Click here to monitor SSC

Tony Davis is an Editor with Red Gate Software, based in Cambridge (UK), specializing in databases, and especially SQL Server. He edits articles and writes editorials for both the Simple-talk.com and SQLServerCentral.com websites and newsletters, with a combined audience of over 1.5 million subscribers. You can sample his short-form writing at either his Simple-Talk.com blog or his SQLServerCentral.com author page. As the editor behind most of the SQL Server books published by Red Gate, he spends much of his time helping others express what they know about SQL Server. He is also the lead author of the book, SQL Server Transaction Log Management. In his spare time, he enjoys running, football, contemporary fiction and real ale.

Katmai: Ship date or ship-shape?

Published 29 January 2008 8:11 am

By now, most of you will have read the news of the delay in releasing SQL Server 2008. It all seems fairly typical Microsoft, though I wouldn’t want to be overly critical, even if the bizarre blog announcement, written in strangled Dalek-speak, was almost beyond parody (although Phil Factor had a good attempt).

 

In reality, it isn’t a huge delay, and the sense that the product is slipping has probably been exaggerated by the fatuously-entitled ‘Heroes Happen Here‘ launch wave, which attempted to roll three “launches” into one. This meant that the VS2008 launch was too late, the Windows Server 2008 launch was about right, and the SQL Server 2008 launch was much too early.

 

Compared with the two-year delay to SQL Server 2005, mostly caused by security issues, the current progress seems acceptable. Nobody wants a product with bugs in it and a subsequent trail of fat service packs. The only time to ship a product is when everyone agrees it is ready. Within reason, the date of release isn’t the big issue; it is whether the product will be sufficiently improved to make upgrading an obvious business decision for the companies that use SQL Server.

 

Nobody who lives by producing commercial software can really condemn product slippages. Certainly, Red Gate has slipped a product or two in its time (SQL Prompt 3 being one), but the overall goal was to ensure that the tool, when it finally arrived, was good and fit for purpose.

 

OK, a 2-year delay on SQL Server 2005 was excessive, but when it finally arrived, it was pretty good. Or was it? We all have our bugbears, the parts of SQL Server we like to criticize, such as SSMS. However, it would be fascinating to hear your nominations for the one aspect of SQL Server 2005 that could have done with another year’s development before it saw the light of day, and why.

 

Post your nominations as a comment to my blog. All entries will go into a draw for a $50 Amazon voucher!

 

Cheers,

 

Tony.

12 Responses to “Katmai: Ship date or ship-shape?”

  1. ronaldo_aquino says:

    Can’t fault SQL 2005, we’re still on 2000 ! While cosmetically appealing, we can’t justify the move. None of our vendor apps are demanding a move to 2005.

    The delay of Katmai only means I can’t justify going to TechEd in June. But like everyone else, I don’t mind the delay. At least we’re getting a solid product in the end.

    Best feature of 2005 ? Definitely the Management Studio.

  2. Brad24 says:

    Strange. If there was one feature I’d single out for criticism, it would be SSMS. A strange mutant Visual Studio without intellisense or plug-ins. Vastly overweight. It has all the flab of Visual Studio but without the features. It is as if someone at Microsoft had the idea of creating a version of Visual Studio for SQL Server, but lacked the resources to do it properly. Service packs improved the beat more than somewhat but it is still a beast.

    It is always a relief to go back to Query Analyser.

    The only feature of SQL Server 2005 I use is the Varchar(MAX) and varbinary(MAX). The rest could go up in a puff of smoke tomorrow and I wouldn’t even notice.

  3. RandyD says:

    SSMS stinks. With every upgrade of SQL Server, the GUI requires more clicking and clicking to do what was once a simple task. Go build a multi-step job in SSMS. I’m tired when I’m done. Where are the Next and Previous buttons? No tabbed forms anywhere! What’s 2008 going to do, ask “are you sure” every time I click OK? My feeling is “I can wait, for 2008″.

  4. paschott says:

    1. SSIS – just some things that don’t work in a manner that helps us to be productive like DTS. Easy import/export of objects, supporting various data formats (which now require “advanced” editing to change strings on the connection), Excel support, Unicode/Non-Unicode conversions, not being able to select all/many of the tables instead of needing to write code for each table, and my favorite – poor end of line support if the exporting program (like Excel) didn’t populate all column delimiters to the end of the line for blank columns.

