OpenCities Map: Future development and directions?

Hi,

I am posting this question as a result of several discussions I had recently with customers. I am aware it's not a representative pattern, because I typically serve smaller companies and individuals, but in my opinion the topic is important generally.

My customers use BM (or more often BMPV because of very good price offer in the past) as drawing tool only, with spatial data support. BM was often used because INSPIRE and other data are important, and even when not well supported in V8i versions (WFS, WMTS, local formats...), it was still better than in plain MicroStation. In V8i, it works quite fine, because it's "like MicroStation", so experience and setting can be shared easily and it allows to access data in ad-hoc style and to create output on top of them.

When talking about "BM as GIS", it's much worse. Right now I have only one customer who use BM as GIS tool, all other left Bentley platform towards to competitive (better supported with more advanced functionality).

All these users are now evaluating a migration to CONNECT Edition. They are usually have mixed environments with MicroStation, PowerDraft or civil products. Sometimes competitive tools like AutoCAD or ArcGIS are also used in parallel. The question is: What is the future and directions of OpenCities Map? Does make it sense to move to CE platform, or to plain MicroStation / PowerDraft, or to some competitive platform?

I must admit, that I have had no clear answer or recommendation for them. And I realized I have no idea what OCM product line should be used for, what are development aim, targeted areas and workflows and competitive strengths. Whereas I see priorities and quite clear directions for products like MicroStation (the platform in general), OpenRoads Designer or GenerativeComponents, because there are active discussions, webinars, SIG sessions etc. available, OCM world looks more like "we gave up".

 It is a bit complicated to talk about GIS today:

  • I agree with an idea that "GIS is dead" ... it has become integrated technology, working often in a hidden way.
  • Also, the most of GIS has moved to web solutions (including editing and analysis), which is area what have been abandoned by Bentley years ago.
  • Existing GIS desktop products (ArcGIS + QGIS as ethalons) offer complex advanced functionality in a range from data capture, processing, analysis, visualization and even some cartography production).
  • GIS is spreading as technology to other disciplines and is incorporated, integrated or inspiring other areas (aggregating data, enrich them with spatial info, analyze them in cloud...).

So, what is OCM expected to be?

  • Desktop GIS product? No.
    • Complicated for users, nothing is "click and use or drag-and-drop".
    • Weird (even powerful) GSA, not flexible.
    • Very limited functionality with a lot not well connected tools.
  • Data capturing tool? Not quite.
    • To limit OpenCities Map PowerView to be spatial database read-only client, in time when spatial databases are core of spatial data world, force users to use other cheaper solutions.
    • When for basic data capturing (or editing) another technology is used, why to use full OCM, which increases operation and maintenance costs?
  • Data processing tool? No.
    • Still fragile weak support of spatial databases.
    • When there is an issue, not enough materials.
    • No migration, data cleaning, data quality etc. tools available.
  • Development platform? No.
    • Comparing to MicroStation CE API, which is consistent and well designed (and pretty well performant), it's ... well, it cannot be used.
    • Geospatial Extranet (the only source of some information) is not functional anymore.
  • Map production tool? No.
    • Map producing tools are limited, basic functionality only (well, Update 5 offer a bit better labeling).
    • Complicated styles / sings definition, limited, fragmented between MicroStation and Map-specific tools.
    • No cartography tools at all.
  • Client side of server solution (ArcGIS Pro + ArcGIS Server, QGIS + PostGIS/PostgeSQL)? No.
    • There is nothing available for main spatial DB (Oracle, SQL Server, PostGIS/PostgreSQL) that optimize the platforms for usage of OCM client.
  • Client side of web publishing solution? No.
    • GeoWebPublisher is history, weak and dysfunctional.
    • PostgreSQL + GeoServer + OpenLayers + ... can be alternative, but again painful to solve e.g. line styles, because no conversion tools or best practices availability.

So, what is unique feature of BM and what is its development priority?

Maybe it's 3D GIS, but it's something very specific, not interesting for 99% of users. Which does not mean it's wrong, because when old users will be lost, new ones can be found. But now, what users tell me, this platform cannot be trusted as stable for such expensive decision like the migration. Even when the migration to competitive platform is typically quite expensive (data conversion, users training etc.), in long term perspective it seems to be cheaper, because: directions are known, the solutions are well supported locally, there are companies available to develop local specific tools and there are best practices how to go for web, which is crucial.

