Beta testing for Microstation Connect Update 12?

Hi Robert,

RobertArnold said:

is it possible to get some early-release versions and get things fixed?

during What's New in MicroStation CONNECT Edition Update 11 webinar Samir Haque touched also beta testing question (and as developer, I agree with his explanation): Three months release cycle makes impossible to use a separate external beta testing approach, there is no enough time to build public version for testing purpose only. It's also the reason why "Technology Preview" features are available in a standard versions (some publicly and described in documentation, some are hidden and discussed e.g. in Technology Preview community).

Interesting information, and I appreciate it, was information that Update 12 is planned to be focused primarily to finish running implementation and bug fixing, without new features added. So maybe Update 12 will bring much longer list of fixed Defects (and linked Service Tickets) ;-)

With regards,

  Jan

Hi Jan,

Thanks for the video link.  It points out some of my issues with Microstation!

1) Bentley is committed to four releases per year, which they have been doing on this release for the past three years.  As you mentioned, there is not enough time for doing a Beta program.  The point is that Bentley will give us quantity, rather than quality.  (A bunch of xxxx, instead of a little good.)
2) The last update of the year is focused on bugs/defects.  The whole reason why incremental updates was hoped for was because bugs/issues would be found in the current release and then it would take a year or two for the next release to come out with the fix.  Bugs/defects should be the first priority.  New features should not be more important than having something working properly.  In the world of cars, this sounds like, "We installed a new radio in your car, but we still haven't fixed the broken transmission."
3) I didn't catch what the subject was, but there was something in Updated 11 that fixed an Update 10 workaround.  Why was the product released with a "work around"?  Again, quantity instead of quality.
4) Lots of pushing to have people upgrade from V8i to Connect.  Update wizard, fast track help, videos, etc.  BUT, there are "thousands of ways to configure Microstation", so they have concentrated on the main core of users and the "edge cases we have left alone".  Again, quantity rather than quality.  If you're an edge case, depending on a reliable platform, oh well.
5) Bentley View, based on the same engine that Microstation is base on, is finally coming back.  But on the day it comes out, the previous version will be retired.  In other words, they are releasing an untested piece of software (with no Beta testing) and killing "old reliable" before anyone has a chance to the issues.

I would ask for a "limited access token", except that I am more interested in making sure that the core product is working well. Once that is done, then I can start looking at new features.

During the Beta phase, a pre-release version would be tested by users, bugs/defects were found and reported back.  The next version of the pre-release would come out, the same bugs/defects would be tested again and found not to be working as expected.  Then the next iteration would finally get it fixed.  (I even had cases where I was sent an internal build to work out a bug before the next pre-release.)  So now when you find a bug, you get to deal with a tier one support person, then hopefully get a Defect Report filed (which you never hear or know the status of), and wait until the end of the next year to see if it works, then file a new Defect Report to wait for another year to again see if it is fixed...

Bentley seems to be in love with Microsoft.  They use Microsoft modules instead of writing their own.  When they don't work properly, "It's Microsoft's fault."  (For instance V8i text editor, Connection Client forgetting passwords, etc.)  When Microsoft changes to ribbons, Bentley has to follow, even though it does not help the productivity of their product.  When Microsoft goes to a regularly scheduled Windows 10 release, Bentley follows.  (In fact, what happened to Windows 9?  What happened to Microstation 9?)  Microsoft, in their pursuit of features and quantity of releases, has produced a great Windows 10 October 2018 update that will potentially delete all your personal files.  They also have released their update that will overwrite your files without warning the user.  They also have released and update that tells users that their copy of Windows is no longer valid.  Will Bentley follow?  Oh wait, we've got to get this product out, and there is no time to Beta test!

Linux likes to have a lot of regularly timed releases, but they have an LTS version.  "Long Term Support".  The goal of this version is stability and reliability.   The user has an option--go with the reliability of the LTS version, or go with the latest and greatest with the possibility of problems.  Why are we forced to follow Microstation's new and problematic line?  I guess it is because they have not worked hard enough to call the Connect version stable and produce a "LTS" version.  But at the same time, just as Microsoft is determined to get everyone on Windows 10, Bentley is determined to get everyone on Microstation 10.

I apologize for the ranting, but when we are at the 13th official release of a product and it still has significant issues, there is a problem.  It's time to stop and figure out what the future of Bentley's product should look like.  Is it something reliable, or is it something new? 

--Thanks,
--Robert

Hi Robert,

thanks for collecting some important issues and topics, but personally I dsagree with some of yours conclusions, because in my opinion despite of based on correct fact, they are wrong. But of course my perspective nad priorities based on 25+ years long experience from software industry could be very different from users' perspective.

