RSS feedPlease find below a list of questions that we've collected in the last three years thanks to more than 100 customers. We've tried to order the questions from more general towards more specific. Please contact us and tell us about your question if you don't find it on this page.
Or you can put it in another way:
Free software is a matter of the users' freedom to run, copy, distribute, study, change and improve the software. More precisely, it refers to four kinds of freedom, for the users of the software:
There are quite a few (at least 300) commercial project management
systems in
the market.
There are two basic ways you can inquire for additional features.
a) Request your feature at sourceforge.net :
We use sourceforge.net to manage feature requests
from ]po[ users. When updating our development road map we evaluate those suggestions
and try to consider as much as possible. Such requirements are subject to the condition that
the feature serves a broader amount of users.
b) Sponsored Development
In case you don't have the time to wait for approval and implementation as described above,
you can always have us developing particular features based on your specific needs.
In case the requested changes fit into our product strategy the necessary resources are
available to deliver the functionality in the required time frame, we usually sponsor such
type of development.
However, depending on the complexity of the implementation and the need for further
specifications to be defined in cooperation with the customer, these projects
might be in some cases not in line with the budgets of smaller companies or single users.
Please use our contact form to inquire about sponsored development.
We can provide you with remote professional support for your installation or your production system. Please contact us for details. For free support please use the open-source "Discussion Forum".
More and more companies do so. Please apply your standard vendor evaluation matrix and you will see that open-source doesn't score that badly. However, you will have to adapt your matrix a bit and replace "Vendor Size" by "Community Size and Professionality" and so forth.
We believe that Open-Source is a great fit particularly for small companies (even though we've already got our first Fortune-500 customers...).
]project-consulting[
and ]project-translation[
are specific products, while
The main difference is between ]project-translation[ and the other tree products (]project-consulting[, ]project-lifecycle[ and ]project-consulting[) is the type of supported "project tasks":
Apart from that they are the same product. You can actually have "Timesheet Tasks" and "Translation Tasks" mixed in the same system, for example as subprojects of a common parent project.
]pc[, ]pl[, ]pa[ are very similar to each other and share the same code base. However, the configuration (types of projects, types of ..., other packages that are installed by default) vary. They are really more "variants" of the same product, than different products. However, we have given them different names because we plan to differentiate them more in the future by adding extension modules.
Yrjan from Norway asked:
> I downloaded "project-open-v32Beta2.zip" and installed
"VMware Server".
> But are I supposed to run my company on a virtual machine?
How can I upload and
> install ]project-open[ to my webserver and make it online?
Hi Yrjan,
Yes, one good option to run ]po[
is as a virtual machine on your server (fileserver, ...). Fileservers
usually already exists, they usually are integrated into a backup
system and their CPUs are mostly idle during the day. I have seen
many small companies running on such a combination. These companies
usually forward port 80 of their aDSL routers to the ]po[ VMware,
either on port 80 (configured to run Pound by default) or directly
to the AOLserver (by default on port 8000). Then get an account
at www.dyndns.com for external
DNS. This configuration may be the nightmare of an Internet security
consultant, but AOLserver provides you with with great protection
against "script kiddy" attacks, and for everything else
your own employees are your biggest security thread...
The other option is a "dedicated server" at a hosting site. For 10 concurrently active users at peak time (an organization with ~50 members) you'll need 1Gyte RAM, a 2GHz CPU and 20G disk space, a recent SuSE, Fedora or CentOS system and root access in order to install the software. We had negative reports from virtually hosted servers, so you might have to go for a dedicated server.
We also do offer a ]po[
server as SaaS (Software as a Service). Please check http://www.project-open.com/en/services/project-open-hosting-saas.html
for details.
Advantages of a "home" installation:
Disadvantages:
Advantages of SaaS:
However, we have just released RPM installers for Fedora, CentOS,
OpenSuSE. Using these installers you should be able to install ]po[
on your web server if you really want...
It's quite easy to to move a ]po[
installation from one server to another (just perform a backup -
restore), so it may be a good option to start with a "home"
or "SaaS" installation and then move it "home"
if you really see the need for it.
Probably Yes. English is our default development language, and
we directly support Spanish and German. Adding any arbitrary other
language just takes you a day of work to translate the localization
strings. For example we have already received translations for French
and Brazilian Portuguese from our customers. Please see the PO-Configuration-Guide
for details about the
Yes, all recent Window versions (XP, 2000, 2003) are supported. ]project-open[ doesn't run on Win-95 and Win 98 though. Checkout the Windows download pages.