    2. The GUI – Why? SQL 6.x had some really easy methods to handle permissions and the like. SQL 7 was a little worse, but still manageable. SQL 2000 was worse than 7. SQL 2005 just flat out causes issues. I end up using the GUI to generate a script and tweaking the script because that works more often than the GUI. Even then there’s no easy way to work on multiple objects at once and I have to use the mouse constantly.

    3. Keyboard support for the GUI – sure we have some commands for the TSQL/Command editors, but to actually use the GUI is just painful. I can understand VS’ IDE for SSIS and Reports, but the general usage of some of the same for SSMS is just painful. I shouldn’t have to take my hands off of the keyboard that often.

    While not related, I think this SQL 2008 Launch Event could have used about another 6 months or so before being released. I understand marketing and all that, but they have to realize that this looks worse than if they just waited to hold another Launch event. Who wants to go to an event where you’re told “Look what’s coming in 6 months” when the other products get “Look what you can get right now.”? (And I’ll have to remember that dialogue from Phil’s article. That was truly inspiring.)

    Oh, and for Randy’s comment on the number of clicks – I’m sure MS can come up with something better than “Are you sure” prompts. Perhaps more expandable sections with no hotkeys, more searching for various “types” of objects, perhaps even one per search, coupled with some of the neat features from Vista’s UAC perhaps (so we don’t run SSMS under an account that’s a SQL Administrator, of course)? Note – If anyone from MS is reading this – please take that as humor, not a serious request.

  5. DavidBSQL says:

    Definitely agree that SSMS needs work. Nice to have add-in features from 3rd party vendors that make it a little more robust but you would think this would come out of box if they are going to make it as bloated as it feels.

    However, the biggest area of contention that I have is with the install. Rather than making the install for 2005 more robust and flexible while maintaining ease they made it quite a mess. The uninstall process can be quite annoying and un-intuitive and if there is ever an install that gets messed up, well, hang on for the fun…. Ultimately I would have thought that this would have been one of the more solid parts of the release. Hopefully with Katmai….

  6. Philip Kelley says:

    2005′s SSMS is one of my favorite punching bags [I want my ctrl+b back!] but since everyone else has already taken a whack at it, I’d like to add SQL Profiler to the lineup. My main issue is (of course) the user interface. There’s a load of options (events and data columns), and they do a very poor job of managing their presentation, selection, and configuration. Parts are better than 200′s interface, but others are worse, and the whole seems like a small step back.

    For example, I dislike how when they list columns, they set them at all the same width, which is not wide enough to display the full column title–so you end up with (for example) three “Object…” columns (ID, Title, and Type, I think), and the only way to figure out which is which is to expand each column individually, which requires double-clicking or click-and-dragging on two–pixel wide lines. Multiply for all columns. Joy, not.

    Once you get it working it does work as well as Profiler ever works–which honestly is pretty good, even if it takes 10 minutes to fine tune and drill down to the problem at hand, by which time that problem on Production has long since disappeared.

    To mention, Joel Spolsky (of “Joel on Software” fame) presented a few relevant insights on MS and their recent user interfaces in an essay a month or so back–it’s towards the end of this article, but parts 2 and 3 are still worth reading.

    http://www.joelonsoftware.com/items/2007/12/03.html

    Philip

  7. stic says:

    Already working on technical materials about SQL Server 2008 I have to admit that I am not convinced that the product as whole really need this new release.

    For me it is just the “good old” 2005 done correctly. All the ‘new’ things just fill gaps in functionality that IMHO should be shipped with previous version. From a developer (me) point of view, almost all of these features we already have – support for ADO.NET Entity Framework – we already have it in 2005 version, the same goes for support for LINQ (which isn’t in fact so closely related to db engine used, anyway). Hey, even Microsoft Synchronization Services is already out there to pickup and use…

    Of course, these new things are nice, probably will make our work a little easier, but shouldn’t these be shipped as SP3, just slipped over door some beautiful sunny day? IBM with DB2 9(Viper) gives great set of new XML oriented feutures, Oracle is building great set of data processing/management services and applications around they 10/11g databases without really shipping anything new (ok, they solves bugs and issues all the time, but who doesn’t?).

