chapters which haven't been made public yet, I'll work tomorrow to putting them into LaTex too.  I'll send it out for printing on monday, proof/fix in a while loop.

A second edition will be available on Amazon by mid Febrary.  I'll announce it here when it is available.

Erick

> (Before counting me off as some Microsoft basher, let me assure you that Excel and Word are my two favorites pieces of software of all time, period. I truly think they are a monument to human ingenuity.)

I can tell you I was seriously thinking about it a few times in the last few months, as LaTex is something I've used years ago and a close friend just rediscovered it last year and reminded me.  But I had already invested a lot in the Word version and didn't know a good way to convert Word to LaTex.  I figured that would wait until the next project.

Well, I'm ready to renew my Knuth membership card today.  I still have all the books.

The only thing I liked about Word was the grammar checking.   But the autocorrections were killing me as I cut and paste code.

> There are ready-made classes to create your book. For example, I took a sample chapter provided earlier in this thread, opened it with Word, saved as .docx, then used docxtolatex to convert the file. After some _very_ minor editing, got the following:

I will definitely pursue this option.  Thanks for some excellent advice and the effort you put into this.

I wrote:
"So far I've written chapters about Sharding/De-Normalization..."

and that first six page chapter is now available.

It, and future chapters as they come available, can be downloaded at http://www.erickengelke.com/ewbmormot2.zip

The same files are included in my large ewbmormot.zip, but that file is huge and I don't expect everyone to keep downloading it every week..

If you can send your suggestions to my Email address, I will communicate with you to refine the topics in Email before I post a complete article to the web.  Try to include a little description of each so I can understand why it seems confusing.

Thanks
Erick

I updated the file the clarify a point a few hours later.
Erick

Hi,

I'd say it depends on your client.  You said you are getting my book, it shows using mormot to output JSON, and using web clients based on Elevate Web Builder which accepts that JSON as a dataset and lets you walk through the JSON as a TDataSet in terms that feel very Delphi-like.

I would recommend using the SOA RPCs in the book, as they are the most flexible way of accessing data.

Good luck.
Erick

The formatting was a problem.  I agree generally, though we will all have personal opinions on some aspects.

I will be significantly reformatting this book in 2017, probably in March as I will have a month at home for a medical leave.  There may be a few minor edits for clarification too.

That doesn't help you directly, but it does benefit the Delphi/NewPascal/Mormot/EWB communities we all care deeply about.

I will update this thread when the new revision is available.

It's new to me too, but it makes sense instantly!

My suggestion of using multitenancy is because people like using on application where they can, rather than a bunch of smaller applications they might have to sign into.

In practice, I find I struggle between making small single-purposed apps that are well defined, and later addressing larger needs with a single-mega-app.

If you have single sign on (SSO), like CAS or OpenID, it makes less of an inconvenience for the user to use multiple apps as they don't have to sign in each time.

Erick

>I've ordered Erick Engelke's new book, but it won't get here until next week, and I have no idea if it covers >this. And I'm impatient. So I'm asking here.

Hi David,

Since you've ordered the book, you'll be happy to know it comes with source code that auto-adds people.  In my implementation I first check that they pass an active directory or NT-styled password challenge, but you can change the code to support whatever addition requirements you desire.  I wrote it to support EWB clients in a company, but it works just fine for Delphi clients too.

If I remember correctly, you authenticate with a pseudo-guest account to gain access, then you send whatever credentials to the server using the guest account.  It processes the request and auto-adds a userid. Then you log in with that userid and you are all set to do anything. 

Merry Christmas,

Erick

I'm sorry you are disappointed.  As I listed in my blog, the first few chapters are to get new people started that are not familiar with mORMot.  The steady stream of beginner questions on the board indicates there was a need for such a section.   But once you get past that... hopefully not too much more disappointment.

I disagree that there is no SOA as there are both Delphi and EWB SOA examples.

