Design to avoid reflection

Hello:
I'm creating an application which loads other classes at runtime (plugins) and it must expose some of these plugins methods to the exterior. All the plugins implement a common interface (ICapsule), however, the methods that need to be exposed are not part of ICapsule and are not known until runtime, and they are specified by the plugin author in a config file provided with the plugin.
I know that if I have the method name and it's signature I can call those methods by reflection, but I would like to avoid using it. My question is: what can I change in my design to avoid reflection? does anyone knows of a pattern I could use? Right now I have come up with some solutions, but none seem optimal:
1.- Instead of a method, create a new class for each exposed behavior and make it implement Runnable. Then store the classes in a hash, with the name of the class as the key and the Object as value and, when needed, look one up and call it's run() method.
2.- Use a switch construct inside each plugin. Let the plugin provide an enumerated type to be compared inside the switch operator and act according to the received value.
Right now I lean towards #2. Does anyone thinks of a better solution or should I definitely go with reflection?

Cesaro wrote:
When the system loads a new plugin, it stores the instance in a list and assigns it a unique ID. These plugins are intended for hardware access, and what actions the user can do with the hardware is determined by the external actions defined in the plugin. When a new plugin is loaded the system registers those actions and the user can discover what actions are available by calling a special function (actually this is a webservice).So the user software gets a list of method names through this webservice (which comes from the config file), which looks maybe something like this:
LEDPlugin.turnLedOn(int whichLed)
LEDPlugin.turnLedOff(int whichLed)
etc...Then the user sends a request back, with the method name, i.e. LEDPlugin.turnLedOn(2). Is that the basic idea?
I don't know of any way to actually call those methods other than using reflection.
All the plugins adhere to a common interface as you said: ICapsule. However this interface only defines a few methods (like init, start, stop, send) but the plugin author can choose to add commodity methods (like turnLedOn) that send predefined commands to the hardware so the user don't need to bother with the internals of how to turn on a LED attached to the parallel port, he/she just calls turnLedOn and it gets done.
How does the application know what to call based on action names is part of the design. If using reflection the name should be the fully qualified name of the class, along with it's signature to make a successful call. If using arbitrary names defined by the plugin author, I would have to come up with the means of assuring that no two action names are repeated and that a given action points to that method of the Object containing the action.

Similar Messages

Changes made to cell reference and cell formula in designer are not reflected in analyser?

Dear Experts,
I have made few changes to Bex Report in Query designer by creating cell reference for a cell and the same cell reference used by the formula created on the adjacent cell.
I changes are working for 1 Report where as , the exact chanegs made in 2nd Report are not reflected in the Analyzer upon execution of Reports by giving exact inputs as fpr the previous report.
Could you please help me resolve this issue ?

Hi Loed,
Please find the screenshot of the query result below.
The cells marked in green are where the cell reference has been created and the ones in yellow will have the formula pointing to the cell reference value.Ideally, the cell values in green and corresponding green cell should be identical(as seen in the 2nd screenshot).
Screenshot1.
screenhot2

Designer does not reflect Inspector changes though source does

Folks,
Don't expect I'm the first to report this but I didn't find anything about it from a forum search so better to report it then ignore it.
Making changes in Inspector for an assign, that is, deleting 2 of 7 copy statements that were initially created through the Designers/Process Map/Assign icon "add copy" function. After creating the 7 copy statements, decided to remove two, used Inspector and deleted the two. Glanced at Designer graphic of the Assign and the two statements were still reflected there. Double-checked Inspector and they were gone, as expected; checked source and the statements were also gone from source, as expected. Went to Overview, hit refresh, waited for rebuild to finish. Went back to Process map, opened Assign statement and there were still 7 copy statements rather than the expected 5, Inspector had 5 as expected, checked source, source had 5 as expected. So, overall the project has the required 5 but Designer is still showing 7. Designer is 0.9.13, Eclipse is at Version: 3.1.0
Build id: I20050627-1435
Thought you might want to know.