Yes, all major Linux distributions are supported. Please checkout the Download pages.
Also, there are users running ]project-open[ on Mac OS-X, *BSD, Sun Solaris, AIX. We have even heard about a port to OpenVMS...
No, and it probably will never work with Apache. ]project-open[ is based on AOLServer and the OpenACS platform. AOLServer is "the backbone of the largest and busiest production environments in the world".
AOLServer is easy to install, easy to understand, fast, reliable and little know to hackers and worm writers. We haven't had a single (successful) attack in more then 5 years.
Not yet... We've chosen the OpenACS platform (written in TCL) as a base for different reasons. (see Key Benefits). A system developed based on OpenACS is less vulnerable, easier to maintain and more skalable.
You can learn TCL in about 2 hours if you know PHP.
We're using mainly PostgreSQL. However, we can offer support for Oracle if desired.
No, and it probably won't in the near future. MySQL has improved in since 2005, but it still doesn't offer triggers, stored procedures and several other features that are frequently used in ]project-open[. Also, MySQL still has difficulties with incremental and online backups etc., which limits its usefulness for a system like ]project-open[.
Yes you can.
Our Windows installer includes a particular version of PostgreSQL for "CygWin" (7.5.X). This version has prooved to work very well in our tests and this is the version that we are using for our internal test & development servers.
Since the version 8.0 in March 2005 there is a "native" version of PostgreSQL for Windows. This version is know to be some 30% faster then CygWin Postgres, and is supposed to be more stable in future. However, there are some issues when moving a backup dump from 7.5.x to 8.0.x, so that we won't deliver this option as part of our free Windows Installer yet. However, you can try it yourself or you can contract us to upgrade your system.
We recommend an upgrade to 8.0 on Windows for companies with more then 10 "concurrent users" (=> users that access the system _at_the_same_moment_). Apart from that we would recommend you to stay with the CygWin version that comes with the installer.
> What is the actual meaning and import of :
> "In this configuration, the database will allow full access
to all data for
> all local users of the server computer while blocking the access
for
> anybody not working locally on the computer.
>
> This setup is very convenient for our ]po[
demo server where we can’t
> predict the name of the local users. However, you may want
to change
> these settings for a productive installation."
Please see http://www.postgresql.org/docs/7.3/static/client-authentication.html and the PO-Configuration-Guide for details.
The basic configuration allows all users of your local computer access to all of your data. Is that right for you? In general yes.
However, there are configurations where normal users have the right to log onto your server computer. That is the case for example if you use Windows Terminal Server. In this case you will have to modify your PostgreSQL security configuration.
> "]project-open[
supports the integration into infrastructures of larger corporate
> environments via LDAP"
Correct. "Larger corporate environments" frequently use a LDAP (or Windows Active Directory) as a centralized authentication and "directory server". The LDAP server includes all information about users, and provide authentication services to many other applications to provide "single sign on". This means that you can enter all applications with the same password, and that you only need to change your password once.
So: You an integrate ]project-open[ with such an environment. However, such an integration will come with a certain price of 1-2 days of consulting/sysadmin work (unless you know how to administer an LDAP server and you know how to read the OpenACS forums). So you have to decide wether you want to pay this price...
This message means that a "message key" has not been translated ("localized") yet. Please go to http://localhost/acs-lang/admin/ and click on "Translation Mode: ON". The go back to the page with the message and click on the red symbold to add the missing translation.
Here is how the relationship between OpenACS and ]project-open[ works:
Let's take the "Calendar" as a practical example: We've just added the standard OpenACS 5.1.3 "calendar" module to our default ]po[ distribution. The module worked fine as a standalone personal calendar, but we wanted to map ]po[ events such project deadlines and tasks into the calendar. So we've created an "intranet-calendar" module that defines a menu item for "Calendar", a pluggable "calendar overview component" for the homepage, and a few database triggers so that ]po[ objects are "duplicated" as calendar-items. So the "intranet-calendar" module doesn't really add new functionality, but it provides the "glue code" to link the new module with the system and to provide end users with a pleaseant and integrated user experience.
