Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Programming Books Media Data Storage Handhelds Book Reviews IT Technology Hardware

Server CE Database Development with .NET 162

Posted by timothy from the select-where-as dept.
William Ryan writes "SQL Server CE is Microsoft's preferred database backend for its Compact Framework initiative. Compact Framework is a cool technology, but it's still in its infancy. This book does a lot to help you get started using SQL Server CE." Read on for the rest of Ryan's review, or revisit his November review of a related book, The Definitive Guide to the Compact Framework .
SQL Server CE Database Development with the .NET Compact Framework
author Rob Tiffany
pages 450
publisher Apress
rating 10
reviewer William Ryan
ISBN 1590591194
summary A start to finish discussion of using SQL Server with SQL Server CE on the Compact Framework

This book comprises 10 Chapters in just under 450 pages, including the indexes and usual book stuff. Since CE supports ASNI Sql, there are a few chapters that discuss using SQL to manipulate databases, which are probably not necessary for most database programmers. Chapters 4-8 are dedicated to DDL (data definition language), DML (data manipulation language), and taking advantage of metadata. It's a good discussion of these subjects, and I guess an author must include them to be thorough, but if you aren't that familiar with SQL, you probably have some learning to do before diving into data-driven PDA apps.

Enough about the background, though. The book really excels in two areas, one of which I think is probably useful to any developer, even if you don't use the Compact Framework or Microsoft Products. That area is security.

Far too many developers blow off security concerns, or claim to care but do little or nothing about actually increasing security. Let's face it: no matter how secure your OS is, no matter how killer your firewall is, today there are a lot of people trying to break your app and they aren't always outside of your company. Tiffany points to a GIGA Information Group article criticizing the industry for ignoring security on mobile devices.

A lot of what he says is focused on security issues that are 'common sense,' and yet ones that people ignore all the time. It's kind of a shame that a writer needs to explain the benefits of using 'Strong Passwords,' but let's face it, no matter how well you write your app, it won't be secure if you leave the front door open.

In no way am I saying that the author's discussion of security is limited to such elementary topics, but he does a great job of bringing many issues into focus and suggesting ways to deal with them.

The other area that this book really excels in is getting you through replication. This is not a fun topic if you don't know what you are doing and there isn't a lot of literature out there to help you get through PUSH/PULL subscriptions and the like. Pragmatically speaking, of the topics this book covers, Tiffany's coverage of replication is probably going to benefit people the most, because if you can't sync your PDA with your server, you are effectively out of gas. If you aren't a Sql CE user you won't appreciate the value of this chapter, but love MS or hate them, the newsgroups and forums are filled with folks with the same sorts of problems that the author works diligently to get you through.

It's hard to know what will and won't work yet on the Compact Framework and CE. It's quite helpful to have a list of common functions that are supported listed in depth -- another thing I liked about this book.

What else? Well, the text was well written, very similar to his last book on Pocket Access (Pocket PC Database Development with eMbedded Visual Basic) and easy to read. If you are a total newbie to CE, you can use this book and hit the ground running. Everything that you need to write professional apps is included, and I can't find anything that Tiffany omitted.

I really appreciate the fact that the author wrote an entire book on such a niche subject. Many areas, particularly the Compact Framework, don't have a lot of literature on them and if you are writing SQL Server CE, you are on your own...until now.

If you develop in CE, or plan to, this book is a Must Have.

You can purchase SQL Server CE Database Development with the .NET Compact Framework from bn.com. Slashdot welcomes readers' book reviews -- to submit a review for consideration, read the book review guidelines, then visit the submission page.
This discussion has been archived. No new comments can be posted.

Server CE Database Development with .NET More

Server CE Database Development with .NET

Comments Filter:

  • can't wait for the anti-.net comments (Score:2, Insightful)

    by Anonymous Coward
    no, really, i can't. bashing things out of ignorance is fun and easy.

    • Re:can't wait for the anti-.net comments (Score:5, Insightful)

      by Anonymous Coward on Wednesday December 17, 2003 @12:08PM (#7746088)
      "no, really, i can't. bashing things out of ignorance is fun and easy."

      (Posting anonymously to protect myself and my employer).

      No ignorance here. I worked on a .NET Compact Framework (via VB.NET) + SQL Server CE application in the earlier part of 2003. Note that while they integrate with their namesakes, .NET CF and SQL CE's implementations appear to share almost nothing with their desktop counterparts.

      I cannot emphasize strongly enough how much this stuff sucks.