Thanks for the bug report. I will forward to the dev team. -Edwin

Need to change infopackage design to avoid overheads

Hi Gurus,
In brief, every weekend the SNP data from the planning area in live cache is extracted into a PSA/Infocube in APO, Then the data from this infocube in APO is loaded into the PSA/Infocube in BW. All together this data extraction process takes more than 12-13 hours. This is not acceptable to the business users as the BW reports after this extraction process are available only on Monday late evening for the APAC region.
To rectify this the user now wants to split the selection for each region.
User Requirement:
This extraction is expected to be regional based on the selection criteria of APO locations and products outlined below. Also the extraction is expected to be split into two periods using a ABAP routine (Fiscal year / period) as given below. The extraction should begin from step 1 and proceed sequentially ending with step 6.
Step 1. APAC : u2013 (ABAP Routine: -31days to +91days)*
APO Locations: 6451, 1515, 1435, 1325,8251, 8451, 8412, 3551, 1517, VS00 to VS99
APO Products:
Info Package 1. 000-0000 to 033-2488
Info Package 2. 033-2489 to 661-2500
Info Package 3. 661-2501 to 922-0000
Info Package 4. 922-0001 to 999-9998
Info Package 5. 999-9999 to ZZZZ-9999
Info Package 6. AA000-0000 to ZZ999-9999
Data loading into Info Packages 1 to 6 should happen parallel .
Step 2. EMEA : u2013 (ABAP Routine: -31days to +91days)*
APO Locations: 7210, 7250, 7254, V200 to V299
APO Products:
Info Package 1. 000-0000 to 033-2488
Info Package 2. 033-2489 to 661-2500
Info Package 3. 661-2501 to 922-0000
Info Package 4. 922-0001 to 999-9998
Info Package 5. 999-9999 to ZZZZ-9999
Info Package 6. AA000-0000 to ZZ999-9999
Data loading into Info Packages 1 to 6 should happen parallel .
Step 3. AMR : u2013 (ABAP Routine: -31days to +91days)*
APO Locations: 5651, 5653,V100 to V199
APO Products:
Info Package 1. 000-0000 to 033-2488
Info Package 2. 033-2489 to 661-2500
Info Package 3. 661-2501 to 922-0000
Info Package 4. 922-0001 to 999-9998
Info Package 5. 999-9999 to ZZZZ-9999
Info Package 6. AA000-0000 to ZZ999-9999
Data loading into Info Packages 1 to 6 should happen parallel.
Step 4. APAC : u2013 (ABAP Routine: +92days to +365days)*
APO Locations: 6451, 1515, 1435, 1325,8251, 8451, 8412, 3551, 1517, VS00 to VS99
APO Products:
Info Package 1. 000-0000 to 033-2488
Info Package 2. 033-2489 to 661-2500
Info Package 3. 661-2501 to 922-0000
Info Package 4. 922-0001 to 999-9998
Info Package 5. 999-9999 to ZZZZ-9999
Info Package 6. AA000-0000 to ZZ999-9999
Data loading into Info Packages 1 to 6 should happen parallel.
Step 5. EMEA : u2013 (ABAP Routine: +92days to +365days)*
APO Locations: 7210, 7250, 7254, V200 to V299
APO Products:
Info Package 1. 000-0000 to 033-2488
Info Package 2. 033-2489 to 661-2500
Info Package 3. 661-2501 to 922-0000
Info Package 4. 922-0001 to 999-9998
Info Package 5. 999-9999 to ZZZZ-9999
Info Package 6. AA000-0000 to ZZ999-9999
Data loading into Info Packages 1 to 6 should happen parallel.
Step 6. AMR : u2013 (ABAP Routine: +92days to +365days)*
APO Locations: 5651, 5653,V100 to V199
APO Products:
Info Package 1. 000-0000 to 033-2488
Info Package 2. 033-2489 to 661-2500
Info Package 3. 661-2501 to 922-0000
Info Package 4. 922-0001 to 999-9998
Info Package 5. 999-9999 to ZZZZ-9999
Info Package 6. AA000-0000 to ZZ999-9999
Data loading into Info Packages 1 to 6 should happen parallel .
The flow of data from APO to BW is also needs to be changed. Currently the data for the BW Cube is feed from the APO cube, Now the BW cube will get data from 2 datasources Instead of fetching data from the APO cube.
Proposed Design:
For the above requirement, Totally we have 6 steps. For each step we need to create 6 infopackages, therefore totally 36 infopackages needs to be created for one datasource. Finally we land in creating 72 infopackages for 2 datasources which will cause overheads while data load.
Please review the requirement and suggest us a robust solution so that the no.of.infopackages can be reduced to some extent.