A brief look into the source code of "intranet-calendar" will reveal to you all necessary steps for such an integration. It's just some ~500 lines of (mostly SQL) code.
In case you do not have IT expertise to install and configure all components necessary, you might want to make use of our professional services. We can help you finding the right hosting provider / package for you and install your system on a server of you choice.
You can upgrade ]project-open[ manually using CVS. Our CVS server is publicly available with the user "anonymous" on berlin.dnsalias.com.
Upgrading from version 3.1 to version 3.2
In Version 3.2. we needed to make major adjustments to some basic 'pillars' of the application in order
to prepare ]po[ better for future requirements such as integration
in leading ERP systems.
Therfore this particular upgrade can only
be done semi-automatically. More details can be found here: http://sourceforge.net/forum/message.php?msg_id=4212495
Upgrading from version 3.2 to version 3.3
Upgrading from version 3.2 is quite straight forward. Just cvs update using the stable branch 'b3-3-patches'
See
http://sourceforge.net/forum/forum.php?thread_id=1862803&forum_id=295937 for more details.
This is explained in detail in the PO-Configuration-Guide.
Basicly, you use the built-in PostgreSQL commands to backup and restore your database. There is a "wrapper page" in ]project-open[, but this page basicly calls the PostgreSQL commands:
To backup: pg_dump projop -c -O -F p -f c:/backup.YYYY-MM-DD.sql
To restore: psql -d projop -f c:/backup.YYYY-MM-DD.sql
These commands work both on Windows (CygWin shell) and Linux. Please note that the "forward slashes" are being used in both cases.
Please note that the "restore" operation may write out a lot of error messages if you load a backup dump into an empty database. This is because the backup file contains SQL commands to delete the data in an existing database. You don't need to worry about these errors.
However, there is a small issue with PostgreSQL databases with version 7.5.4 or below. These version create backup dumps with a wrong ordering of the data. As a result, a few "SQL Views" are not created. But there is a workaround: Just issue a "psql -d projop -f c:/ProjectOpen/preconf/poall.post-patch.sql" to load these missing views into the database.
Setting up a new project in ]project-open[ requires you to go through several steps, because the reality can be very complex. We have though about simplifying the setup of a sample project using a type of wizard, but we haven't found a customer yet who would be willing to pay for it... :-)
Here are the steps in their order:
The main point is that all the information entered are stored in the database and available for reporting later, for an analysis of quotes that have become project (or not) etc.
Correct, the "percent completed" variable is not (yet) connected to any other element in the system. In future, it will be part of a "Earned Value Metric" package to track the progress of long running projects and possible be connected to Translation and other workflows.
Currently, the "percent completed" variable provides the Project Manager with a way to report the completion status of his project to his boss. Sorry, no magic... :-)
That's difficult. Basicly, there is no GUI for something like this.
The main reason is that large companies can have literally thousands
of employees, and showing such a list would just be too much.
However, there are several alternatives:
Please also see:
This message means that you haven't configured your financial system yet. Please see the PO-Configuration-Guide. You need to add a company of type "internal" with the "Company Path"="internal".
]project-open[ currently accepts numbers in American format only with a "." decimal separator and without thousand separators.
This will likely be changed in one of the next versions.
The default configuration of ]project-open[ only includes 5 default currencies: USD, EUR, CAD, AUD and JPY. In order to enable more currencies you will need to manually enabling your currency using a command like (for example for Russian Rubles 'RUR'):
update currency_codes set supported_p = 't' where iso = 'RUR';
This command should give you currency support in all major ]project-open[ modules. However, you will also need to go to the /intranet-exchange-rate/ module and add atleast a few exchange rates in for the past month in order to obtain reasonable results in the project "Profit & Loss" view and for some reports that use currency conversion.
In the "Admin" -> "Parameters" section of ]project-open[. Check the "PO-Configuration-Guide" for details.
> I see that it's possible to add multiple project to
> each invoice. Is it also possible to create a Customer
> Bill with multiple projects?
Yes.
Financial documents can be assoicated with any number of projects. However, we don't recommend this, because the ]project-open[ controlling subsystem doesn't know to to assign the invoices to the projects, so the controlling data become inconsistent.
However, the inconsistency is limited to the projects contained in a single financial document. It's OK to have inconsistent controlling information on subprojects, if all subprojects are children of the same main project, because the main project will be fine again.