      Let's display a form with maybe 10 elements (a few text fields, a few static text labels, a few command buttons). It takes maybe half a second to a second, highly visible control-by-control drawing that the user can see.

      Let's perform a simple query, like SELECT id, name FROM items WHERE parentid = , returning a dozen or so rows. Yep.. half a second. From what I could see a lot of overhead was involved in passing the data from SQLCE to .NET, representing it all as ADO.NET or whatever objects along the way..

      By the time you're bringing up a form, performing a couple of queries and throwing that data in said form, the user is waiting a couple of seconds. Every. time. they. tap. something. It is not good enough. The resulting application is a joke.

      Databases can be big, so you might want to put them on external media ie a compact flash card. Except you point SQLCE at a db on a card and it just doesn't frickin work!

      This is all based on a fastish (400MHz XScale CPU iirc) Dell Axim. Don't even think about throwing this stuff on an older device.

      By the way, Microsoft .. requiring a total of about 8 megs of runtimes (.NET CF + SQL CE) is not something handheld users are accustomed to.

      • Works fine for me on a 400 MHz PXA255'd XScale. It's running WinCE.NET 4.1, which includes the .NET CF. I've never used SQL CE, I've used the built-in db system. You must be working with older software, PPC2k or 2k2. Things move forward- yeah, it's a bitch, but would you have us all writing code in assembler instead?

        If your company has any other .NET CF jobs, send 'em my way- I'd be happy to do some consulting.
      • Have you tried caching data or other tricks to try and improve performance? Where you repeatedly opening and closing the data connection? No doubt that .NET is notoriously slow, but there are some things (maybe you've already tried them) to help mitigate this.

      • It's a friggin' file-based RDBMS - if you can even call it that - on an extremely compact piece of hardware. What do you expect? There is no service that runs and takes care of memory and performance issues, it loads the DB file when needed and pulls data directly from the file.

        Frankly, I agree it isn't a great RDBMS for what you're probably used to, but it's great on a Pocket PC.

        • And to add a little to that, the overhead of adding a SQL compliant query engine in such a tiny space is hard work.

          I have been using PERST [garret.ru] targeted at developers who need an embeded database. But it's not relational, it's OO, which can be a benefit in some cases. You have to establish your own indexes and relationships between these objects. it's not a full blown SQL server, far from it. But on a small device overhead is everything and this thing is bare bones. It can be configured easily for the platform,
      • Take a look at Sybase SQL Server Anywhere [ianywhere.com]. It is a very fast embedded SQL DBMS. Very nice.

      • Sadly, the 400 MHz XScale PocketPCs are about 2x SLOWER than the older 206 MHz Strongarm PocketPCs. I've done C development for both. One test case takes 15 minutes on the 206 MHz Strongarm PocketPC. On the 400 MHz XScale PocketPC, the same test case take 30 minutes. So the XScale is about 4x slower per MHz than the older Strongarm processor.

        I even used a beta version of Intel XScale-optimizing compiler. The object code as larger yet slower!!?
        • Have you tried something with the latest and greatest pocketpc 2003 os? I've noticed my application performs significantly faster on a new dell axim with pocketpc 2003 than on the slightly older dell axims with pocketpc 2002.

          • I have only done some basic testing on PocketPC 2003. It felt slower to me, but I was using a pre-production beta version of the new HP iPaq. I heard that it should be much faster because Microsoft finally created an XScale-optimized build of the PocketPC OS. Before Microsoft was only building for generic Strongarm and forcing XScale vendors to run the Strongarm build.

            Microsoft builds CE for about a dozen of differenct processors. Why would separating XScale from Strongarm be so much more work? Maybe part

  • compact? (Score:1, Funny)

    by bash-2.02$ ( 732038 )
    regardless of microsofts ability(or inability, whichever) to produce good software, it is never small. how can "compact" and "microsoft" be in teh same sentence...

    okay, perhaps, "microsoft's focus on security is compact."

    before you flamebait me all to hell, its a joke and i use windows almost exclusively because it runs the software i need to run.

  • SQL Server CE (Score:1, Funny)

    by Anonymous Coward
    This book tells me how to run an enterprise DB on my PDA? I had no idea they were that powerful now.