So, end of whining. May I ask for some insight what is the future? When the current surviving mode will change to "we are on edge, these features are breaking news and we are able to offer bigger value than our competitors, including supporting local regulations and specifics"?

With regards,

  Jan

Parents
  • Hello,

    thanks very much for Your questions!

    I have for example two bad experiences in my company:

    -  in 2016 we start using GeoWebPublisher, today when we still don´t have all done in intranet, is clear that next year we will stop paying Select for this product (without future); I was hoping that Bentley would offer some help with another solution, no offer

    - is year 2020 and our software extension for Bentley desktop software still don´t use Connect version – Bentley Select price (discount) is about how is using Connect version – but our systems needs at least V8i SS10.

    Best regards

        Pavel Zobal

  • I was hoping that Bentley would offer some help with another solution, no offer

    It seems Bentley gave up any activity in web publishing. In fact, there are not many companies now developing full featured GIS publishing (of course ESRI, but it's huge complex solution). There are rational reasons behind such decision, because every installation is often highly custom, but the other companies offer best practices, tools etc. how to implement web server using open-source stack (typically PostgreSQL/PostGIS + GeoServer or MapGuide), using QGIS Cloud or similar.

    Unfortunately there are no tools for anything :-( ... data processing / cleaning / quality check (especially when XFM is stored in DGN, not in DB), data migration etc.

    is year 2020 and our software extension for Bentley desktop software still don´t use Connect version

    Based on my experience (as I have participated on several MicroStation V8i > CE migration projects), the migration is "controlled" by several requirements and conditions, and all have to be (partially at least) fulfilled:

    • Platform existence: Bentley Map in CE version was released with a huge delay comparing to MicroStation, and the first versions were pretty buggy.
    • SDK existence: Geo SDK exists, but whereas MicroStation API was cleaned, modernized and restructured (downside is you have to learn completely new API), geo SDK looks fragmented and just weird.
    • No migration tools and best practices defined: When spatial database is used, it's simpler, but e.g. to migrate hybrid solution (DGN + mslinks to DB) is nightmare.
    • Complicated "migration stack": To migrate BM solution, you have to know at least something about migration MicroStation (GUI, workspace, code) and add Bentley Map migration knowledge and complexity on top of it.

    In this situation, as I expressed in my original question, it's quite understandable question of some customers "Why to migrate to CE platform?", when there is no clear future and even when the migration to another platform is more expensive (but the difference is often no so big), but operation / maintenance cost is just fraction (because of available of tool, knowledge base etc.)?

    With regards,

      Jan

  • Hi. I'm sorry to hear that you find the OpenCities Map SDK fragmented. It is true that the Microstation SDK has been transformed quite a bit from the previous generations. Unfortunately, with all those changes comes a higher cost to migrate. When porting Bentley Map to OpenCities Map CONNECT Edition, we took the decision to keep the API as close as possible to the previous generation so that our users and 3rd party developers would have an easier time porting existing applications. At the time we thought that was the best decision. 

    When moving to CONNECT Edition, the file format has not changed. We do not force any data migration to spatial databases, projects can continue to work with their data sources and repositories as before for the most part. If you are interested in migrating to a spatial database, I would like to hear more about your needs. 

    I would also like to hear about the issues that are preventing your applications to be ported to the CONNECT Edition. Maybe there are things we can help on?

    Bentley is developing tools that can work efficiently on the web. You can have a look at iModelJs for example. 

      

  • Hi Martin,

    thanks for your answer. Honestly, I expect there will be faster response than 3 weeks for such basic question :-(

    Unfortunately you did not answer my main question: What is future direction and priorities of Bentley Map development. Without such justification, it's hard to discuss anything else, because it can be limited to particular features or bugs, but without any prioritization or broader context.

    I'm sorry to hear that you find the OpenCities Map SDK fragmented.

    Sorry, but it is. No doubts.

    There is not single piece of documentation / intro / tutorial, explaining concepts, structures, types of APIs (several separate of managed, old C native, VBA...). Just a few chm files. No use cases, real example (not API examples, but example demonstrating anything more complex).

    It's like to take several puzzle games (native, COM and NET APIs for different functionalities), to take only some pieces from every set (old DB, DGN based, spatial connection, new MapManager features), and expect such game can be played (puzzle will be built). But no information what pieces can fit together, what not, what is the context and relationships. So just a mess.