So you'll have to decide whether fine-grain controlling is useful for you or whether you prefer to reduce administrative overhead by including many small projects in a single invoice.
> I get an error when trying to upload/download a file
Please check the "intranet-filestorage" section in your Admin -> Parameters menu and see the PO-Filestorage-User-Guide for more details.
The ]project-open[ filestorage works by looking at an existing directory, there is no "magic". To your directory please login as "projop" (or whatever user that is running the ]project-open[ server daemon) and try to access the directory. You'll probably get a permission error at some point.
> How to make "Company base path" variable point to another computer file storage?
The Filestorage module allows you to access the contents of a Windows or Linux file server via a web interface. The root of the Filestorage is defined in the "Admin" / "Parameters" section. The "PO-Configuration-Guide" contains a detailed explanation on how to integrate the ]project-open[ Filestorage into your file server.
Please check if the filestorage pathes are correct in "Admin" -> "Parameters". Make sure that all files in the subdirectories have the same user and group as the AOLServer process. Please note that under Linux/Unix a daemon only has the rights for the specific user/group. Daemon process permissions are not affected if you add the daemon user to another group.
> I would need a new folder called "reference" for
every
> new project that I create
Starting with V3.2 ]project-open[ supports a new concept called "user exits". These user exits are BASH or PERL routines that get called every time a certain action happens, such as the creation or deleting of a project, a user etc., so you could write your own code to create new subdirectories for every new project. Please see the documentation of "user exits" for more information.
> I have successfully installed ]project-open[,
but I don't see the
> "Reporting" tab in my server
The "Reporting" (intranet-reporting) package belongs to a number of closed-source extensions that we're selling on top of the free ]project-open[ product. We have got the polica that all important functionality should be free. However, we have found that larger companies are not very sensitive to costs, so we have split some modules from the main product as closed-source.
You can use Crystal Reports or any other reporting tool to get your reporting. Please see the "PO-Data-Model" guide for details on the data model.
> When I add a new customer, my name becomes an
> "employee" of this new customer. Why is that?
That's correct. The customer's ViewPage shows "Customer Contacts" (the customer's employees) and "Our Employees". Our employees are a kind of "key account managers" and have write access to the customer's file. That is necessary because not everybody in the system may have write access to the customer's file and we want to make sure that the creator of a customer's file does have this right.
So: You can delete this entry if you've got the right to edit customers anyway.
Please also see:
By Ryszard Kasprzyk from www.abovetranslations.com:
> We often have 10+ stages of QC. (See our Web
site, section | Projects for the full QA).
> What to do, if there is only a 3-stage process provided in
your system? Why 3 only?
Hi Ryszard,
there is a difference between a business process and a workflow. You may have a 10-step process, but this process may be implemented by a workflow with only three different roles.
For example, a project manager may do a final review before sending out the documents to the customer. However, you don't need(!) to create a new task for the PM, because he is responsible (and accountable) to do this step anyway (atleast in theory...). So basicly, you try to keep your workflows simple, because otherwise you get a lot of overhead and users will start to complain that the system is too difficult to use...
A software testing lab would be a different siutation. Such a lab may requires a completely different workflow because there are several types of "linguistic", "cosmetic" and other tests that are assigned to different persons.
However, even such processes can be implemented in ]po[ using Translation Tasks with a task type "other", even though not very elegantly.
To summarize: ]project-open[ only supports a standard workflow that fits 90% of all translation projects in the world. You would need to extend ]project-open[ to cover more specific requirements. But we need to make our money somehow... :-)
> How does one set the delivery date for subcontactors?
> How to track people's progress, like Gantt's chart in Microsoft
Project?
Translation progress is tracked using the "Project Status" display on the main project page. (The project needs to have a "Translation" type such as "Trans+Edit+Proof" in order for the "Project Status" to appear - it only makes sense with translation projects...)
Please see the PO-Trans-WorkflowGuide for details.
> How are users alerted to tasks?
> Are they sent mails when a new file (task item) appears?
No, we once tried a system like this, but users started to complain about too many emails. So we decided to alert a user (freelancers, employees, ...) only once when assigning them to a project.
We have found that translators keep up with the project news once they are assigned. Translation Projects normally tend not to be very long.
> Is there a central todo tasks list on home site for each member? I could not see any
There is no central list of translation tasks, no. The lists are visible only in their projects. However, each translator will see his list of _projects_ on his home page.