  • Sql Server CE? (Score:5, Informative)

    by Hayzeus ( 596826 ) on Wednesday December 17, 2003 @11:36AM (#7745797) Homepage
    Geez -- a crippled db platform if there ever was one. Calling it "SQL Server" is really a major misnomer -- the two are not even remotely similar in even basic features. Obviously, this would be expected to some extent -- but the name tends to mislead. I had the misfortune of working for an employer who wanted the same application working on both SQL Server and a handheld version on CE. They expected significant code to be shared between the two, and ultimately got a nasty surprise when this proved infeasable, given the alotted dev time frame.

    • Re:Sql Server CE? (Score:1, Interesting)

      by Anonymous Coward
      The name was chosen by the branding bozos, not by the product development team.

      SQL Server CE is called "SQL Lite" internally at Microsoft. Different team, different thing than SQL Server.

      • "SQL Lite" would be a far more fitting name -- it is obviously NOT the same product. It's not really a bad product as far as it goes -- but calling it "SQL Server" was a really unfortunate decision (I would have guessed that their marketing dept. came up with this one).
    • The fact that its a database app that supports ANSI sql, in my book, are basic features that are remotely similar.

      • If you read the small print, you'll find that SQL Server is only claimed to support SQL-92 Entry, which is a tiny subset of SQL-92 that doesn't even include VARCHAR. Yet it still fails to conform to that; it follows SQL-89 in some cases where SQL-92 differs. If you care about ANSI conformance I think you'll be better off with PostgreSQL.
      • so then why not call it Access CE or something similar?
      • Heh, that's what I was thinking. The fact that it serves databases out to apps and uses real-live SQL seem to be the most important features. I'm sure it doesn't do any of the fancy-pants enterprise stuff of MS SQL or DB2 or Oracle, but it is still an SQL ... server.
    • Informative? Really??? The documentation spells out very clearly the similiarities and differences - you did read the documentation right? Not just the marketechture brochurementation? There is a great deal of Transact SQL that is compatible between the two, and they are ANSI compliant. What more did you expect out of a db engine running on this class of processors? I'm sorry your management is gullible, but that isn't SQL CE's fault.
      • What more did you expect out of a db engine running on this class of processors?
        Oracle, DB2, Postgres, Sybase and so on used to run quite happily on machines with far less processing power. The SQL92 standard dates from 1992; I'd be surprised if today's handheld is not at least comparable in processor speed and RAM size to a 1992-vintage Unix box.
        • My bad. I should have mentioned IO constraints as well. How well does oracle run on a pocket pc, it being a speed demon and all?
          • Sure, IO performance may suck, but that's not a reason to support only a subset of ANSI SQL. Not if you already have a database engine that does the job properly, and reduced speed is the only difference moving it from a server to a handheld. For small databases RAM should be big enough to cache most things anyway, so I/O speed doesn't matter that much. You could probably put the transaction log in RAM too.
            • There is no ram on most pocketpcs. or rather it's all ram. You see persistant storage and ram both reside on the same medium. The ram is the same speed as the storage, so it's pretty slow.
      • Uh -- yes, although apparently the project planners didn't. To be fair, this was NOT 2.0 -- on the platform they had targeted major features were missing (like stored procedures, although cascading deletes were available).

    • Re:Sql Server CE? (Score:4, Insightful)

      by Unordained ( 262962 ) <unordained_slashdotNOSPAM@csmaster.org> on Wednesday December 17, 2003 @01:15PM (#7746684)
      That's why I look forward to the completion of a few last (for this) features in Firebird [ibphoenix.com] (the database server) -- they have an 'embedded' version that sits in a .dll and acts exactly like the server version as far as the calling program is concerned. It's the same engine code, so the same SQL is supported, the same features available. It can still load UDF libraries (the same ones it loaded as a server) but won't accept connections from other programs. If you want several programs to access the same file at the same time, you'll have to do that yourself. It's already great for building single-user programs to be distributed to clients who wouldn't want a "full-blown" database server running (really, it is, but at least it's not listening to any ports.)

      Firebird/Interbase has always been rather small -- 10 meg install, very little memory usage. The main problem porting it to PDA's and other embedded devices was its loopback / inter-process-communication systems. Its shared-memory method wasn't completed (though from what I hear, they have it working in development versions, and the new method speeds up the server quite a bit as well) and it relied on 127.0.0.1 loopback. Windows CE didn't allow for that (no ethernet interfaces at all by default, not even lo) and thus the server couldn't talk to itself. (It tends to use multiple processes.)

      Regardless, Firebird will most likely be available for embedded devices soon, run quickly and quietly, and won't be a stripped-down version at all. Code written for large multi-user environments would still work, but you might have to wait a bit longer for your huge reports.