    It is true that the Microstation SDK has been transformed quite a bit from the previous generations.

    It's not important whether it was transformed or not. When stable and without bugs, the quality of API is often defined by quality of documentation, not whether API structure is the best in the world.

    It's one from core feature of the whole product, not API only: Different tools/technologies (data access, data creation, topology cleanup, spatila analysis, scripting...) work only with some data and projects (mslinks, XFM DGN, XFM spatila, plain DGN data).

    We do not force any data migration to spatial databases

    My experience from earlier versions has been very different: hybrid (DGN + mslinks to DB) did not work, plus some features (I reported some in the past) do not work with data structure at all. Plus, because you do not deliver any data quality / data cleanup tool, the only way how to clean the data, keep maintenance possible and ensure at least something will work fine, is to migrate data to spatial database.

    projects can continue to work with their data sources and repositories as before for the most part.

    It's nice theory, but too close to marketing promise, but far from reality.

    I am too tired and annoyed to test every released version, to report blocking bugs, find migration steps by myself because not documented. I have only one customer customer using BM as GIS, and he sees no reason to pay for testing, reporting bugs and finding workarounds for things that is expected they will work and will be documented properly. Especially when local support is just zero and competitive platforms are localized (using correct terminology, not such **** as Czech Bentley Map GUI, translated by somebody who never heard about GIS, official GIS terminology and probably used copy-paste from Google/Bing translators), and typical workflows and situations are described.

    So not, the migration is painful.

    If you are interested in migrating to a spatial database

    Personally I am not interested, because it's not win-win business case for me. To learn all specific aspects, how it's implemented in BM, especially using user-agressive Geospatial Administrator GUI, is just time and money lost when there are no business opportunities, but only one customer.

    But it seems to be the only way in long term planning ... or to go move to another solution.

    I would like to hear more about your needs. 

    It's quite simple: From very old (GeoGraphics based) project, converted somehow to XFM mslink based, to transform into SQL Server spatial data. I have never ever seen any best practices + tool for such process, that I am sure (based on my experience from some projects based on open source) will involve e.g:

    • Analyze defect in DGNs (often messages like "Duplicate current version XFM root-feature element" are displayed, have never found any information what does it mean and what is workflow/tool to fix it).
    • Analyze XFM features in DGNs (what types, how many, versions, some are native, some inferred).
    • Analyze DB records (structure, integrity...).
    • To find orphans on both sides.
    • Analyze necessary changes in current SQL Server data model.
    • Analyze what feature can be migrated and what changed (collections
    • Analyze how existing annotations can be converted to DB (to maintain existing position is crucial)
    • Analyze how XFM project has to be changed
    • Design and implement all necessary scripts (DGN, DB, migration)
    • Do the conversion

    I think these steps, best practices and workflows are common and should be integral part of any GIS product documentation. But because there is not Bentley Map knowledge base available...

    I would also like to hear about the issues that are preventing your applications to be ported to the CONNECT Edition.

    I have nothing Bentley Map specific to be migrated.

    Long time ago I gave up my effort to develop anything on Bentley Map API. There have not been enough time and budget to learn everything from scratch without proper documentation and good tools: Even simple GIS tool with workflow select feature + another feature, display dialog, enter values, process them and based on result start drawing is just nightmare: How to set method / operation (and what they mean, how they should be used?), what should be implemented in code, how to work interactively with XFM dialogs etc. Documentation is "only about API", describing concepts not described anywhere or in just few sentences, written by somebody who already knows complete context, relations and dependencies.

    There is even single one demo how to write application (not how to write code snippet solving one task, without any context) from beginning to end, what API to use to what etc. I have no time to analyze everything myself from code and examples (as I did e.g. for EC data in the past in V8i, but there are many business opportunities).

    Bentley is developing tools that can work efficiently on the web. You can have a look at iModelJs for example. 

    Is it a kind of joke? I asked for web for GIS. Moreover iModelJS, that I monitor carefully from beginning and I tried to wrote some code already (but it's not priority and my TypeScript and web environment is not good enough), but it's technology, not product.

    Product can be iTwin Design Review or i-model Hub, but they are very good examples how Bentley are typically excellent in technologies, but very bad in designing end-user products and their promotion and explaining to users So far all people do not understand how it works, what is workflow (because it's seriously overcomplicated) and where benefits are.