By Neil Gauld from Brightlines (slightly edited):
> We are having a problem with PO.
> Here is what we need to do and then what PO forces us to do:
>
> 1. Send text and usually a PDF of original doc to the translator.
> 2. The translator translates.
> 3. Send translated text, original source text and pdf to the
editor.
> 4. The editor edits the text
> 5. The editor returns the Final translation, word doc to us.
>
> Here is what we have to do with PO:
>
> 1. "Zip" source text and pdf together so we can upload
it as one task.
> 2. The translator downloads and unzips and translates.
> 3. The translator has to "Zip" translated text, pdf
and source text for upload.
> 4. The editor downloads the zipped file and edits text.
> 5. The editor uploads the edited text - but also has to make
sure that it is zipped
> as the task is named as a zip.
>
> If the Editor just uploaded a word doc then PO would change
the name to the
> original task name, which was a zip file. So when we download
the file we actually
> get a word doc masquerading as a zip file. Zip icon and zip
suffix.
>
> Any ideas? How do other people get around this problem?
> It would be nice to have a bit more flexibility in this area.
I see what you mean. Let's rephrase it: You would basicly want
to work based on "jobs" instead of "individual documents".
Is that right?
We have seen many customers, particularly from Localization facing this issue. In L10n you frequently get a long list of very small files, sometime just containing a single word or phrase. In this case there are very few alternative to working base on "jobs", because it would be impossible for translators and edtiros to download/upload each file individually.
Unfortunately, there is no siver bullet for this question...
The short answer is no, you can't. The main Project ListPage ("Projects") for freelancers only shows the freelancer's main projects. So we need to make the freelancer member of the main project.
In order to avoid this situation you could setup another sub-project to contain any critcal information that a freelancer shouldn't see.
The main project is frequently the location where discussions and knowledge management should happen, as opposed to the sub-projects. That's because users are much more likely to enter into a main project than into a sub-project, and participation is usually the biggest issue if you want to use forums for KM.
This questions belongs to the more general question about how you want to map your particular data and processes to the ]po[ application. We'd recommend that you define a few rules for your company on how to use ]po[.
Here are some ideas about subprojects:
Subprojects add a level of complexity for both administrators
(to setup sub-projects) and for users
(to find the location of information). So you should setup sub-projects
only if really necessary. In general we recommend that you use the
forum and filestorage only of the main project. Otherwise your users
might not see that there is information "hidden" in one
of the subprojects.
In ]project-translation[:
We recommend the use of sub-projects if you've got translation projects with more then 20 tasks. This could either be because of the sheer number of documents to translate (such as in the case of a website localization) or because you have to translate a number of files into many languages.
There is a button in the "Translation Details" section to "Create Language Subprojects" which automatically creates a new subproject per language for you. Also, the function copies the "source_xx" files into the subprojects.
In ]project-consulting[:
Sub-projects are automatically generated as part of the project hierarchy when using the GanntProject integration. However, we recommend that you try to reduce the number of tasks and subprojects as much as possible in order to avoid administrative overhead. We usually recommend that a "task" should correspond to 1 day to 5 days of work, and a sub-project to atleast to week of work.
]project-open[ has a built-in "DynField" architecture that allows you to add fields to Projects, Users and Companies (we'll extend this list in the future). DynFields modifies your data model, so it should only be used by experience SQL developers. Please search for "project-open dynfield" in Google for more information. However, it is probably best to post a message in SourceFourge Discussions or to contact us with your particular requirements.
Installing ]po[ successfully in a company has to do a lot with change management. Please contact us for more material and our change management methodology (PowerPoint slides with best practices).
I frequently need to look at the list of "Delivered Projects"
in order to create invoices. Is there a way
to get there quicker then going to "Projects" and then
selecting the corresponding Filters?
You can bookmark most pages in ]po[, so you may just create a shortcut in your Web Browser. This works particularly well for ListPages and reports. However, it doesn't work for pages designed to create new objects.
The customer's project reference ("Customer Project Nr") is included in the full-text-search.
In order to provide your customers with this "privilege" you have to go to "Admin" -> "Profiles" and select the checkbox "add projects" for the profile "Customers".
However, you may want to restrict the ability to create new projects only to "power customers" or a similar group, so we'd recommend you to setup a new Profile for your advanced customers.