      (And yes, I feel fine writing this without being too concerned about vaporware -- it's already proven technology, it's actively worked on, and the Microsoft thing isn't any closer.)
      • The Firebird embedded DLL sounds really interesting. I looked at sqlite [sqlite.org] but was put off by its almost total lack of data validation (for example, you can't even declare a column to hold integers). Which is claimed as a feature, not a bug, but for me it defeats one of the main reasons for using an RDBMS (foreign key constraints and check constraints are also not supported). Embedded Firebird might be a saner if slightly more heavyweight alternative. I wonder if there is a similar thing for Postgres?
        • I forgot to mention, though likely nobody cares, that no stable version of Firebird-embedded has been released yet, as it's a branch of the current 1.5 series, which is currently at RC7 (last I checked.) It feels stable to me, but ... watch for a slashdot headline about 1.5-stable being released. (Wait, the slashdot crowd would think a new mozilla browser had come out ...)

          Also to note is the recent (last week?) merger of Firebird and Yaffil, its russian (commercial-product-for-a-while) counterpart. Yaffil
      • Now, wait. I'm confused...

        You are talking about the browser, right?
    • Hmm.

      SQL Server CE != SQL Server
      Pocket Word != Microsoft Word
      Pocket Excel != Microsoft Excel
      .NET Compact Framework != .NET Framework
      Windows CE API != Windows 32 API

      See a trend there? Microsoft marketing tries to make them seem all the same, but no CE technology or app is even close to its Windows counterpart.

      Which is a shame, Pocket PCs could be great mobile computing platforms if Microsoft dared more.

    • I agree, the last time I checked they didn't even have basic things like views, stored procedures, or triggers.

      Our company uses Adaptive Server Anywhere on the Compact Framework. Not only does it perform better but we can replicate directly to Microsoft.

  • To hell with SQLite! (Score:1, Funny)

    by Anonymous Coward
    Now we've got SQL Server CE!

    No bloat there, I'm sure (*cough* MSDE *cough*).

  • Technology in infancy... (Score:5, Insightful)

    by heironymouscoward ( 683461 ) <heironymouscowardNO@SPAMyahoo.com> on Wednesday December 17, 2003 @11:39AM (#7745824) Journal
    Do we really need more technologies in their infancy? Is it not one of the biggest problem in this industry that we keep inventing new ways of doing the same stuff, throwing out entire generations of products and developers each time?

    We need ways to make existing good code last longer, not new frameworks to waste time on.

    But... sigh... I don't expect that kind of attitude from Microsoft. One more reason to avoid such platforms like the pest.

    • Re:Technology in infancy... (Score:4, Insightful)

      by GoofyBoy ( 44399 ) on Wednesday December 17, 2003 @11:57AM (#7745995) Journal
      >Do we really need more technologies in their infancy?

      >But... sigh... I don't expect that kind of attitude from Microsoft.

      Why do you think that this is a MS issue?

      What about IBM? Sun? Oracle? What about over half OpenSource projects?

      (Or perhaps I should be used to the mindless MS bashing around here)
      • I agree about many software vendors.

        The Microsoft bashing is definitely present, and I would argue well-deserved, but it is not mindless. Microsoft are easily the world leader in the invention of new software development platforms that provoke huge investment in time and work by millions of developers, and which are then discarded after 2-4 years. Sun produced a couple of versions of Java. IBM... some C++ platforms, and then websphere. Oracle? OK, a bunch of platforms but mainly just Developer and Des
        • But PDA's are in their infancy. And that's what the compact framework is for. People are driving the market for smaller and more powerfull handheld type devices. The software is just following...

          Regarding the rest of the bashing, you are being mindless. How can you compare the last 10 years of sw development in 3-4 sentences? (Not worth a shoe.)
          • And yet they are as powerful as computers 10 years ago. What is the problem? Basic operating system, simple database, tight, efficient programming language... all these things exist and do not require more new platforms.

            I do not accept that the resolution of the screen or the size of the device requires a new development platform (especially at the database level).
            • I don't know about you, but there is no way I'm going to be running what your average DOS PC was running 10 years ago on my PDA, in the area of OS, networking/web or databases. Sorry.

              SQL is nothing new. What is so scary about having an embedded SQL server for CE, Linux, etc? There are number of them, it's not like MS thought of it.