    My point is, lets create something good, stable, a kernel database engine that will rest long. Something similar (but faster, more secure, scalable and reliable :-) ) to SQL Server 2000, which already support well a lot of customers requirements. Then create all this fancy ‘* Services’ around this base. Wouldn’t that make more sense?

    It is somehow ridiculous that e.g. something that important as a mirroring support was shipped in SP1 of SQL2005, whereas at the same point of time this product was full of ‘business level/ready applications’.

    Summarizing, form my point of view, I would like to ask for more database oriented 2008, less Reporting Services with great KPI dashboards, more support for performance, scalability, etc. In today, web oriented world, at the end of the day it is the db engine speed that matters …

  8. Cheval says:

    Sql Server 2005 Replication. Especially with re-publishers. We’ve got support cases open for over 2 months with MS China, US, India and Australia trying to figure out why the merge agent goes nuts. We’ve even had to manually edit the replication stored procedures and conflict tables when a simple addition of a non-foreign key field was added to a non-filtered table.

  9. JJEugene says:

    There is plenty that I like about SS05. CTEs are great. I have often used the new Output and FOR XML syntax. (The FOR XML syntax comes in handy when needing to concatenate rows.) Varchar(max) is another helpful feature. I’m not the least bit sorry we upgrading. Having the above features available to me helped speed up the release of a major application. Being more productive (on a stable, powerful platform) is the whole point for me.

    Which brings me to SSIS. I don’t mind a steep learning curve. So, having to buy two books, read multiple on-line tutorials and spend 3 months converting 4 DTS packages would be fine with me — IF all that time was spend productively. Reading all those glowing reports about SSIS, I worked really hard almost that whole time to keep a very open mind and positive attitude. There are a couple nice things about SSIS, but overall SSIS is super buggy. I like the definition of a bug as being something that doesn’t work the way a user think it should. I found a couple bugs of the normal type and million bugs of the type where SSIS was counter-intuitive. There are at least two completely different ways to write expressions depending on where you are in the application. Who needs that kind of unnecessary complexity?

    In the end, I decided it was not just a matter of a steep learning curve. The user interface is sub-par. Another thing that was troublesome about SSIS is the difficulty in documenting. Working with the text boxes is a serious pain – the text does not even wrap. And there is no note field for the variables. Etc. This may sound like a minor issue to some people, but documenting is vital to me, especially when I use an application so little.

    I love my job in general, but hated the time I spent with SSIS. No software product has ever had me hating my job before. It gave me such pain that when one of my packages recently stopped working for no reason whatsoever that I can figure out, I couldn’t stomach having to go back into the package. It was the final straw in making me recommend to my management that we simply stop pulling that data into our database. Wish I could do that with the other SSIS packages. But alas, we really need the data.

  10. db042188 says:

    I’m relatively new to SS so outside of SSMS’s tree refresh quirks, I was pretty happy with Management Studio.

    I spent a lot of time in SSIS and was a bit disappointed with its performance on low volume, trickle movement of data. This limited reusability for us. I also thought that in spite of the graphical look, it became unwieldy rather quickly in spite of sub package use etc.

    The most painful experiences for me so far had to be the sliding window scenario and RS2005.

    The sliding window scenario is a meta data approach to “in and out” movement of partitioned data. It was as if the LEFT and RIGHT approaches to this scheme were developed by two different software companies.

    My first experience with RS was in the 2005 product. I can imagine that such a product can only do so much but I was floored by 1) the performance issues that are accepted as “known issues”, 2) quirky behavior in response to user parameter setting , 3) random changes by the software to parameter characteristics, 4) non dynamic nature of parameter labels

  11. SAinCA says:

    Report Builder – premature. Non-intuitive even for folks who’ve had to publish User-Schemas before (Oracle Discoverer and Noetix Views). Columns I added to the “schema” (in quotes because it doesn’t behave properly!) were “invisible” to Report Builder… Wassup with that??? Even simple “publications” that replicate a canned query with a half-dozen reference table expansions were such a pain I junked it and went back to SSRS. Good job we didn’t tell the User Community about Report Builder… It may look easy to build SIMPLE reports, but the controls an experienced BI User would want are lacking and the limitations on presentation styles are frustrating. Definitely too little too soon… Here’s hoping 2K8 will kick it out of “version 1 limitation territory”.

Leave a Reply