    Moreover, nothing from these technologies support 2D drawing now, it's not even possible to display simple 2D design (map) in iTwin Design Review. And when talking about GIS web output: I am not aware of any support for interactivity (select feature, display complex query result), support for thematic mapping and resymbolization etc. And of course, it's not live or pre-cached publication from database, so cannot be used i GIS are at all (but maybe in asset management, but in such case I do not believe such critical data can be uploaded to cloud).

    Plus, I do not see anything like "Bentley Map iModelBridge" in the list.

    BTW Bentley often fights with partners (becase we are Bentley, we can do everything ourselves), limits every possible development (even to be hobby in-house developer, you have to sign agreement, which block 99% of people interested in to test to learn coding, BDN program is just disaster, all competitors offer much much much better) and also because of lack of local staff, there is zero knowledge about local specific, rules et.c

    Conclusion: I do not see any future for Bentley Map, when compared with what customer need and competitors offer (only to use BM as simple drawing tool with limited support of spatial data sources). It does not mean this product cannot excel in some specific areas, but it's not clear where they are, they are not explained and demonstrated.

    With regards,

      Jan

  • One more example, why I am so pessimistic and I think the situation is just a tragedy (plus I am supported in this mood also by my customers):

    There are:

    • Bentley Map
    • Bentley Map Mobile Publisher (but not available for OpenCities Map yet accordingly to documentation)
    • Bentley Map Mobile

    Three puzzle pieces, who normal customer (user or manager) is not able to put together in the right and expected shape.

    I do not recall any documentation, which I treated as absolute basic (and has been asked for by customers often), describing e.g.

    • typical scenarios (e.g. utility company, data in XFM + spatila database, maintenance workers)
    • how licensing works in this case (for customer it's chaos in SELECTserver and SES licensing, does Mobile Publisher requires license, how it's with Bentley Map Mobile)
    • how license should be implemented (e.g. shared devices versus named users in SES licensing)
    • what data can be accessed in Map Mobile
    • what is typical workflow (publish, download to e.g. tablet, review, incorporate notes/comments back to project)
    • how conversion should be configured
    • description of individual steps
    • what licenses and subscriptions, in what price, are required for the described scenario

    It's end-user solution, but nobody care to identify benefit, explain it and really sell it to customers. And all customers in the past interested in at least testing of mobile access, stop this initiative because of lack of clear documentation and information (how Map Mobile is licensed has been changed and unclear so many times, to find the cost even now is complete disaster).

    Regards,

      Jan

Reply
  • One more example, why I am so pessimistic and I think the situation is just a tragedy (plus I am supported in this mood also by my customers):

    There are:

    • Bentley Map
    • Bentley Map Mobile Publisher (but not available for OpenCities Map yet accordingly to documentation)
    • Bentley Map Mobile

    Three puzzle pieces, who normal customer (user or manager) is not able to put together in the right and expected shape.

    I do not recall any documentation, which I treated as absolute basic (and has been asked for by customers often), describing e.g.

    • typical scenarios (e.g. utility company, data in XFM + spatila database, maintenance workers)
    • how licensing works in this case (for customer it's chaos in SELECTserver and SES licensing, does Mobile Publisher requires license, how it's with Bentley Map Mobile)
    • how license should be implemented (e.g. shared devices versus named users in SES licensing)
    • what data can be accessed in Map Mobile
    • what is typical workflow (publish, download to e.g. tablet, review, incorporate notes/comments back to project)
    • how conversion should be configured
    • description of individual steps
    • what licenses and subscriptions, in what price, are required for the described scenario

    It's end-user solution, but nobody care to identify benefit, explain it and really sell it to customers. And all customers in the past interested in at least testing of mobile access, stop this initiative because of lack of clear documentation and information (how Map Mobile is licensed has been changed and unclear so many times, to find the cost even now is complete disaster).

    Regards,

      Jan

Children
  • Hi Jan. From your tone I see that there is nothing I can say that will make things better :-) Still, I will provide some additional information.