              We could go back to running DOS database apps. Paradox! dBase! YAY! Knock your socks off, do as ye will- but I want a more modern platform, with a more modern API and language

        • FUD (Score:3, Insightful)

          by GeckoX ( 259575 )

          Microsoft? Dozens, maybe as many as fifty huge platforms produced, promoted, sold, and then abandoned in the last ten years. Visual Basic. C++. Office. ASP. COM. .NET. SQLServer. Access... ad nauseam. And every single version of every platform incompatible with the last.

          This is pure and simple FUD.
          Not ONE of these technologies listed is obsolete or not used any more.

          Bash MS all you want, when you can back it up.
          When you bash MS and don't back it up, it kinda null and voids all of your other arguments.

        • No really this is mindless bashing.

          >Visual Basic. C++.

          Two languages which are incompatible as with any company. How many langauges has IBM pushed? And the end product are insanely compatable since you can create DLL/COM/ActiveX objects and work with each other.

          How many OpenSource FORTRAN, LISP, PROLOG, Perl compiler/translator projects out there? How much better compatiblity do they have with each other?

          >Office.

          Their SINGLE office suite for the past 10 years. What makes you think that they ar
          • A friend of mine wrote an application in VB, some years ago, on Windows 3.1.

            A nice application, complex and sophisticated. It used numerous components (VBXs) and he spent about 4 months solid working on it.

            He sold this for a year or so, then VB4 came out. He started on an upgrade, found that several of his VBX vendors had vanished, stayed with the old code.

            Then VB5 came out. Completely incompatible.

            He stopped maintaining the product, and swore to never use Microsoft's tool again.

            Now: it may have the
            • >He started on an upgrade, found that several of his VBX vendors had vanished, stayed with the old code.

              Is that MS fault that the third party vendors had vanished?

              Oracle has tonnes of drivers that work with only one verion. Sun and IBM have lots of integrated products that they will only support a particular version combination. I ask you again, what is so special about MS?

              >it may have the same name (VB) but it is at least 3 totally incompatible products.

              You seriously need to get some first hand
        • OpenSource? I'm building my applications using Linux (standard POSIX), using MySQL (very standard SQL)
          MySQL may be a useful program for many applications, but it certainly isn't 'very standard SQL' [sql-info.de].
          • Why, there's a list of almost 20 "gotchas" with MySQL?

            Almost 20??

            The SQL in my applications runs unaltered with MySQL. Now, perhaps I'm just not the kind of person who tries things like "INSERT INTO... SELECT FROM". Yeah, I should do more of those things.

            To take a counter example, in an Oracle database, it is not even possible to get good performance on large tables without resorting to non-standard extensions such as hints. (Disclaimer: it's been a while since I Oracled, this may have been improved
            • I think most of the gotchas are related to data integrity and things like declaring a foreign key constraint which is then ignored under some circumstances, or columns which get assigned values you didn't explicitly ask for. A particular app may run fine, but when you ask the database to enforce something like a foreign key constraint, the value of that is that the database makes sure it will always hold no matter what. If you can't trust the system not to do random things with your data (and the gotchas
              • You can't pick a particular example, say 'it works for me', and use that to conclude the problems are not serious. If that were true we'd have to class MS VC++ 6 as a standards-compliant C++ compiler or Netscape 3.0 as a web browser that supports CSS.I

                Mumble grumble. You're right. Moderators, help unleash the hounds of hell on my humble head.
    • Do we really need more technologies in their infancy?
      Everything has a beginning. Second, yes we do need new technologies. The day we stop innovating and inventing we've become no more intelligent than the rest of the species on this planet. Mind as well we should start walking backwards.

      Is it not one of the biggest problem in this industry that we keep inventing new ways of doing the same stuff, throwing out entire generations of products and developers each time?
      I would disagree. It is not the big
      • We agree that progress is good. No-one regrets the passing of days when we hand-coded interface languages because XML had not been invented.

        But you make the mistake of many people, namely to assume that change is equal to progress.

        Change is only part of the cycle. The other part of the cycle is the refinement of techniques and maturation of knowledge that lets you exploit change to the fullest.

        Imagine if every three years we had to discard languages like HTML and XML and start with new, incompatible co
        • We agree that progress is good. No-one regrets the passing of days when we hand-coded interface languages because XML had not been invented.
          Part of my point.

          But you make the mistake of many people, namely to assume that change is equal to progress.
          Change in itself can be tied to progress. Changes can happen based on progress, especially when it comes to new technologies (in their infancy) like the internet. The invention of HTML (the most simplistic example) lead to the first sharing of information
    • Do we really need more technologies in their infancy?