But first of all, I think such ranting is not bad think when formulated in this way.

And now, why I think some conslusions are incorrect:

The point is that Bentley will give us quantity, rather than quality. Software quality does not depend in any way how many versions you will publish. Many software is released even more often (e.g. Firefox every 6-8 weeks), servers are typicaly updated daily or faster (Amazon deploy their software every few minuts). Are these code more buggy than e.g. Windows with 1/2 year update cycle? I am sure not. Software quality is about development management, not about releases (and even not about beta testing programs).

A bunch of xxxx, instead of a little good. This is how waterfall project management works. But it does not work for software, it has been confirmed by 20+ years of software development. On the other hand, agile approach with increasing popularity for last few years, when not managed properly, does not product good software too. But again, it's not about the style (one big release versus many small releases), but about how the development is managed.

As you mentioned, there is not enough time for doing a Beta program. Samir mentioned it, I quoted it only ;-) This formulation, cut off from its context, is wrong. The beta program still exists, but realized in a form of Technology preview, both public and internal. And do not mix "beta testign" (generalized approach how to test software) and "beta program" (prepare software to be tested externally).

New features should not be more important than having something working properly. Yes and no, it's always about a balance, because you never have enough resources. But often I do not understand how MicroStation development is managed, because even when big bug is discovered, discussed here in forum, defect number and tickets are created, never information when (in what update) it will be solved, is provided.

Why was the product released with a "work around"? I do not recall this topic, but it sounds to me like a misunderstanding. Workaround mean something was discovered as disfunctional in the version and another solution (maybe longer, more complicated or whatever else) was found. When a new version remove the workaround, it means the bug was corrected and the longer workflow is not needed now. But software are not released with workarounds, such formulation does not make any sense.

they are releasing an untested piece of software (with no Beta testing) Based on what information (because nothing like this was published as far as I know) did you write this summary? Bentley View is based on the same power platform as MicroStation, so maybe about 99% of the code is the same. Of course e.g. installer can be buggy, but the rest is equal to MicroStation itself (whatever it means in terms of quality ;-). And beause no public beta testing program was published, you assume there has not been any testing done?

and killing "old reliable" before anyone has a chance to the issues. It was not told in this way.

I would ask for a "limited access token" Join technology Preview community, that's simple.

During the Beta phase, a pre-release... With all respect, there are too many "would" and it looks to me more like very optimistic and naive view how software testing works. I do not want to tell that testing by users is useless, it's not for sure, but my experience it can provide  just a friction (10%?) of inputs required for software development. And betatesting exists in MicroStation, it's only not called beta, but technology preview. I agree with one thing here: Bentley does not motivate users to test these features enough, the prepared features are not visible enough in my opinion.

They use Microsoft modules instead of writing their own. Thats normal approach in today software development. The question whether it's not good or not requires much wider discussion. See any software in some open source repository like GitHub or GitLab and you will see, that often tens of different modules are used by even simple applications.

When they don't work properly, "It's Microsoft's fault." Which is correct. the problem here is not "to use Microsoft modules", as you wrote, but to choose functional one, regardless they come from Microsoft or not. My preference is to use open source solutions, because you see what is inside and there is (not always, but often) a community of other users maintaining the code.

When Microsoft changes to ribbons, Bentley has to follow, even though it does not help the productivity of their product. You are mixing pines and apples. I am very sure to use ribbon was more management / political decision, not technology (maybe because of certification?). And I remember some overview during early acess to CONNECT Edition, somebody provided overview what CAD software uses ribbon and the result was: many, I guess it was the most of them. But the difference was that other solutions extend ribbon feaures to new controls, so the ribbon fits better to typicall workflows. Contrary to it, Bentley sticks very much to standard (and limited) ribbon.

(In fact, what happened to Windows 9?  What happened to Microstation 9?)  Seriously, who cares about naming? Why at the same time you don't ask what happened to MicroStation 6? And to blame Microsoft from everything bad was cool maybe 10 years ago. MicroStation CONNECT Edition is pretty new animal, rewritten probably more than ported, nwely structured with new API, so it's not surprise it's represented as new version with new naming.

Microsoft, in their pursuit of features and quantity of releases... Please, stop blaming Microsoft here, this is MicroStation forum. Whatever Microsoft does well or not, has nothing to do how MicroStation is developed and should be discussed in other forums I think.

Why are we forced to follow Microstation's new and problematic line? I know you think everything bad come from Microsoft, but try to attend any professional software engineering conference (not development one, but the conference focusing how software is engineered and maintained) and you will see "a dark side" of agile development is discussed often (and Microsoft is not mentioned at all). Whereas in average agile development brings better results than waterfall management, unfortunately it's also sensitive to proper planning and prioritization, otherwise the result is "constant beta" ... which is what MicroStation seems to be now unfortunately.