Hi Jackie,
Since with every infopackage you have changed only in time period and APO locations.....also since data source is the same and there is no restriction on the time at which data is available for loading...you can try to combine the data selections together...
If it results in huge amount of data volume then may be you may you can store these selections in the a Z table as a variant and use them in the infopackage dynamically through infopackage level routine...but that will be too much overhead and tough to maintain in case of failures.
Thanks
Ajeet

How to avoid zombie in biztalk?

Hi Biztalk expert!
I need find out the solution for biztalk zombie (very urgent), I have read many article about zombies, but I can't resolve problem.
I'm finding the approach to change the orchestration ( not use external tool/script as zombie management tool or WMI script).
below image is my orchestration.
any idea is good appreciated! If you can change this orchestration to resolve zombie, please give me screen shot.
sorry if my question duplicate in this forum.
Thanks in advance.

Maybe the queues would help?
It works if you process messages individually and don't have to compound some date from one message to another (like creating batch or calculate some aggregates like message counter or some amount from the message sequence). Seems your processing could be implemented
without orchestration at all, only with pipelines and maps on ports
For example if you need only an ordered delivery.
The possible design is:
Your messages go to the MSMQ queue (or ServiceBus Queue)
Send port(s) consumes messages from queue, this port is with ordered delivery on. There is no convoy orchestration and no orchestrations at all.
In this design you avoid convoy completely hence avoid the zombie problem and all problems around the singleton orchestration (the orchestration status, the complexity of the error handling).
Leonid Ganeline [BizTalk MVP] <a href="http://social.technet.microsoft.com/wiki/contents/articles/20258.biztalk-integration-development-architecture.aspx">BizTalk Development Architecture</a>

How can I design Load Balancing for distant Datacenters? without single point of failure

Dear Experts,
We are using the following very old and passive method of redundancy for our cload SaaS but it's time to make it approperiate. Can youplease advise:
Current issues:
1. No load balancing. IP selection is based on primary and secondary IP configurations. If Primary fails to respond, IP record for DNS changes to secondary IP with TTL=1min
2. When primary server fails, it takes around 15 min for clients to access the servers. Way too long!
The target:
A. Activate a load balancing mechanism to utilized the stand-by server.
B. How can the solution be designed to avoid single point of failure? In the previous example, UltraDNS is a single point of failure.
C. If using GSS is the solution, how can it be designed in both server locations (for active redundancy) using ordinary DNS server?
D. How can HSRP, GSS, GSLB, and/or VIP be used? What would be the best solution?
Servers are running ORACLE DB, MS SQL, and tomcat with 2x SAN of 64TB each.

Hi Codlick,
the answer is, you cannot (switch to two web dispatchers).
If you want to use two web dispatchers, they need something in front, like a hardware load balancer. This would actually work, as WD know their sessions and sticky servers for those. But remember you always need a single point for the incoming address (ip).
Your problem really is about switchover groups. Both WD need to run in different switchover groups and need to switch to the same third software. I'm not sure if your switchover software can handle this (I'm not even sure if anyone can do this...), as this means the third WD needs to be in two switchover groups at the same time.
Hope this helps,
Regards,
Benny

Questions VLAN design best practices

As per best practices for VLAN design:
1) Avoid using VLAN 1 as the “blackhole” for all unused ports.
2) In the local VLANs model, avoid VTP (use transparent mode).
Point 1
In a big network, I'm having VLAN 1 as the blackhole VLAN. I'd like to confirm that, even if we're not complying with best practices, we're still doing fine.
a) all trunk ports on all switches have the allowed vlans explicitly assigned.
b) about all ports on all switches are assigned to specific data/voice vlans, even if shutted down
c) the remaining ports (some unused sfp ports for example) are shutted down
d) we always tag the native vlan (vlan dot1q tag native)
So, no data is flowing anywhere on VLAN 1. In our situation, it is safe to use VLAN 1 as blackhole VLAN?
Point 2
Event if we're using local VLANs model, we have VTP in place. What are the reasons of the best practice? As already said, we allow only specific VLANs on trunk ports (it's part of our network policy), so we do not have undesired layer 2 loops to deal with.
Any thoughs?
Bye
Dario