      Yes. It keeps geeks like myself employed.

    • I'm sure I'll be modded down for 'flaming' for this, but it needs to be said, despite my love for open software.

      Do you really think that the attitude of replacing old frameworks is exclusive to MS, who does it for financial gain? No. Open source communities are -constantly- reinventing the wheel - look at all the open source projects out there just for word processing; don't you think they could share a little bit more of a common codebase, say, for the inport/export of 3rd party file formats? That would m
      • No-one minds upgrading to a new platform when there is a crying need for it.

        No-one should pay to upgrade to a new platform because the vendor decides it's timely.

        Every responsible OO project starts by defining standards that will last years, decades (look at OOorg's document formats). Every responsible OO project sweats blood and tears to ensure that new releases are compatible with old ones.

        How many versions of PHP have been incompatible with previous ones?

        How many versions of Perl? Of Apache?

        Now th

  • Somebody please port MySql :) (Score:4, Interesting)

    by Mean_Nishka ( 543399 ) on Wednesday December 17, 2003 @11:39AM (#7745828) Homepage Journal
    I'm so desperate to see a Mysql port to Windows CE.. I've been running this along with PHP/Apache on my Sharp Zaurus [techfornewbies.com] and that platform proves it's possible to do this stuff with open source software and 206mhz Strongarms.

    While I haven't been interested in shelling out the $$$ for Microsoft's server product (getting data back to the desktop database is important to me) I can say that I haven't found any database solution that operates as quickly as MySql on a PDA.

    My Zaurus is running as a full fledged server and its rock solid. Even the desktop ODBC connector is able to pull data into my windows apps at a surprisingly fast rate of speed. In a word it's simply awesome. I just wish my PHP skills were better.

    For the mortals Microsoft offers an Access compatible format (pocketAccess) that's horribly limited and slow. There are some great PocketPC database apps [kaione.com] that are crippled by this horrific database solution that could be killer apps with a free backend.

    • Try this instead:

      http://www.sqlite.org/

      Enjoy,
    • Some one please don't. It's in a half stage of trying to be big while not being small. It has a lot of flaws in it that don't make it a large RDBMS. PostgreSQL comes a lot closer, but there are a lot of larger databases that can pound both of them out for reliability and overall performance.

      Please don't post that link about mysql's supposed better performance. It likes to choke after dealing with much MUCH larger sets of data. But I digress.

      I'd imagine something MUCH smaller, like mini sql [hughes.com.au] or SQLLite
    • As a longtime user of PDAs, Newton, WinCE and Zaurus, I feel I can safely say for most apps, MySQL is pretty sucky on the Z. It wastes precious resources on the Zaurus- especially precious considering how much the setup- Linux+Qtopia- eats up as it is. Far worse than CE.

      I can't say I've used this particular MS SQL for CE product- but I have used the built-in database, which allows happy queries. As well as SQLlite on the Zaurus. Both seemed a lot faster and provided more enjoyable solutions for me, the de

  • ..if you leave the front door open... (Score:3, Funny)

    by burgburgburg ( 574866 ) <splisken06NO@SPAMemail.com> on Wednesday December 17, 2003 @11:50AM (#7745933)
    let's face it, no matter how well you write your app, it won't be secure if you leave the front door open.

    It will be secure if you leave the front door open, but have a tiger pit cleverly disguised just inside the doorway.

    Thieves tend to go to the path of least resistance, and unless they see one of their criminal compatriots plummeting into the tiger pit, or hear the horrible screams that follow, this is their most likely point of attack.

    Or was that just a metaphor?

  • free and high quality alternative (Score:4, Interesting)

    by e4liberty ( 537089 ) on Wednesday December 17, 2003 @11:51AM (#7745939)

    The very high quality embeddable SQLite database has several ports to .NET -- see the web site [sqlite.org].

  • So let me see if I get this straight... (Score:3, Insightful)

    by Dr. Bent ( 533421 ) <`ben' `at' `int.com'> on Wednesday December 17, 2003 @11:55AM (#7745970) Homepage
    After all, unless you are simply playing games on the PDA, it probably needs to interact with a database somewhere and I can assure you, just about every common task that you may encounter is discussed in depth. The show you how to bind controls to data, retrieve it from a Web Service, retrieve it from a SQL Server on a local network, use SQL CE to take advantage of replication and using XML as a Data Access technology.

    So, if I'm reading this right, the way this is supposed to work is:

    1) You bind your GUI controls to a SQL server database using the SQL CE library installed on your PDA.