It's time to stop and figure out what the future of Bentley's product should look like. I agree. That's the problem plans and priroties are not shared and discussed, so nobody can optimize MicroStation deployment. I do not know whether it's because Bentley think it's top secret or nobody knows what will be development focus and priority in next 2 montsh, but I know about companies and software producers that share agendas and targetings for next few release cycles, so it's easier, quicker and cheaper to decide when it's a right time for installation or e.g. limited testing.

With regards,

  Jan

Hi Robert,

thanks for collecting some important issues and topics, but personally I dsagree with some of yours conclusions, because in my opinion despite of based on correct fact, they are wrong. But of course my perspective nad priorities based on 25+ years long experience from software industry could be very different from users' perspective.

But first of all, I think such ranting is not bad think when formulated in this way.

And now, why I think some conslusions are incorrect:

The point is that Bentley will give us quantity, rather than quality. Software quality does not depend in any way how many versions you will publish. Many software is released even more often (e.g. Firefox every 6-8 weeks), servers are typicaly updated daily or faster (Amazon deploy their software every few minuts). Are these code more buggy than e.g. Windows with 1/2 year update cycle? I am sure not. Software quality is about development management, not about releases (and even not about beta testing programs).

A bunch of xxxx, instead of a little good. This is how waterfall project management works. But it does not work for software, it has been confirmed by 20+ years of software development. On the other hand, agile approach with increasing popularity for last few years, when not managed properly, does not product good software too. But again, it's not about the style (one big release versus many small releases), but about how the development is managed.

As you mentioned, there is not enough time for doing a Beta program. Samir mentioned it, I quoted it only ;-) This formulation, cut off from its context, is wrong. The beta program still exists, but realized in a form of Technology preview, both public and internal. And do not mix "beta testign" (generalized approach how to test software) and "beta program" (prepare software to be tested externally).

New features should not be more important than having something working properly. Yes and no, it's always about a balance, because you never have enough resources. But often I do not understand how MicroStation development is managed, because even when big bug is discovered, discussed here in forum, defect number and tickets are created, never information when (in what update) it will be solved, is provided.

Why was the product released with a "work around"? I do not recall this topic, but it sounds to me like a misunderstanding. Workaround mean something was discovered as disfunctional in the version and another solution (maybe longer, more complicated or whatever else) was found. When a new version remove the workaround, it means the bug was corrected and the longer workflow is not needed now. But software are not released with workarounds, such formulation does not make any sense.

they are releasing an untested piece of software (with no Beta testing) Based on what information (because nothing like this was published as far as I know) did you write this summary? Bentley View is based on the same power platform as MicroStation, so maybe about 99% of the code is the same. Of course e.g. installer can be buggy, but the rest is equal to MicroStation itself (whatever it means in terms of quality ;-). And beause no public beta testing program was published, you assume there has not been any testing done?

and killing "old reliable" before anyone has a chance to the issues. It was not told in this way.

I would ask for a "limited access token" Join technology Preview community, that's simple.

During the Beta phase, a pre-release... With all respect, there are too many "would" and it looks to me more like very optimistic and naive view how software testing works. I do not want to tell that testing by users is useless, it's not for sure, but my experience it can provide  just a friction (10%?) of inputs required for software development. And betatesting exists in MicroStation, it's only not called beta, but technology preview. I agree with one thing here: Bentley does not motivate users to test these features enough, the prepared features are not visible enough in my opinion.

They use Microsoft modules instead of writing their own. Thats normal approach in today software development. The question whether it's not good or not requires much wider discussion. See any software in some open source repository like GitHub or GitLab and you will see, that often tens of different modules are used by even simple applications.

When they don't work properly, "It's Microsoft's fault." Which is correct. the problem here is not "to use Microsoft modules", as you wrote, but to choose functional one, regardless they come from Microsoft or not. My preference is to use open source solutions, because you see what is inside and there is (not always, but often) a community of other users maintaining the code.

When Microsoft changes to ribbons, Bentley has to follow, even though it does not help the productivity of their product. You are mixing pines and apples. I am very sure to use ribbon was more management / political decision, not technology (maybe because of certification?). And I remember some overview during early acess to CONNECT Edition, somebody provided overview what CAD software uses ribbon and the result was: many, I guess it was the most of them. But the difference was that other solutions extend ribbon feaures to new controls, so the ribbon fits better to typicall workflows. Contrary to it, Bentley sticks very much to standard (and limited) ribbon.