We are currently using VTP version 3 and migrating from Rapid-PVST to MST.
The main reason for having VTP in place (at least for use) is to have the ability to assign ports to the correct VLAN in each site simply looking at the propagated VLAN database and to manage that database centrally.
We also avoid using the same VLAN ID at two different sites.
However, I did find something to look deeped: with MST and VTP, a remote switch can be root for a VLAN it doesn't even use or as active ports into, and this doesn't feel right.
An example:
1) switch1 and switch528 share a link with allowed vlan 100
2) switch1 is the root for instances 0 and 1
4) VLAN 100 is assigned to instance 1
5) VLAN 528 is not assigned to any particular instance so it goes under instance 0
6) VLAN 528 is the Local Data LAN for switch528 (switch501 has VLAN 501)
swtich528#sh spanning-tree vlan 528
MST0
Spanning tree enabled protocol mstp
Root ID Priority 24576
Address 1c6a.7a7c.af80
Cost 0
Port 25 (GigabitEthernet1/1)
Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec
Bridge ID Priority 32768 (priority 32768 sys-id-ext 0)
Address 1cde.a7f8.4380
Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec
Interface Role Sts Cost Prio.Nbr Type
Gi0/1 Desg FWD 20000 128.1 P2p Bound(PVST)
Gi0/2 Desg FWD 20000 128.2 P2p Edge
Gi0/3 Desg FWD 200000 128.3 P2p Edge
Gi0/4 Desg FWD 200000 128.4 P2p
Gi0/5 Desg FWD 20000 128.5 P2p Edge
switch1#sh spanning-tree vlan 501
MST0
Spanning tree enabled protocol mstp
Root ID Priority 24576
Address 1c6a.7a7c.af80
This bridge is the root
Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec
Bridge ID Priority 24576 (priority 24576 sys-id-ext 0)
Address 1c6a.7a7c.af80
Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec
Interface Role Sts Cost Prio.Nbr Type
Should I worry about this?

In pages, the graphic inspector doesn't allow for reflection.

When I open a new page in pages the graphic inspector highlights shadow as selectable but no reflection. I don't know what might be interfering with this.
I would like to be able to design words with reflections and shadows.

Seems to be a Preview oddity. It search entire words. If you search for reflections, it find it !
Thank to your post, I begin to understand why so many users miss infos which I find easily.
I don't use it but Adobe Reader which bring to page 154 as 2nd occurrence.
It's more powerful than Preview when used to search something.
Jerrold Green1 wrote:
Further, the line on which the term exists, as shown in your post, does not allow the selection of an individual word or character.
The pointed page give a list of objects upon which we may apply the effects. Text box is one of these objects, words or single characters aren't. Isn't it a logical behavior ?
If you drag and drop the screenshot on your Desktop, you will see that there is a reflection effect below the word term.
It is enclosed in a text box so the effect apply to the box.
In an user guide, every word is important as well as a bridge pier. Remove one of them and the construction collapse.
Yvan KOENIG (VALLAURIS, France) mercredi 25 mai 2011 18:42:49
Please : Search for questions similar to your own before submitting them to the community
To be the AW6 successor, iWork MUST integrate a TRUE DB, not a list organizer !

All reflection and no type safety makes Jack a dull boy...