    When I pointed to iModelJs I just wanted to show that there is a new iModel platform and technology stack that is web-ready and can help users build digital twins. A web gis application is one type of digital twin application. I believe the iModel platform will have better capabilities than what GeoWebPublisher could do in the past. It already supports BIM data and we are adding geospatial capabilities. The platform will support display of external sources like WMS/WMTS for example. We are also building a geospatial iModel Bridge. We have SHP file support in production now and are adding spatial databases. That means the GIS data can be brought into the iModel. The iModel display on the web is efficient, with streaming and caching of 3d "tiles". The Design Review application might not be well suited to an operations digital twin but the iModelJs api is well defined and well documented. It is possible to build your own application. Bentley will also develop other applications on that platform, we are working on some now. We will communicate externally when they are closer to being ready. 
    You can also look at OpenCities Planner, that supports a bunch of geospatial formats already and does great rendering on the web. You are correct, none of these are direct replacements for GeoWebPublisher. However both of these show that Bentley is still very much in the geospatial and web areas. 

    Apart from product documentation, being a SELECT subscriber gives you access to our support services. They can help you when things aren't clear, and especially around application usage. Bentley also provides consultants, services and success plans to help users choose the right software and architect thew overall solution. 

    I will happily relay your data quality/validation and migration comments to the OpenCities Map product management.

      

  • From your tone I see that there is nothing I can say that will make things better :-)

    Until my main question is answered, probably not...

    When I pointed to iModelJs I just wanted to show that there is a new iModel platform and technology stack that is web-ready and can help users build digital twins

    Unfortunately "digital twins" has become overused marketing *** terminology. A week ago I attended a conference (2 days web casting because of current restrictions) and it was substantial part of one great presentation: What digital twins means, what they really are, whether they are just crap or is there real use-case and benefits behind them etc.

    But iTwins and iModelJS are very different topics from my original question, so even it can be valuable to discuss it, it does not solve any from my current pain.

    I believe the iModel platform will have better capabilities than what GeoWebPublisher could do in the past.

    I do not think and I do not believe. Other capabilities, targeted to other scenarios and users? Yes, no doubts, there is a huge potential. Whereas Map has been promoted as GIS / geospatial solution (but I think it's not anymore and it shifted to another direction and is not competitive GIS), iModelJS with iModelBridges and iModelHub sounds so far better suited for AM/FM (where I think Bentley can provide great technology to pass gap between GIS and BIM).

    Two (simplified) typical scenarios have presented by my customers:There are spatial data, often quite complex both graphically (utilities maps with plenty of hierachical features and many annotations, dimensions etc.) and in their structure (lookup tables, dependent tables, IDs to external system). When these data have to be displayed, they are merge of original data, enhanced by parcels, ortophoto etc.

    • When element is located (clicked), based on data in spatial DB plus queires to external system, information is displayed.
    • When system is queried, map is displayed and feature, buffer, part of topology is highlighted.

    I do not think this is scenario where iModelJS can help anyhow.

    Second scenario is more interesting: Similar to first one, but to be able to access data outdoor (typically off-line). Here iModelJS briefcase sounds interesting, because it's not about live data. Not sure how enhancement / enrichment spatial data by data from external system (what was a role of i-model Transformer in the past) when data are streamed from i-model Hub (?).

    But I can imagine, it will be implemented out of i-model Hub ecosystem, because no transactions are required in this case.

    but the iModelJs api is well defined and well documented.

    Yes, it's best Bentley project in history. Whereas Bentley has fought against 3rd party developers last several years and the development has become expensive and painful, iModelJS project is what Bentley were long time ago, when they were top class technology company and openly support 3rd party developers to be able to solve local requirements, that never can be addressed by global company (even when now many people in Bentley think that global Bentley solution is the best for everybody, don't accepting a fact every country and discipline is different).

    My feeling is that that one from big problem of Bentley: Providing (sometimes excellent) technology, but trying to sell it as end-user product, optimized for local usage. In the case of iModelJS, there is not such mistake done, the roles, rules and business approach are clearly defined from early beginning.

    Apart from product documentation, being a SELECT subscriber gives you access to our support services.

    Historically I have not very good experience with Bentley Support for many reasons. And without knowledge of local standards, no good application support can be provided.

    Bentley also provides consultants, services and success plans to help users choose the right software and architect thew overall solution. 

    Yes, in US rates, English or maybe German speaking people, without knowledge of local terminology, standards and requirements. How it can be applicable in local conditions, where the complete documentation is in (in my case) Czech language and users + managers often do not speak English (completely or very limited). Plus, budge is limited to just a few days every year.

    Regards,

      Jan