(In fact, what happened to Windows 9?  What happened to Microstation 9?)  Seriously, who cares about naming? Why at the same time you don't ask what happened to MicroStation 6? And to blame Microsoft from everything bad was cool maybe 10 years ago. MicroStation CONNECT Edition is pretty new animal, rewritten probably more than ported, nwely structured with new API, so it's not surprise it's represented as new version with new naming.

Microsoft, in their pursuit of features and quantity of releases... Please, stop blaming Microsoft here, this is MicroStation forum. Whatever Microsoft does well or not, has nothing to do how MicroStation is developed and should be discussed in other forums I think.

Why are we forced to follow Microstation's new and problematic line? I know you think everything bad come from Microsoft, but try to attend any professional software engineering conference (not development one, but the conference focusing how software is engineered and maintained) and you will see "a dark side" of agile development is discussed often (and Microsoft is not mentioned at all). Whereas in average agile development brings better results than waterfall management, unfortunately it's also sensitive to proper planning and prioritization, otherwise the result is "constant beta" ... which is what MicroStation seems to be now unfortunately.

It's time to stop and figure out what the future of Bentley's product should look like. I agree. That's the problem plans and priroties are not shared and discussed, so nobody can optimize MicroStation deployment. I do not know whether it's because Bentley think it's top secret or nobody knows what will be development focus and priority in next 2 montsh, but I know about companies and software producers that share agendas and targetings for next few release cycles, so it's easier, quicker and cheaper to decide when it's a right time for installation or e.g. limited testing.

With regards,

  Jan

Hi Jan,

Thanks for your comments.

I think that the bottom line for me is that I'm caught between a rock and a hard place.  Microstation V8i (SS3) no longer can handle the DWG files that we are getting from newer Autocad versions.  Microstation V8i (SS4), immediately showed printing bugs related to reference files with clips.  (I don't remember what else I found in the two weeks of using it before uninstalling it and returning to SS3.)  Microstation Connect, on it's 12th commercial release, still has significant bugs that prevent our daily workflow. 

I FEEL like I am putting in a lot of time into creating SR's that become defect reports or feature requests which go into a black hole.  I don't know if they will be scheduled for the next version, or if they will never be gotten to.  I would like to be able to continue setting up and testing Microstation Connect to distribute to everyone in the office, but I am unable to be productive in it without messing up our working DGN files.

That is where the frustration comes into play.  I don't have a good platform to move on to, but my "old" platform is becoming obsolete both from the functional side and from the support side.

--Thanks,
--Robert Arnold

P.S. I was just thinking that I should give a couple examples:

1) Issue wtih Update 11:  Open a DWG file, Save As a DGN file.  Day One and Day Two of using Update 11, I have encountered a DWG file(s) that gets SaveAs'ed a READ ONLY DGN file.  This is confusing for the users, and breaks one of my daily VBA routines (Open DWG, Save As DWG, delete unneeded elements.)  This did not happen with Update 10.  So even if I were doing "new feature" testing, this change had no chance of end-user input.

2) Enter Data Fields turn white when you enter data into them.  This was in Update 10 and reported, and now continues to be in Update 11.  This messes up the DGN file--not just an operational glitch.

3) When re-attaching a reference file with the path defined by a configuration file, Microstation 7 would keep the configuration variable if you stayed in the same folder.  Microstation 8 lost that feature.  I filed a bug report (or two--different releases).  I filed the same bug report for Microstation Connect Update 10.  Finally, the support person found my 9 year old old bug report and let me know that it was marked as WAD.  So my bug report is now a feature request. 

4) It's a minor annoyance that the "file open" dialog box randomly moves to the top left corner of the screen.  I reported this early on during the pre-release phase, but it's still here.

I have more issues (some more important than listed here), but I wanted to point out different situations.  1) A bug introduced into the core product (not a new feature) during an update.  (I probably will not distribute Update 11 to my office.)  2) A core feature that is broken and has not been fixed with the update, which is one of the things preventing us from using Connect.  3) A bug report that went into the black hole which, after 9 years, I find out that it went into the trash can.  4) A bug blamed on a Microsoft module, which has not gotten fixed in 12 commercial releases, even though reported during the Beta phase.  (I don't care whose bug it is, if it doesn't work right, I have paid Bentley to fix it.)

--Thanks,
--Robert

With Bentley's new release cycle we are implementing a release certification inhouse before it can be rolled out to users. So when release 12 is published its new functionality will be tested and the old features regression tested before deployed to user. This is the state of software today.

I used to supervise a small department of CAD techs (5 or 6), now it's just me, so switching to Connect was easier.

From reading your posts, I suggest you stay on v8i for as long as possible. In the meantime you'll want to evaluate your work flow, some tasks will need to be done differently.