There has got to be a way of refactoring which takes reflection into account, right? As my project grows larger it's getting crazy trying to remember what needs to be changed when I modify a method name, etc. What are people doing about this? I have code that looks like this:
UpdaterProcessor canUP = new UpdaterProcessor("setCancelSentTime", new Date().getTime() ) ;
Filter f1 = new EqualsFilter("getStatus", OrderStatus.NEW); // only req cxl on NEWs
ConditionalProcessor cp1 = new ConditionalProcessor(f1, canUP);
Order.cache.invoke(o.getID(), cp1);Are people just coding stuff like this by extending abstract classes to avoid reflection or something?
Thanks,
Andrew

There is no type-safe way around this, if you don't want to register every one of the type-safe processor or filter or whatever classes you have to send through the network in the POF configuration (which IS the future).
The safest way to use these processors is to create constants for every method you want to access and use only those constants as extractor names. Alternatively you can create enums instead of constants, and create your extractors instances with the enums passed instead of the method names, and put in a checkstyle or pmd rule which ensures that you don't directly instantiate filters or extractors or etc. with the method names.
Best regards,
Robert

Avoid conditionally suprressed sub report to fetch data from DB.

Do we have any option in crystal designer to avoid a suppressed subreport to fetch data from DB?
I have two sub reports in Report Footers say 'a' and 'b'.Report footer sections have a suppression formula based on parameter input.
Hence based on the parameter input only one sub report will get printeded.However I noticed that only one sub report is getting printed but both sub reports are fetching the data from DB.
I have also tried putting the suppression formula in suppress formula field of Sub report using format sub report option but it did not help.
Any idea how can we avoid a subreport that is suppressed to fetch data from DB.
Regards

Hi,
Suppress can only suppress the data, it cannot stop the query attached being running from the database, One thing you can do only if your input parameter is a DB level parameter is you can have a case statement in the where clause of your query to avoid reading the records there by increasing the performance of the query.
Say, your parameter value 1 specifies that you want to suppress the report and in other words, you dont want to read the records from the DB.
Select feild1, field2,....from Table1
where 1 = case when @param = 1 then 0 else 1 end
Hope this helps,
Thanks
-Azhar

Do the new Time Capsules have a different design?

Having had my TC die, but out of serial number range for the paid repair, I did the job myself and it now works again - but for how long?!!
The fault was obviously overheated capacitors, the same as the articles on the internet show, so I'm not sure whether it is sensible to buy a new TC.
If the basic flaw has not been fixed with a new design that avoids these capacitors overheating then the same thing will likely happen again.
Does anyone know if the new ones have been redesigned?

Millions of later generation Time Capsules have been sold and we hear of a failure having to do with the power supply every once in while. Based on that, my opinion would be that the design is just fine. If my memory and observations are correct based on forum posts, I would also say that over the last 18 months, a higher percentage of AirPort Express devices fail than Time Capsules, even though the Express is a much simpler device.
Having looked inside at both an early generation Time Capsule and a later generation Time Capsule, there are definitely different types of capacitors used. Not having a schematic of the circuitry, I can't tell you whether there were any design changes in later versions in the power supply section or not. That would be a question for the Apple product engineers.

Representing coax cable in Multisim

I am trying to simulate a 50 ft coax cable connected to a signal generator on the left-hand side and a 50 ohm load on the right-hand side. How would I build this using Multisim?

Hi,
The best option for you is to model the coax line as a 50ohm lossless transmission line and just calculate the delay it will introduce in second. You can use the following formula:
delay time = length of the cable/ TEM velocity (3*10^8m/s)
If you want to include losses you have to calculate the RLCG parameters of the coax cable and include them in the component Lossy Transmission Line.
You can learn more about this here.
I attached a design file that includes these components, always make sure that your load is the same as the transmission line characteristic impedance (matched) to avoid reflection.
Hope this helps,
Mahmoud W
National Instruments
Attachments:
coaxial TL.ms11 ‏113 KB

IMac ergonomics?

I can't find any real information online on any blogs or on this site, on how I should set up my iMac 24" on my desk.
It seems to me, that the usual advice on what height your eyes should be (in line with the top of the screen) can't apply to the 24" iMac. Can anyone point me to any ergonomic resources for the iMac? I am a web designer and work at my computer 7 or 8 hours per day, so it's really important that I am sitting properly at my desk. Thanks - my first post!