There are also examples of SQL and ORM methods depending on which model you need to use for your project.

Starting page 127 to about page 190, it should all be new material to you because it covers REST and SOA RPCs with Elevate Web Builder that is definitely not documented anywhere on Synopse and I think the Web is the right direction to head for market acceptance. 

It includes a new library which supports REST+SOA with mORMot from EWB, which does auto-reconnect of sessions and supports Active Directory for authentication on all web browsers.  We're using it in production here for a couple of months so it's stable and opens new opportunities that didn't exist prior to the publishing.

I will take your recommendations on DDD, etc. and try to incorporate them into a future edition.  That doesn't help you right now, but it will help others in the future.

Amazon is a world-wide company.  My copy was printed on premium Canadian paper, but they demand-print for US, EU and UK and quality may vary by market, the only choices they offer the author are cream or white paper  and glossy or matte cover.  I picked them because shipping to EU and UK is expensive and past experience indicates people are keen for fast and, I believe, free delivery.   I've never seen a Polish book so I have nothing to compare.

The font is a readable 12 point recommended by the publisher.  Apparently 10  or 11 are hard-to-read for many people.  To compensate for the font size, the book uses a large page size of 8x10.   Book sizes are standardized for shipping purposes.

Some of the text boxes were reformatted by the publisher, and I agree there is sloppiness introduced in that last section where they messed up the frames.  Probably I'm to blame for having used MS-Word with text boxes.  This didn't show up in the pre-publishing copies because it had to do with PageMaker or whatever they use.  I do apologize for that.  But every time you submit the document they reformat it.  I thought it was gold this time, I was wrong.

I know it's in line with prices for timely US technology books I buy.  I do not know the price of European books.  Approximately 1/3 of my books published on Amazon are sold in the EU so I think it must not be too outrageous.

On-Demand publishing is more expensive, but it is perfect for things like mORMot which gets extended regularly and which is so niche that no regular publisher would touch it.

This book has been a passion for a year and a half, with countless hours going into documentation, writing stripped down working samples,  and also building the next technologies which are unencumbered open source in the accompanying download.

I hope most readers will look past the production problems and see it as a valuable and timely resource.  I could have spent another year finessing the text and layout, but AB and I thought timeliness was critical in our Emails, I was starting to think he might be giving up on me :-).

The price reflects two facts: one the market size is currently very small - meaning its hard to recoup my expenses, and two, the fact I have to legally turn some of this into money to qualify for writing off expenses on my taxes.  Otherwise it's just an expensive time-consuming hobby to the tax man.  I promising you, writing a computer text about a wonderful but niche product is not a lucrative deal.

I've been heavily contributing to the open source comunity for almost 30 years, sometimes as programmer, sometimes as book author, often both.   And I think you have to look at how something like this book will help grow the project, and will save many people many days or weeks of effort, and in the case of EWB, open up new opportunities which didn't exist in that form prior to my work.  In my case, my books subsidize my development environment and activities.  I hope it will benefit you.

I guess people might wonder why pay for something related to open-source, it's all free, right?   To the contrary, I hope people reward AB by trying to help answer questions to save him him time, and by contributing to the project however they can, including donating source and helping financially when they can.

Sorry, I've simplified it a lot a few seconds ago.  GPRS... I remember that.

Erick

Yes, the source code url is in the book.  I don't have it handy at the moment but it's on page 17 under "Acquiring the software".

The cover art is for the zen-like peaceful feeling one has when the database is working hard and many people are using it, and everything is just humming along.  You look at the netstat to count sessions, you look at the disk lights, and the process stats, and you just know that people are being served their data and nobody is complaining.

Erick

It is available in Germany, and they've already reported a few sales there this week. 

https://www.amazon.de/dp/1517516005/ref … +databases

Erick

I've figured out almost everything else I need, but this escapes me.

14 maps AllowRemoteExecute to what?  It's not function #14.  Why is it not 13, or 12.  What does 14 mean?