    2) When a GUI control needs to update the data, or refresh it's view of the data, it performs a query using SQL CE.

    3) SQL CE then transforms that request into an XML document, sends it over a (presumably) wireless network connection, to SQL Server.

    4) SQL Server then processes the request, generates an XML response, and then the whole thing runs in reverse until the response gets back to the control, which can then update itself with the new information.

    Is this really the way this thing is supposed to work? Please somebody tell me that I'm reading this wrong.

    SQL statements are great when you're trying to optimize for query efficency, but they're awful when you're trying to optimize for network latency/bandwidth. This is because SQL is based on the idea that you perform operations over a persistant connection, because establishing that connection and performing the query is usually the bottleneck.

    But any wireless developer will tell you that latency/bandwidth is the major bottleneck in a mobile application. When you're sending data over CDMA/GSM networks, the best data rate you can hope for is 192kbps, and forget about low latency. This means you have to package your requests in a big bundle and do them all at once. Which means you need something a little more coarse-grained than SQL.

    I suppose if you assume that your application will always have 802.11 access, this could be a good model. But otherwise you're going to be waiting 5-10 seconds every time you click a button, change a setting, or scroll down through a list. Which means nobody is going to actually use your application because it's a grand waste of time.

    • You may not be reading this wrong, but it is wrong nonetheless.

      SQL CE data resides on the Pocket PC, however it can be replicated to a grown up SQL server as needed, and provided proper connectivity.

      A typical use of this technology is a driver in the field. They get their data replicated down from the SQL Server, then they head off for a day off deliveries. Transaction are then done by the driver on the SQL CE side. He/she brings it back to the loading dock at the end of the day and replicates his/he
    • No, you're reading it wrong. See how he mentions getting data from SQL Server on a "local network," and then use the SQL CE server from when you're away from said local network. You know, like when you're out of the office, away from the office LAN or wifi network. You use the built-in server for the time you're away- the latency and bandwith issues are moot. You come back to work the next day with the new data you collected and new data from the mothership- you sync and that's that.

      It's not a new approa
    • SQL statements are great when you're trying to optimize for query efficency, but they're awful when you're trying to optimize for network latency/bandwidth. This is because SQL is based on the idea that you perform operations over a persistant connection, because establishing that connection and performing the query is usually the bottleneck.

      Assuming the steps you stated above are correct and the data itself resides on a server and not the PDA, I don't really see the problem. In the infrastructure you des

  • I recently clustered 26 SQL CEs (Score:2, Insightful)

    by Anonymous Coward
    to store the english alphabet.
    Performance was good unless you wanted the results sorted :)
  • When I was working on a mission critical application for an insurance company, the customer wanted applications that would enable folks to continue to handle essential business functions even if their network or database servers were down. The network connections to CE type devices are largely wireless-and you just can't assume they'll always be available. In that situation, having a database that will cache important items locally in a way they can be actually used is quite valuable and allows you to updat
  • So far many of the comments seem to be along the lines of, "SQL Server for a PDA?! WHY?!" Well, why not? Think about where PDA's could be headed. Why not have the PDA be the server for your Personal Area Network?

    PDA's these days are getting more and more powerful. They're already as powerful as what was on the desktop less than a decade ago, and they're catching up fast. It's not unreal that in 5 years we'll have PDA's capable of running Half Life 2 and Doom 3. I mean, today's PDA's can already run Quake 1 [pocketmatrix.com]
    • Can you describe one of these "really powerful mobile apps" that require a SQL server on a PDA?
    • Because people are pigheaded. Ignorant. etc.

      A PDA should be using a database as it's central store as it is. CE already includes a database system, though the SQL CE has more features. The NewtonOS and PalmOS both do very well by having a database at their core. I don't care if it's a sql-based relational or object database, it just makes more sense.

      The PDA is a chance to start over. With different OSes and CPU archs, compatibility isn't a big deal as it is on the desktop, where it has held back develop

  • Comment removed (Score:3, Funny)

    by account_deleted ( 4530225 ) on Wednesday December 17, 2003 @12:12PM (#7746130)
    Comment removed based on user account deletion

  • pipe dreams (Score:1, Interesting)

    by Anonymous Coward
    Anyone that works with wireless apps will tell you, dream on. Not because it's not feasible to sync with a remote server, hardware limitations of devices or that .NET is good/bad. The reason is wireless service sucks and transferring large amounts of data blows. By blows I mean data more than 2-4kilobytes won't transmit fully and will most likely get corrupted.