Hey Jane,
Well those OSHA standards were set before the larger monitors were standard. I would say follow the standards for seating position and keyboards then set the monitor at a comfortable height.
Here's Apple's site about ergonomics:
www.apple.com/about/ergonomics
This is from the 2009 iMac manual which you can get here:
http://support.apple.com/manuals/#desktopcomputers
"Position the display to minimize glare and reflections on the screen from overhead lights and nearby windows. The stand lets you set the display at the best angle for viewing, helping to reduce or eliminate glare from lighting sources you can’t move.
Thighs tilted slightly
Shoulders relaxed
Screen positioned to avoid reflected glare
Forearms and hands in a straight line
Forearms level or tilted slightly
Lower back supported
Feet flat on the floor or on a footrest"
The picture is identical to the OSHA picture except it's a drawing
I think the seating position and arm position is the most important. Since you're short you might need a special workstation to be comfortable.
You know in the old days we were all trained on proper typewriter posture and technique. I see people pounding on their keyboards all the time, usually out of anger, then they wonder why their arms and wrists hurt.
Richard
Message was edited by: spudnuty

What do i do when external optical drive doesn't show up in finder? how can i import software, music, etc. when the drive isn't recognized/mounted/whatever?

the internal optical drive on my early 2006 20"imac has mechanical problems ejecting discs. i've gone thru all the recommended steps to eject discs, and have luckily been able to eject the last 3 discs but only after spending 30-60 minutes going thru all of the steps over and over, and today just trying "eject disc" in itunes since i had itunes open. i don't want to put another disc in - the drive clearly has problems with the mechanical eject mechanism, and from what i've read that's not unique to my computer.
i bought an external optical drive: lite on, clamshell design to avoid problems with the eject mechanism; i contacted the mfr who said it's mac compatible.
the drive plays dvds, which is all i've tried so far, but doesn't show up in finder. so i don't know how to use it to load software, or music, or anything besides watch dvds.
is there something i can do to get the drive recognized/mounted/whatever? i would prefer to not spend the $220 which was quoted to me by an authorized mac repair company, even tho i think that's probably a reasonable cost given the delicacy of the operation as shown on ifixit.com.
thanks!

I have an external LaCie burner and it will not show up on my Mac unless I insert a CD or DVD - in that case, the media will show up (I have these settings in the Pref Pane):
If you want to burn files or music to blank media, you can use Finder or Disk Utility for files, but should use an iTunes playlist/iTunes for music and a movie burn app for videos/movies (such as iDVD, Roxio Toast, Burn, etc).
If you simply want to drag files over as a backup, an external hard drive is a much better solution than a burner.

How do I move Elements 8 projects (slide shows) from a Windows XP computer to new Windows 7 computer that also has Elements 8 on it so that the slides shows will work (Elements 8 works well on the Windows 7 computer & I have done several new slide show pr

How do I move Elements 8 projects (slide shows) from a Windows XP computer to new Windows 7 computer that also has Elements 8 on it so that the slides shows will work (Elements 8 works well on the Windows 7 computer & I have done several new slide show projects on it)? I tried backup & restore without success. I have also found the .psess files and copied them to the catalog in the new computer but when the sides shows are opened none of the images or music is there. Only the text that I entered shows. I have even tried to replicate the file paths from the XP computer to the new computer to see if that will help.

UPDATE: By a stroke of luck I happened to find a catalog link (a PSE8D file) in the 'My Catalog' file where the imported data was stored. When I clicked on it to investigate what is was 'magically' all of my transferred projects appeared in Organizer. Upon further investigation I found that the projects transferred from my XP computer were stored in a hidden catalog that had to be accessed through the Custom Location option in Catalog Manager (accessed by clicking File then Catalog then Custom Location). No Windows or Elements search revealed this fact. I am appreciative that I just got lucky. The Adode/Photoshop company was absolutely of no help. If fact to me their structure is designed to avoid contact with their customers unless you're buying a new product. Their products have many good features (even though user unfriendly) and it's a shame they don't value their customers more once the sale is completed.
Bottom line: All of the projects (sideshows) have been recovered and work in my new Windows 7 computer. I hope my experience helps others.

Design to avoid reflection

Similar Messages

Maybe you are looking for