Also, the CRUD numbers - I'm assuming a bitfield for 0-256 tables ordered from the database tables.

So I tried an experiment:

// grp.AccessRights was = '14,3-256,0,3-256,0,3-256,0,3-256,0';
grp.SQLAccessRights.Edit(1, True, false, false, false);
TSQLLog.Add.Log(sllTrace, 'accessrights % ',  [UTF8toString(grp.AccessRights)]);

to see how the bits get changed.  But they didn't.

20161124 16542458 res   SynSQLite3.TSQLDatabase(027CAC30) [{"ID":5,"Ident":"EWBlogin","SessionTimeout":60,"AccessRights":"14,3-256,0,3-256,0,3-256,0,3-256,0"}]
20161124 16542717 trace accessrights 14,3-256,0,3-256,0,3-256,0,3-256,0

No change.  So I clearly don't understand this.

Thanks in advance
Erick

No, it does not.

TLS < 1.2 is  not secure anymore IIRC

Okay, my SSL cert doesn't work with IIS either.  So that will help me debugging, thanks.

Erick

I'm thinking of doing this by opening a websocket.  The server will  (probably) close the socket when it goes down gracefully, or the socket will close in time if the network connection fails (assuming keepalives).

E

Just an update, I'm expecting the mormot book to be ready before Christmas, actually aiming for late November.  It covers essentials about mormot, and building web clients with Elevate Web Builder.  So it's end-to-end pascal!

I'll post some example videos soon.  For the time being, here is a sample of the recaptcha done with
Elevate Web Builder, showing how to weed out humans from the many bots that will try your code.

http://dark.uwaterloo.ca/temp/recaptcha.html

The book will be technically published in Canada, but available through Amazon EU, UK and US.  Today's Canadian European Trade Agreement may have an impact on price for European readers.

Erick

Take care
Erick

Apple's documentation reminds developers that their mobile versions should have different goals than desktop or web apps, it's not just a screen real estate issue but also a matter of thumbs and shorter attention spans on mobile, so make it dumbed down.

Erick

I think that doing RPCs to mormot encourages one to do their business logic in the server, and leave the client for handling presentation only, which is how I'd recommend anyone program - separating logic from presentation. 

a) I recognize the SMS in the example you gave.  That one looks like SMS, it would take *me* about 15 minutes to code it in EWB except for the circled images, that would take longer because circle images aren't a native thing in EWB.  I don't know how long it would take other people, I think I might
have more experience than many.

b) Like many compiled languages, Delphi and FPC included, the executable is larger than hand-coded assember/javascript.  In fact, the first thing most people say when they compile for FPC/Lazarus is WTF, why is it so hug?  But the overhead grows linearly from that point.  For example, I have a huge 30 form application I'm doing in my job, it compiles to UNCOMPRESSED

       871,368 faculty.html
    1,672,593 faculty.js
      857,422 libsodium.min.js
          5,015 sha256.js
        68,988 shield.png
               5 File(s)      3,475,386 bytes

When compressed, the faculty.js file will be much smaller than lib sodium.min.js.  And the png and jpg images I download will be the real bandwidth hogs.

When you call services-methods or RPCs, you have a completion handler, which is the other half of your subroutine.  SMS handles that differently
than EWB - and again, it's a preference matter as to which is superior.  I think completion functions are cleaner for me to look at.  Some other people
rave about .. I forget the name... it's been a long day... but writing nameless functions, or whatever they're called.    I liken them to 8 bit assembler where you could only jump a short distance and were stuck with intermixing your code.  I think it's hideous and error-prone.  But that's how javascript people like to do it.

c) I'm not the author or owner of EWB, I've paid my subscription fees just like everyone else.   I don't know the exact price, but it was half what I pay for Delphi, and I find I'm using it more than Delphi these days.  No, it does not have code insight or code completion at this time.

The code generated for user functions is efficient, but there is overhead for the windowing library.

Erick