    There are already wireless devices for field force applications, which have a database on the device. And guess what, they don't use Sql server. All

  • .NET.
    Compact.

    (pause)

    Bwahahahahaha!!!!! BWAHAHAHA!!!

  • Why? (Score:2, Interesting)

    by BaronAaron ( 658646 )
    Build your application as a web app using ASP.NET/PHP or whatever. Use whatever SQL backend you want and have it run on a REAL server that is meant to do that work.

    On the PDA, throw a 802.11b card in when at the office and some sort of cellular card for when you are out and about. The only software you need on the PDA is a good web browser. Bandwidth is much less of a issue because all you are transfering back and forth is simple HTML.
  • It took 10 chapters and 450 pages! Damn I can do it in 4 words: Do NOT do that. Yup, MSSQL a ten's of megabytes application, let's dumb it down and use it manage a few hundred megabytes of compact flash on a handheld or embedded device, yup that's a plan. I guess it's like the man said: when all you have is a hammer, everything looks like a nail. I guess Microsoft needs a bigger data management toolbox.
    • I guess it's like the man said: when all you have is a hammer, everything looks like a nail.

      Remember, this is Microsoft we're talking about. It should go, "When all you have is a hammer, everything looks like your thumb."

  • Free hint for the reviewer: this is not a high school assignment, there is no minimum word count.

    The other area that this book really excels in is getting you through replication. This is not a fun topic if you don't know what you are doing and there isn't a lot of literature out there to help you get through PUSH/PULL subscriptions and the like. Pragmatically speaking, of the topics this book covers, Tiffany's coverage of replication is probably going to benefit people the most, because if you can't syn

    • Well, after the flaming I took for dare reviewing another compact framework book and taking feedback from it, I decided to speak to a few issues in more detail. I do think it's relevant that the guy's book talks about stuff that people have a lot of trouble with. With that said, it's only my second review here, but with all of this feedback, I'll get better.

      Guess it just goes to show you can't make everyone happy.

      • Hi, dude.

        I'm not trying to discourage you from writing reviews. I can tell you enjoyed this book, but your writing style distracts from your message. Please omit phrases that take up space without adding meaning.

        Enough about the background, though.

        the text was well written [...] and easy to read

        Do you work for the Department of Redundancy Department?

        This is not a fun topic if you don't know what you are doing

        How many times did you say "let's face it" in this article? Once is too many.

        H

  • This just seems ridiculous to me... (Score:3, Interesting)

    by AmazingRuss ( 555076 ) on Wednesday December 17, 2003 @01:17PM (#7746697)
    ...why separate the database engine from the development tool on a palmtop, where efficiency is paramount? Do they expect people to use a palmtop as a db server or something?

    Something like Foxpro or Access, with integrated database support, would make a helluva lot more sense. The latest version of Foxpro for the PC has a smaller runtime than what they are cramming on there for SQL CE for chrissakes! It's fuggin silly. What gets IN to those people?

    10 years ago, I was putting the entire foxpro for dos development environment on 1 meg HP xxLX palmtops, and it worked really well. Full portability between desktop and palmtop, good performance, and good stability. Why have we taken such a giant step backwards that despite the hardware being 100 times better, we can no longer build decent software to run on it?

    Everything MS does anymore should be accompanied by clown music.

    I am truly starting to believe that MS is going to be a niche player like Apple in 10 years...an idea I would have laughed at a few years back. I don't LIKE this idea...it means I have to spend a lot of time retraining...but the writing is on the wall.
  • I've developed a couple of data driven apps on Palm OS that interact with backend databases using jdbc, I've used these with both Oracle and MySQL. The MS approach sounds rather heavy wieght, you can use Waba on the client side, and you can serialize the data which allows the data transactions to be very fast and efficient. XML seems a poor choice for the PDA environment being how it requires much nore band width and CPU cycles than Object serialization .NET seems to heavy and clunky for even the PC envi
  • Isn't it amazing how seriously slashot readers really don't take .NET?

    60 posts talking about a Microsoft fledgling technology here says so much more than 300 slashbots flaming each other on the other "Microsoft's New Core OS Team Learning from Linux" thread.

    Microsoft is a smart company. No really, they must be. If you have shed loads of money, you can pay lots of smart people enough money to make them work hard enough to roll out products which will do something "acceptably well" at worst, "reasonably wel

Slashdot Top Deals

The herd instinct among economists makes sheep look like independent thinkers.

Close